~- e .' - ~neg!Prd 'and Ivestment Study ndustry Operations Divlsion etna Caribbean Region- A . .~ . ... > .......... q X tz4 3T,3 -ýé i, 44 -33 , 3 3 3 3"4 V,3 ' 3'~-;3. -'' - ea t . t 3; 3 '.~ ' ' 333~33 ~ 3' "' 3 3W' . . - 3>."3~.- 3' . 3 -' ~ '~ - - >"~$3- ' ~ . ~ 33-33 .~>3 3 .433 33'3 5 ~ 3 ~ . ~ ~ 33 3 - - ~' 3 -' ' Mw 314 Xr. b 4 '3~ 3 ~3 ~ '334 3 3' ~ ~ .>~3 é- '' 33 ~3' , 33 ~ > ~~ .33333 ~ - 3M - ¶ --- iBMIL ENERGY PRICING AMD lNVESTMENT STUDY CURRENCY EQUIVALENTS Currency Unit = Cruzeiro (Cr$) US$1 = CR$50.75 (May 1, 1990) Cr$1580 (February 27, 1992) UNITS OF MEASURE 1 Metric Ton (m ton) = 1,000 Kilograms (kg) 1 Metric Ton (m ton) = 2,204 Pounds (lb) 1 Meter (m) = 3.28 Feet (ft) 1 Kilometer (km) = 0.62 Miles (mi) 1 Cubic Meter (m3) = 35.3 Cubic Feet (cu ft) Mm3/d = Thousand cubic meters per day MMm3/d = Million cubic meters per day boe = Barrel of oil equivalent bpd, b/day = Barrels per day BCM = Billion cubic meters MBPD = Thousand barrels per day MMBbl = Million barrels CMD = Cubic meters per day 1 Barrel (Bbl) = 0.159 Cubic Meter 1 Barrel (Bbl) = 42 US Gallons 1 Metric Ton of Oil (API 300) = 7.19 Barrels 1 Kilocalorie (kcal) = 3.97 British Thermal Units (BTU) 1 Ton of Oil Equiv. (toe) = 10 Million kcal (39.7 mil. BTU) kV (kilovolt) = 1,000 V (volts) kW (kilowatt) = 1,000 W (watts) kWh = kilowatt hour 1 Megawatt (MW) = 1,000 kilowatts MWh = megawatt hour 1 Gigawatt (GW) = 1,000,000 kilowatts GW.. = gigawatt hour 1 Terawatt hour (TWh) = 1,000 gigawatt hours 1 Cubic Foot of Natural Gas = 950 BTU 1 Kilowatt-hour = 3,411 BTU FOR OFFICIAL USE ONLY ENEr Parcme Ann NaMn=T aI=Dr LIST O ABMEVIATIOMS AND ACRONYMS AIC Average Incremental Cost SAU Business As Usual Scenario BNDES National Bank of Economic and Social Development BPT Billed Power Tariff BST Bulk Supply Tariff CBER Companhia Betadual de Baergia Eletrica CEO Companhia etadual de Gas do Rio de Janeiro CBMIG Companhia Inergetica de Minas Gerais CBNAL National Alcohol Executive Commission CESP Companhia Energetica de Sao Paulo CHESY Companhia sidro-eletrica de Sao Francisco CRAL National Alcohol Council CNE National Energy Commission CNP The National Petroleum Council COMAS Companhia do Gas do Sao Paulo CONPET Program of Conservation and Rationalization in the Production and Use of Petroleum Products COPEL Companhia Paranaenase do fnergia. CPI Consumer Price Index CRRMS Energy Matrix Coun-vion DNM3 The National Department for Water and Electric Energy DRC National Fuels Department DNPM National Department of Mineral Production EGTD Guaranteed Energy for Fixed Tem LTROBRAS Centrais Bletricas Brasileiras, S.A. ELETaoNORTU Centrais Bletricas do Norte do Brazil, S.A. ELETROSUL Centrais Eletricas do Sul do Brasil, S.A. 8MP Environmental Master Plan RPS Efficient Pricing Scenario FGV Gtalio Vargas Foundation FINSOCAL Social Investment Fund PRD National Development Fund FUP Uniform Price Rate FURNAS Centrais Xletricas Furnas, S.A. PRI Foreign Exchange Index GASMIG Companhia de Gas de ina Gerais GCOl Coordinating Group for the Operation of the Interconnected Systems GCPS The Coordinating Group for System Planning Go? The Global Guarantee fund GPI General Price Index MV High Voltage IAA . Sugar and Alcohol Institute IBAA Brazilian Institute for the Environment and Renewable Natural Resources (Ministry of the Interior) IBF Brazilian Institute for Forestry Development (Ministry of Agriculture) This document has a restricted distribution and may be used by vecipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. INS Imposto do Circulacao do Mercadorias a Servicos IPI Tax on toduatrial Products XUCL Tax on Combustibles and Lubricants IVE8 Electricity Tax ImC Tax on Wholesale of COmbustibles LPG Liquifted Petroleum Gas LRHC Long Run Marginal Cost Ur Low Voltage MCL Minimum consumption level MDRE Evaluation Model for Energy Demand HIC Ministry of industry and Commerce MNW The Federal Ministry of Mines and Energy MOP Ministry of Finance HDI Ministry of Infrastructure PAEP Assistance Program for Public Service Employees P800 Program for Economizing Fuels PETROMISA Subsidiary of PETRoBRAS to evaluate hydrocarbon and mineral potential of sedimentary basins INTEPIRAS Subsidiary of PETROBRAS for the import and export of commodities PETROPERTIL S"haidiary of PETROBRAS for the production and sale of fertilizers and raw materials BRASPUTaM Subsidiary of PXTROBRAS for the overseas exploration and product.Lon of hydrocarbons and the provi'sion of services and technical assistance PETROBRAS Petroleo Brasileiro, S.A. PXTROQUISA Subsidiary of PMTROBRAS for the production and sale of petrochemicals PETRO8RAS/BR Subsidiary of PETROBRAS for retailing of petroleum products Pis Social Integration Program PPT Published Power Tariff PROALCOOL National Alcohol Program PROCEL National Electricity Conservation Program PRODUL The Voluntary Prooram to Bconomize Diesel and Lubricants PRORN Program to Rationalize Nnergv RENCOR National Compensation Reserve for Remuneration REVISE Power Sector Institutional Study RPT Received Power Tariff SRAP Special Secretariat for Supply and Prices 8ST Secretariat for the Control of Public Enterprises (Ministry of Finance) S0 State-owned Enterprise THS Seasonal Tariff BRAZIL ENRGY PRICING AND INVBSTMENT STUDY Tble of Contents PORWORD . . . . . . . . . . . . . . . * * * * * * * * * * * * * * z PRSVACX .. . . . . . . . . . . . . . . ABSTRACT. . . . . . . . . . . . . i........ a EXECUTIVE SUMMar , . . . . . . . . . . . . . . . . . . . . . . . . . Uii CHAPTHR I: AN OVERVIT OF THE ENERGY SECTOR . . . . . . ..* . .... .* . 1 1.1 Introduction .... .... .... .... ... I 1.2 Supply and Demand . . . . . . . . . . . . . . . . . . . . . 2 1.3 Energy and the Economy . . . . . . . . . . . . . . . . . . 8 1.4 Income and Price Elasticities . . . . . . . . . . . . . . 9 1.5 Energy Resource Requirements . . . . . . . . . . . . . . . 11 1.6 Government Policy . . . . . . . . . . . . . . . . . . . . . 14 CHA.,PTSR Ils THB INSTITUTIONAL STRUCTURE OF THE ENERGY SECTOR . . . . 16 11.1 Introductions The Role of the Federal and State Governments . . . . . . . . . . . . . . . . . . .. 16 11.2 Sector Institutions . . . . . .*. . . .. . . . . . . .*. 17 11.3 Institutional Aspects of Pricing .. . . . . . . . . . . . 20 11.4 Conclusions and Recommendations for Change . . . . . . . . 22 CHAPTER 1s1t ELECTRIC POWER . . . . . . . . . . . . . . . . . . . . . 27 111.1 The Electricity Market . . . . . . . . . . . . . . . . . 27 111.2 Electricity Demand Forecasts . . . . . . . . . . . . . . 27 111.3 Power System Planning, Expansion and Investment . . . . . 33 111.4 Long-run Marginal Cost . . ............... 36 111.5 Electricity Tariffs . . . . 9..... . 37 111.6 Financial and Fiscal Aspects .............. 44 111.7 Conclusions and Recommendations for Change . . . . . . . 46 CHAPTER IVs PETROLEUM PRODUCTS, ALCOHOL AND NATURAL GAS . . . . . . . 49 XV.1 The Liquid ftels and Gas Market . . . . . . . . . . . . . 49 IV.2 Liquid Fuels Demand Forecast . . . . . . . . . . . . . .. 50 iV.3 Petroleum Sector Development Strategy . . . . . . . . . . 53 IV.4 Refining of Petroleum Products . . . . . . . . . . . . . 55 IV.5 The Pricing of Petroleum Products, Alcohol and Natural Gas . . . . . . . . . . . . . . . . . . . . . . 56 IV.6 Financial and Fiscal Aspects . . . . ... . .. . . . ... 66 IV.7 Co clusions and Recommendations for Change . . . . . . . . 67 CHAPTER Vt NERGY CONSERVATION AND DEMAND MANAGEMENT . . . . . . . . 71 Vo, Introduction . . . . . . . . . * . . . . . . . . . . . . . 71 V.2 Energy Intensity in the Brazilian Economy . . . . . . . . . 71 V.3 Instruments of 8CDM Policy .. ....... . . .. . . . 74 V.4' Conclusions and Recomendations for Change . . . . . . . . 76 CHAPTER VI: CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . 78 VI.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . 78 VI.2 The Issues * . . . . . . . . . . . . . . . . . . . . . . . 78 VI.3 The Strategy . . . . . . . . . . . . . . . . . . . . . . . 81 VI.4 -Future Action . . . . . . . . . . . . . . . . . . . . . . 84 ANNEXES AM= I 1.1 Domestic Inflation and Exchange Rates 1.2 Energy Resources 1.3 Gross Domestic Production of Primary Energy 1.4 Gross Domestic Consumption of Primary Energy 1.5 Total Consumption of Diesel Oil 1.6 Automotive Fuels in Transportation 1.7 Export and Consumption of Gasoline and Alcohol 1.8 Evolution of Sales of Vehicles in Brazil 1.9 Energy Demand Forecasts and Methodology ANNE III III.1 Power Generation Forecast (ELETROBRAS) III.2 Power Market Structure (ELETROBRAS) 111.3 Power System Planning, Expansion and Investment II.4 1989-1998 Power Plant Commissioning Program III.5 1989-1998 Transmission/Substation Expansion Program 111.6 1989-1991 Distribution Expansion Program III.7 Power Sector Investment Program III.8 Probability of Energy Shortages 1989-1995 111.9 Electricity Tariffs III.10 Sources and Applications of Funds 111.11 Suamary Tablet Interregional Flow of Funds Under the GQF and RENCOR ANNEX IV IV.1 Petroleum Sector Development Plans, Investment and Strategy IV.2 Petroleum Balances IV.3 Reserves Evolution and Cost of BBL Found and Developed IV.4 Refining Capacity IV.5 Yield of Refineries IV. PETROBRAS Historical Investment Program IV.7 Natural Gas Reserves IV.8 Crude Oil and Natural Gas Supply Projection IV.9 Petroleum Balance and Gas Supply Forecast IV.10 Petroleum Balance Forecast and Cost (Option 1) XV.11 Petroleum Balance Forecast and Cost (Option 2) IV.12 Petroleum Balance Forecast and Cost (Option 3) IV.13 Petroleum Balance Forecast and Cost (Option 4) IV. 14 Road User-Charges I15 Evolution of Real Prices Indices for Gasoline and Diesel Oil (1954-1988) 1V.16 (Part 1) Estimates of Energy Consumption by Road Vehicle Fleets Based on Existing Dieselization of Fleet IV.16 (Part I) Estimates of Energy Consumption by Road Vehicle Fleets Assuming no Further Dieselization of Fleet After 1974 IV.17 The Economic Cost and Netback Value of Natural Gas IV.18 Some Social Considerations in Gas Pricing IV.19 Net Transfers to the Federal Government From the Petroleum Sectors 1980-1988 IV.20 International Comparison of Petroleum Taxes ANNEX V V.1 Energy Conservation Programs MAPS IBRD 22230 Natural Vegetation 22231 North-Northeast Transmission-Network 22232 South-Southeast Transmission Network 22233 Map of Brazil 22234 Natural Gas Pipelines 22235 Power Generation Sources in Northern Region 22236 Principal Regional Electrical Interconnections 22237 Principal Coal Deposits 22238 Interregional Power Flow 1. A draft version of this report was issued on may 9, 1990 and sent to the new Brailian administration, which had just taken office on March 15 1990. The document was subsequently reviewed by the Brazilian authorities, and detailed conmunte were sent to the Bank in February 1991, after prolonged internal discussions. A Bank mission, comprising Messrs. 8. Ettinger and R. Bates, discussed those cownents with the Brazilian authorities on March 21, 1991. Although some limited changes were subsequently made in the body of the text, the Government a couents have been addressed primarily in this Forewcrd. The Foreword alec selectively updates information on the energy sector, to reflect the situation at the time of the March discussions. The Energy Pricing and Investment Study is being issued now in order to make the underlying analysis available to a wider audience. We have not undertaken the massive task of updating the text itself, as we believe that the substance of the report's conclusions has not been significantly affected by developments since the May 1990 version. ii. The overall philosophy and policy of the present Government -- for example, as contained in the President's August 1991 discussion document, Brasil* Um Projeto de Reconstrug&io Nacional" -- has been generally consistent with the recommendations of the reports (a) improved fiscal responsibility; (b) public sector investment and pricing based on economic criteria; (c) carefully-targeted subsidies to the poor; (d) a greater role for the private sector; and (e) more competition. The report's findings have also been largely reflected in the draft Energy Matrix, recently submitted to President Collar by the *Commission Responsible for the Usexamination of the Energy Matrix "(CRRE). iII. We fully agree with the Brazilian authorities that several important influences outside the energy sector have exerted a profound effect on energy policy, and constrained the Government- a actions, especially in the short term. Among the more important of these influences were: (a) continuing high inflation; (b) regional development policy; (c) the 1988 Brazilian constitution; (d) the severe financial difficulties faced by power companies in the Nkrth and North- East; and (e) existing contracts with large energy consumers. The report recognizes these constraints, but it also concludes that Government policies have not always been well-suited to dealing with them, and that they are generally not binding in the longer term. In particular, the report argues that: (a) Low energy prices did not materially help to combat inflation; on the contrary, they increased the public sector deficit, and hence, aggravated the inflation problem, while simultaneously bringing the power seutor to a financial crisis (paras. 1.34, 2.27 and 3.58); (b) Uniform national energy prices are not an economically efficient way to promote the Government's regional strategy ((paras. 3.47, 3.61, 4.27, 5.19 and 6.14-6.15). The resulting subsidies are not transparent; they distort prices and demand, and thereby encourage inefficient energy consumption and discourage efficient energy produ--ion in remote areas; and the financial/economic flows are not even always in the intended direction); (a) Despite the constitutional monopoly of the State (exercised by PETROBRAS) in most areas of the hydrocarbon sector, more could be done already to introduce private sector participation (paras. 2.6-2.7 and 2.29); furthermore, when the constitution is revised (expected by 1993), particular attention should be given to the monopoly of PETROBRAS and to the role of the private sector (paras. 21 and 4.60); and (d) The power companies of the North and North-East do not benefit from their lower long-run marginal cost (LRMC) of generation, because of the uniform national bulk supply tariff; while they suffer from the uniform national retail tariff, due to their higher distribution costs (paras. 3.47, 3.56, 3.57 and 3.61. iv. Given the inter-relatedness of the issues, it is essential to introduce a package of policy measures in the energy sector. not merely action on energy pricing. The other key elements of the package, as identified in the report, comprises more private sector participation (paras. 20, 21 and 6.21)! a strong and independent regulatory agency (paras. 18, 21 and 22)1 less direct Government control of tho sector agencies (paras. 22 and 6.21); and continued emphasis on least-cost planning and operation, to avoid passing on to consumers, through higher prices, the inefficiencies of the electric power, alcohol and pqtroleum sectors (paras. 11, 14, 16 and 6.22). Eneray Pricina v. The underlying methodology used in the report with regard to energy pricing relies upon Z&RC as the appropriate reference point for pricing in electricity supply and natural gas, which are generally national monopolies (at the distribution end) and not traded (Sections 111.5.2 sad IV.5.5); and on international (border) prices for petroleum products and alcohol, which are tradeables (Sections IV.5.3 and IV.5.4). We do not agree with the view that LRMC pricing per se is more cost-inflating than other methods, as it tends to transmit to other parts of the economy the inefficiencies of the electric power sector. The problem of costs relates to the lack of competition or other incentives for efficiency, not to the pricing rule. In the long run, to the extent that IRMC pricing promotes an efficient pattern and growth of consumption, costs should be lower than otherwise, not higher. Furthermore, the fact that the structure of petroleum products' consumption and production in Brazil differs from the international market does not invalidate the use of international prices. Of course, we are not suggesting that Brazil should rigidly follow short- term (e.g. day-to-day) fluctuations in international petroleum product prices; and while an alternative market to Rotterdam could be chosen, the results of the analysis -- for example regarding the prices of petroleum products and alcohol, and the strong economic arguments in favor of greater penetration of natural gas -- would not alter. In particular, naphtha would still be shown to be one of the most hea'ily subsidized petroleum products (paras. 15 and 4.25). vi. We were informed that steps have recently been taken to introduce limited price differentiation for petroleum products, to allow for differences in transport and distribution costs from the "secondary bases." We also recognise the Government's position that, in the longer term, the importance of the issue may be reduced, if regional cost differences decline, as the distribution system for petroleum products is improved, for example through a more extensive network of product pipelines. Our own judgement, however, is that cost differences among secondary bases will remain sufficiently important that they should be reflected in prices for a long time to come. vii. In electricity aupply, the recent decisions to change the system of uniform prics by incorporating differences in distribution costs into prices is an important first step. viii. Extensive institutional changes have been made by the Governwent of President Collor. Notably, the Ministry of Mines and Energy has been subsumed within the Ministry of Infrastruzture; while SEAP, IAA, CNP and CNN have been replaced or eliminated. Nevertheless, we believe that these changes do not affect the substance of the report's recommendations. Power Svstem Exoansion Proaram ix. The report strongly supports the continued investigation of a number of alternatives for the expansion of the power system, consistent with least-cost development (paras. 7. 11 and 3.21-3.26). According to information made available to us in March, 1991, especially the latest investment program ("Plano Decenal 1991-2W4O," by the GCPS, undated), the power sector is takirg several initiatives which are fully compatible with those recommendationes (a) no new expenditures are projected for hydroelectric plants in the Amason or tua nuclear power (beyond the possible completion of Angra II); (b) efforts are underway to develop projectr with neighboring countries, e.g., to import gas or gas-based power from Bolivia and to interconnect with the power system in Argen!ina; (c) the amount of thermal plant may be increased, using domestic coal from open- cast mining in the South, heavy fuel oil in S&o Paulo and Minas Gerais, and natural gas, principally from Urucd in the Amason; and (d) cooperation with the alcohol producers is being sought to increase the amount of power generation from bagasse. x* We fully recognize NLTROBMS' position that, under present cost calculations, hydroelectric projects are the least-cost way of generating power, notwithstanding transmission and environmental costs. However, we continue to urge the Brazilian authorities to consider the use of some imported as well as domestic natural gas and coal for power generation, although we emphasize that any development of coal-fired plant, whether using local or imported coal, would have to make provision to reduce the pollution to an acceptable level, which would impose a cost penalty on coal in the least-cost analysis. Of course, we re4 agnize that a reduction in the alcohol program, which would ensue if the Thes doerment confirm that negotiations with Argentina and Botivia hae ben initiated for the lort of natural gas. pricing recommendL%tions in this report were followed, would reduce the availability of bagease for power generation. lectricity Demand Prcast. Eneray Conservation and DEMand ManaOement ai. The report considers in detail, in Section III.2, demand forecasts for the power sector, and attempts to estimate the impact of efficient pricing, i.e., based on LRMC. Given the substantial changes which have taken place in the SLTROBRAS forecast, as a aormal part of revising its investment program, we have also made new dematct forecasts, utilizings the more recent information, provided by ITROAPAS, on the actual evolution of the power market, electricity tariffs and GDP in 1989 and 1990; and the same figures for future GDP growth as ELATROBMAS. The assumptions regarding future GDP growth are in Table 1 belowl the new forecasts are in Table 2, which replaces Table 111.4 in the report; and paras. xiii-xiv discuss the results, replacing Section 111.2.2 and para. 3.60 in the report. It should be noted that ELETROSMS does not accept the Bank I a demand forecasts. xii. The "Business as Usual" (BAU) scenario assumes that the tariff structure is unaltered and remains at the 1990 level, in real terms. In the 8fficient Pricing Scenario" (BPS), residential and rural electricity tariffs are brought into line with LMC by 1995, implying increases of 157% and 185% respectively. Industrial and commercial tariffs are increased by 58% and 37%, to reach LRMC by 1992. According to the scenarios, the impact of these very substantial price increases is (not surprisingly) significant, rliducing electricity demand by nearly 29% under BPS compared with BAU in 1395 and 2000. Tabl. Ahernatve Forecat Sceara Pic Convad 1990 Rica Reach level of LRMC by: Impich 1992 for industrill and an 1995 for meidenia and na oasumm. GODP Gowth rate 1991: 2% Same as BAU Same as RAU 1992: 4% 1993-1994:5.5% 1995-2000:6.0% The BUTROBRAS forecast is between the BAU and BPS scenarios, being 6.2% lower than those of BAU in 1995; and 7.8% lower by 2000. These results confirm that the new BLaTBRAS forecast makes some (implicit) allowance for price increases in real terms (apparently to an average level of US$S4/MWh, expressed in prices of July, 1989, as agreed between the Government and the Bank, as compared to $71/XWh estimated LRMC used in the BPS scenario); and for energy conservation under PROCaL. We infer that the implementation schedule or the lagged effect of price increases (or both) may be longer in the BLETROBRAS model; and the long-run price elasticity is lower. aiii. The difference of nearly 29% between the IPS and BAU forecasts, by 2000, entfArms the repartle conclusions that there is a cructal role for efficient pricing i energy omeservation and demand management. Indeed, the report' a earlier calctalation of potential pt 'ngs of 10-15% by 2000, attributable to efficient pricing (para. 3.60), may actually be conservative. of course, all three forecasts are dubject to a range of uncertainties, e.g., concerning the GPP growth rate and the price elasticity of demand. nevertheless, even recognising. this and the iterative relationship between price, demand and LAtMC (para. 3.14 of the report), we judge that a determined move to LVtC pricing, in level as well as structure, could out demand by at least 20% from the BAU forecast in Table 2 by the year 2000. A reduction of 20% correspondAs to annual savinge of tS$2.8 billion for 1991-2000, assuming maintenance of the system lad factor at 70% and valuing the cost per kW saved at US$2000; which exceeds the result in para. 3.60 (i.e., a range of 1881.2-1.8 billion on a comparable basis). Having incorporated some of the pri effect of tariff increases in its demand forecast, BLETOORAS is allow4ing for nearly one-third of these savings, i.e., about US$0.9 billion pet year for 1991-2000. Compared with the investment program undeilying the demand forecast in our May 1990 draft, discussed in para. 3.29, ZLVTROSRAS is now assuming a drop of US$1.5-2.0 billion per year, a substantial sum, although the diterence is due to slower GDP growth as well as Alternative Demnd Forecasts, 1991-2000 (Tn; 1980 1985 1988 1989 1990 1991 1992 1993 1994 1995 2000 1980 ELETIOCRMS Aesidential 23.2 32.7 40.6 43.7 47.9 50.0 52.9 56.1 59.5 62.9 83.0 7.5% 5.6X 5.7% Industrial 61.6 79.7 99.5 102.8 99.8 108.3 115.1 122.3 129.2 134.8 180.4 5.0% 6.2% 6.0 others 29.5 40.9 48.6 50.5 52.8 55.3 58.5 61.5 4.9 68.4 88.4 6.0% 5.3% 5.3% Total 114.3 153.3 188.7 197.0 200.5 213.6 226.5 239.9 253.6 266.1 351.8 5.8 5.8% S.7% ALTUATIVES IAM Residential 47.9 47.9 50.5 53.3 56.4 59.8 83.4 4.5% 6.9% Indutrial 99.8 113.7 127.4 141.3 155.2 164.9 218.4 10.6% S.8 Othere 52.8 54.1 5.3 56.5 57.7 58.9 79.6 2.2% 6.2% Total 200.5 215.7 233.2 251.1 269.3 283.6 381.4 7.2% 6.1% GPS Residential 47.9 47.4 49.4 51.6 53.9 56.4 76.4 3,3% 6.3U Industriat 99.8 101.1 96.6 95.1 95.8 98.3 130.9 -0.3% 5.9% Others 52.8 49.6 44.5 45.8 46.8 48.0 64.0 -1.9% 5.95 Total 200.5 198.1 190.5 192.5 196.5 202.7 271.3 0.2% 6.0% 2 In para. 3.13 of the report, the caqoertson sa made betueen the EPS and ELETROORAS forecasts, to deduce the fapct of efficient pricing, since the ELETROURAS forecast exceeded that of BAU. The ELETACRAS forcest now includes an implicit attowance for price increases, so the relevant comparison to between the EPS and AU forecasts. higher tariffs. our revised results more than confirm these in section i.2.. and pars, 3.60, and we therefore repeat the recommendations of paras. 3.14-3.15, thats (a) ELETROBRAS incorporate explicit analysis of alternative price effects in its studies of future electricity demand; and (b) electricity tariffs need to be adjusted to keep the level as well as the structure in line with LRMC in real terms, for financial as well as economic reasons. xiv. The report calculates the delivered IZPWC of ethanol at US$53.20 on average for BraxtI, compared with an equivalent gasoline cost of about US$25 per barrel, based on export value, or about US$30 per barrel, based on import value. By definition, LRMC includes a capital cost component, to represent the replacement of the capital investment in the alcohol distilleries in the long term. At the request of the Esrazilian authorities, we have examined the effect of excluding all "sunk" costs from the calculation, and of including provision for productivity gains and the economic value of by-products. An estimate of the economic cost of alcohol production, using shadow pricing of all inputs, was made in 1988. by Ser&o da Hotta and da Rocha Ferreira.3 Their lowest-cost result, based on the variable costs of an annexed distillery in Slo Paulo, yields a comparable delivered cost of US$48 per barrel, in March 1989 prices. Even allowing for "substantial gains in agricultural and industrial productivity, and to compensate for the revenues gained from the respective by-products," the comparable delivered cost falls only to US$41 per barrel in S&o Paulo; and it would still be as high as US$57 per barrel in the North/North-Bast. The former figure is close to that of another estimate based on assumptions which are highly-favorable towards the alcohol program.4 Hence, exclusion of "sunk" costs from the economic cost of alcohol would not alter our recommendations for alcohol pricing (paras. 4.56-4.59). xv. Paras. 4.56 and 6.20 of the report point out that we were not able to evaluate the environmental JAplications of our proposals for alcohol. We therefore recommended (paras. 18 and 6.28) that investigations be carried out into these "downstream" effects, as well as *upstream" impacts on the sugar industry. We emphasize this point again here. In particular, one of our scenarios includes the reduction of the proportion of alcohol in gasohol from 22% to 12% (paras. 4.7 and 4.57). While the analysis that has been carried out in the context of the CRRME confirmed the technical feasibility of supplying gasohol with an alcohol content as low as 10%, it is necessary to allow for the environmental impact of such a reduction, taking into account national standards for vehicle emissions. CrOde Oil. Snly and Cost xvi. PETROBRAS provided us with its revised investment progra and the corresponding production of crude oil (Table 3). The revised program is konatdo Sese do Motta and Lo do Roche Ferreira, "The Brazlitan National Alcohot Programe", Eagg LcONM July 1988. 4 Morld Bank, "Brait: Pubtic Expenditure, Subsidy Policies and Budgetary Reform", Renort No. 778-B Decebr 15, 1989, which also excludes usuna capital costs and refers to distilteries in the South-East. [awi. a Projections of Crude Oil Production and the PETROBRAS Investment Program, 1991-20V Oil Production (million b/d) PETROBRAS Bank Investment Program PETOBRS smuna AC (US biogn) 1991 700 770 2.3 1992 740 850 2.9 1993 780 930 3.4 1994 850 1020 3.9 1995 1000 1115 4.5 2000 Q MQ n.a. Notes: n.a. = not available Bank Scenario BC is an average of Scenarios BB and CC from Anne IV.8 equivalent to US$3.4 billion per year on average, of which roughly US$2.4 billion per year is for exploration and developments; it anything, it is more ambitious than the range of US$1.9-3.5 billion, which the report judged to be high relative to past achievements (paras. 14 and 4.60) (the range being a function of the growth rate of demand and the degree of oil self-sufficiency which the Government wishes to achieve). Nevertheless, we have attempted to assess the likelihood that these production targets might be achieved, in light of the report's analysis. An investment of US$2.4 billion per year for exploration and development would lie between our scenarios B and CC (see para. 8, Annex IV.1). Taking the average production from these two scenarios (Annex IV.8) gives the intermediate Bank production scenario shown in Table 3, which we term OBank Scenario 9C..6 Even accepting the PETROBRAS scenario of lower production than the Bank Scenario BC in the period to 1995, our analysis suggests that crude production is unlikely to be much more than 1300 million barrels per day by 2000, in contrast to 1700 million barrels per day assumed by PETRMBRAS. xvii. The report estimates the average incremental cost of Brazilian crude oil production at about US$25 per barrel, in 1988 prices (para. 4.13) . PSTROBRAS believes the cost to be lower, mainly in light of the expected large increases in reserves from the Marlim and Albacora fields. In particular, PETROBRAS expects the reserves-to-production ratio to improve so that, in PEOMIAS' view, the use of historical ratios to predict future production is too conservative. Reportedly, studies on future oil cost now under way in PSTROBMS suggest an PETROGRAS, Petroleu Production Goat in Brasit", Departamento do Produao, Septaer 1990, p.3. 6 We regard Scenario SC as optimistic regarding crude oit production, to the extent that the PETRORM program is closer to the toer figure assumed by scenario 9 (USS2.3 billion) than to CC (US$3 bittion). - v111 - incremental cost of US$15-20 per barrel. Unfortunately, we were not provided details on these studies, so we have no basis for revising our original calculation. PETROBRAS also expressed the judgement that there is a national securi'y argument for domestic crude production, to provide some buffer against unforeseen adverse international developments in the oil market. Such a view is not unusual, but there is no evidence that the policy is cost-effectives a strategic petroleum reserve, for example, could be built up while international crude prices are relatively low. xvii. Table 4 presents more recent information on future natural gas production, provided to us by PETROBRAS, corresponding to the investment program in Table 3. According to PETROBRAS, natural gas production could exceed 80 million m3 per day by 2000. We show also in Table 4, as in para. xvii, an intermediate "Bank Scenaio BC," which represents an average gas production from scenarios BB and CC, in Annex IV.8. Natural gas production is essentially identical under the PETROBRAS and BC scenarios up to 1994, after which they diverge sharply, and gas production in the latter would attain less than half the PETROBRAS level in 2000, i.e., a little over 40 million m3 per day. The reason for the divergence is almost entirely explained by PETROBRAS' assumed discovery of a non-associated gas field after 1995, amounting to 35 million an per day. Our best estimate is that, to anticipate such a discovery, PETROBRAS would need to set aside additional investment funds, perhaps in the order of US$1.5-2.0 billion, i.e., about US$0.5 billion p.a. in the period 1995-2000. No such investment program appears to be planned. xix. The report underscores (para. 15 of the Executive Summary) the advantages of natural gas and the case for expanding its use (Section IV.5.5). We suggest, Inter alla, exploring further the option of importing gas, e.g., from Bolivia or Argentina (para. 4.37), and we are pleased that the Brazilian Governmnet and PETROBRAS have now agreed in principle to import natural gas from Bolivia. The Brazilian authorities stated the view that the volume of gas to which Brazil would have to commit itself in the case of Argentina would be too large for the market in the South; while the cost of bringing it to the South- East would be too high. Furthermore, Brazil would face a considerable risk in also entering into a contract with Argentina at this time, as there is promise that significant reserves of gas may be found within Brazil, e.g., at Campos, Santos and the Bacia de Parant, which are closer to the important markets in the South and South-East. We are pleased that further study of the import options is planned. Next Stelps xx. The report has benefitted substantially from our dialogue with the Brazilian authorities over an extended period of time. We hope that it has contributed to the Government's difficult task of formulating energy policy; and that, in conjunction with the recently-completed Energy Matrix and with proposals being developed in Brazil for restructuring the power sector, it provides a basis for future discussions on energy policy within and between the World Bank and the Brazilian Government. We especially welcomes (a) the renewed attention now being given to the expansion of natural gas supply, and look towards the rapid resolution of any remaining institutional obstacles between the States, the - in - Projections of Natural Gas Production, 1991-200 (Million m3 per day) PETR)BRAS Bank Associated Non-Associated Total Grand Scenario BC Existing New Existing New Existing New Total Total 1991 n.a. n.a. n.a. n.a. n.a. n.a. 19 25 1992 n.a. n.a. n.a. n.a. n.a. a.a. 26 28 1993 n.a. n.a. n.a. n.a. n.a. n.a. 31 32 1994 n.a. n.a. n.a. n.a. n.a. n.a. 34 35 1995 24 2 12 6 36 8 44 37 2000 24 12 12 35 36 47 83 41 Notes: n.a. = Not available I Bank Scenario BC is an average of Scenarios BB and CC from Annex IV.8. private sector and PVTROBRMS; (b) the continued focus on environmental issues; and (c) the progress being made to reduce the intervention of the Federal Government in the energy sector and to encourage greater participation by the private sector. Among other things, we continue to urge that progress be made with regard to the level and structure of electricity tariffs, including the recognition of regional cost differences; and that policy towards the alcohol program fully recognize its economic costs, after an objective quantification of the net environmental benefits. These changes, along with the other recommendations in the report, would provide the foundations for a stronger energy policy in Brazil, which could be supported by the World Bank, through its leading program. -x- PREFACE This report is based mainly on the findings of a mission which visited Brazil in April-May, 1989. The mission consisted oft R. Dates (Mission Leader, LA18I), L. Gutierres (Energy Economist, LhTIE), C. Khelil (Petroleum Specialist, LATIE), J. Larrieu (Power Engineer, LAI8I), T. Joyce (Gas Specialist, Consultant) and A. Rodrigues (Enerrd Economist, Consultant). The following also contributed to the reports T. Markus and M. Sheehan (Petroleum Finances, L&IIN), C. Veles (Electricity Tariffs, LAlIN) , J. Vietti. (Electricity Finances, LAI), G. Bodely (Petroleum Transport, Consultant), and P. Motoke (Petroleum Taxes, Consultant). Messrs. T. Byer (OEDD2) and M. Munasinghe (NNVPR) were the Peer Reviewers. Some limited updating of the information took place during subsequent operational missions, carried out by R. Bates in July and October, 1989. A new forward was prepared and some minor updating introduced following a mission, consisting of S. Ettinger and R. Bates, which discussed a draft version of the report with the Brazilian authorities in March 1991. - xi - This study identifies three central issues in the energy sector in Brazil, namelys (i) inefficient resource allocation, because prices do not reflect economic costs, and investment programs deviate from least-cost solutions; (ii) fiscal and financial problems, caused by inadequate pricing, taxation and investment policies; and (iii) income distribution, in both regional and individual terms. Environmental issues, which raise concerns far beyond the energy sector and merit separate study, are given only limited consideration in this report, mainly in tne context of the electricity investment program. The study finds that, while the electricity sector made progress in implementing a tariff structure based on long-run marginal cost (LRMC), the low tariff level discouraged energy conservation and over-stimulated demand and investment. The resulting economic subsidies to consumers were some USS5.4 billion in 1989. In contrast, the domestic price level for liquid fuels was generally at or above border parity, but the structure of petroleum product prices was distorted, leading to economic subsidies of US$1.1 billion to consumers in 1989. Natural gas found it hard to compete with subsidized fuels, despite its low cost and environmental advantages. The price of alcohol has been close to its LRMC, but a substantial implicit economic subsidy of US$1.8 billion in 1989 arises out of the difference between alcohol's LRMC and the economic cost of gasoline, a close substitute. In terms of investment, Brazil's goal of greater independence in energy supply led to (a) a nuclear power program, which has produced virtually no electricity, despite massive investments; (b) continuation and expansion of the high-cost alcohol program; and (c) relative neglect of transmission and distribution investments. The low level of electricity prices destroyed the power sector's self-financing capability, and forced the Government to contribute increasing resources (reaching US$6 billion in 1987) from its own over-stretched budget. While petroleum contributed US$1.6-3.4 billion annually to public sector revenues in the 1980s, delayed price adjustments and the need to cross-subsidize alcohol and other petroleum products caused the sector's financial situation to deteriorate and undermined its tax-raising potential. In addition, by giving insufficient support to private sector participation, Government policy, combined with the inadequate legal and institutional framework, deprived the energy sector of a potentially significant source of investment financing. Uniform national energy prices are inappropriate for promoting regional development, income distribution and political obiectives, because of the resource misallocation which results. Wasteful energy consumption is encouraged,.while regions are deprived of the incentive to exploit comparative advantage. Personal income distribution objectives were pursued by providing residential electricity and liquifLed petroleum gas (LPG) at subsidized prices, but most of the subsidies (which amounted to US$2.5 billion and US$0.6 billion respectively in 1989) probably did not reach the poor. The study's key policy recommendations are for the Government to (a) ensure that energy prices reflect economic costs, in level as well as structure; and (b) establish the conditions and framework for the minimization of costs. In particular, the price of alcohol should be set equal to or above that of gasoline, in terms of energy equivalence, including all taxes; - xii - alternatively, a free market should be established in alcohol. Regional energy prices should reflect regional differences in production, transport and distribution costs. To address the basic needs of the poor, energy pricing should be targeted more efficiently. The sise of the first block of residential electricity consumption should be increased from 30 to 50 kWh per month, and sold at a subsidized rate, with all subsequent residential consumption charged at least at LMC. LPG, however, should be sold at its economic cost, as it is difficult in practice to target consumption to the poor. Adjustments in energy prices over time must be made with sufficient frequency to safeguard the financial health of the sector enterprises; and an increase in the tax rates on petroleum products, alcohol and natural gas is recommended. To further support energy conservation efforts in the longer term, steps are required to (i) encourage more competition in BraziliaL. industry, (ii) reduce market imperfections related to the lack of information on energy efficiency savings and the availability of financing for energy conservation, and (iii) as stated above, reduce cost in the public sector. We further recommend a sharp reduction in Government intervention in the energy sector, in favor of a greater role for competition and private participation. Such a role can be expected to support cost-reduction efforts in the public sector. Energy investment programs need to be more sharply directed towards least-cost solutions, by giving further attention to thermal alternatives to domestic hydroelectric power generation, in particular domestic and imported coal and natural gas; and imported hydroelectricity. Further investment in domestic crude oil production is supported, provided that ongoing efforts to cut costs, through inter aJJ the upgrading of PETROBRAS' costing system, are vigorously pursued. To assist with the implementation of the study's recommendations in the longer term, we propose studies related tos (i) ELETROBRAS' planning methodology; (ii) the development of PETROBRAS' profit centers; (iii) the effects of reducing the alcohol program; (iv) the scope for improving petroleum legislation, to attract more private sector participation; and (v) the fiscal potential for taxing petroleum products, alcohol and natural gas. Although the issues, findings and recommendations of the study are discussed in the context of the energy sector, they are relevant to the Bank's broader concerns of public sector reform, the scope for privatization and private sector participation, and the impact of the legal and institutional framework on economic incentives. A secondary area of focus is poverty reduction, both in terms of regional income distribution and the better targeting of subsidies towards the lowest-income groups. * Xit - EXECLTIVE SUMMARY Introduction: The Raeray Problem 1. The central conclusion of this study is that there has been a serious failure on the part of Government energy policy to maintain the level and structure of energy prices in line with economic costs. The consequences have been profounds energy demand was overstimulated, while operating aad investment outlays were higher than necessary. The financial situation of the sector deteriorated, and it became incapable of mobilizing the resources necessary for its development, which in turn imposed an increasing burden on the Government's fiscal resources. The underlying forces which led to this crisis have beent (a) the quest for greater energy self-sufficiency, without sufficient regard for the economic costs; (b) the subordination of energy policy to short-run efforts at price stabilization, despite the adverse repercussions on the energy sector enterprises eAd on the rate of inflation itself in the longer term; and (c) the pursuit of social and regional goals through inefficient financial and economic subsidies. 2. While these broad social and economic concerns are properly matters for the Government to decide, after weighing the costs and benefits, the impact on the energy sector was extremely serious. This, in turn, aggravated Brazil's macroeconomic problems, through higher public investment and debt and through a larger public sector deficit. Now is the time for a determined application of economic pricing and investment policies in the energy sector. These policies should be coupled with institutional changes, in terms of a more clearly defined and less interventionist government role, an adequate regulatory framework, and increased private sector participation. Brazil has relatively abundant energy reso,-cest they need to be husbanded more effectively. The anergy atrix, recently completed under the auspices of the Ministry of Infrastructure, in conjunction with this Report, should provide the basis for a continued dialogue on energy strategy for Brazil. The Oricins of the Problems A Diagnosi 3. A key need of Brazil's macroeconomic policy is to reduce the overall public sector deficit. The energy sector has exerted a major claim on scarce public sector resources, and accounts for a significant share of total investment. The peak in energy investment occurred in 1981-1983, following the second oil price shock of 1979, and the implementation of a national strategy for greater energy self-sufficiency. At that time, the energy sector was absorbing 20% of all investment in Brazil and over 4% of GDP. Energy investment declined thereafter, but it has still been averaging around 15% of total investment. In addition, PETROBRAS and the ELETROBRAS group account for about 12% of the total external debt of the public sector, due to the direct and indirect needs for foreign exchange to implement their heavy investment programs. 4. The significant claims of the energy sector on the Brazilian economy were a function of the steady growth in energy demand, which closely tracked but grew faster than GDP over the past decade. Stimulated by inefficient pricing and substitution policies, the energy intensity of commercial energy coniumption in general and electricity consumption in particular climbed significantly from 1970 - xfV - to 1987, while most developed countries were beginning to reduce the intensity of energy consumption. 5. Prior to 1978, the electricity sector enjoyed a long period of robust financial performance. But as the real average tariff fell thereafter, and the financial needs of the sector grew, electricity's claims on the public sector budget increased, from US$400 million per year on average in 1977-1982, to US$2 billion per year from 1983 to 1985, US$5 billion in 1986 and US$6 billion in 1987. It has fallen since then only because sector investments have dropped precipitously. Still, more than half the combined deficit of all public sector enterprises during the last five years was due to the electricity sector. Neither has the petroleum sector escaped. Revenues (after sales tax), not profits, and investment levels all declined after 1986. However, the price level of petroleum products was maintained at a more reasonable level than electricity, so that the hydrocarbons sector as a whole contributed US$1.6 to US$3.4 billion annually to the revenues of the public sector during the past decade. Nevertheless, while PETROBRAS generated substantial tax revenues for the public sector, the average tax on the retail price of petroleum products is low compared with other countries, although it rose from less than 10% in the early 1980s to nearly 20% after implementation of the new constitution. 6. The costs of inefficient energy pricing policies are magnified when account is taken of the effects of the financial and economic subsidies necessitated by the alcohol program. Approximately US$836 million appears in the 1988 budget under costs related to sugar and alcohol, but the figure does not capture the sacrifice of tax revenues brought about by the loss of gasoline sales, which were replaced by alcohol. We estimate the economic lose attributable to the alcohol program in 1988 to be about US$1.8 billion. 7. Investment programs were not governed fully by least-cost principles. Thermal options, such as natural gas, imported coal and bagasse, were relatively neglected, while investment in alcohol, domestic coal and nuclear power departed seriously from least-cost planning. The nuclear power program has produced virtually no electricity, despite investments of several billion dollars, and there has bean controversy over its safety as well.1 8. In the petroleum sector, distortions in the price structure fostered an increased consumption of alcohol, diesel and LPG; Brazil embarked on a policy of manufacturing alcohol vehicles; greater dieselization of the truck fleet was stimulated; and increased imports of LPG were encouraged, not only for residential cooking but also clandestinely to displace gasoline in vehicles. Low fuel oil and LPG prices made it difficult for natural gas to compete; and the greater consumption of diesel and alcohol necessitated, on the one hand, refinery investments to increase the yield of diesel, and on the other, the search for suitable export markets in which to sell the growing surpluses of gasoline production over consumption. The Secretaria do Assuntos Estrategicos rejects the report's conclusion an nuclear pomer, arguing that the nucleer pomer progra is justified not simply an the basis of power generation but also on strategic grounds. * XV* A Proaram for Actions Slectricity 9. The critical financial position of the electricity sector, in conjunction with the growing demand for electricity, underlines the importance of Amplementing rational pric"g policies. The excellent work now under way on applying a tariff structure based on long-run marginal cost (LtMC) should continue unabated. Simultaneously, the level of electricity prices should be brought at least into line with economic costs, to ensure a sound financial situation and proper demand management practices.2 Tariff adjustments must occur with sufficient frequency to safeguard the sector's financial health. An increase in the average tariff billed to US$55/XWh (in September 1989 frices), represents a minimum acceptable financial target for the near future. It is also consistent with LRC pricing, if taxes are increased to at least 30%, provided that inflation subsides, so that the effective tariff received by ELETROBRAS approaches the LRMC of US$71/MWh. Particularly large increases are needed for residential consumers, who pay less than half the LRMC. The consumers who have benefitted most from the subsidies were not lower-income groups. The poor could be protected by extending the size of the first residential consumption block, which enjoys a subsidized rate, from 30 kWh to 50 kWh per month. 10. Full LRMC pricing could cut total electricity consumption by 10-15% by the year 2000.4 It would entail price increases (compared with September 1989) of 32% for industrial consumers; 3% for commercial; and 124% for residential. The ensuing drop in system capacity requirements in the year 2000 would be between 6,500 MW and 13,600 MW, resulting in savings of US$1.0 to US$2.5 billion per year (compared with ELETROBRAS' investnent program of some US$7 billion per year). 11. Our numbers are preliminary. We recommend strengthening 3L3MRMBRAS planning methodology, especially to deal more rigorously with risk and uncertainty. In demand forecasting, better simulation of alternative pricing, demand and taxation scenarios and energy conservation and the implied effect on energy demand and LRMC, are needed. GDP scenarios should be made consistent with those of PETROBRAS. Better methodologies are also needed on the supply side, to deal with alternative reliability standards and the optimization of transmission and distribution systems. In terms of generation, international hydroelectric schemes need continued investigation. Thermal complementarity is a controversial question, but natural gas for power generation, coal imports and bagasse warrant examination as alternatives. Finally, the investment program needs to give adequate attention to transmission and distribution; improvements are already occurring in that directien. 2 te note that the Departamento de Abeastecimento e Precos contias to undertfre the Importance of national development arguments and the absence of competition In setting electricity prices. 9 A tariff of US$55/Mih, expressed in the prices of September 1989, corresponds to US54/MA in prices of July 1989. The latter was identified as an appropriate financiat target for the electricity sector in Jamary 1990 negotiations between the Government and the Bank. In Septeer 1991 prices, the equfvalent f Igure is about USS67/M, due to rest appreciation of the cruzeiro over 1989-91. Further studies of the adequacy of this Level are needed. 4 These figures reflect the antaysts in the main text. Updated estimates are in paras. xiI-xiII of the Foreword. * XVI . A Proram for Actions Petroleum and Alcohol 12. Economic priciag of alcohol, petroleum products and natural gap is needed to foster efficient inter-fuel substitution, assist in financing the expected investment requirements in the hydrocarbons sector and reduce crude oil imports. The required pricing and investment decisions are part of a complex of policy issues concerning* the tAxation of petroleum products; the alcohol program; the quest for greater seLf-sufficiency in crude supply; and greater priviAte sector participation in hr,drocarbons development. 13. Annual investments by PETROBRAS are likely to be US$1.9-3.5 billion to the year 2000, according to the degree of self-sufficiency waich the Goverinent wishes to achieve. This range is generally above PETROBRAS' average eiac-9 1983 (US$2 billion). We believe that further investment by Brazil in its domebtic crude oil production is justified, provided that parallel efforts are made to cut costs, given the narrow margin between the cost of imports and domestic crude at existing and projected international price levels. 14. While the impact of economic pricing on the demand for petroleum products would not be as high as in the case of electricity supply, because rr.zil has maintained an average price level for petroleum products generally at or above the economic cost, we believe that there could be lemaad cut of 5% by the year 2000, which would eventually translate into annual cost savings of US$0.3 billion. The suggested price increases for individual products over March 1969, ares 73% for LPG, 40% for fuel oil, 22% for kerosene, and 300% for naphtha. We estimate that subsidies of US$1.1 billion p.a. would be eliminated. We are not persuaded by the social argument for subsidizing LPG. Higher-income families typically receive larger subsidies than lower-income families. At the same time, the low psice of LPG encourages clandestine use in vehicles, and discourages the use of cheaper natural gas. Economic pricing also means reducing the price differential between diesel and gasoline. Natural gas would be priced on the basis of the opportunity cost of substitutes, in thermal equivalence, which would provide stronger market incentives to expand its use, given the high netback values we have identified and attractive environmental features. As in the case of electricity, price adjustments must take place frequently enough to ensure the financial health of the petroleum sector. 15. 2ho policy changes which we recommund for alcohol require difficult and politically sensitive decisions, but substantial resources could be saved by reducing the huge implicit economic subsidy now enjoyed by alcohol. The average retail price of alcohol reflects its LRMC, but the production cost '.a double its present opportunity value as a substitute for gasoline. We recommend setting the alcohol price equal to or above that of gasoline, including all taxes and in terms of thermal equivalence, which would effectively eliminate the demand for new alcohol cars.5 subsidies to alcohol would continue but decline parl paeas with the rundown of tho alcohol fleet, as it ages. The subsidies would no longer, however, be hidden within the energy sector: they would become transparent. An alternative approa:h is to eliminate altogether subsidies on alcohol as a fuel, by establishing a free market, which is consistent with the fact that production is in the hands of the private sector. The price of alcohol would rise in the short term, and to accommodate concerns over fairness, the 6 If not, the alcohol price should be raised further or a tax penalty Introduced an alcohol vehicles, due to the higher economic cost of alcohol coopered with gasoline. - xvi I Government could instead subsidize part of the cost of converting alcohol care to gasohol, using some of the savings in subsidies on alcohol as a fuel. 16. Projections of the effect of our recommendations on alcohol subsidies depend on the dynamics of the LRMC for alcohol, as the production level and regional pattern change; and on the movement in international oil prices. Using the World Sank's oil price projections and assuming no change in the LRMC of alcohol production, if the sales of new alcohol care cease after 1990, the implicit alcohol subsidy could fall by US$1.5 billion p.a. by the year 2000, compared with the present level of US$1.86 billion, and US$2.7 billion p.a., compared with the level which would otherwise obtain, which could translate into higher tax revenues for the public sector. These proposais on alcohol pricing would reverse the decline in gasoline sales and increase tax revenues. We also recommend an increase in the rate of tax on petroleum products and alcohol: the amount of the increase depends on macroeconomic judgements beyond our scope, but our estimated price elasticities for petroleum products are low enough, in both the short- and long-run, to permit significant additional tax revenues. 17. Several important studies are needed to assist with the implementation of the above recommendations and to lay the basis for further change. First, PTROBRAS' efforts to develop profit centers and to upgrade its costing system will provide a better basis to evaluate alternative policies regarding domestic crude exploration and production, including ways of remaining competitive in the face of uncertainties in the international oil price, given the relatively narrow cost advantage which domestic crude now enjoys. They will yield more detailed and accurate figures for the economic cost of natural gas and petroleum products. Second, to improve the analysis of different policy options, updated information is needed on the LRMC of alcohol, taking into account possible improvements in productivity, a more efficient allocation of output between producers in different regions, and the expanded use of bagasse for electricity cogeneration. The "upstream" impacts on the sugar industry and "downstream" effects on the automobile manufacturers and the environment also need to be examined. Third, further work is needed on the fiscal potential for taxing petroleum products, alcohol and natural gas, recognizing demand elasticities and the financial implications for PETROBRAS. The results of these studies will provide vital inputs to a general evaluation of the demand and supply balance for petroleum prc'ducts, alcohol and natural gas, incorporating the optimization of refinery operutions, and covering alternative: domestic pricing and taxation scenarios; international oA.l price assumptions; investment levels for PETROBRAS; and domestic crude supply. Institutional ChanMe: Privatisation and Intearated Plannina 18. The actions recommended above on energy pricing should be complemented by institutional changes, notably increased participation by the private sector is energy supply, the establishment of an independent regulatory process which would operate according to clearly defined and transparent principles, less intervention by non-sectoral agencies, and a more restricted role for MMR (now part of the Ministry of Infrastructure--MOZ). 19. The 1988 constitution restricts the range of alternatives available to Brazil to permit private sector investment in petroleum. The ownership and exploitation of oil and gas reserves has been vested exclusively in the Federal Government, and PMTROBRAS is barred from entering into new risk contracts with *XVif1* foreign companies for the exploration of either oil or gas. The impact of the new constitution on the electric power and coal subsectora is, however, less restrictive. Aside from the arguments in favor of private sector participation to increas. efficiency, the sheer volume of future Investment in the energy sector will necessitate a search for alternative sources of financing and owership. 20. Ia oil and gas, the Government should easurage greater private sector participation in the areas of PatRoaRAS which are not protected by its monopoly (e.g. subsidiaries) and encourage the maximum prirate participation in State-level gas companies, such as COMGAS. Furthermore, although we fully recognize that any amendments to the Constitution are the prerogative of the legislature, when the constitution is revised, we suggest that particular attention should be given to the monopoly of PETROBRAS and the role of the private sector. In electricity supply distribution, private participation at the State level also offers scope for tapping new sources of capital. a variety of ownership and financing models exist to permit private participation in generation, inclading cogeneration, for example using bagasse in the sugar industry, and c4ncession contracts for entire isolated systems and for thermal and small hydroelectric plants. Legislation is needed to give more active support to such options. 21. The application of economic pricing across the energy sector implies replacing price control by the Ministry of Economy with clear rules administered by an independent regulatory agency. This would make a major contribution towards the creation of a regulatory environment to foster private Initiatives. It would also improve energy planning, as decisions would be taken on a more cohereut basis. Brasil starts from a relatively favorable position, in that oversight for all the commercial fuels is centralized within K01, except for alcohol. This exception need not create difficulties if, as we suggest, alcohol is forced to compete fully with its substitutes on the market, notably gasoline. The broader strategic and political decisions relevant to the future role of alcohol within the sugar industry would then, correctly, be isilated from the energy sector. Raising the prices of all petroleum products to, at least, their economic costs also implies a greatly changed role for CNP. Government's main activity would be *auditing" the proper application of this pricing policy at the ex-refinery level; and the regulation of PETROBRAS as a transporter of natural gas, crude oil and petroleum products. In conjunction with the elimination of uniform national energy pricing (para. 25), competition among distributors can be left to determine retail prices. 22. The role oi MOX needs better definition and should be less interventionist. M01 would concentrate on establishing and maintaining the overall policy framework; ensuring the application of a consistent planning methodology in the publicly-owned parts of the energy sector; helping manage the reduction of public and promoting more private involvemen' coordinating energy sector policies with other Ministries; and providing informational support to the pu.lic and private energy firms. Xnera. Conservation: the. JMnortance of Pricina 23. Brazil has evolvud a plethora of energy conservation programs. Bven the most promising, PROCEL, has a limited chance of success without concomitant action on energy pricing. The Government should also tackle directly a variety - XX - of market "sperfections, including the impediments to industrial competition, as they have reduced incentives to conserve energy. Efficient pricing and energy conservation require tariffs which reflect the level as well as the structure of economic costs. We believe that action on pricing and dealing directly with market imperfections are more appropriate for energy conservation than program. which may compound existing market imperfections. Of course, efforts to reduce losses in production and improve the utilisation of energy by the public sector, and to inform the private sector about the energy costs of alternative investments, are justifiable; and externalities (such as environmental impacts) may have to be dealt with in special cases with non-price action, although even here costs external to the sector can often be "internalised". 24. Finally, Brasil's policy of uniform national energy pricing, and the cross-subsidisation which ensues, promotes wasteful energy consumption and distorts patterns of development away from activities that would exploit regional comparative advantage (e.g. development of local and regional energy sources, such as natural gas, biomass, coal and minihydro), and promotes uneconomic development of the Amason and other remote areas. Local private initiatives, in particulbr, are discouraged. Regional uniform tariff have also been a serious financial handicap in the case of electricity supply, since the tariff equalisation scheme combined with the non-functioning of the mechanism for compensating the lese/unprofitable power companies, has helped bring the sector to a financial crisis. CHAPER I AN OVERVIEW OF THE ENERGY SECTOR 1.1 The objective of this study is to propose a strategy for the commercial energy sector in Brazil, which could be supported by the World Bank through its program of lending and, as appropriate, further economic and sector work. In so doing, we have identified a number of key issues, on the basis of an analysis of the recent past, the present situation, and alternative projections of possible resource requirements in the medium-term future, i.e., 1990-2000. Finding ways to tackle these issues effectively, within a realistic time horizon, lies at the heart of our stzategy. We propose that this report, in conjuaction with the Energy Matrix, which has recently been completed by the Brazilian authorities (see par&. Ii of the Foreward), provide the basis for an early discussion of energy strategy in Brazil, between the Brazilian authorities and the World Bank, and in the World Bank itself. 1.2 The importance of an effective energy strategy can be seen in light of the absolute size and rate of growth of commercial energy demand in Brazil. In 1980, Brazil's primary commercial energy consumption was the third highest among developing countries. Since then, commercial energy consumption has grown at more than 4% p.a., significantly above the average for upper middle-income developing countries. In terms of resource allocation, commercial energy investment accounted for 16% of total investment in the 1980s and 3% of GDP (Table I.5). The resource mobilization impact was commensurate: in 1987, the power sector alone accounted for more than half the public sector deficit, equivalent to 1.3% of GDP. The study focusses on the main areas of commercial energyt electric power, petroleum products, alcohol and natural gas. Coal is relatively unimportant as a fuel and plays a minor part in our discussion. 1.3 The level ad structure of energy prices are central elements in any energy strategy. Working through demand and investment, they determine not only the physical allocation of resources to the evergy sector as a whole and to the various subsectors. They also govern the extent to which financial resources are mobilized by the sector enterprises, or through the Government's budget, on their behalf. Decreases in real energy prices in the 1980s had a major impact on resource mobilization, contributing to the substantial net flow of financial resources from the Government to the electricity sector, and to the recent deterioration in the financial performance of the petroleum sector. 1.4 Institutions issues also are fundamental and are examined, especially as they relate to the pricing and investment framework. The consistent application of economically efficient pricing across the energy sector is a prerequisite for an adequate regulatory environment, which would foster greater private participation in the energy sector. It would also bring about less Government interference in decision-taking in the energy sector and promote more integrated energy planning. 1.5 The high priority of environmental issues warrants their treatment in a separate study. Their resolution also needs to be addressed in a wider context than the energy sector alone. We have included them only in our analysis of power investment planning, although the closely related topic of energy *2- coaservation and demand management -- which has direct and indirect environmental benefits -- is treated in detail. 1.6 A major complicatlon in preparing this report arose from the effect of inflation and movements in the official exchange rate. Annex 1.1 shows the price indices and exchange rates that we have used in our analyoris. They exhibit a rapidly escalating rate of inflation and (with the exception of 1983-1985) an almost constant decline in the real exchange rate since 1974, undermining intertemporal comparisons. Particular caution is required in comparing prices, costs, investments and borrowings, even when expressed in US$ terms. Prescriptions for the future fare continued uncercainties over the possible course of inflation, its level and rate of change. Nevertheless, we believe that our main conclusions and recommendations hold with regard to the energy sector strategy which the new Government will need to articulate, as part of its broader macroeconomic policy, if it is to make a successful assault on the problem of the public sector deficit and inflation. 1.7 The plan of the report is straightforward. The remainder of Chapter I provides necessary background on supply and demand in the energy sector; and establishes the main linkages between the energy sector and the economy. Chapter II analyzes the institutional structure of the energy sector, particularly regarding pricing, energy planning and private sector participation. Chapters III and IV identify the pricing and investment issues in electricity supply and in petroleum products, alcohol and natural gas respectively, based on concepts of economic cost and opportunity value. Chapter V brings together our energy supply and demand analysis, through the particular question of energy conservation and demand management. Finally, Chapter VI summarizes the Report's conclusions, in terms of the key issues, a strategy to deal with them and some proposals for further action. I. 2 SUPPLY AND DfMMD 1.2.1 EneXay Resources 1.8 Brazil is relatively well-endowed with energy resources, both renewable and non-renewable (see Annex 1.2). The economic costs, however, have increased, due to a variety of factors, which are discussed in this report. Notably, the long-run marginal costs of both domestic hydroelectricity and hydroca:-bons, which have assumed a dominant role in Brazil's energy matrix, are rising as the best opportunities are exploited. Constraints and costs have also been introduced, to deal more effectively with environmental and resettlement questions. 1.9 The principal renewable energy supplies are hydroelectric power and biomass. Most of the hydroelectric potential outside of the northern part of the country, has been developed. For the future, the greatest hydroelectric potential is in the north of Brazil, notably in the Amazon region. The two major sources of biomass are sugar cane, yielding bagasse and alcohol, and fuelwood. While better utilization of bagasse for energy purposes is warranted and supported in this report, we argue against the continued large-scale production of alcohol as a gasoline substitute, due to its relatively high economic cost. Brazil' s fuelwood base has been seriously eroded, and the main existing forestry resource is the Amazon (see map of natural vegetation). The scope for developing the Amazon for energy purposes is severely restricted, for economic as well as environmental reasons. .3- 1.10 Brazil's chief potential in ea-renewable energy is in oil, gas, coal and uranium. The country has been successful in finding and developing both oil and natural gas, principally off-shore in the South-East part of the country, although the Amazon region holds promise. Much of the natural gas is associated with oil, e.g., in the Campos basin, but non-associated gas exists in several areas, such as the Santos field, off-shore from the State of Sao Paulo. Coal is located in the South (see map of principal coal deposits), but is not economically attractive, with high sulphur and ash content. As far as uranium is concerned, we see no role in the foreseeable future, given the costs and safety aspects that are involved with nuclear power. 1.2.2 Primary Enerv 1.11 Data on primary energy production and consumption in Brazil, for 1970-1987, are in Annexes 1.3 and 1.4 respectively. Primary energy consumption increased by 5.3% p.a. from 1970 to 1987. Primary energy production expanded more rapidly, at 5.6% p.a., indicating a slight decline in Brazil's relative dependence on imported energy. The percentage of total energy demand imported increased from 24% in 1970 to 37% in 1979, declined to 16% in 1985, and then rose to 21% in 1987. 1.12 In terms of composition, Annex 1.3 shows that the iicrease in primary energy production during the 1970s was provided mainly by hydropower and, to a lesser extent, coal and sugar cane. Between 1980 and 1987, the contribution of petroleum and sugar cane increased dramatically, although hydropower continued to meet a substantial part of the growth in energy demand. The relative shares of the different energy sources in final primary energy consumption are detailed in Annex 1.4 and summarized in Table I.1, with respect to 1970, 1978 (just prior to the second oil price shock) and 1987, the latest year for which data are available from the National Energy Balance. 1.13 Although imported energy as a whole declined only slightly relative to total energy demand after 1970, as mentioned above, dependence on petroleum fell markedly. Petroleum imports repre3ented 67% of petroleum consumption in 1970. After reaching a peak of 85% in 1979, the degree of dependence on imported petroleum had fallen to 46% by 1987. In absolute terms, petroleum imports were 340 thousand bpd in 1970, reached a peak of nearly 1 million bpd in 1979, and fell to a little under 500,000 bpd in 1987. Diversification of the source of petroleum imports has also occurred, with dependence on the Middle East falling from over 90% in 1979 to 80% in 1987. 1.2.3 Sectoral Ener=v Consumotion 1.14 The evolution of final energy consumption by sector in Brazil, over 1970-1987, is portrayed graphically in Figs. X.1 and 1.2. These graphs reveal important structural shifts which were taking place in the Brazilian economy, away from primary activities towards the industrial and service sectors. Most apparent is the large increase in the industrial and (to a lesser extent) the commercial components, along with sbarp drops in the shares of residential and agricultural consumption. By 1987, aside from the consumption of the energy sector itself (8.8), the main consuming sectors weres industry (42.6%), transport (19.9%), residential (16.5%) and agriculture (4.9%). The data for 1987 are in Table 1.2. *4. A. The Idustrial Sector 1.15 The structure of industrial energy consumption has undergone a complete transformation since the first oil price shock. In 1970, electricity, fuel oil and wood accounted for over 70% of industrial energy consumption, very roughly in equal proportions. By 1987, electricity had come to dominate, with nearly half the total; while the rest wae fairly evenly distributed between wood, bagasse, fuel cil, coal and charcoal, at close to 10% each. In the case of electricity and fuel oil, the inter-fuel substitution resulted from the relative price effect, as preferential electricity tariffs were offered to certain industrial consumers. Sectoral Energy Consumption 1970 - 1987 1000 TOE 160 140 - 120- 100 ......... 90 50 40 20 0 1970 1972 1974 1976 1979 1980 1982 1984 1996 Industrial Transport Residential Energy Agriculture 2=clal Figure I.1 -- Sharea in Total Primary Energy Consumption dr ..e.trkIety .15.5 '.7..9. Aled4L4 DA 17. .ae. C o 4. 5 j. 7 40%4 WoetGa04 0,7 2.2 Uranhim 0.0 0.0 . Structure of Sectorat Energy Consumpt ion 100% 80% 60% 40% 20% 1970 1972 1974 1976 1978 1980 1982 1984 1986 IndUStr181 Transp ort Resldentfal Energy Agriculture fllf Comerial 1 ju r e 1 . 2 \ *6- Consumation by sector, 1987 B. The Tranort sector Setor Corouptfon Ibt 1.16 Almost half the petroleum tEMY 0#7 transport 30,07 19.9 products and over 90% of all alcohol 1idenis 25.467 16.5 are consumed in the transport sector. Energy 13,446 8.8 Auricuatwe 7.04 4.9 Government policies aimed at Cm5e" 4.2 substituting imported gasoline by PbI ic 4,765 3.1 domestic ethanol and, to a lesser Total imam . extent, the relatively low diesel soumrc2 manal EneM BatInce. Uo. price compared with gasoline (see .__ Chapter IV), have resulted in a high level of penetration of ethanol in the transport sector and caused gasoline consumption to fall from 15. 6 billion liters in 1979 to 9.1 billion in 1988. Diesel oil consumption increased steadily, from 12 billion liters in 1975 to over 24 billion in 1988, as a result of both the rapid growth of the road transport industry and the dieselization of part of the truck fleet (Annexes I.5 and I.6). As discussed in Chapter IV, the two trends combined have led to a costly imbalance in the consumption and refinery profiles of petroleum products, forcing Brazil to refine imported crude oil for internal diesel oil consumption and to export gasoline in large quantities (Annex 1.7). 1.17 The importance of alcohol in Brazil's energy matrix -- and indeed its worldwide interest, due to its unique scale -- warrants further discussion here of the National Alcohol Program, known as PROALCOOL (Pr- grama Nacional do Alcool). Launched in 1975, with the object of encouraging the production of fuel alcohol to assist in reducing Brazil's dependence on petroleum imports, PROALCOOL has been remarkably successful in expanding alcohol consumption (see Table 1.3).1 Some US$7 billion were invested in the Program up to the end of 1987 (Table X.5). The Bank supported PROALCOOL under the Alcohol I Project (Loan 1989-BR), with US$250 million financing, in April 1981. While the target was originally to expand capacity to 14.3 billion liters by 1989/1990, both through new distilleries and expansion of existing plants, the boom in alcohol demand in 1986, during the Crumado Plan, led to an urgent decision to increase capacity. By December 1987, 661 projects to install or expand alcohol distilleries were approved by CENAL, amounting to a potential productive capacity of 16 billion liters. This capacity was divided equally between annexed and autonomous distilleries and was only utilized to the extent of 63% in 1987. Over 52% of the investment in distillation is in the State of Sao Paulo. Nuch of the information in this section on Brazt's alcohot program is from the Mortd Bank's Suar Sbsmctor Revies Report No. 7589-R, dated Nay, 1989. -7- TAML I. 3 Alcohol-ConLMaWtion 1976-1988 (Million Liters) .1976 384 172 45 172 1980 3,549 W2.53 63 429 12 2,682 1985 Vb235 2,121 23 5932 64 8053 1986. 1905 2 44a at 8,25 69 10"6? 1967 14,830 2136 18 8,772 74 1090 1988. 12,5W9 l9ga 16 9,644 77 11,630 Source a uar*" Sbsecto0 RevieW, May, 1989, Note: Other than as a fut, alcohot to uWd In the chemical industry for potable prposs and in perfu~ei SooU mo c also epoed. thes other Usee nw accot. for about 71 of the tet consu t Ion. 1.18 The official consumption of 12.6 billion liters of alcohol in 1988 is almost certainly an underestimate, perhaps by as much as one billion liters, due to an active black market. Ethanol is consumed directly as a fuel (hydrous), in blends with gasoline (anhydrous), as a petrochemical substitute, and in other traditional uses (potable alcohol, perfumes, etc.). By far the greatest demand for ethanol is as a fuel, accounting for 93% of total ethanol consumption (Table 1.3)1 and most of the fuel alcohol consumption (83%) consists of hydrous ethanol, used in ethanol-powered cars. Anhydrous ethanol, which is mixed with gasoline in gasohol-powered cars, accounts for the remaining 17% of total fuel alcohol consumption. Anhydrous ethanol consumption, like gasoline, has shown a declining trend. 1.19 Ethanol demand was stimulated by attractive official pricing policies and tax advantages for the purchase of ethanol-powered vehicles (Chapter IV). These vehicles have dominated new car sales since 1983, accounting for over 90% of all passenger care (Annex 1.8). We estimate that the size of the fleet of passenger cars, pickups and vans at the end of 1988 was 11.3 million vehicles, of which one-third was alcohol powered; 65% was driven on gasoholl while 2% was diesel. For the gasohol vehicles, anhydrous gasoline is blended with gasoline to the level of 22% and Brazilian car-makers adapted automobiles to this level, both to save on gasoline imports and to enhance the octane rating of domestic gasoline. Gasohol engines can run on a lower mixture of alcohol and in March 1989, to cope with shortages of alcohol and to decrease gasoline exports, the Government authorized a reduction in the mix to 18%. The mix was further reduced to 12% in January 1990. C. The Residential fector 1.20 The importance of electricity in the total energy consumption of the residential sector has grown as significantly as in the industrial sector, from 10% in 1970 to 44% in 1987. Electricity is typically used for lighting and electric appliances, notably refrigerators, televisions and radios. The principal competition from other energy sources arises mainly in cooking, lighting and (to a much smaller extent) water heating, where fuelvood (including charcoal) and LPG continue to be important, accounting for 38% and 174 of the residential market respectively. Town gas and kerosene account for the residual .g. of 1%. The importance of LPG is actually greater than these numbers would imply, if allowance is made for its superior efficiency in use over wood. Furthermore, the LPG market has grown steadily since 1970, encouraged by uneconomic pricing, tripling its share of the market by 1987. The fuelwood share dropped continuously, meanwhile, from 82% to 38%. Penetration of town gas has largely been confined to Rio de Janeiro and Sao Paulo. The rapid growth in LPG usage has caused problems in supply, leading to substantial imports of this expensive petroleum product.2 In 1988, PETROBRhS imported over 2 million of LPG at a cost in excess of US$140 million. Some LPG was certainly used illegally as a substitute for gasoline. D. The Aricultural Sector 1.21 The evolution of energy consumption in the agricultural sector has illustrated a history of rapid growth in diesel oil usage (13 % p.a. from 1970), from 7% of consumption in 1970 to 41% in 1987, and contraction in wood, from 91% to 36% over the same period. Electricity has also grown substantially, to account for nearly all of the balance of 23% in 1987. Lying behind these powerful movements towards commercial energy consumption and away from traditional energy forms has been the modernization of the agricultural sector, encouraged by Government policy, leading to mechanization and the concomitant expansion in diesel tractors and in pumps. Over and above the direct consumption of energy in the agricultural sector, there has of course been a parallel extension of demand for fertilizers, especially since 1970, which led to a substantial indirect use of energy. I .3 ENERGY AND THE SCONOMY 1.22 The growth in energy consumption in Brazil has closely tracked GDP growth for the last two decades, with a correlation coefficient of 0.99 between the two growth rates. Total energy consumption grew at an average rate of 6.4 % p.a. during the 1970-79 period, while GDP grew at 7.6 % p.a., giving a ratio of 0.85 between the two growth rates.3 Between 1980 and 1987, dEspite a temporary fall in energy consumption after 1979, due to a drop in GDP, energy consumption and GDP on average increased at 4.0 % and 3.3 % p.a. respectively, with a consequent increase in the ratio to 1.2. Figure I.3 below shows graphically the close relationship between the rates of growth of real GDP and energy consumption since 1970. This tendency for the ratio between the two growth rates to increase when GDP growth slows has been observed elsewhere, suggesting that energy consumption is less responsive, in the short and medium term, to the downside of economic performance. 2 The LPG pricing issue is discussed in Chapter IV. The ratio is often termed "the elasticity of energy consuaption with respect to GDP," but it is slapty a statistical relationship, obtained by dividing the growth rate of the former by the growth rate of the latter. It Is not to be confused with the income elasticity of energy deend, which is estimated in section 1.4, taking into account price as wett as incUme. Real GDP and Energy Consumptlon Growth Rates a I ' I ' l I I I l I I 1971 1973 1975 1977 1979 191 983 1985 1987 Energy Conumption tel GDP FIgure 1.3 1.23 Table 1.4 compares Brazil's energy consumption with a number of developing and developed countries, in terms of selected indicators, using World Bank data. Brazil has experienced a rapid growth rate of energy consumption but energy consumption relative to GNP, while generally higher in Brazil than in the developed economies of Europe and North America, remains considerably lower than a selection of other developing countries. The growth rate of energy consumption as a ratio of the growth rate of GDP is substantially higher in Brazil than in the developed countries shown in Table 1.4, but comparable with or lower than other countries in Latin America, with the exception of Mexico. I.4 INCONS .AND PRIC LASTICITIS 1.24 The estimated short-run and long-run price and income elasticities of demand for each main non-fuelwood energy product are presented in Annex I.9, as part of our discussion of energy demand forecasts and methodology. These results play an important role throughout this study, especially in the work on petroleum and electricity demand forecasting in Chapters III and IV and in the discussion of energy conservation in Chapter V. The statistical analysis showed relatively inelastic demands with respect to both price and income changes for most fuels in the short run. However, the long-run elasticities are substantially higher and,. indeed, demand becomes income-elastic for nearly all fuels as time is allowed for consumers to adjust. 1.25 The importance of prices for inter-fuel substitution is also shown in Annex 1.9 by the decline in the price elasticities with the level of aggregation i.e. energy demand as a whole is less price elastic than the individual fuel components because of substitution. The demands for LPG, kerosene, diesel, natural gas, and residential and commercial electricity are not very sensitive to price variations, indicating a moderate effect on demand and significant revenue effects as prices are adjusted upwards, which has -10* implications for taxation policy.4 Alcohol and gasoline, of course, are close substitutes, at least in the longer run, which explains the higher price elasticities and industrial electricity demand in Brazil is the most price- elastic of all the fuels in the long term, indicating a rapid and effective response of industry to relative price changes. 1.26 Practically all the individual fuels are sensitive to income variations in the long run. LPG, like charcoal, is well regarded as a cooking fuel in Brazil, even among the higher-income groups.5 Similarly, there is evidence that the demand for diesel oil is income elastic, because the pattern of consumption of goods diversifies as incomes grow, through the import of commodities from other geographical regions by truck. 1.27 The history of the Brazilian energy sector during the last two decades has manifested clearly these price and income elasticities in operation. The importance of income reveals itself in the close link between energy demand and macroeconomic growth, identified in Section 1.3. The fact that energy demand is also responsive to changes in the relative prices of alternative fuels is the main explanation for the striking success of the Government policy, described in Section 1.6, in inducing the substitution of domestic hydroelectricity for imported oil; alcohol for gasoline; and, more recently, natural gas for fuel oil and other energy forms. For example, Brazil's success in modulating its load curve by implementing a tariff structure based on long-run marginal cost (LRMC) is attributable to the high price elasticity of industrial electricity demand. Two facets of that pricing policy have came to be known as tae "tarifa hora-saxonal" (THS) and the program of substitution of electricity for fuel oil, called BGTD (Energia Garantida por Tempo Determinado) . The TRS, desig-ed chiefly to reduce peak load, helped to achieve a system load factor higher than 70 percent; and the program was so successful that -- coupled with the problems related to regulation of river flows in the Brazilian hydroelectric system and the slowdown in investint in generating capacity -- the power system shifted from one characterized by capacity shortages to one of energy shortages (paras. 3.38.3.39). The EGTD program, based on a concessional tariff, also contributed to energy shortagess consumption under the program expanded rapidly from 0.7 Tith in 1982 to 12 TWh in 1985 (i.e., 13 percent of total industrial demand). * These elasticities are preliminary. Further work Is needed on cross-price elasticities. Uhl te the price and incme elasticities for fuetood were not calculated, It can be noted that the share of charcoat in the total cararption of energy has Increased since 1970, in contrast uith firewood, wAich has steadl ty dropped. *11* Eneray Consumation and Economic Growth in Selected Countries di (2) (3).(4 Greath Growth aP .(i X p.O. 2 p.O. Divided by ..(gD . 12-al a ) us 0.1 3.1 0.39 0.0 Jaan 1.7 38 0.21 0.45 Canada 0.9 2.9 0.60 0.31 Germany 0.2 1.6 0.31 0.13 France 0.6 1.6 0.29 0.38 Austratlio 0.6 3.2 0.43 0.19 Ux 1.1 2.6 0.37 0.42 Italy 0.0 2.1 0.26 0.00 Ve0euets 2.3 0.2 0.74 II.$0 Argentina 1.5 - 0.3 0.62 5.00 Hhngary 1.1 1.7 1.37 0.65 Uruguay - 2.0 - 1.3 0.35 1.54 reaill 4.0 3.3 0.41 1.20 Mexico 0.6 1.0 0.71 0.60 Chit* 1.5 1.0 0.63 1.50 India 6.0 4.6 0.69 1.30 China 4.4 10.4 1.81 0.42 Source: OWrnd DeWeoprent Report. 1989. Notes: Inergy ConsuiptionfGNP Is in kg. of oft equivalent per USS 1987. I S 1M 5fQR SMQ.S~I1g ~IAMM 1.28 The public sector deficit and the external debt were major concerns of the Federal Government during the 1980s and continue to dominate macroeconomic policy at the opening of the 1990s: the former was a driving force behind inflation and the latter represented a continuing commitment of foreign exchange. Financing the resource requirements of the energy sector aggravated these concerns. Brasil' a public enterprises as a group accounted for a large part of the public sector I s operational deficit during the last five years. For the period 1983-1987, the deficit of the Federal public enterprises averaged 2.2% of GDP, representing 176 of their operating revenues and 70% of their investment. The deficit was financed primarily by Government transfers and subsidies (two- thirds of the total) and by the accumulation of additional debt, largely short- term.6 More than half the deficit was due to the electricity sector (1.3% of GDP), while the petroleum sector contributed a modest surplus. 1.29 The preceding data do not take into account the full effects of the financial and economic subsidies necessitated by the alcohol program, much of which escapes the Goverament's budget.7 Approximately US$836 million appears in the 1988 budget under costs related to sugar and alcohol, with the bulk of See "rasit - An Assesement of the Current Macroeconomic Situation", Report No. 7540*8R, Deceber, 1988. ' See "ralt: Public Expenditure, Subsidy Policies and Budgetary Refortf, VorLd Bank Report Ne.. 7738*8R, Deceber, 1989. this expenditure being for credit operations, financing the purchase of sugar for export. The figure does not capture the sacrifice of tax revenues brought about by the loss of gasoline sales, which were replaced by alcohol. The economic lose attributable to the alcohol program in 1989, as shown in Chapter IV, is in the order of US$1.8 billion. Hard decisions need to be taken concerning an apprapriate level for the alcohol program, given the uncertainty about future movements in international oil prices, the likely extent of economic subsidies and the benefits to the Brazilian economy. We return to this issue in Chapter IV. 1.30 A key linkage between the energy sector and the Brazilian economy worked through the tavestment programs of the former. Table 1. 5 shows investment in petroleum, electricity, coal, alcohol and the nuclear power program for 1973- 1987, expressed in terms of US$ ut the exchange rate obtaining for the year in question. Also shown are estimates of i (I) investment in electric power and petroleum as a proportion of total energy investment; (ii) total energy investment as a ratio of all investment in Brazil; and (iii) total energy investment as a ratio of GDP. 1.31 The peak in energy investment occurred in 1981-1983, following the second oil price shock of 1979 and the implementation of the national strategy for greater energy self-sufficiency, described in Section 1.6. At that time, the energy sector was absorbing one-fifth of all investment in Brazil and over 4% of GDP; while investments peaked in absolute terms (at over US$1. billion), as a proportion of all investment and relative to GDP. Energy investments declined thereafter, although some recovery took place in 1987. The lower share of electricity and petroleum in total energy after 1975 reflects the growing importance of the alcohol program, investments in which were concentrated in the 1980-1984 period, with the crest being attained in 1981. Within the energy sector, plectricity has traditionally absorbed double the investment by petroleum. Both the power and oil sectors have given the most emphasis to projects for basic production: generation accounted for 50-60% of investment in electric power for the last 20 years; while the weight of petroleum investments went to crude oil and natural gas exploration and produ.,tion since the second oil price shock of 1979. The energy sector will continue to account for a significant part of Brazil's public investment, as energy demand and the economy grow. Later in this study (paras. 3.60 and 4.60), we estimate that energy investment could exceed USS10 billion per year during the coming decade, of which LETROBRAS alone would represent US$7 billion, unless determined efforts are made to implement effective energy conservation measures, including realistic pricing. Demand management and energy conservation could cut this by US$1.2-2.8 billion annually over the next decade, of which USSO.9-2.5 billion would emanate from the power sector. -l Investment in the Eneray Sector 1973-1987 moat t. u Tot. Y P ) :2 it ) (4) 0) (6) (7 1974 0* ,1 1 0 0 ,5 9 18 1975 9 3. 1974 19 L0 11 114 i4 470 94 13?f. 1977 1,528 3.722 1o 378 145 5.785 1 1. 1978 1,8 4,566 7 85 28 6,831 O 5 1979 1,728 6 708 246 300 6,984- 2 14 3 4 73 4 8 13 84 210 .3 6 3 1981 18S 1 9 ,636 S27 10 7 ' 198913 ~6,084 '107 887 703 1164 86 1 .J 1985 2,5 '700 "'97 868 437 4,67? ' O 3 1984 14 5J417 7? 1,t31 41? 6,407 5 1.' < 1985 170 3,89 42 154 30$ 3,1 1486 15 45 261 6 158 5oLi~Gw.... . Mr~pe i dii,fAb 4. qwm9, M~ U, p . ?.~# for'thsas ? 1% *(Ii45 1.32 Since energy prices were held down to contain inflation, the energy sector was encouraged to borrow abroad to expand investment beyond the limits of its profits. This led to an extensive accumulation of external debt, particularly in electricity, and to a rising debt-service burden. Data on the medium- and long-term external debt of PETR.OSMAS and the =LTROMMAS group for 1983, 1985 and 1987 are in Table 1.6. The weight of PZTROBRRS and the WAETROMMAS group in the total external debt of the public sector is now about 120, down from 17% in 1983, due in part to the Federal government assuming direct liability for a portion of the external debt of Brazilian enterprises. 1.*33 The resource requirement* of the energy sector were also high in terms of foreign exchange. Table 1.7 suaaLzes the evolution in the imports and exports of crude oil and petroleum products for the ten-years 1979-1988. AIfter approaching a maximum of US$10 billion in 1981 and 46% relative to total imports in 1983, just before the rapidly escalating investments in energy had time to bear fruit, net imports fell sharply to US$2.5 billion and 17% of total imports in 1988, as Brazil's energy import substitution policy, coupled with falling international oil prices, started to pay dividends. The Enerov Sector And En 17.3 18.0 12.4 and SEAP, in particular, attached a subordinate role to the efficiency soure«. k frat. and fiscal losses which their regulation imposød on the energy eector; and gave higher priority to the implementation of counter- inflationary measures, including price "freezes" (see Chapter II). This has seriously undermined the financial and economic basis of the energy sector, the resource mabilization requirements of which contributed markedly to Brazilian inflation and foreign debt. At the sam time, quantitative vork carried out by D~NAE, for example, suggests that, while energy price increaseø have a modest i~ediate Lnflationary impact, the longer-run consequences are minør. TABLE .7 Imvortp and Ex=rts of Petroum (UG$ millione) 1.6GOVERNMT POLI£Y 1.35 The sharp increases which took place in international ... oil priceø in the 1970. profoundly 1979 434 .1 2.. 3.4 affeoted the Government's policy in 50 9 the energy eector. Brazill s 1982 19,12* . 419 707 dependence on imported petroleum had increased in every year since 1 1970, reaching a peak of 85% of 1986 #7 # 34. total petroleum needs in 1979, making Brazil the largest net. . .. importer of petroleum among the developing countries. In such . ......... circumstances, the major increaseg in the price of imported oil from 1973 had a serious impact on Brazil's economy and balance af payments. In consequence, a national energy strategy was articulated for implementation during the period to 1985 -- known as the Brazilian Znergy Model -- which contained three basic aims: (i) to improve energy and especially petroleum conoervation, including the reduction of certain petroleum use, through inter- *is- fuel substitutioni (ii) to increase the production and the proven reserves of indigenous crude oil and natural gas and (iii) to maximize the utilization of other indigenous energy resources. 1.36 1* ters of supply, greater emphasis was placed on: (i) investments in petroleum exploration (especially offshore and including the opening up of selected areas to the international oil companies), production and ref ining; (ii) hydroelectric and nuclear power plants, with Itaipu and Tucurui notable among the former and Angra I and II the latter; and (ULi) alcohol production from sugar cane, to replace gasoline in the Otto Cycle segment of the automobile market. Some studies were conducted in alternative energy sources, for example vegetable oil to replace diesel, but with no practical results. fore recently, natural gas production and use has been expanded, to replace petroleum products. On the demand side, a relatively favorable price has been set for alcohol, along with tax incentives, to encourage its substitution for gasoline, while a high absolute level was established to cover the much larger economic and financial costs of alcohol; a relatively low diesel price has been set to stimulate the diesel fleet, again in substitution for gasoline and to hold down freight costs; wood and coal burning were promoted in ceramics and in cement and steel respectively, to substitute for fuel oil; and tariff incentives were provided for the consumption of hydroelectric power, at a time when it was in excess supply, again to reduce fuel oil use (the EGTD). 1. 37 The effectiveness of these programs and policies in reducing Brazil's dependence on imported petroleum, aided by falling international oil prices, has been described in Sections 1.2 and I.5: imported petroleum fell from 85% of total petroleum consumption in 1979 to 47% in 1987, with the value of net imports dropping from 46% of total imports in 1983 to 17% in 1988. However, as shown in Section 1.5, the success of the Government's policy was achieved at a very high price, since the resources needed to increase the domestic supply of petroleum, hydroelectricity and alcohol required a major increase in the level of public expenditures, which contributed significantly to the country's macroeconomic problems, particularly in terms of the foreign debt, the fiscal deficit and inflation. The policy led to imbalances in the petroleum sector: falling gasoline demand, along with rising diesel requirements, meant that refinery investments became necessary to adapt the product slate more closely to the demand structure; and there has been controversy over the safety as well as the economics of the nuclear power program, which bas produced virtually no electricity, despite investments of several billion dollars. However, the Government's judgment has been that strategIc aspects of the alcohol and nuclear policies override strict economic considerations. 1.38 Finally, as we argued in the previous Section, Brazil's escalating macroeconomic problems were exacerbated by the resource mobilization requirements of the energy sector and at the same time undermined the Government I a policy in the energy sector, placing the latter increasingly in a subordinate role vie-a- via aoroeconomic stabilization policy. Distortions in the energy price structure were initially intended to promote, for example, alcohol and hydroelectricity, for reasons of energy self-sufficiency. As macroeconomic z%cerns came to predominate, these distortions were aggravated, for example to restrain the prices of LPG, diesel and residential electricity, and the energy price level was also held down. Consequently, energy price increases and investments were constrained and it has become increasingly difficult to discern any clear, consistent or integrated strategy towards the energy sector. -16- CHAPTER II THE INSTITUTIONAL STRUCTURE OF THE ENERGY SECTOR 1I.1 ;NTRODUCTIONs THE ROLE OF THE FRDEAL AND STATE GOVERNM=5T 2.1 Even under Brazil's new constitution, which was promulgated in October 1988, the regulatory influence of the Federal Government in the energy sector is pervasive, especially in the matter of pricing. Only in the case of natural gas do the State governments play a direct role in pricing. Each State government has an appropriate agency to approve the retail tariffs for natural gas, although, even then, the Federal Government enters into the picture, both directly and indirectly. The Ministry of Finance (MOF) has a direct authority over the final approval of all prices in Brazil, not only energy prices, to ensure that they are in line with macroeconomic policies. The Federal Ministry of Mines and Energy (MME) has an indirect influence over retail natural gas prices, through its power to approve wholesale gas prices, on the one hand, and through its control over the retail prices of close substitutes (electricity and petroleum products) on the other. 2.2 The Federal Government also has the responsibility for the coordination of energy policy, through the MME and its National Energy Commission (CNE). In practice, energy planning, decision-making and policy implementation have been characterized by a lack of coordination and clearly-defined goals. In addition, the Government has failed to follow any consistent long-term objectives, having been typically swayed by short-term annual budgetary and immediate anti-inflation considerations. Two of the major players in the energy sector are large and powerful State-owned Enterprises (SOEs), ELETROBRAS and PETROBRAS, under the jurisdiction of the MH. At the same time, there are numerous privately-owned companies or individual entrepreneurs in the fuelwood and alcohol subsectors, which are subject to different or no central control. The Ministry of Industry and Commerce (MIC), for example, oversees alcohol, while fuelwood operates entirely under private enterprise arrangements. Although the Federal Government, through the Secretariat fcr the Control of Public Enterprises (SRST), in the Ministry of Finance, approv-e the budgets of all public utilities, to ensure their consistency with the crerall level of public spending, in practice it cannot exercise this mandata fully, due to shortages of staff and a relative lack of the requisite technical knowledge. 2.3 An important initiative to improve energy sector planning was launched by the Government towards the end of 1990, in the form of the Energy Matrix Commission (CRRME), working under the coordination of the Ministry of Infrastructure (Foreword, para. ii). The Energy Matrix assesses energy sector strategies; policies for energy pricing, energy conservation, and the environment; and investment planning and related policies regarding major supply sectors, such as electric power, oil, gas, alcohol, coal and biomass. The Bank has not yet had an opportunity to review the Energy Matrix, but it appears to provide an excellent vehicle to continue the dialogue with the Brazilian authorities. -17- II.2. SECTOR .INSTITUTIONS 11.2.1 01 and Gas 2.4 The 1988 Constitution gives the Federal Government a monopoly int (i) the prospecting and exploitation of deposits of petroleum, natural gas and other fluid hydrocarbona (it) the refining of both domestic and imported petroleum; (iii) the import and export of products and basic derivatives resulting from the activities in (i) and (ii); and (iv) the maritime transportation of domestic crude oil or derivatives as well as the pipeline transportation of crude oil, its derivatives and natural gas of whatever origin. This monopoly is exercised through Petroleo Brasileiro, S.A. (PETROBRAS), which was established in 1953. Legally, the Government must hold at least 51% of PETROBRAS' voting shares. Ownership of the remaining shares has been closely controlled to prevent foreign domination. As of June 30, 1989 the Government held its minimum legal quota of shares, with the remainder being held bys private individuals and companies, through publicly-traded shares (26.4%); other entities (12.3%); and the National Development Bank (BNDES) (10.3%). 2.5 Since 1976, PETROBRAS has permitted foreign oil companies to prospect for petroleum under risk contracts, which require foreign oil companies to set up Brazilian branches or subsidiaries. Payment for services rendered and costs incurred is to be made in US dollars, based upon a negotiated percentage of the net income from any oil eventually produced. A total of 243 risk contracts have been signed, with the participation of 22 foreign and four domestic firms. These companies have invested US$1.8 billion so far in exploration and development activities, drilling about 200 wells and "shooting" 165,000 km of seismic lines. In the face of concern that such contracts violated PETROBRAS' constitutional monopoly, the 1988 constitution banned all future risk contracts for both oil and gas. However, risk contracts already in existence at the time of promulgating the new constitution were permitted to continue. Additionally, "farm-ins" by foreign companies with Brazilian enterprises are permitted on existing acreage.1 Finally, the new constitution extended PETROBRAS' monopoly to the refining but not the distribution of petroleum products, although existing refineries were permitted to continue in operation. 2.6 Apart from its monopoly in exploration, development, refining and transport, PNTROBRAS has extended its operations into other areas, through six subsidiaries, to cover the production and sale of petrochemicals (PETROQUISA); the retailing of petroleum products (BR); the overseas exploration and production of hydrocarbons and the provision of services and technical assistance (BRASPETRO); the production and sale of fertilizers and raw materials (PETROFERTIL); the import and export of commodities (INTERBRAS); and the evaluation of the hydrocarbon and mineral potential of sedimentary basins (PETRONI4SA). While PETROBRAS is not protected by monopoly in these activities, it dominates both the petrochemical and fertilizer markets (over 80% of market share) and is a major participant in the domestic retail distribution of petroleum products (40% of market share). 2.7 The Federal monopoly over natural gas does not extend to retailing opeistions, as the State Governments have the exclusive right to distribute gas, directly or through State natural gas retailing companies. To date, Brazil's Etf Aquitaine, the French petrolen company, has recently taken advantage of this provisfan. -18- natural gas system has developed in localized regions, primarily to deliver locally-produced gas to local industries. It now consists of five independent sections scattered along the Atlantic Coast (see map of natural gas pipelines). 2.8 Gas distribution networks have been operated in Rio de Janeiro and 8ao Paulo for many years. They are used to distribute "town" gas, which will be replaced by natural gas as it becomes available and the distribution system is upgraded. In Rio de Janeiro, the distribution network is operated by a state- owned company, Companhia Estadual de Gas do Rio de Janeiro (CEG). At the present time they receive natural gas from PETRODBRAS and reform half to medium BTU gas for distribution through the existing pipeline networks. The rest is sold directly to industrial customers and a few residential consumers. Until PETROBRAS started to make deliveries recently, there was no consumption of natural gas in Sao Paulo. Only manufactured gas was distributed, derived from naphtha and refinery gases. By 1992, the Sao Paulo State gas company, CONGAS, expects to replace town gas entirely with natural gas, with the assistance of World Bank financing (under Loan 3043-BR) . A gas distribution enterprise, GASMIG, has been established to initiate planning for gas supply in Belo Horizonte and Juis de Fora, in Minas Gerais. It is currently a department within the electric power company, CVMIG. 2.9 The National Petroleum Council (CNP) was established in 1938 to regulate the supply of petroleum and its derivatives. It sets standards for petroleum and alcohol and oversees their distribution and marketing. The CNP establishes the prices of all liquid petroleum products, alcohol and the bulk supply prices of natural gas. It also determines transport rates for the distribution of liquid fuels and allocates supplies of alcohol to PVTRosRAS and other vendors. 11.2.2 Electric Power 2.10 The Brazilian power sector is one of the largest and most complicated in the world. At the end of 1988, total installed generating capacity in Brazil had reached mearly 50, 000 NW. About 80% of the total was in the South and South- East (including Itaipu) and over 90% was hydroelectric. The country has two separate transmission systemes one interconnects the South, South-East and Center-West Regions; and the other the North-East and part of the North Regions (see maps). At the end of 1988, the transmission systems had a length of about 53,000 km at 230 kV and above. 2.11 As with oil and gas, the Federal Government has the exclusive right to the ownership of all hydroelectric facilities, the exploitation of which can only occur through governmental concession or authorization. The Federal Goverament also has a monopoly overt (i) the exploration, mining, reprocessing, industrialization and commerce in nuclear materials and their derivatives; and (ii) electric power generation and distribution, although these can be delegated to concessionary companies. In practice, the Federal Government has granted concessions to companies with shares owned by Federal and State governments, as well as private interests, to carry out generation and distribution. Large-scale generation and high-voltage inter-regional transmission of electricity (at 500, 345, 230 and 115 kV) are under the control of four Federal concessionst FURNAS, CHESP, ELETRONORTS and ELETROSUL. The State, municipal and private concessionaires are responsible for "supplementaryu generation and transmission and (especially) distribution to final consumers. CUP (Sao Paulo), CEMIG (Minae -19- Gerais), COPEL (Parana) and CEEE (Rio Grande do Sul) are also responsible for power generation within their concessions. 2.12 In 1F62, Centrais Bletricas Drasileiras (ELETROBRAS) was established under MME as a mixed-economy corporation. Legally, the Federal Government must hold at least 51% of the company's voting stock. in fact, the Federal Government currently holds 99.8% of total company shares and 100% of the voting stock. ELETROBRAS serves as a holding company for the four Federal concessions mentioned above and for two distribution companies, viz. LIGHT and ESCELSA. ELETROBRAS holds the 50% interest of the Federal Government in Itaipu Binacional (the agency set up by the Governments of Brazil and Paraguay to build and operate jointly the 12,600 14W Itaipu hydroelectric facility on the Parana River) and owns minority stock interests in the State-owned utilities. Under a decree in August 1988, ELETROBRAS was assigned the task of overseeing the completion of construction of the Angra II and III nuclear plants, through FURNAS. ELETROBRAS serves as a technical support for IME; analyzes expansion plans for major generating and transmission facilitiesi and coordinates and supervises the inter-connected public power system. 2.13 The National Department for Water and Electric Energy (DNABE), within MME, is the regulatory agency for the power sector. It grants licenses for generating plantsp assigns concessions approves expansion plans I and sets tariff levels and structures (see Chapter III). In some areas of responsibility, the approval of other Federal authorities is also required, e.g. that of the President for major licenses and concessions and of the MOP for tariff increases. 2.14 The institutional anomalies and over-lapping functions in the power sector have been examined under the Power Sector Institutional Study, REVISE ("Revisao Institucional do Setor Eletrico"), which was initiated as a condition of the Electric Power Sector Loan (Loan 2720-BR), under terms of reference acceptable to the Bank. Pricing issues and the potential for increased participation of private interests in the electric power sector were also analyzed in the study. A final version of the study will be issued after the new Government has had the opportunity to review its recommendations and propose specific action plans. The Government has agreed with the Bank to exchange views on these matters from time to time. 11.2.3 Alcohol 2.15 The production of alcohol is entirely in the hands of private producers. Over two-thirds of Brazil's cane area is devoted to alcohol and not sugar production. The bulk of alcohol production capacity is in the South-East Region. About 80% of the cane in the South-East, South and Center-West is used for ethanol production and 20% for sugar, while only 45-50% of cane in the North- East produces alcohol. 2.16 The sugar industry is regulated by the Sugar and Alcohol Institute (IAA), which is subordinate to MIC. Other important agencies are the National Alcohol Council (CNAL), an interministerial body charged with setting overall policy directives on the national alcohol program, and the National Alcohol Executive Commission (CENAL), an interministerial commission which is located in MIC and which approves projects for the installation or expansion of alcohol production capacity. The key institutional link between alcohol and the energy sector is CPN, which sets retail alcohol prices and controls the marketing of alcohol. .20- 11.2.4 Coal. Wood and Charcoal 2.17 As in the case of other natural resources, coal exploration is a monopoly of the Federal Government. This monopoly is exercised through the National Department of Mineral Production (DNPM), in M. Production is carried out under a system of concessions and is almost entirely in the private sector. The coal industry is regulated by CNP, which fixes prices, production quotas, imports, transport arrangements and the consumption quotas which have been imposed from time to time on large industrial consumers (e.g. steel and metallurgical coke). 2.18 Wood and charcoal production and marketing is highly decentralized and totally privatized. Until recently, however, the Brazilian Institute for Forestry Development (ISDP), in the Ministry of Agriculture, was charged with the oversight for exploration and use of both coal and charcoal. In February 1989# IBDF was eliminated and replaced by the newly-created Brazilian Institute for the Environment and Renewable Natural Resources (IBAMA), in the Ministry of the Interior. 11.3 INSTITUTIONAL bSPCTS OF PRICING 11.3.1 IntrXodction 2.19 With the exception of wood and charcoal, energy prices have been subject to a tightly-controlled system of Federal Government regulation. The existence of a number of different regulatory bodies, alongside numerous privately-owned companies or individuals, sometimes free from Government control, has given rise to a highly-fragmented and uncoordinated situation. The principal Federal agencies are DNANZ, CNP and ZAA, with the first two being under MHZ while the third is subject to MIC. The prices of gas are established by the individual States. These agencies have pursued disparate policies in establishing the prices of electricity, petroleum products, natural gas and alcohol. Further distortions are introduced by the Secretaria Especial de Abastecimento e Precos (SAP), which must examine or approve retail energy prices in light of the Government's anti-inflationary policies, which are often inconsistent with prices based upon sectoral economic considerations. 11.3.2 fA= 2.20 The one common point of reference for all energy prices is SEAP. However, in exercising its role, it relies not on any concept of energy policy, but purely on macroeconomic concerns, particularly the Government's overall anti- inflationary policy. SEAP has not confined its intervention to keeping down the level of energy prices. It has played a key role in promoting or at least continuing distortions in the price structure. Notably, it has examined the impact of proposed price increases for residential electricity, gasoline, kerosene and LPG on the cost-of-living index; and consistently tried to moderate those price increases, in the belief that thereby cost-push inflation would be contained. In contrast, it always left aside or underestimated the inflationary impact of its pricing decisions on the fiscal deficit and hence (indirectly) on inflation as well. -21- 11.3.3 g 2.21 Proposals regarding the level and structure of prices of petroleum products, natural gas and alcohol are made by CNP for approval by SEAP. These prices are all fixed on a uniform basis geographically. The general guidelines of pricing policy are based on the Government's primary objective of reducing the country's dependence on crude oil imports, and specify thatt (I) the pricing policy should be flexible, so that the Government can promote the production and consumption of locally produced fuelst (ii) the prices of all petroleum products should be adjusted gradually but continually, consistent with the anti- inflationary policy of the Government. to reflect the level of international petroleum prices and exchange rate policy; and (iii) the prices of nationally- produced fuels should ensure an adequate return on capital as a priority instrument for mobilization of the private sector, which will be subsidized when necessary. As a result of the above policy: (i) the price of alcohol to the consumer was to be no higher than 65% of the price of regular gasoline, a limit which was raised to 69% in July 1988, and to 75% in January 1989; (ii) the price of diesel was to be increased gradually to reduce the differential with gasoline; (iii) the price of fuel oil was to be adjusted gradually to achieve parity with alternative fuels; and (iv) the price of LPG was to be adjusted to a level that would promote conservation by consumers. 2.22 The 1988 Constitution established that the States, not CNP, will be responsible for decisions on gas prices to consumers. CNP's role is now limited to fixing the transfer or bulk supply prices from PETROBRAS to the distributing companies at the city gates along the main transmission gas pipelines. Recently, in the context of the Sao Paulo Natural Gas Distribution project, CMP confirmed its intention to base those transfer prices on the economic cost of supply, with due allowance for macroeconomic and social considerations. As long as natural gas is in limited supply, that means that transfer prices should be set with reference to the opportunity cost, i.e. the cost of substitutes. Since CNP continues to fix the prices of these substitutes, notably fuel oil and LPG which compete with natural gas, this indirectly puts a cap on natural gas prices. It is recognized, however, that a 20% premium on natural gas prices will not affect its competitiveness, given its superiority in terms of cleanliness, ease of use and lack of storage requirements. 2.23 CNP also has a role in setting coal prices, but recently this has become minor. Since coal is used mostly as a substitute for fuel oil in industry, the maximum coal price is fixed at 80% of the fuel oil price in thermal equivalent. In the past, the mine-mouth price was also fixed -- at 40% of the fuel oil price -- and transport costs were subsidized as necessary to ensure that the 80% ratio in the final price was not violated. The transport cost subsidy was terminated in January 1989. The coal price is now set fully by the market, up to the 80% limit in terms of equivalence with fuel oil. 11.3.4 IM 2.24 IAA sets all sugar and producer ethanol prices through a complex system designed to ensure parity between the two products. There is no link between domestic and world prices and it appears that prices are set on the basis of production costs. PETROBRAS buys all the anhydrous alcohol and half the .22- hydrous, the remainder being distributed by other (private) companies.2 The ethanol is sold to PETROBRAS on a basing point system, which assumes that all fuel ethanol is distributed via a major PETROSRAS distribution center. Thus, a distillery that is well located to supply a nearby area whose own ethanol production falls short of local demand is obliged to sell the ethanol to PVTRORAS at a price that is calculated as if the ethanol would be delivered to the final consumers only after being transloaded at the distribution basing point. In this way, the distillery loses the benefit of any proximity advantage that it enjoys in practice. PETROBRAS' margin from the sale of alcohol has progressively declined as the international price of gaboline fell and PETROMS has reportedly suffered financial losses on the sale of alcohol since 1986. 11.3.8 ] N= 2.25 The responsibility for setting electricity tariffs was bestowed on DNAN8 in accordance with legislation which has evolved over many years. The 1asic criterion which was established as early as 1934, in the Water Code, is the inancial principle of service-at-cost plus an adequate return on capital investment. This return is supposed to be 10-12%. Furthermore, since 1974, tariffs are geographically uniform. More recently, in 1982, a major new criterion was introduced, requiring the implementation of a tariff structure based on long-run marginal costs (LRMC). This latest development became possible after 1977, when a technical cooperation agreement was concluded between DNAEU and LZTRODBRAS. With the participation of all the power utilities and assisted by EDP (with the support of the World Bank under Loan 1300-BR), DNAsZ initiated a comprehensive study of electricity tariffs incorporating marginal cost principles. An initial version of the study was completed in March 1979, and the first step to initiate tariffs based on L1ENC was taken in October 1981. 1X.3.6 The States 2.26 The States enter into the setting of energy prices only through their power to determine the retail prices of gas. This power is significant mainly in the States of Sao Paulo and Rio de Janeiro. In the former, a major study of natural gas prices is ongoing, supported by Loan 3043-BR. The State of Sae Paulo has expressed its intention, upon completion of the study, to base retail prices for natural gas on LRMC. However, as noted above, since CUP fixes the prices of the major competing fuels for natural gas (notably fuel oil and LPG), CNP will continue to influence natural gas prices indirectly at the retail level. 11.4 CONCLUSIONS AND RECOMMENDATIONS FOR CHANGE 2.27 The experience gained with the operation of the institutional framework for the energy sector during the 1980s demonstrates clearly the role that the Government should and should not play in energy pricing and investment. Energy pricing must be consistent with the Government ' a broader macroeconomic and social goalel however, as shown in Chapters III and IV, the Government's distortion of the structure of energy prices, and its refusal to allow the price of electricity to reach appropriate levels, had disastrous consequences. The macroeconomic benefits were largely absent. The indirect inflationary effects of failing to raise energy prices, by exacerbating the public sector deficit, are 2 The use of safohot in vehicLes is discussed nd eptained in Chapter I. -23- worse. The design of future macroeconomic stabilization measures should recognize this experience. For this reason, we recommend eliminating the role of SEP, since it has been counter-productive. Greater emphasis in the future would accordingly be given to economically efficient prices, set according to the financial and economics needs of the energy sector. 2.28 As Government reduces its regulation of the energy sector, it could focus its attention and resources according to stricter priorities, where public action is needed, such as in education and health. The private sector has a direct commercial interest in ensuring reliable and low-cost energy supplies; and the public agencies could deliver energy in a more efficient way than at present with lose government interference. However, the application of economic energy pricing is a prerequisite to create a regulatory environment which will foster more private sector participation and more efficient public sector performance. Economic pricing includes introducing regional differentiation. 2.29 The 1988 constitution restricts the range of alternatives available to Brazil to improve the institutional framework which governs the energy sector. The ownership and exploitation of oil and gas reserves has been vested exclusively in the federal Government, with PETROBRAS acting as the legal representative of that monopoly. The constitution expressly forbids the Federal Government from "granting" or "conceding" any participation whatsoever in the exploitation of oil and gas reservee, except with regard to local governments in appropriate circumstances. In particular, PETROBRAS is barred from entering into new risk contracts with foreign companies for the exploration of either oil or gas. Overseas interests can only participate in such activities through "farm- ins" of existing acreage by companies that already hold concessions in Brazil. The monopoly does not include the export, import and marine transport of petrochemicals and nitrogen-based fertilizers; and the distribution of petroleum products, petrochemicals and fertilizers may continue outside the monopoly of PETROBRAS. 2.30 The provisions of the new constitution for the electric power sector are somewhat different. Article 176 states that "mineral resources and hydraulic energy potential constitute property distinct from the soil for the effects of exploitation or utilization, and belong to the Federal Government". The working or operation of these resources can, however, be conceded to Brazilian companies. In fact, although growing government intervention and control took place in the industry after 1960, it was established with a strong private sector orientation. However, in light of subsequent history, the legal framework for electricity supply needs revision, to facilitate renewed private sector participation, as discussed below. 2.31 Apart from the general arguments in favor of greater private sector participation to increase efficiency, the future resource needs of the energy sector, which could amount to US$10 billion per year in the coming decade (see paras. 3.60 and 4.60) will necessitate a search for alternative sources of financing and ownership. Even if Brazil achieves greater success than in the past in raising the real level of prices to generate internal funds, the public sector alone is unlikely to be able to mobilize the resources necessary to expand the energy sector on the requisite scale. With the right legal and institutional framework, direct or indirect financing of that expansion is an attractive option for the private sector, given the serious implications of energy shortages. 2,32 in oil and gas, the Government should encourage greater private sector participation in the areas of PSTROBRAS which are not protected by its monopoly, e.g. subsidiaries such as PETROQISA and PMTROFERTIL, which are not reserved to the Federal Government under the Constitution. The increased participation of the States in gas distribution is to be welcomed and should be pursued with greater private involvement in State-level gas companies. COVMAS shows every promise to be a leader in this regard. 2.33 In electricity supply, the scope for tapping private sector resources exists in generation and distribution, although adjustments are needed in the legislative and institutional framework, to eliminate or minimize impediments. As in gas, private participation in distribution, at the State level, should be expanded. Generation facilities in private industrial or agro-industrial complexes potentially offer significant economies in the supply of electricity to public systems, although autogeneration now accounts for only 8% of total installed capacity. of note is the use of bagasse from sugar production processes, as an input to power generation. A major obstacle to increasing the share is the lack of a clear policy on the guarantees which would be given to the private owner, concerning his own access to the output. Adequate guarantees are also needed regarding payment for the generation which is sold to the public system. Furthermore, while the legislation allows for the transmission of electric power generated by private facilities within a State, laws are required to permit transmission on a wider scale, e.g. between States. Strict constitutional limitations surround the private ownership of hydroelectric facilities, although concession arrangements could be established with private firms, as mentioned earlier. In such cases, provisions would need to be made to facilitate land acquisition, with due allowance for the protection of affected populations and the environment. 2.34 Several different models or options exist for private sector participation in electric power. As stated in par&. 2.28, adequate tariffs are a prerequisite, to create a framework in which investors are willing to risk their capital. The models includes (i) the establishment of consortia between private and public companies; (ii) the assumption by private investors of responsibility for works in progress but arrested by the lack of public funding, including the subsequent operation of the facilities, for example through arrangements to build, own, operate and transfer (BOOT); (iii) the granting of concessions to private interests to operate isolated systems, using competitive bidding procedures; (iv) signing contracts with private interest* to buy energy, for example with full integration of the private generation within the public systeml (iv) signing contracts with large private consumers to guarantee a certain level of future sales, in return for an advance payment to help to finance the facilities; and (v) allowing private investora to purchase public debt, in return for equity participation in the power sector. 2.35 Along with our proposals to extend private sector participation in the energy sector, we recommend institutional changes at the Federal level. As we have mentioned (para. 2.27), intervention by non-sectoral agencies, such as SNAP, should be diminished. Raising the prices of all petroleum products to, at least, their economic costs implies a greatly changed role for CNP (para. 4.54). Government's main activity would be "auditing" the proper application of this pricing policy at the ex-refinery level; and the regulation of PETROBRAS as a transporter of natural gas, crude oil and petroleum products. Competition would play an enhanced role in determining retail prices, taking into account regional differences in transport and distribution costs (para. 4.54). At the same time, -25- the Government would play a more clearly defined and restricted part ins establishing and maintaining the overall policy framework for energy; ensuring the application of a consistent planning methodology in the publicly-owned parts of the sector; help manage the reduction of public and promote more private involvementl coordinate energy sector policies with other Ministries and provide full informational support to the energy sector. Items specific to electricity and to hydrocarbons could be delegated to streamlined versions of DNARE and CNP, respectively. 2.36 A key item which MM should iddress in the overall policy ramework relates to pricing policy. As we show later, much needs to be done to place the structure of petroleum product prices on as rational a basis as electricity tariffs, to support economic efficiency objectives. The price level, especially in electricity, should be moved into line with the economic costs, with additions as appropriate for taxes. The price level must also yield a dsgree of profitability to ensure an adequate flow of private capitals our Ualculations later in this report suggest that a price level based on economic costs is likely to be financially satisfactory. We recommend, on both financial and economic grounds, that a rapid move be made to regional tariff differentiation. 2.37 0B would approve the investment programs of the energy sector on the basis of agreed guidelines for preparation. Notably, such guidelines would cover GDP projections, discount factors and other planning parameters, which are used by the Grupo Coordenador de Planejamento dos Sistemas 916trice (GCPS) in the preparation of sector expansion plans. 2.38 The role of MNE in preparing the ground for the extension of private sector involvement and ensuring the application of correct "rules of the game" would be critical. Legal changes are necessary to deal with the obstacles we discussed above,to ensure a healthy influx of private capital, while giving full attention to protecting the environment. Guidelines must be established for the types of concessions to be introduced and the contracts to be employed. 2.39 Inter-ministerial coordination will be required to implement a sounder policy with regard to alcohol production, which is entirely in the hands of private producers. Coordination will also be needed with the Ministry of Finance if, as we emphasize in Chapter IV, a more effective tax base is to be developed for petroleum products. 2.40 The informational role of MME is not new, but it assumes a different dimension if MME is to act as a catalyst to private sector investment. Evidently, it will be essential to have a central point to collect, organize and distribute the information needed to take economically and financially sound decisions. 2.41 A strengthened version of CNN might form a good nucleus for the activities listed above. It should include a wide representation of interested parties, including consumers, environmental interests and private investors, as well as ELETROBRAS, State and other companies and the Federal Government. 2.42 The Energy Matrix (para. 2.3) will provide several crucial elements for future decisions on an 6nergy strategy for Brazil. It can serve as a focal point to raise the key issues in energy planning, regardless of their political sensitivity, and to bring together representatives of the main agencies active in the sector, thereby permitting the emergence of differences of opinion and .26- contct of interest. While the Bank has not yet had the opportunity to review the Xnergy Matri, it appears to be an excellent vehicle to continue the World Bank's dialogue with the Brazilian authorities. . 27• ELECTRIC POWER 2II.1 TE_ EhETRICITY . MAET 3.1 The sectoral distribution of electricity consuemption in Drazil (Annex 12I.2) reflets the process of increased Industrialisation, with industrial deand abeorbing more than half of total electricity demand (53% in 1989). The perspectives are that this share will remain constant to the year 2000. As shown In Annex 111.2, the regional structure of electricity demand has changed steadily over the past fourteen years. The sharre of electricity coneumption of the less developed regions of the North and the North-East increased, as electricity service was extended more widely to residential consumers and the industrial and agricultural sectore experienced economic growth. Furthermore, expansion in electricity demand was encouraged by the development of large electricity- intensive industries in these regions, which benefitted from special low tarifuf to promote export products derived from alumina (see Annex III.9). Even mo, in 1989 over 80% of electricity demand was concentrated in the South/South-East. TABL III-1 Alectricity Gener&tion: Blatorical. & Pro-leted (TWh) fotlUsere 72.8 224.t 4l0? 8.4 5.7 St*ietf* 32 L 78.0 83 ..11:. S 2. 9t L. SE te 10.9 K7 46.4 4. t.. .. ... 1. 10.. 6o e.. ..9 .4 .. .. outest 47.7 130. 24 7.5 .1 toade 10. 21.7 46-4 7.8 4. 4bt~- ....>Ij ~ L..... . . .. .. . III2 * 2 XIIX.2 =LECCT DEN FRCAT IX1.2.1 The ELTRODRAS Forecast. 3.2 gLETROBRM coordinates the preparation of electricity dmand projections, with the participation of about 35 power setor utilities. The adopted set of forecastö in principle incorporates the sf fect of energy conservatior goals --although without explicit consideration of prices-- and eventuallT serves as the basis for the industry's long-term expansion plan. Projectd global electricity dmand Ig the result of adding the individual projections prepared by the sector utilities, which are estimated on a S28 disaggregated basis. For the short-term, the projections are based on trend analysis of past demand by sectors, correcting distortions derived from actual suppressed demand or outageal information about new industrial loads and expectations about consumer connections. For the long-term, the projections are based on eectoral analysis (residential, commercial, rural, industrial, public lighting and government), taking into account correlation with macroeconomic (gross regional product, industrial product, household income, etc.), and socio- demographic variables (population, number of dwellings, etc.) and expansion plans for large electricity-intensive industries. 3.3 ELETROBRAS' demand forecast for 1989-2000 (Table 111.4) projects an annual rate of increase for 1989-95 (6.2%) slightly below the increase achieved in the nine years from 1980 (6.4% p.a.) and much lower than the average for 1975- 89 (8.4%). The growth of electricity consumption is expected to slow further after 1995. In the case of the more developed parts of Brazil, where greater saturation of the power market exists, such as the South and South-East, electricity sales are expected to increase more slowly than in the North and North-East. As a result, the share of the less developed regions is expected to increase from 19% in 1989 to 24% by 2000. 3.4 To forecast future generation requirements, RLETROBRAS allows for transmission and distribution losses. During 1989, these were in line with commonly accepted values for utilities in developed countries, amounting to about 12.9% of net energy requirements. ELETROBRAS predicts that system losses will decrease as a percentage of generation during 1989-2000 (see Annex 111.1), due to better utilization of the corresponding transmission systems and improvement in distribution systems, in particular through the implementation of the National Electricity Conservation Program (PROCEL) (Chapter V). ELETROBRAS' forecast of electricity consumption and power system losses and the corresponding required generation are shown in detail in Annex 111.1 and summarized in Table 111.1. 3.S ELETROBRAS' methodologies for demand forecasting compare favorably with other utilities. Nevertheless, given the size and complexity of the Brazilian power system, several issues associated with ELETROBRAS' demand analysis, as presented in the Plano 201Q, merit consideration. 3.6 First, the ELETROBRAS methodology is based essentially on a single demand scenariol. Such an approach has merit under stable economic conditions but has given rise to growing deviations between the forecasts and the realized market growth since 1980, when the period of rapid and sustained economic growth cam to an abrupt halt. The consideration of only one demand scenario makes no provision for the macroeconomic uncertainty confronting the Brazilian economy and weakens the process of strategic planning. No consideration is made of the financial con-traints on the sector in implementing its investment program; and further uncertainty concerns the impact of the energy conservation program which is being implemented. 3.7 Secondly, price is not considered as an explanatory variable. Given the prevailing financial difficulties of the power sector and the inflationary environment, it is important to consider the effects on the power market of alternative pricing and tax scenarios. The PlnI 2010 refers to an atermntfve scenario but efopty concludes that the original forecasts are conservatiad m coatibe with historical experience. . 29 - 3.8 & third problem with the demand analysis is the way in which it incorporates the impact of special industrial loads. The regional utilities survey future potential non-marginal loads, which they add to their normal projections. Aside from the uncertainty regarding the emergence of these new loads, there is potential double o.wunting. The projection is based on historical data, which already include the addition of previous non-marginal loads, so that the data need to be screened before projecting and adding the new loads. 3.9 Fourthly, there are problems with the way in which Plama201Q handles energy conservation and demand management, in particular the targets and policies of PROCEL. It appears that the forecasts were conducted independently of the conservation program and it is not clear what effect prices have on energy demand, as distinct from the impact of special conservation programs. Since, increasing prices are not explicitly considered as a policy instrument, PlagS 201Q seems to overestimate the benefits of energy conservation. According to Plano 2010, electricity consumption is reduced by 12.3% in 1995, compared with the level which occurs "without the conservation program"I and by the year 2000 the figure has reached nearly 20%. These results may be examined with the help of the relationship between the ratio of growth of energy consumption and GDP shown in Table 111.2. The ratio of the former to the latter, for 1970-75, 1975- 80 and 1980-85, increases with the decline in the rate of growth of GDP due to the influence of past investment in equipment and appliances (i.e., due to the asymmetrical response of energy demand to upward and downward movements in GDP). For the future, one expects the ratio to decline with improvements in technology, rising energy prices and conservation measures. However, the assumption of Plano 202 that the ratio will fall below unity as early as 1995-2000, seems overly optimistic, without action to increase the electricity price level. In industrialized economies, ratios around unity are the norm for economic growth rates of about 2-3% p.a. Th=TLII.2 Relationshin Between Ratios of Growth of Electricity Demand and GD? in Plano 2010 With and Without neray Conservation with wtout ()(2) Cowervation OP Divided by Divided by 19705 12.3 10.3 1.19 19?S-SB. 12.1 7.1 1.70 1980*85 7.4 1.7 4.3S 198590 .7.3 7.6 6.8 1.07 1.12 I990-95 6.3 7.0 6.1 1.03 1.15 1995-2000 5.4 6.3 5.8 0.93 1.09 2000-00. 5.0 5.4 5.4 0.93 1.00 2005*2010 4.3 4.S 4.7 0.91 0.96 3.10 Finally, ELZTRORAS' forecasts are not consistent with those of PETROBRA in terms of the assumptions and statistical basis used. As early as 1986, PETROBRAS and BLETROMS began an attempt to arrive at a common framework *30 * for planning supply on a consistent basis, in order to avoid bottlenecks and sector imbalances. A working group was set up to define alternative economic scenarios up to the year 2000 and examine the consequences on the evolution of energy demand. However, in 1989 both companies are still forecasting demand with different demand scenarios, rendering their forecasts incompatible. 11.2.2 AlternatAv Deman Scenarios 3.11 In order to address some of the methodological issues mentioned above and to incorporate different scenarios for electricity demand, we have developed alternative forecasts based on simplified analytical models derived from regression analysis (see Annex 1.9).2 Two pricing scenarios were selecteds the basic scenario, which is referred to as "Business As Usual" (BAU), assumes that the tariff structure is unaltered and the level remains constant in real terms. In the second scenario, called the "Bfficient Pricing Scenario* (BPS), residential and rural electricity tariffs are brought into line with long-run marginal cost (LUAC) by 1995; and industrial and commercial tariffe by 192. The underlying analysi- of LRMC and tariffs is presented in Sections 111.4 and III.S. Both scenarios incorporate the same assumptions regarding GDP growth, using World Bank projections. According to the scenarios, the movement towards LJMC pricing reduces electricity demand by 10% under BPS compared with RAU in 1995; and by 16% in 2000. A description of the key assumptions underlying the VtAMRoAS, BAU and UPS forecast scenarios is shown in Table 111.3; while Table 111.4 and Figure 111.1 summarize the results. 3.12 The demand projections of the BAU scenario are somewhat lower than those of ELXTROBRAS. The differences c *n be explained by different assumptions concerning GDP growth rates. BLETROBRAS' projected GDP growth rate is lower than the Bank's for 1990-95 (4.5% p.a. versus 5.2% p.a.), but equal from 1995 to 2000 (6.0% p.a.). Hence, the BAU forecast is 7 percent lower than ELUTROBRAS in 1995, declining to 3.7% by the year 2000. These rasults would seem to confirm the judgement expressed above, namely that BLETROBRAS' forecasts do not incorporate price effects on demand and do not integrate energy conservation in a satisfactory manner. The energy savings of PROCEL seem to have been added to ELUTROBRAS I projections to obtain forecasts "without the conservation program". 3.13 The BPS forecast, in contrast, which incorporates the effect of price elasticity, exhibits striking dissimilarities from BLMTROBRAS. Under RPS, in line with the discussion of section 111.5, the power tariffs for industrial, commercial and otber users would have to increase by 32%, 3% and 78% respectively by 1992, compared with 1989 price levels; by 1996, the residential and rural tariffs would have to increase by 124% and 154% respectively. The impact of these price rises is substantials in relation to ZLETROBRAS, electricity demand would be 16 percent lower under UPS in 1995 and nearly 20 percent lower by the year 2000. 2 ELETRORAS does not accept the validity of thes demand forecasts. - 31 - TABLE Inj. Aternative Forecast Scenarios . .. . ..... .. ....*... ... . Pffices 0satant 1989 Prices Reach 1aeve of 4.#C Syr Uot coaldereC 199 for industr$iaan -n tural by 199 R Dm1990.1995* 5.2. . m2 as 59.4 .990-195. 4.5 1995.2000: 6.02 1995-000 6.02% Alternative Demand Forecasts t 1980 198 1989 1990 1995 2000 9 1 99 199 Ridenti 23.2 40.6 42.2. 44.7 59.4 78.0 8.7. 6.9% SM *;. -Ind~al .61.5 99.9 102.8 108.8 148.6 191.6 8A 2 6.3· 5.2% ··Others 27A - 46.2 50.4 53.3 72.3 947 8.0 - 7.0· 6.2 5.5 Total 12.2 186.7 15.4 206.8 280. 364.3 8 4 6.4 6.22 A4 AL.TERNATIVEU BAU · Rsfduntfat 42.2 44.3 58.9 85.5 . . .w Iustr7i p 102.8 109.0 142.4 185.6 . 5.6% 5.A Other 50.4 52.8 59.7 81.6 2·.9 6.5 p otal 195A 206.0 260d.9 350.7 4.9% 6.1a Residentlisl 42.2. 44.1 55.6 76.6 4.72 6.6. . ndustrLøt . . 102.8 107.3 ¶20.8 146.8 . 272 4,02 .Others -50.4 52.8 58.5 70.5 . 2.95 5.8% Ttal 195A4 204.2 234.9 295.8 3.12 .4.62. 3.14 s in the cass of ELETROBRAS, the BU and EPS forecast are subj Act to a range of uncertainties. One uncertainty concerns the economic growth rate. If GDP grows at 6% p.a. on average during 1990-2000, total electricity denand will grow at 74 p.a. in the AU scenario, and about 5 p.a. in the lPS. However, if economic growth slows down, say to 4.4% and 6% p.a. respectively for the two period. 1990-95 and 1995-2000, electricity demand growth vill drop to 5.8% p.a. over the whole planning horizon in the BAU cass, and 3.5% p.a.in the EPS. A second major uncertainty surrounds the estimation of price elasticity of demand in the EPS scenario, because there are major econometric problems in estimat ing price elasticities from past data and practical problems in projecting them forward for non-marginal changes for as long as a decade. Finally, the EPS scenario showa the outer limit of what might be achieved in energy conservation and demand management through efficient pricing: if demand falls by 20% in 2000, U.=O can also be expected to fall, an*d lower prices would bu appropriate. Nevertheless, we conclude that demand ie likely to lie in a range defined by - 32 - 400? 35O ltrba 300 200 1980 1985 1990 1995 2000 Figure III.1: Alternative Power Demand Forecaste, 1990-2000 ELETROBMS/BAU at the upper end and EPS at the lowerl and our judgement is that demand could be cut by at least 10% and probably 15% by the year 2000 with a determined move towarde LRMC pricing in level as well as structure. We strongly recommend that ELETROBRAS' studies of future electricity demand incorporate econometric analysis of price elastickties and take into account alternative price assumptions and the impact of LRMC. 3.15 The alternative demand scenarios conf irm the importance of adjusting tariffe to keep the level as well as the structure in line with LRMC in real termg and to stop the erosion caused by inflation. They also underline the importance of tariff policy in supporting energy conservation efforte. If tariffe deteriorate in real terms, both demand and investment requirements will be subøtantially larger, financial problems will be aggravated, the quality of service will decline and the risk of supply interruptions will increase. We shall return to these points again in Section 111.6, following a discussion of LRMC and tariffe in Sections 111.4 and 111.5 respectively. . 33 111.3 POER SYSTEM PLNNG, EPANSION AND INVSTMNT 111.3.1 Planning Mhodlagg 3.16 BLETRODRAS has the primary responsibility for the coordination of planning, which includes the preparation of generation and transmission expansion plans. Each utility is responsible for distribution planning, but ELETROBRASI participation in the global assessment of the utilities' investments is intended to ensure cosistency. As we saw in Chapter II, ELATROBRAS' plans are subject to the approval of MME and the overview of DNAR8. Since 1981, SEST (within MOP) has approved the sector' s budget, to ensure consistency with the level of public investment spending. Details on the sector's planning methodology are in Annex 111.3. 3.17 The sector's planning methodologies are generally reasonable, and certainly better than in most other Latin American countries, although there are ways in which they could be strengthened, particularly regarding the treatment of risk and reliability. In generation, a 5 risk of deficit has been accepted as the norm, without any clear basis; at least, ELTROBRAS should evaluate alternative reliability levels, in terms of the economic cost of power outages. In transmission and distribution, risk is incorporated in a much less sophisticated manner, through design standards based on norms and experience; more use should be made of probabilistic methodologies and (as in generation) outage costs need to be considered. Furthermore, alternatives to bulk supply from the main power system need more careful analysis, such as local generations the promotion of private sector participation in generation, which we support in Chapter II, represents a rich potential to augment alternatives to public supply. 111.3.2 Environmental Asnects 3.18 Since the mid-1970s, the Brasilian power authorities have made significant progress towards implementing environmental and social safeguards in the power sector. Notably, they have taken initiatives to mitigate the harmful side effects of large hydroelectric projects on river basin ecosystems and human populations. In connection with the Bank's 1986 Power Sector Loan (Loan 2720- BR), BLETROBRAS has completed the Environmental Master Plan for the Power Sector, which sets the framework for: (i) national environmental policy, legislation, regulations and guideliness (ii) specific environmental guidelines for the power sector; (iii) the preparation of environmental and social action plans on a project-by-project basis; and (iv) measures to strengthen the sector's ability to implement the plan. Furthermore, SLETROBRAS is developing a methodology to integrate environmental effects directly into project appraisal, through an Environmental Index, which is considered jointly with standard cost-benefit measures. 111.3.3 The Rxpansion Plan 3.19 The power sector's expansion plan is based on Plawn 2010, a comprehensive long-term planning study, completed in early 1988. Details of the generation expansion plan are in Annex 111.4; while the transmission and distribution plans are in Annexes 111.5 and 111.6 respectively. 3.20 Plang.201Q was a commendable step to subject the discussion of the sector's planning and methodology to a much wider audience, thus making the public more aware of the sector's problems and needs. Plano.2010 is based 33'. essentially on developing hydroelectric power in the expansion of generating capacity. The share of hydroelectric installed capacity is projected to remain in the order of 90% until the year 2010. 3.21 A number of issues arise with this solution. Since Brazil has already developed its low-cost hydroelectric energy, the available hydroelectric options outside the North are relatively expensive compared with the past (with costs around US$1700/lW and over as against US$1200/kW for Itaipu). After 2000, most additional hydroelectric capacity would, therefore, come from the Amazon. Those projects are large and require the transmission of huge blocks of energy, equivalent to Itaipu every five years, over long distances, to the consuming centers of the South and South-East (see map of principal regional electrical interconnections). The solution presents a major technical challenge to the industry and raises transmission costs substantially. Equally important, there are serious environmental risks and LETRORAS will not find it easy to reduce the environmental costs to an acceptable level, despite the significant strides forward that it has made in environmental analysis. However, ELETROBRAS' calculations continue to demonstrate the economic merits of hydroelectricity, despite the transmission and environmental costs. 3.22 Although ELETROBRAS has considered thermal options, as in the Plano 2010, we urge their further and continued consideration to provide a transition program, say for the period 1994-2005. The realistic options ares (L) imported coall (ii) imported and domestic natural gas; and (iii) bagasse cogeneration, from the sugar industry. Provided that adequate emissions control equipment is included in the case of coal, thermal solutions not only offer environmentally acceptable alternatives at reasonable economic costj they would provide the system with more flexibility to respond to demand uncertainty, possible new technologies, the lessons to be learned from additional experience with the Amazon, and variable hydrological conditions. 3.23 While domestic coal appears uneconomic and of poor quality (see Annex 1. 2), coal-fired steam generation, using coal imports from Colombia or Australia, for example, presents an interesting option. If equipment is contracted through intexnational competitive bidding, in relatively large units (say 300 MN-700 MW), the capital costs could be kept down to around US$1,000 per kW and total costs to around US$35-40/Wh. The provision of adequate environmental safeguards would raise thee capital costs, perhaps by 10-20%. After gaining experience, Brazil could consider producing its own coal-fired generating units. 3.24 Construction of thermal plants using natural gas, either domestic or imported, for example from Bolivia and Argentina, are another interesting solution. Brazil' s own gas reserves are limited relative to overall potential demand, and some further analysis may be required of the economic value of the gas in alternative uses, along the lines discussed in Chapter IV. The promise of substantial reserves in the Amazon, while subject to some of the location issues we raised with hydroelectric solutions, would be low in capital cost, quick to construct, environmentally less harmful, and face few, it any, competing demands for the gas. Finally, cogeneration, using bagasse from sugar production processes, could provide a worthwhile contribution to total generation requirements, if the institutional and pricing framework provide the correct incentives. Of course, if the production of alcohol declines in the long term (paras. 4.55-4.59), the availability of bagasse for power generation would drop. - 35 - 3.25 In the longer term, hydroelectric options are available to Brazil through international cooperation, especially with Argentina, Bolivia and Paraguay. These are attractive, even allowing for the risks associated with greater external energy dependence. We do not believe that electricity imports are subject to the same political and economic uncertainties on supply and price as petroleum imports. Mutual economic interdependence and more common geopolitical interests make electricity interchanges relatively reliable, as experience has shown elsewhere. 3.26 Finally, we emphasize again the role of energy conservation and demand management as an alternative to supply options, in both the short and long run. Based on our results with regard to price elasticity and demand forecasting elsewhere in this report, we argue in Chapter V that important savings in investment requirements are obtainable by pursuing increases in the electricity price level as an integral part of well-formulated energy conservation program. Energy conservation not only offers a low-cost and environmentally sound alternative to investment in generating plant: it also makes more manageable the risks related to demand growth. 111.3.4 The Investment Proaram 3.27 The sector's investment program is shown in Annex 111.7. It is based broadly on the precepts of Plano 2010, and ELETROBRAS' demand forecasts shown in Table 111.4. However, the sector has had to recognize its limited financial capability to execute its expansion plan, relative to the expected sources of funding. It is also attempting to restrain the growth in demand of electricity- intensive industrial consumers, under subsidized tariffs. 3.28 During the late 1970s and early 1980s, the power sector allocated a large portion of its financial resources to generation, which accounted for as much as 654 of total investment. Consequently, the financing to implement associated transmission and distribution works was curtailed beyond prudent levels, and the risk of energy shortages has climbed above the level of 5% used for planning. While the sector is now attempting to implement a more balanced program, in terms of the mix between generation, transmission and distribution, the risk of energy shortages will continue to rise in the coming years, especially in the South and South-East. Detailed risk analyses for the different regions are in Annex III.8. It is estimated that past constraints on investment, in conjunction with hydrological conditions, have given rise to a 15% to 20% probability of energy shortages equivalent to 10% of projected consumption by 1992 and 1993. 3.29 The estimated cost of the power sector's five-year investment program is about US$39 billion, excluding the investments to complete Itaipu, i.e. a little under US$8 billion per year. Some 53% would go to generation; 23% to transmission: 16% to distribution; and the rest to general investments. The sector estimates that it will continue to invest US$6 billion per year thereafter, to keep pace with demand. The estimates are optimistic, as the sector's ability to generate funds is severely impeded by the low level of tariffs. However, the program could be reduced by effective energy conservation measures, mainly through pricing, and we have made a tentative assessment of the possible impact of LRKC pricing on the sector's investment program later in this Chapter. * 36- III.4 LONG-RUN MARGINAL COST 3.30 Since the 1970s, LasC has provided an important reference point for electricity tariffs in Brazil. The average LRMC for generation and interconnection for Brazil as a whole is therefore updated regularly by DNAZB, using the methodology briefly outlined in paras. 3.38-3.43 and Annex 111.9. 3.31 Marginal costs vary not only through time, with changes in the load profile and system expansion, but also cross-sectionally, according to the type of user and location. For example, rural and residential supplies are usually the most costly; and the cost of low voltage (LV) supply is generally higher than high voltage (NV) . Regional variations in LRMC are caused by many factors s local generation costs may exceed the system average, for example due to a local preponderance of thermal plants; unusually long distances from the generation sources may raise transmission costsl higher sub-transmission and distribution costs are caused by low population density; and there will be regional differences in the quality of service and reliability of supply. Table III.S illustrates these points for the interconnected systems. For Brazil as a whole, the average LRMC of rural supply, at US$124/KWh, exceeded that of all other users. The cost of LV supply (Category 8) on average for Brazil as a whole was about US$99/Kwh, while the LRMC of an average NV consumer (Category A) was US$54/MWh, i.e not much more than half. At the regional level, the lowest LRKC (US$59/XWh) is for the North and North-East systems, which have the greatest concentration of future hydroelectric resources to be developed (it includes the Xingo project). The highest LRUC corresponds to the systems with the thermal program (coal and nuclear), which are also furthest from the principal sources of generations the South (US$82/KWh) and the South-Bast/Center-West systems (US$76/XWh). Of course, the structure of regional supply costs is different if account were taken of the isolated systems, which are high-cost and weigh more heavily in the North and North-East. Lono-Run Marainal Costs by Region 11989 USS ner MWhI AZ 138 kv to 88 kV 48.40 - 52.08. 5L39 A3 69 kv to 20 kV 50.90 39.15 54.8t 4.33 A4 13.8 kv to 2.8 kV 63.99 62.48 6.60 68.52 Average A - Hr & liedium Voltage 54.32 44.91 57.35. 64.03 oi- At Residentiat 8.95 84.47 110.11 107,40 82 Rural 123.67 119.98 137.34 136.77 S3 Other User 96.30 95.51 100.65 99.75 94 Public Lightin 96.40 84.48 95.64 95.9 Average. 9 Lo VoLtage 99.45 89.10 107.60 105.48 Average A and 6 ALL Vottage Levels 71.25 59.02 75.53 8t.25 Sources D ,cuitos marginss d -forneimento a preos Modla peticadoes, October 1989. end Soletto'de tarifat do eergfe 'oletrica Septuber 21, 1989. Ifotesr aR m ortb*fast; SE South-East; Of w Cetertt. Date refer to Intercomnwnted ytqa only. The colmn haded vSraxfi Is the average L.M for the caitry as a %hole. * 3? - II1.5 EX&CTRICXTY TARIFFS 111.5.1 Taiff . Inflation and Taxes 3.32 High inflation, combined with frequent changes in the official tariff schedules and fluctuations in the real purchasing power of the Brazilian currency, dictate a cautious approach to the analysis of the actual level of tariffs compared with LRC.3 Furthermore, the electricity bill includes not only the basic tariff for electricity: taxes and tax-like elements must be included (the "fiscal load"), which make the final price to the consumer exceed the tariff. 3.33 With high inflation, it is necessary to distinguish between the Published Power Tariff (PPT), the Billed Power Tariff (BPT) and the Received Power Tariff (RPT). Under conditions of rapid inflation, the PPT should in theory be raised continuously; and if the consumer is billed and pays continuously, there would be equality between PPT, BPT and RPT. In practice, of course, the situation is different. According to the Brazilian system, when electricity bills are sent out, they are prorated according to estimates of the consumption that has taken place under the various schedules published since the last meter reading. On the other hand, because of the length of the billing cycle --i.e. the time that it takes to reflect fully a tariff increase in all bills -- the system results in average values of BPT which are always less than the average of PPT. In Brazil, where almost 60 days pass before the whole increase is reflected in all bills, and where inflation has remained high for most of the years under analysis, the billing system results in a significant additional subsidy to all electricity consumers. During 1983-1989, for example, BPT averaged 90% of PPT. From January to September 1989, BPT fell to 74% of PPT. Furthermore, RPT is less than BPT, to an extent which depends on the payment lag and the daily rate of inflation. 3.34 Brazilian electricity bills include: (i) the basic tariff, or electricity charge proper, i.e. the metered consumption times the corresponding rate; (ii) a consumption tax on residential, commercial and small industrial consumers;4 and (iii) for large industrial consumers only, a "ccapulsory loan" to the power sector. Until January, 1989, the electricity consumption tax was Federal -- the "Imposto Unico de Rnergia Zletrica", ZUEB -- and applied uniformly to the entire country. About 60% of the IUEE was returned to the sector. As from February, 1989, a State tax has been in effect instead -- the "Imposto de Circulacao de Mercadoriae e Servicos", ICMS -- which varies from State to State. The compulsory loan is obligatory for industrial consumers using more than 2,000 kWh per month. 3.35 The addition of these taxes and compulsory loans to the published and billed tariffs (PPT and BPT) gives rise to what may be termed the Published Power Price (PPP) and the Billed Power Price (BPP). While the average aPP has followed a similar trend to the BPT, the "fiscal load" (i.e. the tax and tax-like elements) has tended to fall, from 28% of the tariff in 1974 to 15% in 1988. Furthermore, because part of the revenues collected through taxation were transferred to the power sector, this policy reinforced the negative effects of * see pare. 1.6. 4 Industries using tess than 2,000 kUh per month. - 38 - tariff reductions, thus aggravating the financial situation of the sector (see Section 111.6). It is estimated that, keeping the compulsory loan component constant at its 1988 level of about 5%, the authority transferred by the new Constitution to the States to set electricity taxes will increase the share of the "fiscal load" in the final price, perhaps to as much as 30%. However, it is not yet clear what part of this increase, if any, will be transferred to the power sector. 3.36 The new fiscal load structures are relatiLvely straightforward, with most States having a unique rate -- generally equivalent to 17%-25% of the bill, subject to two or three exceptions for small residential and rural consumers. Only a few States have opted for more complex fiscal structures that include, for example, different taxes for electricity used by an industry as an input and electricity used by the same industry in other uses or a whole set of tax brackets that varies with the amount of electricity consumed. 1I1.5.2 Existing Tariff PolicZ 3.37 In theory, Brasilian policy with regard to electricity tariffs is exercised through DNB, to achieve four general objectives.5 These are: (i) aconcpic efficiency; (ii) financial stability; (iii) regional uniformity; and (iv) social considerations. Since these criteria often conflict, trade-off a have developed between them over time. Unfortunately, in recent years, the policy of keeping down energy prices in the mistaken belief that inflation would thereby be curbed, implemented through SAP, has tended to dominate all these objectives in practice. A. Boonomic Efficiency 3.38 In 1977, DRNE (in association with BLWTMtOBAS) initiated a comprehensive study of electricity tariffs, incorporating marginal cost principles. The work was assisted by Electricite de France (NDF), as consultants, and supported by the Bank, under Loan 1300-BR. The first step to initiate tariffs based on LR4C was taken in October, 1981; and to date most consumers at high- and medium-voltage levels -- accounting for over half the consumption -- are subject to LMC-based tariffs. Full implementation is expected during 1991. One result of applying the policy is that relatively larger increases have been sanctioned for the energy as opposed to the demand component of tariffs -- the power system has become increasingly energy constrained -- which has produced higher average tariffs for consumers with larger load factors. 3.39 Marginal cost-based tariffs are forward looking, i.e. they reflect the anticipated cost of generating and delivering the next (or future) kilowatt hours of electricity as opposed to the embedded financial cost.6 In Brasil, See, for instance, the 1934 Vater Code, with its emembents; Decree No. 86.463 of October 13, 1981; R. Bitu, "Tarifa residencial de caractor social", Banco Interamericano de Desarrotto (BID) Curse en Analists do Costos y Diseo de Tarifaes d Ague y Electricided, Pueblo (Neico) Abrit 21 a ayo 9, 196; R. Bitu, SR. *. DAE, oe Tarife do energia etetrica: Netodologia e Apticacao, Brasitia, 1965 and aConceftos basicos do nove tarfs d energia eLetrica, Junho 1988; and REVISE, "Relatorio diagnhstico: Problems fundamentals, questes emrgentes*, Rio do Janefro, October 1988, "Relatorio finat doe subgrupos A, Ee F", Septefter 1988, and aRelatorto final -Ninutan, January 1989. 6 See DAEM =-i 1985 for a presentation of the methodology and its application to the Brazilian power sector. - 39 * LRMC is derived from the sector's investment expansion plan but a significant deviation is then made from the true marginal cost structure by suppressing all regional cost variations and producing a unique set of tariffs for the entire country. Since the criteria for estimating peak and off-peak energy and capacity coats were established at the end of the 1970s and the differential bdtween power and energy costs has gradually narrowed in Brazil, to the point that energy shortages are more the norm than the exception, it is time to review the criteria for these estimates, to avoid sending the wrong pricing signals. 3.40 The resulting tariff structure distinguishes between bulk supply tariffs and tariffs to final consumers, dividing the latter into three classes, according to the voltage level at which they are connected to the system. The three related sets of tariffs are color-coded: Blue (HV), Green (MV) and Yellow (LV). The details are in Annex 111.9.. 3.41 The relationship between LRMC and the average BPP (para. 3.35), as of September 19, 1989, is shown in Table 111.6 for each consumer category. The tariff adjustment mechanism, combined with the high rate of inflation, has led to an average BPP which is only 65% of both the nominal price, including taxes (PPP), and the LRMC. 3.42 HV and MV consumers (tariff group A), who account for 63% of all sales, pay an average PPP above LRMC. The A4 category in particular, connected in the range 2.3 to 25 kV, accounts for one-quarter of sales and pays an average PPP 42% above LRMC. When adjustments are made for the billing effect, the effective average BPP falls below LRMC: to 80% and 93% of L%MC for the A tariff as a whole and for &A respectively.? 3.43 The "non-residential" LV consumer, mostly commercial businesses and very small industries (category 23), is charged an average PPP 41% over LRMC. Even the average BPP is close to the level of LRMC. Otherwise, all LV consumers are charged an average PPP and BPP well below LRMC, with rural consumers and public lighting showing the largest deviations. Residential consumers as a group are subsidized heavily in economic terms: the average BPP is 42% of LRMC, while the average PPP is 65. B. Financial Xaulibrium 3.44 Financial equilibrium is to be achieved by setting the average tariff according to the principle of "service at cost", i.e. the sum of: (i) operation, maintenance and administration expenses, including purchases of electricity; (ii) straight-line depreciation of average gross revalued fExed assets in operation; (iii) amortization of interest during construction; (iv) the cost of decommissioning assets; (v) taxes other than for income (mainly property taxes); (vi) reversion quotas (an equivalent tax of up to 4% of assets in operation)O; (vii) transfers to the National Compensation Reserve for Remuneration (RENCOR), 'For this discussion recatt that 8PPsO.65 PPP, i.e. Pppal.54 BPP. SReversion quotas b re introduced to create a reserve fuand which would attow the Federal Govermment to buy the electric power assets of foreign companies. States and Mmicipatitfes at the end of the concession perfods. The fund is adtinistered by ELETPORAS and the resources have been spent on empansion of the power sector. *40* a tariff equalization facility; and (viii) an adequate return on capital invested, currently set at 10-12% of remunerable investment9. 3.45 The financial history of the power sector over the past decade is discussed in Section 111.6. As in the case of the economic efficiency objective, price control was exercised by the Federal Government through SAP, to pursue anti-inflationary objectives rather than through DNABE, in light of the needs of the sector and the situation of individual utilities. Far from achieving equilibrium, the financial situation of the sector was seriously undermined. C. ReNional Uniformity 3.46 Until 1974, power tariffs varied regionally in Brazil, according to the cost of supply. Then the Government established a system of tariffs which was uniform nationally, both in level and structure, regardless of supply costs. The objective was to promote a more balanced regional development, so that consumers would not be penelized by their location. In addition, the Federal Government could stimulate the provision of service to remote rural areas, provide supply at the same price to the same class of consumer regardless of their location, and encourage the localization of industry in underdeveloped areas. To ensure that all utilities would earn the legal minimum rate of return on remunerable assets, the Government created a mechanism -- the Global Guarantee Fund (GGF) (later RENCOR) -- to transfer resqurces from the financially stronger to the financially weaker companies (see Section 111.6 below). 3.47 While the system of uniform tariffs may have provided regional and political benefits in the past, the situation has now changed, as the more economic sources of future generation are in the North/North-East (paras. 1.9 and 3.21). The analysis of Table 111.7, which is derived from forward-looking LRMC estimates, suggests that -- as far as the interconnected systems are concerned - - the better-off regions of the South-East/Center-West and South will receive the largest transfer of resources, expected to total US$ 4.9 billions in 1989. The poorer Nor*h and North-East are expected to receive less than US$0.5 billion in the same year. It must be emphasised that the estimates in Table 111.7 refer only to the economic cost flows related to the interconnected systems and not to the financial flows under the GOP or RENCOR, which are quite different (see Section III.6). On the other hand, there will be economic efficiency costs, through (i) higher or lower consumption levels in individual regions than those warranted by supply costs; (ii) wrong consumer choices of fuel and location, based on artificially high or low relative prices of electricity viz-a-viz alternative fuels, which may cause serious long-run distortions in the pattern of development, away from activities which would exploit regional comparative advantage; and (iii) lower management efficiency of the utilities, as the uniform national tariff and the associated mechanism for the transfer of funds deprive the individual power utilities of the control over and responsibility for certain crucial aspects of their operations. Reamrable fnvestmnt is equal to net revalued assets plus an attownce for inventory and not working oapital aims the dfference betwon the prior years% actual and attowable return minus contributions and grants. g4 g e 4 ug eesma . a $ 9 s a nwna esa sluida Us I as ige e I ae emgxx ;; as a~ ,-r 8 s18 : o sus * sm. : 14 44 ----2 44 3 -42 - Ineconnegted Syntem arina1 ras_and an mic Transfera 12989 205-34r IMh) R*ains b/yrg % aP , osts of NC 1llSm 02 10.0% A.49 ?,25 ,5.% L,m04 o-A W.9% 4.69 59.9: .9.. 9 R 31 15$ 4.69, 59.2- 191 É82 sm& 2-' 1.7% 46.69 82.25. 54.8. .99$ MoCA •p 9 lbed power price. • ntercoected syst onty. D. locial Considerations 3.48 Brazil's tariff structure for residential consumers containg a minimm consumption lovel (NCL) of 30 kWh per household per month (kWh/m) along with a series of increasing block ratos. The KML is a form of Olifeline" rate, being a fixed rate with a right to consume a minämm level of electricity fro. of charge. At the same time, the utility is reimbured by the consumer at least partially for itU customer-related fixed costs. The average price paid by households increases as a percentage of LC from less than 14% at the XCL of 30 kWh/m to, 76% at the highest consumption block of more than 300 kWh/m (Table 111.8). In addition to these subuidies? there are explicit subeidies for rural consumers. 3.49 The above structure of residential tariffg is based on a social policy which has the following elemento: (i) every family reuires a certain minimum supply of electricityl (ii) although poor families spend little money on electricity in absolute terms, it figures prominently in their overall family budgetsj (iii) family income is positively correlated with household eleoctricity consumptions and (iv) rural consumers are poorer than urban consumers. 3.50 In September 1989, the minimum residential bill resulting from the ML of 30 kWh/month was equivalent to less than ik of one minimum galary, a percentage which has fallen steadily for the last 15 years (as has the share of the 100 kWh bill) (see Pig. 11.2). However, there is no explicit guideline relating the desirable minimum bill to the minimm salary or to any other identifled parameter. Furthermore, there is no explanatSon regarding the derivation of the NMh. Studies by DNME10 suggest that the minimm electricity consumption of a typical Brasilian family is between 93 and 117 kWh/month, a rang. which results from combining the needs for lighting (14 kWh), refrigeration (32-42 kNh), ironing (4 kWh), radio and television (18 kWh), and fans, water eeR. gftu, 1986, ggi - 43 heaters and stoves (25-39 kwh).it The data suggest that the else of the MCL may be too small to eatiefy basic needs, which would be about 50 kWh/month. 3.51 It is widely accepted in many countries that there are sound social policy arguments for a subsidized "lifeline" rate. Rowever, Brasil's indiscriminate subsidies to all residential consumers appear excessive, especially as most go to the better-off. Table 111.8 estimates the subsidies enjoyed by the various levels of residential consumption, in total and per household for each level, on the basis of the difference between the LItC and the average BPP. It shows that the consumers who benefit most are those in the higher consumption categories. 3.52 Taking into account this data, it is recommended that only two residential blocks be employeds the first (say, up to 50 kWh/m) would be subsidisedl and the second (beyond 50 kWh/m) would be charged a rate at least equal to LtMC. Subsidization of a monthly consumption level of 50 kWh/m provides relief to the budget of poor families without undermining significantly the finances of the power sectors even residential consumption up to 100 kWh/m is only 3.5% of all electricity consumption and low-income consumers, notably in the Ofavelas", have a good record in paying bills. Of course, the amount of subsidization to be allowed must be carefully chosen. We believe that properly targeted "lifeline" rates are an appropriate way to deal with the needs of the poor and preferable to alternatives, such as not installing meters in low-income households. ZABL IxII.8 Residential LRMC, Average BPP and Subsidies (USS of Sea 89) Consumption Pargina t- Subsidfea -- aber Users -- Cost Tariff Price aUS sant. us Tariff stock Thousands Z amlyr % USS/MA USS/Nm UMh. Tofli. pir hld up to 30. k 4 044. O6.0 848 1.9X 105.92 15.22 15.22 77 19 S1.. to 100 kti 7 30.3% 6,168 13.92 104.65 17.08 21'90. $1.. 47 . 194 to MD kW 685 $4.35 15,032 33.9% 102.20 2.07 27.96 1,116 18 - 201 to 300 kIM 2,951 11.75 9101 20.5% 96.75 32.47 39.12 54 18 . OLer 500.1ih 1,971 7.85 13:141 29.M% 92.25. +8.1* P.05 292.- 148 Total 25,319 100.0% 44,289 100.0% 98.95 34.08 41.45 2,58 100 Sources. . . 04816, goli do tarifes do Pergi etetrics, Septedwr 21, 198k. ad fLWlS estimates of consu"ton. Mtes$ Septeuber 19, 1989 exchange rate: 1SSI.00 a MeC S 3.316. Averag pricea fmd revenues resulting from teriffe, copulsery leens mWd takes, after adjustments for the bitting cycle. " these are average figures and are strongly Influenced by the larger consumption of households in the economicully more prosperous South and South-East Regions. If additional studies shou that there are significant regional variations in the requirements of typical families, the MCL could be made to vary regionally. -44. .....ectr city Bi t 1s and M rnimu,m Wage u. ........ I ..... 1974 1976 1979 1990 1982 1994 1986 198g 4401 1999 Sep 1989 Years -"' 30 kWh b I 1I/Mirn Wg. 100 kWh bi I /Min Wg. Figure III.2t Relationship Between Electricity Bills and Minimum Wage 111.6 FINANCIAL AND FISCAL ASPECTS 3,53 Prior to 1978, the electricity sector enjoyed a long period of robust financial performance, because the Government provided adequate equity contributions and implemented a tariff policy which was both realistic and consistent with the legal rate of return on remunerable assets (10%-12%). After 1978, the Government, faced with difficult policy choices and complex macroeconomic circumstances, including high inflation, undermined the financial health of the sector in the interests of fighting inflation. Borrowing --some contracted on short-repayment terms, inappropriate for a power utility implementing projects with long gestation periods and economic lives-- and then equity contributions displaced internal cash generation as the primary source of investment financing. 3.54 In the early 1980s, the Government agreed with the Bank on programs to increase tariffs, but did not fully implement these programs; therefore, the worsening financial trend of the sector was only temporarily arrested. The real levels of the average tariff and hence of gross internal cash generation, eroded by high inflation, failed to keep pace with the financial needs of the sector. Power tariff legislation provided for the addition of revenue shortfalls (below the level to achieve the legal minimum rate of return) to the rate base and their - 45 - subsequent recovery through ruture tarift increases,12 but such increases failed to materialize. The decline in the average revenue (see Fig. 111.3) occurred while the system LRMC was rising rapidly (see para. 3.30). Aerage Cnstnt et..ty Pr ce. .. u ' ' i ' I ' i9 ' i ' i ' i ..~................ 19. 1M 192 1..4 1978 TM esen 1s. 1se. 1I. . 199 Years rigure I 11I.3 3.55 The financial deterloration of the sector moved swiftly (seAnnex III.10). sy 1987, the Governmänt, notwithetanding so oub.tantial tariff Increase, had becom the major source of gector financing, with Government equity contributions of US$5,975 million, compared with long-term borrowing oz only US$1,826 million. The ratio of coneumer-baued financing was -64.4% and the debt-service coverage 0.44. These symtome of financial deterloration are euwmarized In Table III.9. 3.56 While tariff levels are an Important indicator for assegsing the sector'Is f inancial performance, each Individual utillty has Ite own requirements , which sometimes canno boe atiefied by a common tariff level. When uniform tariffs were Introduced In 1974, creation of the GGF was neceseary (para. 3.46) to mak tranofers from the f Inancially stronger to the weaker companies, so that all the individual comepanies would earn the miýnimum legal rate of return on remunerable agsets (10%). In 1981, because of deteriorating tariffs, revenue shortfalls and rates of return which fell below the legal minimum, the Government changöd the role of the GGFs flinan cial recources were to be trangferred go that all comnLes would earn the actual average rate of return, which was less than the legal minimum. In 1987, after years of inadequate tariffg, the power companieo of the South-et region, under inacructiona from their owners, refused to continue contributing to the GGF, as a gesture of protest. The Government responded to this criois by creating a new transfer mechaniem In 1988, the 2 Am comsred with the tevet of reveue wbich woul have resutted thr~ug the apptication of pomer tariff tegistation, there was en acmunueted revenu deficit et the and of 1987 azounting to ~i$ bittion. *-46-* National Compensation Reserve for Remuneration (RONCOR). Under the new arrangements, no transfer is required from any company with a rate of return below 124. In addition, it was agreed that the rates of remuneration of the four major BLETROBRAS subsidiaries would be less than the legal minimum in establishing their bulk supply tariffs. TABL _III.9 Power Sector Indicators Of Financial Performance. 1977 -1987 1977 1978 1979 198 19E,M 19R 198 4 196 196 . 96 1967 Oebt-srvice coverage (x) 2.09 1.67 1.49 0.92 1.11 0.864 0.75 0.37 9.34 0.4' 0.44 Self-fOnsrping ratio (2). 23.5 17.7 2.5 -4.5 5.9 -9.6 -22.3 *7.4 -111.5 -101,1 *84.0 Caw~uer-based financing (E) 52.9 46.8 48.8 23.8 39.8 17.4 11.2 -36.5 -80.2 *7.9 -64.4 SorrowingrwInvestinnt (s) 51.0 55.7 73.6 86.8 ft.4 65.0 25.9 MS0.7 69.9 56.5 31.1 BoteeW Diservice coverag P debt service divided by grass f1telne cas gneration. Self-finanrcig ratio a net Iwternal casth gerraon divided by -totSt frwestame. Consumer-baed finncing a (net fnternal cash gmnrette plus cOnmaer-based contributions) divided by. tota investMent. SMRCE: ANE III.8 3.57 Annex 111.11 shows the inter-regional flows of funds since 1975 under the tariff equalization mechanisms. The South-East consistently was a not contributor to the scheme, while the North-East, Center-West and especially the North have consistently been net receivers. Although the South on balance received funds until 1985, the amounts became relatively minor after 1980 and the South has been a net contributor to the scheme since 1985. The financial flows run essentially in the opposite direction from the economic subsidies measured in Table 111.7. There are several explanations for this phenomenon. The economic subsidies refer only to the interconnected systems and therefore exclude the effect of the isolated systems, which impose a financial burden particularly on the North and North-East. At the same time, due to the uniform bulk supply tariff, those regions do not benefit comsensurately from their lower future generation costs. Furthermore, the financial flows are affected by different histories of borrowing and the debt service burden which has arisen. Finally, the companies of the North and North-East have a greater regional development focus than the South and South-East, which are more profit oriented. 11I.7 CONCLUSIONS AND RECOMMENDATIONS FOR CHANGE 3.58 Chapters I and II explored the linkages between the Governments macroeconomic polit fe and problems and its efforts, exercised through SEAP, to hold down energy prices, as an anti-inflationary measure. We have demonstrated in this Chapter the adverse financial and economic consequences of that policy, at the level of the power sector. The sector's ability to generate resources for investment was totally destroyed. At the same time, the low tariffs relative to the level of marginal cost hampered the sector in meeting economic efficiency objectives. These factors were certainly more important, in retrospect, than the * 47.* successes arising from implementing the LMC-based tariff structure, in terms of improving the load profile and moderating the increase in investment (see Chapter V). As a result, the Brazilian public sector (Federal, State and local Governments) has had to contribute to the power sector annual amounts which increased from US$400 million on average in 1977-1982 to US$2 billion in 1983- 1985, US$5 billion in 1986 and US$6 billion in 1987. We believe that the larger public sector deficits which resulted were more inflationary, due to the way in which the Government has financed its deficits, than the price increases would have been. Neither were there any convincing social arguments for the way in which residential tariff increases were moderated, as the subsidies were not targeted and instead operated to the benefit of all residential consumers, regardless of income group. Indeed, the income distribution effects may have been perverses the biggest subsidies went to the largest consumers, who are likely to be the better-off the very poorest receive no electricity supply at all. 3.59 A determined move to LRMC pricing, for both the level and structure of tariffs, would be consistent with the macroeconomic objective of deficit reduction, with the sector's financial objectives and with proper demand management principles. There are signs that the Government accepted this position, as tariff recuperation became a main concern for a while after 1987. The average level of the basic tariff increased during 1988 but the level of net cash generation was adversely affected by higher operating costs and debt service. It has been agreed between the Government and the Bank that the average tariff billed in 1988 of US$S4/MWh (expressed in prices of July 1989), represents a minimum acceptable financial level for the near future. Such a target (US$55/MWh in September, 1989 prices) would be consistent with LRKC pricing, provided that the fiscal load is at least 30%, to bring the total to the LRUC of US$71/KWh in September, 1989 prices, and also provided that inflation subsides, so that RPT approximates BPT. Evidently, tariff adjustments must occur with sufficient frequency to safeguard the sector's financial health. 3.60 Full LRMC pricing could cut electricity consumption by 10-15% by the year 2000 (para. 3.14). A "low" estimate of the drop in system capacity requirements in the year 2000, corresponding to a 10% fall in consumption and maintenance of the system load factor at around 70%, is 6,500 NWi a "high" estimate, implying a 15% lower consumption with an improvement in system load factor to 75%, as the Green and Yellow tariffs are implemented, is 13,600 MW. Valuing the cost per kW saved at US$ 1500 to US$ 2000 yields monetary savings in the order of US$10-27 billion over the 11-year period or between US$0.9 billion and US$2.5 billion per year. This should be compared with the current investment program of US$6-8 billion per year for 1990-2000. 3.61 As a final conclusion on tariffs, we recommend moving away from regional uniformity. In the long run, there can be a serious misallocation of resources regional incentives for local private sector participation are reduced; and the economic subsidies will go from the poorer to the richer regions. The financial subsidies have gone in the other direction, but with serious adverse consequences, as we have described. 3.62 The sector's planning methodologies are generally reasonable. We have noted in particular the benefits of the decision to open the discussion on planning to a wider audience; and the strides taken forward in addressing environmental issues. However, the planning methodology could be strengthened in several directions. Demand forecasts should allow for alternative scenarios 48- for different macroeconomic assumptions, and different pricing and taxation policies. Also, large individual loads and energy conservation are not being treated properly; while more consistency with PETROBRAS is called for in basic assumptions, which should be ensured by MWN (para. 2.37). Finally, risk and reliability should be incorporated more effectively at each level of plannings generation, transmission and distribution. 3.63 For its expansion program, apart from developing domestic hydroelectric power for generation, the power sector should explore further the import options, both thermal and hydro. of the domestic options, natural gas and bagasse are both interesting. The latter, along with other private sector initiatives, would benefit significantly from the incentives which would accompany the action we advocate on pricing. 3.64 The investment program needs a better balance between generation, transmission and distribution than was achieved in the past, and changes are occuring in that direction. The movement towards LRMC pricing, in level and structure, would reduce demand, cut investment and mobilize additional resources, thereby also improving the sector' s financial and the Government u a fiscal situation. * 49 - CHAPER IV PETROLEUM PRODUCTS. ALCOHOL AND NATURAL GAS IV. THE LOUID FUES AND GAS NAET 4.1 The growth in the consumption of liquid fuels Cohsumpt i on Fue I s since 1970 has been uneven, 1970 1988 as shown in Fig. IV.1. Following steady expansion Thousam TOE during the 1970s, at about 8.6% p.a., overall s consumption declined sharply in the early 1980s, before resuming an upward trend 3 after 1983. As a result, the average growth rate in the consumption of liquid fuels in the 1980s was only 1.3% p.* a.* This uneven performance 1970 1972 1974 1079 197m 199 0 1902 14 1920 ISM largely mirrors the development of the economy as Me gom R4ol Orl EMAlooftl a whole -- the correlation F ume m in - W&es coefficient between the FISpre IM. growth in demand f or petroleum products and of GDP is 0.9 --and the policy of replacing imported fuels with domestic energy forms. 4.2 As can be seen from Fig. MV2, there has Consumption Fuels also been a significant190 98 change in the shares of- 198 different fuels since 1970. Tndow ....0..... Alcohol hags grown rapidly, ..11- from negligible amounts as a late as 1976, to surpass the share of gasoline; and diesel oo has consistently increased at uI a rate faster than GD, despite a significant increase in real prices. of and kerosene also expanded their shares, from around 10 1 m 1M 1M IM 197 09 1= lam 1M 1m at the beginning of the 1970. - 0m Fiat 011 Almol W to about 1S% by 1988. Fuel I M emlm Wm vM oil demand, after reaching a IV.2 peak in 1979, declined through 1985, after which it recovered slightly in absolute terms, but stabilized it. share in total liquid fuels demand at about I5%. The situation from 1975 to 1988 is summarized in Table I. A discussion of fuel consumption by sector is in Chapter I. - 50 - 4.3 The production of natural gas expanded steadily after 1970, from almost zero to approximately 16 million 3 in 1987 and 1988. From 1979 onwards, production grew at 13% p.a., although, until recently, much of this was either used by PETROBRAS or flared. As shown in Table IV.2, in 1987, PETROBRA8 reinjected 18.5% of the gas producedl used 24.7% in its production operations and 23.6% for refinery fuell and 15.3% was vented or flared. The relatively lov proportion of gas lossas ref lects PETROBRAS' efforts to reduce flaring, which in 1983 amounted to almost one-third of the gas produced. The principal markets for natural gas are industry, fertilizers, petrochemicals, steel-making and domestic fuel for cooking and water heating, taking a combined total of 18% of production. TAuM IV.1 '4P2$ 12 . .......... ....... ..... .... .. . . . .. .... P-0249 123 .3 . .,iset 2k2 42 i Si i igF,e n,t2 192 P-m E4 .11 fr<= 1990 ¢ t ¢ 199 Roan A 7% tarcafer; (mi11ee ab slo ap98 iaio In th rea InternaAcho ensmtionaprc of oill0 a/d cii h a orrend o Indutr andtrnsor In IDV.2.1 ree The o ealnd Forecastar ta I . gvrmn policias vill be to: (i) continue with the substitution of gasolina by alcohol and fucl oll by coal and natural gagl and (1i) waintaln the prevailing relativa prices between fuels. Under A.2, on the other hand, it ta ansumed thats (1) fiscal Incentiveu for alcohol vehiclea will bo ellminated in 1990 and a maximum production level establinhed for alcohol at the existing level of 16 million cubic meters per year (see Chapter I); (1i) relativa fuel priceo will be corrected to reflect their opportunity cootsj and (III) the use of natural gas will be fostered in the transport sector. * 51 * NlaturaL as ConuIon.r 198 Use IlitLon Oid K of Total Field use 3.90 24.7 Reinjection 2.92 18. Industrial fuel 3.74 23.6 Steel production 0.52 3.3 Petroenemicals 0.43 2.7 Fertilizers 1.41 8.9 Domestic consuption 0.47 3.0 Total consuoption U " Lases Ja, 151 Total production 15.6 100.0 4.s PETROBRAS' methodology suffers from some important limitations. First, PETROBRAS employs different GDP projections from ELETROBRAS (see Section 111.2). Second, PETROBRAS' forecasts are essentially trend projections, tempered by judgement. Some of the models consider variables other than time (such as population distribution, number of households, and size of the vehicle fleet) but none has a consistent analytical basis which would allow sensitivity analysis to be undertaken and the direct consideration of the alternative scenarios; and none consider price as an explanatory variable. As in the case of ELZTROBRAS, PETOBRAS' approach handicaps the policy maker in dealing with the high degree of uncertainty that confronts strategic planning. IV.2.2 Alternative Demand Scenarios 4.6 In order to overcome some of these methodological limitations, the mission developed alternative forecasts on the basis of simplified analytical models, derived from regression analysis (see Annex 1.9). As with the power sector (see Section III.2), we developed alternative forecasts on the basis of two pricing scenarios. Under the basic scenario, "Business As Usual" (BAU), the domestic price level for petroleum products is kept in line with the international price level and the existing structure of domestic prices for petroleum products, alcohol and natural gas is maintained unchanged. In practice, this means that the domestic price of each individual fuel will change parl pasu with the international price. The "Efficient Pricing Scenario" (BPS), assumes, in contrast that the relative domestic prices of petroleum products are progressively moved into line with the international price structure by 1995. In terms of the March 1989 retail prices and delivered economic costs (Table IV.6 and para. 4.30), this means an increase of 73% in the price of LPG from 1989 to 1995, 22% for kerosene, and 40% for fuel oil; while gasoline and diesel would decrease by 48% and 8% respectively. For alcohol, which competes with gasoline in the Brazilian market, the price is assumed to attain equivalence with that of gasoline by 1995 (on the basis of thermal content), implying a decrease of 44- 54%.1 For natural gas -- which is in short supply and principally replaces fuel oil at the margin -- the price under BPS is equal to its economic opportunity 1 Depending whether fiport or export parity is used. Equivalence on a theral basis (i.e. equal prices per toe) mens that the price of hydrous alcohot will be 81% of the price of gasoline (see para. 4.30). * 52 * value, v1s. the cost of fuel oil in energy equivalent terms, implying a 52% price increase from 1989 to 1995. 4.7 Both scenarios incorporate the same World Bank forecasts regarding GDP growth and future movements in the international price of oil. The two scertarios also incorporate supply restrictions for alcohol and natural gas. In BAU, the maximum supply of alcohol is equal to the industry's present approved capacity of 16 million m3 per year; and the supply of natural gas is constrained at 17 million m3/day, the level of available supply which appears attainable by the year 2000 (see Annex X1.1). In RPS, the same upper limit is placed on natural gas production but two limits are set for the alcohol supply. In the first, as in BAU, alcohol supply cannot exceed 16 million m3 per year -- sub- scenario BPS(U). In the second -- EPS(L) -- wo assume that no new alcohol- powered vehicles are sold after 1990 and the proportion of alcohol in gasohol is maintained at 12%.2 While the second sub-scenario is politically more difficult to achieve, it is economically preferable and gives a significantly lower limit to alcohol consumption. These scenarios are summarized in Table IV.3. AssuaMtions Used tn ForeAt Seenaris _ _ _ _ _ _Ms. eamstic Prices Constant 1989 PrfeeS Petroleua Products: Internationul Prices by 1995. Natural Gass Fuel oil equivalent price ti 1995. Alcohot s Gaso If prie by 1995. GOP Forecast 1990-1995: 5.22 Same a M1 1995-2000s 6.02 eAlcoho Sply. 16 idflion ff/year maxime EPS(U): Sam as WAU EPSCL)t IIo nel atchot cars after 1990. Matural Gas S5pply 1f .illion a /dIay ansaun Inter. 011 Price 19901 US$ 19.40 bbt tm asi SM. (USS /berrel at 1995 uss 16.65 bbt 1988 prfcms) 2000: USS 21.23 bbt Notes: FM musiness as Usual; IS u Efficient Pricine Scenarfo. 4.8 Table IV. 4 and Fig. IV. 3 present the comparative forecasts. Although there is a similarity in the forecasts for most fuels, the A.1 scenario of PRTROBRAS provides the lowest projection for total petroleum products demand by the year 2000, while the BPS(L) scenario provides the highest. On the other hand, PSTROBRAS' A.2 forecast for total petroleum products demand by the year 2000 is almost equivalent to the BAU forecast, despite the difference in pricing assumptions. A possible explanation for PETROBRAS' relatively low A.1 forecast is that it implicitly incorporates supply constraints (e.g. on domestic production, foreign exchange availability and investment) rather than representing true demand, given the likely evolution of incomes and prices. 'The proportion has been steadily reduced fran 221 to 125 as of Jamary, 1990 (see Chapter 1). - 53 - 4.9 Without the upper limit on alcohol supply imposed by BPS (U) , the Petroleum Products consumption of alcohol would (Thousd TOE) continue to increase as the 199 vehicle fleet increases. 10 According to the PETROMS too BAU A.1 Scenario, alcohol saply would exceed 16 million rb 1994 and reach 24 million m by the year 2000. Under s BPS (L) , the manufacture of so alcohol-powered cars would stop after 1990 and all new 70 vehicles from 1990 onwards so would be gasohol-powered. so The gasoline/anhydrous 198981 IMS 1095 1997 198 alcohol mix would be maintained at the present fOure IV.3 ratio of 88/12. There would be a steady decline in the consumption of hydrous alcohol, as alcohol-powered vehicles are retired, with hydrous alcohol consumption falling from the present annual level of about 10 million m3 to about 2.5 million m3 by the year 2000. Anhydrous alcohol consumption would increase, as part of the increasing demand for gasohol, from around 2 million m3 at present to perhaps 3.3 million 5 by the year 2000. These estimates suggest an aggregate demand for alcohol in the order of 5.8 million m3 in 2000 compared with 12.6 million m3 in 1988. IV.3 PETROLEUM SECTOR DEVELOPMENT STRATEGY 4.10 In November 1988, PETROBRAS issued an action plan, the "Plano de Acao do Setor Petroleo" (PASP), which laid out the broad lines of a development program to the year 1997 (see Annex IV.1). Under PASP, Brazil would reach virtual self-sufficiency in crude oil production by the end of the century. In parallel, natural gas productior would increase to 70 million M3/day. PASP required an investment program of USS3.4 billion per year. However, within months of its publication, PA&P was rendered obsolete by the initiation of the Government's new anti-inflationary program, the "Summar Plan" of 1989. 4.11 At the same time that PETROBRAS issued PASP, a natural gas development plan, PLANGAS, was completed under the auspices of CNN (see Annex IV.1). The production target was in line with PASP, but a more detailed analysis was given of natural gas utilization. PLANGAS envisaged a major expansion in gas supply to domestic, commercial and industrial consumers, along with increased use for power generation, transport (substituting for diesel and gasoline), and (to a lesser extent) petrochemicals and fertilizers. The goals were ambitious and would have required a large investment program, but PLANGAS also became obsolete, in view of the anti-inflationary measures adopted under the "Summer Plan". 4.12 Conrequently, PETROBRAS has been without a clearly-articulated investment program. Our approach was, therefore, to analyze the BAU and BPS demand scenarios in conjunction with an envelope of two investment scenarios. For a *low" investment scenario, we selected maintenance of the 1988 level of investments for exploration and development (about US$1.5 billion per year); and as a "high" scenario, we postulated doubling those investments, to represent the .... ... ... .. 1f a fa 99~~~ Jf-a i i,:r.E.t 39 dfg 1 a - 9 .94i.ggea,..seo 4a E...n i u uo .u*. e ae sasa - W nro df lv 00 " w ýW Wehå Iffil o .e . 4• . . . §~ i> i- - i-; i-4 8~~~~-I M3WM UCUMDMWIMWMAU..D * .d -*0 * 55 maximum which we believe could be achieved in practice. We then traced out the implications of alternative development strategies, in ters of the domestic production of oil and gas and of oil imports. We relied on an analysis of the historical relationships between investment, on the one hand, and thq growth in oil and gas reserves and in production, on the other. After inclusion of the refinery investment implied by the strategies, the "low" and "high" investments corresponded to US$1.9 billion and US$3.5 billion per year, respectively. The details are in Annex IV.1. 4.13 Two main conclusions emerged. First, on the basis of World Bank projections of international oil prices, the average incremental cost of domestic oil, at about US$25/barrel, is marginally (4%) above the cost of imports, for both the BAU and BPS scenarios. Nevertheless, given the margin of uncertainty in our cost and production estimates and the additional security that self- sufficiency provides against unforeseen international oil price increases and supply disruptions, it is appropriate for Brazil to develop its domestic oil production, provided that PETROBRAS pursues its cost-minimization efforts, through the upgrading of its costing system, currently under way. Second, the implementation of an efficient level and structure of pettoleum product prices would reduce annual costs by roughly 4%, for both the high and low investment scenarios, which translate into annual savings of some "SSO.3 billion. IV.4 REFINING O PETROLUM PRODUCTS 4.14 The development of Brazilian refinery output and yield are in Annexes IV.4 and IV. 5, respectively. Output of refined petroleum products has grown steadily over recent years, by about 1.3% p.a. from 1980 to 1987. Output was close to 67 million .3 of products in 1987, of which 93% was produced by PETROBRAS and 7% by privately-owned refineries. Diesel oil accounts for 35% of output and gasoline and fuel oil for 17-18% each. Since 1980, the importance of domestic crude as a feedetock increased from 17% of consumption to ;7% in 1985, but with domestic crude production stagnating ard consumption continuing to increase, the domestic proportion declined to 50% in 1988 (Annex IV.2). 4.15 There have been important changeo in the profile of refinery production, due to the impact of PROALCOOL on the demand for gasoline and the continued expansion in demand for diesel. In 1973, about 28% of refinery output was in the form of gasoline and 22% diesel. By 1988, these percentages had changed to 18% and 35% respectively, due to investments by PETROBRAS which helped the refineries to modify the product slate to match more closely the needs of the market. Iven so, a surplus of gasoline has been produced -- growing from 324,000 m3 in 1977 to 5,173,000 m3 in 1988. Initially, this surplus was exported principally to Africa and Latin America, but, due to changing market conditions, 90% is now sold in the US. 4.16 PSTROBRAS bases decisions regarding the mix of domestic crude production and imports, the growth of refinery capacity and the mix of petroleum products on a production model, in which relative prices are a major determinant. The model predicts a need to increase refinery capacity by 1.7% p.a. to 1997. The product balance is expected to shift, with diesel increasing to 38% in 1997, fuel oil recovering to 23% and a decline in gasoline to 13%. Brazil is expected to be a substantial not importer of LPG and diesel; a net exporter of gasoline; *56 * while fuel oil imports and exports should be roughly in balance. The scenario to 1997 also postulates increased self-sufficiency in oil, rising to 71% in 1997. These predictions and conclusions of the model are highly sensitive to Government energy policy, preeminently with respect to the alcohol program and investment in the hydrocarbons sector. Aditional studies and simulations are urgently required, based on consistent scenarios 3nd policy alternatives for domestic crude supply, petroleum products demand, alcohol production and inveftment. In this regard, as in many others, PETROMMs' ongoing costing study can 1 *ovide key information. IV. 5 THE PRICING or PETROLUM PRODUCTS ALCOHOL AND MATURAL GA IV.5.1 Taxation Policy 4.17 The retail prices of petroleum products and alcohol include a variety of taxes, although alcohol has been exempt from some. Prior to the October 1988 constitution, the fiscal regime was essentially oriented towards the Federal Government, but the old tax system has been largely replaced by a system which transfers the bulk of the receipts to the States and the municipalities, leaving a much lower proportion for the Federal Government. The main features of the present system are& a The ICMS (see para. 3.34), a value added tax charged by the States, typically at 17% of the retail price, although in some States 184, for most products. LPG is an exception, having a tax rate of 12%. PETRoBRAS is exempt from ICNS in the State of Rio, where much of its production is located; o The Imposto sobre Venda no Varejo de Combustivel (IVVC), which is levied by the municipalities. The tax varies locally, but in most municipalities has been set at the maximum limit of 3% of the pump priceI and a The IPI, PIS and FINSOCIAL, which have been retained from the old system, but together account for less than 2% of the price. 4.18 These taxes aggregated about 17% of the retail price, as of the end of - 1989. Forecasts for 1990 suggest an average transfer from the sector to Federal, State and local governments in the order of 20% of the retail price. The "fiscal load" by individual product, as of January 8, 1990, when most of the fiscal reform was in place, is in Table IV.9, and shows a weighted average tax of nearly 20%. Some international comparisons of petroleum taxes are in Annex IV.20. 4.19 Beyond the taxes on fuels, there are various taxes on vehicles and vehicle ownership and there are road user charges, as detailed in Annex IV. 14. Tht vehicle ownership tax discriminates in favor of alcohol vehicles. IV.5.2 Existing Pricing Poli 4.20 Two sets of Federal legislation regulate the level of retail prices for petroleum products. The first establishes general guidelines, including the principle that prices should reflect the cost of imports and that regular adjustments should be made for inflation, consistent with the Government's exchange rate policy. The second set aims to ensure that prices will provide, * 57 at a minimum, the following rates of return on each of PETROBRAS' main activitiest 12% on transport; 15% on refining; and 15% on exploration and production. 4.21 Within the "basketo of petroleum products, there is no formal policy of cross-subsidization; indeed, to the extent that a policy with regard to the structure of prices has been articulated, it has tended to stress the importance of economic factors. Goals have been set to raise the price of LPG to a level that would encourage energy conservation; to close the gap between diesel and gasoline prices; and to price fuel oil on a par with substitutes in thermal equivalent terms. In reality, the Government, through SEAP, has manipulated the "basket" in such a way as to produce revenue for PETROBRAS in a way perceived ta be the most consistent with anti-inflation policy (see Chapter II). As in electricity supply, it has been Brazilian policy to have uniform prices (before taxes) for petroleum products, with regard to both level and structure, regardless of supply costs. 4.22 The price of alcohol was set at 65% of the price of gasoline, until mid-1988 to give it an advantage over gasoline on a volumetric basis. The parity price in terms of fuel efficiencies, i.e. equivalence on an energy basis, is 74% - 81% of the gasoline price, depending on the vehicle model and the size of the engine. Since mid-1988, that price advantage has been steadily eroded, and the volumetric ratio stood at 75% in October 1989. 4.23 In the past, natural gas pricing was geared mainly to encourage its use in petrochemicals, fertilizer production and substitution for fuel oil. This situation is undergoing change and progress is being made in the reform of natural gas retail pricing policy, notably under Loan No. 3043-BR. Under the 1988 Constitution, the authority to set retail prices passed from the Federal Government to the States (Chapter II) and the detailed consequences have still not taken full effect. CNP continues to set the bulk supply prices, for the transfer of natural gas from PETROBRAS to the State distribution companies, and has confirmed its policy of basing those prices on the economic cost of supply. At the State level, the key decisions will be taken in Sao Paulo and in Rio de Janeiro. The former is committed to LRMC pricing and economic efficiency, taking into account social and financial obligations. The latter has still not designed a new policy to make the likely transition in the future from town gas to natural gas. Evidently, given the cross-elasticity of demand between gas and other fuels, notably f.41 oil anA LPG, CNP's power over retail troleum product prices gives it an in:.Lrect influence over retail gas prices. IV.5.3 Petroleum Products 4.24 The economic costs of the main petroleum products (as of March 1, 1989) are shown in Table IV.5 for Brazil on average and in Table IV.6 by region. Naphtha is not sold as a final fuel in the retail market but is used in bulk as a feedstock in petrochemicals and in the reforming of town gas. For this reason it is shown only for Brazil as a whole. The reference point for gasoline is the export price if Brazil is a not exporter, or the import price of gasoline with essentially equivalent combustible properties to Brazilian gasohol. On average, in terms of border prices, the economic costs of diesel and gasoline are quite aAs shoum later, to the extent that gas is a substitute for resideftial electricity, DNAE in theory also exercises some Indirect influence over gas pricing. * s - close to one another, although higher than other petroleum products. Fuel oil and naphtha are the lowest-cost products. For the geographical analysis, we have shown only the import (elf) cost of gasoline and Table IV.6 confirms that the regions with a high concentration of population and close proximfty to the coast, namely the South and South-East, have the lowest delivered costs for each productl while, except for LPG, the Center-West is the most expensive region to supply, with its more dispersed and remote population distribution. 4.25 The retail prices of petroleum products, including taxes, as of March 1, 1989, are also shown in Table !V.6. As stated in para. 4.21, these retail prices are regionally uniform. The average domestic price of a composite barrel of oil products, including taxes (US$ 33.56/barrel), is 9% above the average k*rder price (US$30.76/barrel), including all the costs of transport, O"stribution and sales to the final consumer. However, this price level is achieved by high gasoline and (to a lesser extent) diesel prices, which cross- subsidize other petroleum products. LPG is priced at 58% of import parity; fuel oil at 72%1 kerosene/jet at 82%; and naphtha at 33%. Based on 1988 consumption data, the implied economic subsidies for these products are as followes LPG US$ 594 million Fuel oil US$ 397 milliGn Kerosene/jet US$ 117 million Naphtha USS 26 million Total USA1,134 million Of course, arithmetically, these subsidies are more than offset by the implicit taxes on gasoline and diesel. 4.26 In regional terms, the highest subsidies per unit of product are in the North and Center-West, where the abaolute consumption is low, so that a very small proportion of sales and consumption is involved. Hence, while these regions receive about US$280 million in subsidies, it represents less than 3% of the total revenues from the sales of petroleum products.4 The regional differences between prices and costs in financial terms is thus not a major issue for PVTtoBRAS and is handled internally through a price equalization fund, the Frete de Uniformizacao de Precos (PUP). A component of the retail price of gasoline and diesel is earmarked for this purpose, like a tax. Since PSTROBRAS is a monolithic institution, the regional price equalization fund for petroleum products works smoothly, . n contrast to the ait%kation in electricity supply (para. 3.56). 4.27 From an economic viewpoint, the regional cross-subsidies are more troublesome. As we have pointed out in the case of electricity pricing, they distort patterns of development away from activities which would exploit regional comparative advantage.5 For example, they reduce the incentives to develop and exploit local energy sources in high-cost regions. Furthermore, implicit subsidies lack transparency with respect to resource f1cws and economic losses. To the extent that the subsidies generate activities that depend on the subsidy 4 The detits are as follows. Subsfdy to: North, US 26.0 mition cO.2% of revenue); Center-West, USS 253.9 aition (2.3% of revnue); Subsidy from: North-East, USS 113.8 million (1.0 of reveme); South, uft 177.9 million (1.6% of revenue); South-East, USS 445.1 millIon (8.25 of revenue). Total Net Surplus: USS 956.9 million (8.49 of revenue). ' For further discussion, see also iorld Snk Report No. ?77-5 on subsidy policies, gR. dL Deceber -59 - L,O . . 3 . t...¥, , 4 . ou.2sar 7.2 13.14 2.44 ø.! ,24t .ø LOS 4 e S a 244 4Aaa . for vai 20A4: t.he c.ea 24.ae lob for thi p, .se øf t 64.5 100. 10- 0 ll n ae oe oW t .4 Srad T rue k a1 N.0 I sw.,~ Pol golald, •Pdgge de pø.es pa,a co*eewere Uqadee• t8081, Pc , Weaed oni f..h, l P st G erf, th anrl l N ou y patInti 30 y s pe a #sviefl MdInces nra rcs ewe 94ad184sne V1) h gy,qOOeéW4swufr41Osem dOee (e I t.i.* for viability, they treate a lobby for their perpetuation, regardless af the economic consequences. 4.28 While diesel oil and gasoline are both priced in exces oa their ecoomic coste, the latter is over-priced relative to the former, which has caused some undesirable inter-fuel substitution on the part of small and medium truaks and of pick-upa after the mid-1970s. The demand for diesel oil has increased coneistently faster than GNP over the past 30 years, despite a nearly seven-told increase in real prices between 1954 and 1984 (Annex IV. 15). The growth af the road transport industry, which is nov responsible for 75% ao the consumpti.on ·ot diesel and, to a lesser extent, the development and mechanizationl of agriculture have been the mai.n factors.6 We estimate that the consumptionl shift from gasoline to diesel oil was about 2.0 billion liters in 1988 .ncreases, atter the two oil shocks. of the energy costs relative to capital costs in the cost structures af tri - . - have been a major factor In thio suDstitution, as in many other countriei . aince the higher fuel efficiency af diesel engines justifies their higher manufacturing coste. Nevertheless, the high relative S se nex A .5. * 60- price of gasoline to diesel oil (Annex IV.15), which the Government increased fram about 1.2 prior to 1973 to over 2.2 in 1980 and 2.6 in 1987, has contributed to the acceleration of the dieselisation of the above-mentioned fleets. 4.29 Like petroleum products, alcohol is a tradeable. Following recent shortages in the domestic supply of alcohol, due to the switch by sugar producers from alcohol to sugar exports, Brazil has become an alcohol importer, at a reported cost of about US$ 50 per barrel (cif). The international market is limited, and able to satisfy less than 10% of Brazil's 1988 consumption. For this reason, PETROBRAS is seeking permission to substitute imported methanol for alcohol, so far without success, due to environmental objections. With regard to domestic alcohol production, we have assessed estimates of the economic cost which appeared in various studies, including the Bank's recent Public Expenditure Study and Sugar Subsector Review.? Given the controversial nature of the alcohol program, the estimated costs vary widely. They have also been made at different points in time, so that there are problems in comparing the results. The Public Expenditure Study estimate of US$ 33.9/barrel, expressed in prices of March 1, 1989, should be regarded as the lower limit. The PCR for the Alcohol Project gives a similar result. Again converting to March 1, 1989 price levels, the World Bank Appraisal Report of May, 1985, suggests a figure per barrel in the range US$ 47.54 to US$ 60.42; while the Getulio Vargas roundation (GV) obtained US$ 41.02/barrel, almost the same as an estimate by SOPRAL. Given the uncertainties surrounding the data, we have retained the FGV figure as a "best estimate" of the LtMC of hydrous fuel alcohol in our analysis. Table IV.7 identifies the corresponding delivered economic cost for Brazil on average. We have not examined regional variations in the cost of delivered alcohol: regional data is at least as contentious as average data; and the issues raised are more pertinent to the agricultural sector. However, as we ee in Section IV.7, for policy decisions on the alcohol program it is essential to formulate acceptable data on the regional costs of alcohol production, as part of an analysis of the dynamics of the LRMC. IV.5.4 Algohal 4.30 The retail price of hydrous fuel alcohol is now set at 75% of the level of gasoline. The gasoline price is US$ 71.97/barrel (see Table IV.6), which dictates a price of US$ 53.98/barrel for alcohol, almost equal to our estimate of the LRMC (US$ 53.20). However, this situation is misleading, because the economic cost of gasoline, which is a close substitute, is less. The maximum economic value of alcohol in use is the economic cost of gasolits. on an energy- ' "Brazif ts Public Expenditure, Susidy Policies and Budgetary Reform," Report No. T738-BR and %Brazil: Sugar subsector Review," Report No. 7589-OR. Also, CENAL, OThe National Alcohol Program," Nay 1988; L.C. Nonaco, "Alcohol as an Alternative Fuel," April 4, 1989 (CuatemuLa); and CUE, "Avatiaces do Prograns Nacionet do ALcoot,n Brasi lia, 1987. o The estimate Is the tower limit because it was based aon the est generous cost date and vas calculated to evaluate the economics of the alcohol program ex-post under the most favorable circumtances. - 61 * TABLE IV. 6 (USS6/barret as of Narch wayT) Retal t xc m IMMI~ cos__ LPG 15.50 26.89 41.96 26.50 24.55 25.13 39.75 Gasoline 71.97 37.19 39.66 39.49 36.38 36.87 39.70 Diesel 38.01 35.10 36.15 36.50 33.90 34.08 39.17 Fuel Oil 14.40 20.06 24.11 21.82 18.s8 18.10 33.96 Kere.IJet 26.91 32.70 36.16 33.68 30.26 32.26 47.99 Naphtha 8.08 24.80 Ideighted Avg 33.56 30.76 34689 31.46 28.44 30.49 43.39 Volume (IO) 54.30 3.10 8.60 29.30 9.20 4.10 VoluMe (Z) 100.00 5.50 15.60 54.00 16.90 7.50 Notess Volume sof4hted average and regional data do not include nephtha, which is not sold as s fuel in the retail earket. The average eanomic cost for Brazit is calculated frem the border price of petroleow. prodiuts faported plue internal transport distributian and sales costs, as shoun In Table IV.S. The economic costs for each individkal Region are the border prices plus internal distribution and sales costs, from Table IV.5, plus the Individeal econanic transport costs to each reIon. The transport costs are orld Bank estimates, based on 1986 ata, the latest year for ihich sufficient Information is readily available on product movemets, updated to Narch 1989. equIvalent basis. Using the border price of gasoline imports as a reference, the average economic cost of delivered gasoline from Table IV.5 is USS37.19 per barrel, i.e. US$ 30.12 per barrel of alcohol equivalent on average, after the 81% adjustment for energy equivalence. Based on 1988 consumption data, the implied economic subsidies to the alcohol sector from the sale of hydrous ethanol are US$ ..4 billion per year.9 If the export price of gasoline is used, the subsidy increases to US$ 1.7 billion per yearr using a price mid-way between import and export parity -- gasoline exports are roughly 50% of alcohol consumption -- gives an annual subsidy to hydrous alcohol of US$ 1.55 billion.10 4.31 To this figure must be added the subsidy to anhydrous alcohol. While the LRNC of anhTdrous alcohol is slightly higher than that of hydrous alcohol -- perhaps by 6% 1 -- it can substitute for gasoline virtually on a one-to-one basis in gasohol. For simplicity, we assume that one liter of r%ihydrous alcohol substitute. for one liter of gasohol and has a LUMC equal to that of hydrous alcohol, making the subsidy US$ 202 million on import parity; US$ 283 million on export parityl and US$ 242 million, using the average of import and export parity. a From Table 1.3 198 hydrous alcohol consuption was 9.6 billion liters (60.5 million barrels). The calculated difference between the LANC of alcohol and the opportunity import value is USS23.08 per barrel. 1o From Table V.5, the gasoline price f.o.b. Rotterdam (less ocean freight and port charges) is US$18.99 per barrel. Adjusting for the savings in internal transport and distribution and energy equivalence gives an equivalent alcohol value of US$24.89 per barrel. " Sea iorld Bank Public Expenditure Study, gR. g1U. *62- pa gr battel. (March 1, 1989) sconom1c Cost of Production 41.02 a/ Transportation Costs Braall Averages 4.57 b/ Distribution Margin 2.40 sales Margin 5.21 Average Economic Cost (Brazil) 53.20 votees a/ Pra Getulio Vargas Foundation (PGV) b/ World Bank estimate 4.32 It can be concluded that, taking the FGV figure as our best approximation to the LRMC of alcohol and on the basis of March 1989 gasoline price data, the total economic subsidy to alcohol, relative to 1988 alcohol consumption, was US$1.6 billion - US$2.0 billion, depending whether import or export parity is used for gasoline prices. The mid-value, i.e., US$ 1.8 billion, serves as an acceptable point estimate. IV.5.5 Natural Gas 4.33 We have calculated the economic cost of natural gas as the domestic long-run marginal cost of supply, excluding royalties, taxes or other transfer payments. Import options exist, and talks are already underway with Bolivia, but no contracts have been signed for natural gas delivery and little serviceable information is available on the likely cost of imported gas delivered in a major market, such as Sao Paulo. The LRMC of domestic gas is location-specific, due to the lack of interconnected systems and the high cost of long-distance transmission. We have made a rough calculation of the LRMC of the principal gas- supply regions in Brazil, taking into account, as appropriate, the costs of exploration, development, transport and distribution; and a depletion allowance. The details are in Annex IV.17 and the results are summarized in Table IV.8. 4.34 Important changes are taking place in natural gas pricing, as this clean and convenient fuel begins to penetrate Sao Paulo and Rio de Janeiro. A major atudy of natural gas retail pricing is underway in the State of Sao Paulo. Hence, only broad principles of gas pricing have been articulated by CUP, for bulk gas supply; and by the State of Sao Paulo, for retail supplies. While these principles accept that economic costs will be the basis for future gas pricing, the details remain to be worked out. 4.35 The opportunity or "netback* value of gas is equal to the opportunity cost of substitute fuels, adjusted for any differences in the cost of utilization, such as lower equipment costs or higher efficiency, and the premium .-63 - value of gas, stemming from its greater ease of handling, cleanliness, etc.. The opportunity value establishes the key benchmark for efficient retail gas pricing, when gas is in short supply, to ensure the correct allocation of gas between competing uses.12 In the longer term, if gas is expected to be in surplus, as reserves and infrastructure are developed more fully, and if gas continues to be neither imported nor exported at the margin, the appropriate benchmark would shift to the LPMC. Given the high cross-elasticity of demand between gas and other fuels in some uses, as long as distortions remain in the prices of these competing fuels, natural gas pricing in practice may have to rely on "second beat" rules, to ensure efficient inter-fuel substitution. Thus, it will be necessary to compare the potential demand and supply conditions in the fuel markets to arrive at more precise conclusions regarding an appropriate retail pricing policy for gas. As we discuss later, some additional concerns need to be introduced for efficient bulk supply pricing. 4.36 For the near term, natural gas will compete principally with fuel oil in the industrial sector, and LPG and electricity for domestic and commercial use. By the mid-1990s, it may also compete with diesel oil, gasoline and alcohol in the transport sector. The netback values for these applications of natural gas are in Table IV.8 and the detailed calculations are shown in Annex IV.17. 4.37 The potential for natural gas in power generation is still unclear but we believe that it could also be a viable option by the mid-1990s. Its netback value should be at least as high as heavy fuel oil in industry.13 Furthermore, if thermal plants are required in densely populated areas, natural gas is much less polluting than heavy fuel oil. It is our judgment that the domestic supply of natural gas will probably not be sufficient in the coming decade (see para. 4.7) to support a substantial program of gas-fired generation, given the economic value of alternative uses (see para. 4.38), although important local options could be available, for example in the Amazonian region. Discussions on the import of gas from Bolivia and Argentina have been initiated, and we urge the Government to continue to pursue these options. As stated above, we do not have datf on the economic cost of gas imports delivered to a major market, such as Sao Paulo, but we recommend (in para. 3.24) that ELETROBRAS investigate fully the use of natural gas among other thermal alternatives, to moderate the dependence on hydroelectric power in its generation expansion program. 4.38 Table IV.8 yields some broad conclusions on ratural gas utilization for the applications of natural gas which have been evaluated. Given the high "netbacks" relative to LRMC under the assumed conditions for most uses, the development of Brazil's natural gas resources in the markets considered would result in substantial positive net economic benefits to the economy. It can, therefore, be expected that natural gas will be in short supply, relative to its opportunity value, at least until the year 2000, in light of the substantial demands which have been projected for its substitutes (fuel oil, gasoline and diesel).14 Howevr, except in the South-East, the economic cost of CNG is over 2 Gas fa in short supply if doemnd would exceed supply at a price equal to LaNC. ' The notbeck would have to be calculated for specific applications and could be higher or Lower, for example depending on the location of the power plant and its ut iation relative to the lead curves for both the natural gas and electricity supply system. 14 In fhort supply because demand would exceed supply with price a LNC. -64 - 80% of the netback value of gasoline and diesel oil. This differential may be sufficient only to justify the conversion of high-mileage vehicles, such as buses and trucks, to CNG. TABLE IV.8 EconMIc Cost Ar% 1ethod VAlue for iVaturet asm fuss nR 1811ionAMn Raev of oetback Co%=r costs Nirns L.aim IJ Notbacd YA1u .._gs_ karge Indstry 2.21 2.15 1.34 3,00 0.88 * 1.75 anIl industry 2.97 2.51 1.70 4.21 1.64 2. 1 Residential 4.85 4.79 3.98 7.31*29.30 2.46 - 25.32 Transport .469 4.A3 3.82 S.34-5.58 0.6 - 1.76 041t I caoulattafria see AIex IV.Mf. The wvy high upper 11m1t for the netbek value Mn residentiat we reftets the v6lue of natural gas as a substitute fO tou-voltage residential electrieity, whith reqdres a large capital cost to ouet i short-duration load, esentisity r cooking and tighting. 4.39 Table IV.8, in conjunction with Table IV.6, points to some conclusions on retail gas pricing, taking into account the indirect influence which CNP and DWMB exercise through their role in the pricing of substitutes. The high retail prices for gasoline and diesel relative to "netback" values mean that the benchmark for "first-best" efficient gas pricing with limited supplies can be followed, namely a gas price based on "netback" value.15 Fuel oil is priced below its economic cost, but the ratio is sufficiently high, given the premium value of gas of at least 20% over fuel oil, to suggest that the principle for efficient pricing could still be pursued. In the case of residential electricity supply, which is priced at less than 50% of its LItC, the ratio of netback to the LRMC of gas is so high that using the retail price of electricity as the benchmark, to ensure efficient inter-fuel substitution under the criterion of the "second best*, would provide a gas price well in excess of the LRMC of gas. The most problematical case is LPG. The price of LPG is so heavily subsidised for social reasons that -- to ensure correct interfuel substitution, under the rule of the "second best" -- the price of natural gas would have to be set at roughly 60% of the opportunity value, i.e. around US$ 4.21/million BTU, which is below the LRMC of gas. LPG pricing will be taken up again in para. 4.45. 4.40 Clearly, more detailed analysis is required before an efficient retail gas pricing policy can be implemented in practice. The studies carried out as part of PLANGAS provide useful input data, but some critical information, such as the cost of developing gas offshore and onshore in the regions remote from markets, should be refined. The upgrading of its costing system, which PETROdAS is now conducting, will represent an important move in that direction. 4.41 Bulk supply gas pricing needs to take into account some additional concerns. The high "netbacks" to natural gas relative to its LRMC imply that substantial economic rent could accrue to the sale of natural gas, even allowing * The discussion in this paragraph is based an the prmise that anetbacka value exceeds LIUC. - 65 - for distortions which exist in the retail pricing of petroleum products. An important issue is therefore how this rent should be shared between the Government, PETROBRAS, the State distribution companies and the consumer and the implications of that sharing for the long-run incentives which are provided to PEToBRAS to find and develop gas reserves. 4.42 The development of higher-cost non-associated natural gas has lagged behind that of associated gas, as shown by the high reserve-to-production ratio of non-associated gas (44 years in 1986, see Annex IV.7). In terms of natural gas supply, out of a total supply of 6.6 billion m3 in 1988, only about 1.1 billion me (17% of total gas supply) was contributed by non-associated gas reserves. Careful attention therefore needs to be given to the adequacy of the bulk supply price which PSTROBRAS receives for natural gas as an incentive to develop non-associated gas reserves. The bulk supply price would need to be established above the average incremental cost of exploration and production plus transport costs, to give PETROBRAS an appropriate share of the economic rent. Another benefit of the upgrading of PETROBRAS' costing syste is that it will provide necessary data to help to resolve LiAs issue. 4.43 Aside from the level of the bulk supply price, CNP needs to establish guidelines on the structure. Distribution companies will require a higher quality of service than many direct industrial consumers and their load pattern will vary. A range of options on structure exists, involving, for example, the use of fixed and variable charges, demand and commodity charges, take-or-pay etc.. Finally, there will be a need for geographically-differentiated bulk supply tariffs, based on local production costs and transportation distances, which could lead to widely divergent prices between regions where onshore associated gas predominates and regions where offshore non-associated gas must be produced. IV.5.6 Social Considerations 4.44 Both residential electricity and LPG are priced below their respective economic costs, to provide subsistence quantities of energy to lower- income consumers at an affordable cost. Hence, to promote efficient inter-fuel substitution and induce consumers who use LPG and electricity for cooking to switch to natural gas, the price of electricity and LPG would have to be increased or the price of natural gas would have to be set at a competitive level, below its opportunity value. We have argued, in Chapter III, that restructuring of the residential electricity tariff is desirable in order to target the needs of the lower-income population more etficiently. A suggested approach was to have a alifeline" (subsidized) rate for a monthly consumption of 50 kWh per month or less and to charge all additional consumption at full LRMC. A similar "lifeline" tariff could be established for natural gas consumers. 4.45 The case of LPG is diferent. First, the social argument for subsidizing LPG is weak. Based on a national survey of family expenditure,16 families with incomes exceeding seven minimum salaries consume on average three times as much LPG as lower-income families, i.e. with two minimum salaries or less. At the same time, the low price of LPG relative to other liquid fuels, notably gasoline, has encouraged its clandestine substitution in motor vehicles; 'Oftud NaconMet do Despese do Faltia (ENDEF). See A.P. Rodri gues, N.G. Rodrigues end N. Arcute, OGLP: os Fatsos Argusentos de Politica do Proece" So Pauto Srrgfa, No. 53, June, 1989; and N. Aroue end A.P. Rodrigues, "The structure of Energy Demnd In the Residential Sector in Brazit, CPPE, June, 1988. * 66.* and discouraged the use of gas as an alternative. From Table IV.6, LPG is priced at US$ 15.50/barrel (US$ 4.21/million BTU) # 58 of the average economic cost of US$26.89/barrel, which means that natural gas could not compete with LPG outside the South-East, unless priced below LRMC. In the South-East, where the gas is predominantly associated with oil and therefore has a low LRM%, it might also need to be priced below cost, given that our estimate (US$3.98/million BTU) is approximate, less than 10% below the retail price of LPG, and has a margin of uncertainty. On the other hand, increasing the price of LPG is unlikely to have any major harmful impact on lower-income familiess it has been estimated that only 1.5% of the family budget is spent on LPG for families earning up to two minimum salaries.17 There are, therefore, powerful arguments in favor of raising the price of LPG in line with its full economic cost. The energy needs of lower-income families can be met with alternative energy supplies, including fuelwood as well as commercial energy; and with electricity and gas, there are well-tried techniques to ensure that this takes place in a reasonably efficient manner. Such techniques were discussed for electricity in Chapter III. Some details on pricing gas for low-income supplies are in Annex IV.18. IV. 6 FINANCIAL AND FISCAL ASPECTS 4.46 The financial performance of PETROSRAS during the early 1980s was healthy, with a net operating income in the range US$2 to US$3 billion. With this revenue, the jompany was able to generate sufficient funds internally for over three-quarters of its total funding requirements. Generally, investment- related borrowings did not exceed 25% of needs. Despite the large investment program of this era, PETRoBRAS consistently ended its fiscal year with cash surpluses. 4.47 From 1986 to 1988, financial performance deteriorated. Between 1986 and 1987, revenues (after sales tax) fell 11%, from US$13.0 billion in 1986 to US$ 11.5 billion in 19871 and by a further 9% in 1988. Operating costs in 1987 were some 28% higher than in the previous year. In 1988, while the operating costs were still 19% higher than in 1986, compared to 1987 they improved by 7%. In 1987, PUTROMEAS' not profit, less than US$0.2 billion, was only 10% that of 1986. The situation in 1988 improveds net profit was in excess of US$0.6 billion but still only about one-third of the 1986 level. The financial decline was due primarily to the effects of increasingly high inflation, which caused real erosion in domestic price levels, due to delayed or inadequate price adjustments. Futhermore, in 1987 the situation was aggravated by the need to enter into more expensive short-term purchase contracts for crude oil, a situation which was largely remedied in 1988. 4.48 Because of its decreased earnings and its prudent financial policy, PETROBRAS reduced its investment program from US$2.4 billion in 1986 to US$1.8 billion in 1988 and maintained a relatively high level of self-financing, by borrowing only about 25% of its now constrained program. PETROBRAS' gross revenues for 1989 are estimated at US$14.3 billion and net profits about US$850 million, a significant improvement over 1988 results, although still below earlier levels. 4.4q The hydrocarbons sector as a whole has contributed an estimated US$1.6 to US$3.4 billion annually to the revenues of the Federal Government M See A. Rodrues . gJ., g. gft * 67 - during the past decade (see Annex IV. 19), generated mainly by sales taxes on petroleum products sold by PVTROBAS, its retail distribution subsidiary (BR) and other private distributors. Further elements of total transfers include income tax, other duties and dividends paid by PBTROBRAS and BR to the Federal Government. 4.50 Total transfers from the sector over the decade fluctuated, due to changes in the fiscal policy of the Federal Government, principally regarding sales taxes on petroleum products, which represent the bulk of transfers. During 1980-1982, the taxes reached an annual average of US$2.4 billion. After 1980, however, the Government implemented a gradual reduction in sector taxation, in order to reduce the burden of the second oil price shock on the economy. A further and sharper decline in sales taxes occurred during 1983-1984, following the elimination of certain taxes. Thus, annual sales tax revenues for 1983-1986 averaged about US$1.8 billion, or about 75% of the 1982 level. In late 1986, the Government instated a temporary compulsory loan, to be generated through a 28% tax on gasoline and alcohol prices, again increasing sales taxes to an average of about US$3.1 billion during 1987-1988. on average over 1980-1988, the PETROBRAS group generated around 81% of total sector transfers, which amounted to some 10% of its own gross revenues. 4.51 It can be deduced from Annex IV.19 that the average ratio of taxes to gross revenues over 1980-1988 has typically been less than 10%, although increasing to 170-18% in 1987 and 1988. However, this data depicts the tax regime prior to the changes arising out of the new constitution (see paras. 4.17- 4.19). Table IV.9 summarizes the most recent information available, when the fiscal load represented only: 22% and 19% of the average retail price for gasoline and diesel, respectively; 8% of the fuel oil price; and on average, 19.7% of the retail price. The weighted average of taxes (as a proportion of the final retail price of a composite barrel of petroleum products), although nearly doubled from its historic levels, is still low compared with other countries (see Annex IV.20). IV.7 CONCLUSIONS AND RCOMAENDATIONS FOR CHANG 4.52 The Government's substantial fiscal deficit, the expected requirements for large investments by PMTROBMS and the prospect of substantial continuing expenditures on crude oil imports in 1990-2000 make it necessary to reexamine carefully the Government's policy towards: the pricing and taxation of petroleum products; the alcohol program; the quest for greater self- sufficiency in petroleum productio.; and the potential for greater private sector participation (both local and foreign) in the financing of hydrocarbons development. While further research is required in certain areas to resolve various quantitative questions in the longer run, we make recommendations here which we believe can be acted upon immcdiately, given the large minimum reforms which are necessary. 4.53 Brazil has taken reasonable steps in the last five years to keep the price level for liquid fuels at or above the border parity equivalent. However, the tax element was low by world standards and there are distortions in relative prices. These basic distortions -- due to Government pricing decisions, including vehicle tax incentives -- fostered increased consumption of alcohol, fuel oil, diesel and I&C in particular. Brazil embarked on a policy of manufacturing alcohol vehicles; greater dieselization of the truck fleet was *68 - ABEIV,.9 taxamto on Petro~lu an Aleohot 0194M.e Pro* f M-Jv1.1b L A ggg.ag Tax as # of Priceb LPi 3, ... Gaeoline 22.3. N)aphtha .6 Shel OLE - 8.4 Diesel 19.1 Kerosene 18.S Alcohol 20.4 Weighted.Average 19.7 stimulated; and increased imports of LPG took place, not only for residential cooking but also (unintentionally) to displace gasoline in small vehicles. Natural gas had difficulty in competing with fuel oil and LPG due to their low prices; and the greater consumption of diesel and alcohol has necessitated, on the one hand, refinery investments to increase the yield of diesel, and on the other, the search for suitable export markets in which to sell the gasoline surpluses. 4.54 We recommend raising the prices of all petroleum products to, at least, their economic costs; and, as in the case of electricity (para. 3.59), price adjustments at sufficiently frequent intervals to secure the financial health of the petroleum sector. We include in this recommendation the dismantling of regional price uniformity, as economic cost should reflect regional differences in transport c.3st (see para. 4.27). If PETROBRAS' prices are regulated at the ex-refinery level, competition among distributors can be relied upon to determine retail prices, with full reflection of regional cost differences. As noted in para. 2.35, CNP's role would thereby be greatly reduced. We estimated in Section IV.2 that the demand for petroleum products could be cut by 5% by the year 2000, through improvements in the price structure in the ZPS(U) as opposed to the BAU scenario. That would translate into annual cost savings of roughly 4% to the year 2000, equivalent to US$0.3 billion p.a. for both the highest and the lowest investment scenarios. At the same time, based on 1988 consumption data, estimated subsidies of US$1.1 billion would be eliminated on petroleum products which are being sold at prices below economic coat (see Section IV.S.3). These actions should be integrated with the policy decisions on natural gas prices, both wholesale and retail, which need to be taken in the immediate future, as market development occurs in Sao Paulo and elsewhere. Economic pricing of petroleum products will facilitate efficient gas pricing, based upon the economic cost of substitutes, which in turn would provide stronger market incentives to develop natural gas usage, including imports if appropriate, given the high netback values we have identified. 4.55 The next step is to decide on the extent to which taxation will raise or keep petroleum product prices above economic cost, to pursue fiscal objectives. Notably, some or all of the implicit tax on gasoline (equal to the - 69 - existing excess of price over cost) can be converted into an explicit tax, as it will no longer be necessary to crose-subsidize other products in the "basket". Furthermore, our proposals on alcohol pricing will reverse the decline in gasoline sales which, along with the low tax rate, contributed to the lose of tax revenue. These decisions on taxation involve macroeconomic judgements, which we cannot make from an energy sector perspective, although our estimated price elasticities (Annex 1.9) of petroleum products, alcohol and gas are low enough in the short- and long-run to anticipate substantial tax revenues. 4.56 One of the more difficult areas for recommendations relates to alcohol. Ideally, we need to consider it within a cross-sectoral framework, incorporating "upatreamo impacts on the sugar industry and "downstream" effects on the automobile manufacturers and the environment.18 While we recommend further investigations in this area, as minimum steps the Government should immediately rstse the alcohol price to parity with or above that of gasoline (including the same taxes as gasoline and in energy equivalent). We would expect that this measure would effectively eliminate the demand for new alcohol care.19 As a rough comparison of the policy implications, we considered three scenarios for alcohol penetration, assuming the LRMC of alcohol remains constant (see para. 4.9). o Government policy continues as in the past, alcohol consumption follows the PETROBAS A.1 Scenario (to reach 24 million m3 by 2000) and the jxport price of gasoline represents the opportunity value of alcohol. o No further investment in alcohol production capacity occurs, alcohol production is held to an upper limit of 16 million m3 as in BPS (U) and an average of the export and import price of gasoline is taken as a measure of alcohol's opportunity value. o No new alcohol care are sold after 1990 as in XPS(L), alcohol consumption falls to 6.8 million m3 by the year 2000 and the import price of gasoline measures alcohol's opportunity value. 4.57 To evaluate these scenarios under the most optimistic assumptions for alcohol, international petroleum prices are assumed to continue at the March 1989 level or increase by 10%, 30% and 50%. The results are in Table IV.10. They show, for example, that a combination of the World Bank projection of oil price increases (about 40%) with the cessation of new alcohol car sales, would reduce the annual alcohol subsidy by about US$1.5 billion compared with the present level and US$2.7 billion compared with the level which would otherwise obtain (from US$3.0 billion to US$0.3 billion) by the year 2000, assuming no change in the LRMC of alcohol production. 4.58 The above scenarios are illustrative. They do not allow for the impact of policy changes on the sugar industry. As the alcohol industry contracts, LRMC should fall, through the survival of lower-cost producers, improved productivity, and more efficient electricity generation from bagasse. 1 Alcohot ia envfrormentatty tees harmful than fossit fuets. * If not, the aLcohot price would be raised further or a tax penatty introduced on alcohot vehicles, due to the higher economic cost of alcohot compared ilth gesolin. -70 - TBEIV.10Q 074 006 4.59 Other scenarios for alcohol pricing are possible. Of special interest is the deregulation of the price of alcohol, which would result in a price increase in the short-run, given the cost of switching from alcohol to gasohol vehicles, thus easing the present acute supply shortages. The demand for new alcohol care should fall immediately to zero. Deregulation would be more in keeping with the fact that alcohol production is entirely in the hands of the private sector. Since the sharp price increases from deregulation would be regarded as unfair by motorists who were encouraged to buy alcohol vehicles, through Government policy, subsidies could be offered to motorists to convert their alcohol vehicles to gasohol, to accelerate the rundown of the alcohol fleet. One-time subsidies for conversion might be more efficient than continuing (albeit diminishing) subsidies on consumption, depending on the expected present values of the two subsidy streams. However, we would expect the same long-run equilibrium under deregulation as in para. 4.56, i.e., price equivalence with gasoline, as conversion and replacement of alcohol vehicles to gasoline occurs over time and as the LRUC of alcohol is driven down through productivity improvements. We recommend examining these other alternatives after parity with gasoline has at least been achieved. 4.60 Our analysis of alternative investment strategies in Section IV.3 suggests that annual investments by PSTROBRAS could be in the range US$1.9-3.5 billion through the year 2000, according to the rate of growth of demand and the degree of oil self-sufficiency which the Government wishes to achieve. While the investment figures would not be significantly out of line with the levels in the period 1980-1987 (US$1.6 - 3.9 billion), the range is generally above the average that PZTROBRAS has achieved* since 1981 (US$2 billion).2 'e believe that Brazil should invest in its own domestic crude oil production, provided that parallel efforts are made to cut costs, given the narrow margin between the cost of imports and domestic crude at existing international price levels (para. 4.13). Furthermore, when the constitution is reviewed, particular attention should be given to the monopoly of PETROBRAS and the role of the private sector. *Se Table I.S. * 71* !CHAPER V ENERGY CONSERVATION AND DEMAND MANAGEMENT V.1 INTRO U 5.1 Energy pricing and investment provide the unifying theme for this study, with energy demand as the key linkage between the two. It is thus convenient to consider the issue of energy conservation and demand manageme-t (3CDM) in this Chapter, bringing together conclusions and recomandatione which cut across the separate analyses of electricity, in Chapter III, and petroleum products, alcohol and natural gas, in Cnapter IV. We start by examining the energy intensity of the Brasilian economy, at the level of consuming sectors, to establish past trends. V.2 ENERGY INTENSiTY IN THE BAZILIAN ..CONOM V.2.1 The Overall Exoerience 5.2 Trends in energy intensity at the aggregate level and at the level of individual fuels are difficult to interpret in Brazil, due to the impact of a variety of programs for inter-fuel substitution, i.e to promote alcohol and electricity at the exW nse of petroleum products. Furthermore, trends in energy intensity have been masked by substitution between traditional and non- traditional energy supplies, as a natural consequence of economic development, resulting in a steady decline in the share of fuelwood (including charcoal). Such a shift typically reveals itself in an apparent decrease in energy intensity for the economy as a whole, because non-fuelwood energy can be utilized with more useful output per unit of energy input. Aside from shifts in the mix of different energy sources, changes have occurred in the sectoral composition of final energy consumption, as the weight of industry and transport grows in the total, with different energy requirements relative to value added. 5.3 Many of these broad changes in the pattern of supply and demand in the Brazilian energy sector were highlighted in Chapter I, in the context of the linkages between energy and the economy. The high degree of interdependence between energy demand and GDP Ad the apparent increase in the income elasticity of energy demand from the 1970s to the 1980s -- due tc the asymetrical response of energy demand to upward and downward movements in GDP -- translate into a decrease in aggregate energy intensity from 1970 to 1980, followed by an increase to 1987. However, the data in Table V.1 illustrate that the apparent reduction in aggregate energy intensity between 1970 and 1980 was due to a substantial decrease in the fuelwood elements the intensity of non-fuelwood energy consumption increased steadily, measured relative to the value of output. 5.4 Within the total, the behavior of electricity -- one of the fuels favored under Government energy policy of the 1980s -- is strikings the specific consumption of electricity (in toe per unit of GDP, at constant prices) increased by ov-r 80% from 1970 to 1987. International comparisons are also interesting. While Brazil's performance since 1970, in terms of energy demand per US$ of GDP, compares well with Mexico and Argentina, in both of which specific energy consumption increased sharply, Brazil did not experience the substantial gains * 72 * Indices of Eneray Intensity. 1970-1987 (19800100) Fuetwood consBUIon NtonwFustood Conption113 Aggregae fneot &wwerg124 100 106 Electricity COne*tlap 71 lo0 130 Industrfel don*fuetwood Cons. 82 to0 M Transport Liquid Fuels 129 100 103 4ot"I: tram intensity Is anmsu in te per CM,1.O, of GDP at IM0 prices, except for tranoport., shore the base is the sector's vaLue added. Tranport liquid fuels are diesel. gasoline ad -elcohoi.: in conservation recorded by developed countries, such an the U.SLA*, Canada, France, Italy and Japan. Electricity consumption in kWh per US$ of GDP (at 1980 prices) was higher in 1984 than a wide selection of developed countries as vell as countries in Latin America.1 V.2.2 The Industrial Sector 5.5 The behavior of the industrial and transport sectors bears pafticular earination, since their share in total energy consumption has increased from about half to nearly 60% ince 1970. With regard to .ndustry, the intensity of non-fuelwood energy consumption increased steadily from 1970 to 1987.* While the experience naturally varied between different types of industry, specific consumption rose in a wide cross-section of Brazilian industries during the 1970s and 1980s. Trends in the energy intensity of a selection of Brazilian industries from 1973, accounting for 85% of industrial energy consumption in 1987, are summarized in Table V.2. RnMMa Intensity of figlected Industries. 1273-19817 (1980-100) ... ....... ..... . . . .. '. . .:..:. So oeed Inaifu Tt metallurgical 10. 100 10 Textiles 6G 100 108 123 100. 93 Iceat- . 121 to .: . 98 *otes gnoregy intenalty measut rabycolssaptiein toe pr thwsndvons-of p1yiflat prOitlan. 5.6 Reductions in specific energy coasumption were experienced in a number of Brazilian industries since the early 1970s, with cement and chemicals anong them, accoreing to Table V.2. The case of chemicals is ambiguous, to the extent that it apparently excludes the consumption of naphtha as a non-energy U mlty, Germany, Frence, Italy, Japan the U.S.A., Australia, Uruguay, Naico and VenezueLa. See C. Feu, "IvestimOentos Energetices, Movembr, 1988. 73 feedstock. If naphtha consumption is added to the energy cs4.aumption of the chemical industry, its energy intensity goes up from 122 in 1973 to 127 in 1987 (1980-100). The experience with cement, however, is worth noting, as it experienced a reduction of 25% in eergy intensity between 1973 and 1987. Three reasons have been given, based on technological factors related to the production process.2 First, there were substantial economies of scale in the clinker- maktng process, as the average size of installed furnaces in Brazil increased by 75% between 1973 and 1982. Second, there was an increased market share of types of cement to which active ingredients could be added, which permitted significant savings of fuel in cement production. Third, the market share of "dry" process cement increased from 31% in 1973 to 46% in 1982. These "dry" processes use less energy per ton of clinker than "wet" ones. 5.7 Nevertheless, such declines were more than offset by the increase in energy intensity of the metallurgical industry, accounting for nearly 45% of the consumption of the group of selected industries in Table V.2 and 38% of all industries; and, to a much lesser extent, of other industries, such as textiles. A recent survey by the energy companies of Sao Paulo concluded that in a large metallurgical company located in metropolitan Sao Paulo, the monthly electricity bill could be reduced by 28% through adjustments in its maximum demand for electricity and its monthly energy consumption. These reductions would come principally from better lighting and insulation. V.2.3 The TranaDort Sector 5.8 The potential for conservation of diesel, gasoline and alcohol is high in transport.3 The consumption of these fuels per unit of GDP, after falling between 1970 and 1980, turned upwards during the 1980s (Table V.1). In freight transport, the potential for conservation lies: (i) with changes in the trucking industry itself (e.g., by rationalizing routes and schedules, consolidating cargoes and retrofitting truck fleets to improve engine efficiency and the aerodynamic performance of trucks); and (ii) through intermodal shifts, in particular if the railways can capture much of the increasing inter-regional transport, e.g. by offering efficient intermodal services based on the operation of direct unit-trains between specialized intermodal terminals. 5.9 In passenger transport, cars Are responsible for 70% of the total energy consumed, although they account for only 24% of total passenger-km; while buses, with 70% of the passenger-km, use only 16% of the energy. Lower-income households own a substantial share of passenger cars, apparently due, in part to deterioration in urban public transport service. Public transportation could attract private car users with improvements such as better traffic management and the creation of bus lanes and corridors in urban areas; greater frequency of service and reduced bus occupancy ratios; a more realistic use of public transport tariffs, to maintain them at levels which would promote service quality and respond more effectively to demand; and, where justified, through improved mass transit rail systems. Finally, the use of natural gas in urban areas might be feasible, in particular for buses and taxis, and should be further studied. N I.T. Totmasquim, "Energy Demand and Adaptation to the ORL Shocks - Brazil 1973-1985", 10th International IAEE Conference, Luxembourg, July 4-7, 1988. 8 Sea Staff Appraisal Report, "Brazit: Nanagement and Rehabilitation Project,a December 29, 1969. V. 3 InSTRNaTS OF ECDM POLICY 5.10 The level and structure of energy prices are the critical elements in any effective BCDM etrategy.4 They can induce the desired entrepreneurial behavior, for example, with regard to the choice of type of fuel, technological process and investments in energy saving measures. As the prices of all the main energy products declined in real terms after their peak in 1980-1981, energy consumption rose in Brazil.. The fall in the prices of gasoline, alcohol, LPG, diesel and fuel oil since 1980-1981 (see Fig. V.1) may well explain why the index of energy intensity for liquid fuels used in the transport sector rose between 1980 and 1987 after experiencing a decline in the 1970s (see Table V.1). The experience with electricity supply is no differents the prices of both residential and industrial electricity fell after 1981 (see rig. 111.3), while the index of electricity consumption increased sharply (Table V.1). In contrast, reductions in the energy intensity of electrricity usage among residential and industrial consumers in the U.S.A., Oapan, Germany, France, the U.K., Italy and Canada over 1973-1982 were associated with real price increases.5 5.11 The econometric analysis in Chapter I corroborates these observations. Price elasticities of demand for energy products, though typically less than unity, are higher in the long run than in the short run; and for individual energy products than for energy as a whole. DNAZZ has also pointed out that, in the case of electricity, elasticity is highest for the largest and the poorest consumers. 5.12 The argument is frequently heard in Brazil, as in other countries, that energy costs represent too small a proportion of total costs in some industries to interest companies in energy conservation. The argument has most validity in industries which are effectively protected from competitions they are better able to set prices on a "cost-plus" principle and pass on to the consumer the cost of inadequate energy management practices. A recent World Bank report has demonstrated how policies and institutional arrangements in Brazil shield producers from domestic and international competition, and argued that barriers to competition and structural change have muted the price and other incentives to Brazilian managers to introduce cost-reducing measures.6 Certainly, if competitive forces were working properly, the absolute size of the potential for cost savings through reduced energy demand would seem to be large enough in most sectors to produce a noticeable response to energy price increases. A survey of 390 industrial companies in Sao Paulo, carried out by FIESP in September, 1988, shows the average cost of electricity ranging from 1% to 6% of sales. These figures would indicate a profit&ble role for ene-gy conservation. 4 Demand managament is not synonpmus with energy conservation. Efficient demmnd mnen may case no chwe in total energy donnd in physical units - for exeople by rearranwing the profile of a given total demand in time or in space so as to reduce the totaL costs of supply -- but substitute cheaper for dearer energy foram. Similarly, energy conservation can operate strictly on supply conditions, e.g. by reducing system production and transport losses. For a further discussion of the analytical and coreptual underpimings of energy conservation economics, see N. Nunasinghe, "Third World Energy Policies - Demand Nanagement and Conservationa, Energy Policy, Vol. 11, No. 1, March 1983. aInformation from ELETROBRAS. 0 See Morld Sank Report No. 743-1, "Industo fat Regulatory Policy and Irrestment Incentives in BrazfI, June 29, 1989. 7 uAgenda para Apicacao de Energia," Ano 2, No. 10, Narco-Abrit de 1989. * 75.* Constant Fuel Prices Constant Cz $ of 1988 10 160 Gascine 140 120 100 T 80 - Kerosen 60- 40- Fuel O0l 20P 1970 1973 1976 1979 1982 1985 1988 Figure V.1 5.13 The Brazilian authorities have pursued a variety of non-price approaches towards energy conservation, including technical methods, legislation, education and promotion. Several of these energy conservation programs are described in Annex V.1. In 1981, KIC launched an ambitious program (CONSERV) promoting the rationalization of energy use in industry. CONSERVE was principally a program to substitute petroleum products with electricity rather than a broad program of energy conservation. More recently, the Government launched a National Program of Electric Energy Conservation (PROCEL), elements of which the Bank is supporting through its electric power leading program. ELETRoAS has the responsibility for overall management of PROCM. Other conservation programs were: the Program for Economizing Puels (PRO); the Voluntexy Program to Economize Diesel and Lubricants (PRODEL); and the Program to Rationalize Energy (PROEN). Given the importance of energy conservation and the plethora of programs, asE has proposed an umbrella agency to bring about some degree of consistency and coordination. According to CNN's proposal, a Program of Conservation and Rationalization in the Production and ge of Petroleum Products (CONPRT) would complement PROCEL and subsume other programs such as PRODEL. 5.14 All of these non-price approaches carry a clear risk that they may not be cost-effective. Furthermore, they introduce additional market distortions beyond the ones they are designed to redress. In our view, the Government should address these market imperfections directly and create a policy framework in . 76- which energy prices can secure an efficient allocation of resources. Evidently, efforts to reduce the costs of public enterprises are a normal part of efficiency improvement, e.g., the reduction of line losses in the electricity industry. Furthermore, to the extent that environmental considerations influence energy conservation efforts in electricity supply, the power sector should fully incorporate ("internalise") environmental costs in its LRMC calculations; and least-cost programs to disseminate information abot-t energy conservation possibilities can also be justified, if they are properly conceived. V.4 CONCLUSIONS AND RECOMNENDATIONS FOR CHANGE S.15 The data make it difficult to conclude unambiguously that prices were the primary factor causing the observed increase in energy intensity in the Brazilian economy discussed in section V.2. However, clearly the failure to use energy pricing effectively as an instrument of ECDM during the 1980s had serious adverse consequences on energy efficiency. From the economic point of view, efficient pricing requires the implementation of tariffs which reflect the level as well as the structure of economic costs. In the case of petroleum products as a group, the price level for a composite barrel of products has been kept reasonably close to the international level over a long period of time, but the price structure has serious shortcomings (see Table IV.6). The case of electricity supply is the opposites substantial progress has been made in implementing a sound tariff structure, but the price level is too low (see Table 111.6). 5.16 Benefits were obtained from the gradual introduction of LRMC-based tariffs in electricity supply, following more than ten years of efforts. By 1989, as a result of the implementation of the new tariff system, more than 2000 4N of load have been displaced from peak to off-peak hours (see Fig. V.2). This change is estimated to have saved more than US$1.5 billion, as a result of deferrini the installation of peaking capacity. It is a conservative estimate, as it was measured relative to existing consumers only: new consumers will be connected to the system with loads already modulated in response to the new tariffs. Benefits will continue to accrue to the power sector from the further implementation of the LRMC-based tariff -- 500 MW of load can be modulated in each of the Blue and Green tariffs (by January 1992) and in the Yellow Tariff (during the first six years of application), i.e. 1500 MW in all, equivalent to more than US$2 billion. However, the really substantial future gains from tariff action in the electricity supply sector will come from increases in the tariff level, in contrast with the petroleum sector. 5.17 We estimated in Chapter III that, by the year 2000, electricity consumption could be cut by 70 TWh, or nearly 20% of the total, if efficient pricing were introduced, worth 6,500-13,600 NW or US$10-27 billion over the 11- year period. The estimate takes into account the calculations of price elasticity of demand derived in Annex I.9, which showed that industrial electricity demand is particularly sensitive to price. Indeed, it is this price sensitivity which explains the success of the LRMC-based tariff structure in improving system load factor. However, Chapter I indicates that residential electxlcity will also be affected by price increases, with a long-run elasticity of -0.22. Our "top down" or econometric result is fully supported by the "bottom up" or microeconomic approach of Geller, Goldemberg et. al. in a paper published in 1988.8 That paper concluded with cimilar estimates of energy savings potential, based on a detailed analysis of the existing power market and the use of technologies that ar- technically and economically feasible and, in many cases, already available in Brazil. As in our discussion, the paper also identified the need to raise the average tariff level. Given the capital- intensive and environmentally risky nature of approaches which place the main emphasis on increasing supply to meet increasinS electricity consumption over time, rather than working directly on the growth in consumption itself, BCDX represents an attractive alternative. 5.18 Benefits in the petroleum sector are achievable mainly from the efficient inter-fuel substitution which would result from improvements in the structure of natural gas, petroleum product and alcohol prices, i.e. switching from higher- to lower-cost fuels. Notably, correcting the distortions between the prices of LPG and natural gas, on the one hand, and alcohol and gasoline on the other would reduce the total cost of supply. In aggregate, there could be a St reduction in petroleum products consumption, with eft,icient pricing. S.19 Concerning Government policy towards regionally uniform energy prices, we believe that there are potentially serious consequences for resource allocation. We have shown, in Chapters III and IV, how cross-subsidization occurs on a regional basis, due to this pricing policy. Wasteful energy consumption is likely to ensue. Also, as the World Bank Public Expenditure Study has pointed out, the subsidy distorts patterns of development away from activities that would exploit regional comparative advantage.9 For example, regional cross-subsidization of all forms of energy reduces the incentives to develop and exploit Iccal and regional energy sources, e.g., natural gas, biomass, coal and hydroelectricity. The outcome is to deter the conservation of energy resources which are scarce and high-cost in a national context. a N.S. Geter, J. Goldeberg, J.R. Noreira, A. Mtka, C. Scarpinetta and N. YachUoua, OElectricity Conservation in Bradtl: Potential and ProgressO, Energy, Vot 13, No. 6. * "Brazil: Public Expenditure, SUbsidy PoLicies and Budgetary Reform," Report No. 7738-tac p. 62. CHAPTER VI 6.1 This study has identified three central issues in the energy sector in Brazil, namely: (i) the economic efficiency of resource allocationj (ii) fiscal and financial questions; and (iii) income distribution. The efficiency issue arises because prices do not reflect economic costs and investment progranis deviate from least-cost solutions. Inadequate pricing and investment policies also led to serious financial problems in the energy sector enterprises, which had direct fiscal implications, via the public sector deficit; and indirectly reduced the revenues of the public sector, by undermining its capacity to raise taxes through energy prices. Income distribution issues arise both in regional terms, because of the effect of uniform energy prices on inter-regional subsidies and cross-subsidies; and at the individual level, as energy prices determine the affordability of certain basic services to the lower-income groups. Aside from the environment, which raises concerns far beyond the energy sector and merits its own study, these are the three most important issues facing the energy sector. They are discussed in more detail in Section VI.2; and a strategy to deal with them is proposed in section VI.3. The need for further action is outlined in Section VI.4. VI.2 TR ISU VI.2.1 Xconomic Efficiency 6.2 While the impact of a variety of programs for inter-fuel substitution complicates the analysis of the relationship between changes in energy demand and prices in Brazil over the last two decades, the failure to use energy pricing effectively as an instrument of energy conservation and demand management during the 1980s clearly had serious adverse consequences on energy efficiency. As the prices of all the main energy products declined in real terms, after reaching their peak in 1980-1981, energy consumption rose. Our econometric analyjis confirms the responsiveness of energy demand to prices, especially in the longer term. 6.3 Efficient pricing requires the implementation of energy prices which reflect the level as well as the structure of economic costs. In the case of petroleum products as a group, the price level for a composite barrel of products was kept reasonably close to the international level over a long period of time, but the price structure has serious shortcomings. The case of electricity supply is the opposites substantial progress was made in implementing a sound tariff structure, but the price level was too low. 6.4 The gradual implementation of LRMC-based tariffs in electricity supply, over a ten-year period, enabled the sector to modulate the power system load curve, leading to a more efficient use of the given investment. By 1989, more than 2000 KW of load had been displaced from peak to off-peak hours, saving more than US$1.5 billion in investment, as a result of deferring peaking capacity. On the other hand, the tariff level was too low, which discouraged *79 - energy conservation and over-stimulated demand and investment. We estimate that the implicit economic subsidies, measured by the excess of the LRMC of electricity over its price, amounted to US$5.4 billion in 1989. 6.5 Brazil has taken reasonable steps in the last five years, aided by falling international crude oil prices, to keep the domestic price level for liquid fuels at or above the border parity equivalent. The result was achieved primarily by setting a high gasoline price, because the price structure has serious shortcomings. The prices of LPG, fuel oil, naphtha and kerosene are all below economic cost, thereby stimulating consumptions the implied economic subsidies, taking the 1988 consumption data, amounted to US$1.1 billion. The price of diesel was generally kept in line with its economic cost, but it is under-priced relative to gasoline, which caused some undesirable inter-fuel substitution. Further distortions in the pattern of fuel demand arose in the case of natural gas, since the low prices of competitive fuels made it hard tor gas to compete, despite its relatively low economic cost and environmental advantages. 6.6 The retail price of alcohol has been close to its LRC, but resource misallocation arose because the price of gasoline, a close substitute with a much lower economic cost, was kept artificially higher. Vehicle tax incentives further stimulated the switch from gasoline to alcohol. The total annual economic subsidy going to alcohol, measured by the difference between the opportunity value of alcohol (i.e. the gasoline border price) and alcohol's LRMC, is about US$1.8 billion, relative to 1988 consumption. The extent to which this is offset by the environmental benefits of alcohol over gasoline cannot be addressed in the scope of this study (see para. 6.28). Furthermore, we note the position of the Secretaria do Assuntos Estrategicos, that the report fails to give recognition to the strategic arguments for the alcohol program. 6.7 These distortions in the prices of petroleum products, alcohol and natural gas had impacts on investment. First, the greater dieselization of the vehicle fleet and the rapid expansion in the sale of alcohol vehicles affected investment in the motor industry. Second, the petroleum industry carried out refinery investments to increase the yield of diesel and decrease the share of gasoline, because the growing surpluses of the latter had to be exported. 6.8 The Governmenta policy to reduce dependence on imported energy meant that investment programs were not governed fully by least-cost principles. In electricity supply, the development of large-scale domestic hydroelectric projects absorbed a considerable portion of the sector's financial resources; and options to import, by comparison, were relatively neglected. The nuclear power program produced virtually no electricity, despite investments of several billion dollars, and there has been growing controversy over its safety, as well as its eoiomics. The emphasis on domestic production of electricity also diverted a high proportion of investment into generation, at the expense of transmission and distribution, thus reducing system reliability. 6.9 The focus of the petroleum sector on domestic oil exploration and development was economically justified, since the cost of domestic oil was lower than that of imports. At the same time, preoccupation with oil production may explain, on the supply side, Brazila slowness to develop natural gas, for example as a substitute for hydroelectricity, fuel oil and LPG, in conjunction with distorti*ns in the price structure, on the demand side. VI.2.2 Piacal Problems 6.10 The low level o! electricity prices totally destroyed the power sector4s self-financing capability and provoked a critical financial position. Consequently, the Government found itself compelled to contribute increasing resources to the seto,. from its own over-stretched budget, adding to its fiscal problems. Electricity's claim on the public sector budget increased from US$400 million p.a. on average, in 1977-82p to US$2 billion, in 1983-85 US$5 billion in 19861 and US$6 billion in 1987. Using data from the World Bank's report on Brazil's macroeconomic situation, dated December 1988, more than half the real deficit of the Federal public enterprises was due to the electricity sector in 1983-1987. 6.11 Neither has the petroleum sector escaped the impact of Government intervention in pricing, even though the domestic price level for petroleum products was kept reasonably close to international levels and the sector contributed US$1.6-3.4 billion annually #-,m the revenues of the pub4 sector in the 1980s. The ftscal issue arises due to delayed or inadequate price adjustments and to the need to cross-subsidize alcohol and other petroleum products, both of which caused the sector's financial performance to deteriorate -- revenues (after sales taxes), not profits, and investment levels all declined after 1986 - and undermined tax-raising potential. To moderate petroleum price increases, the Government gradually reduced sector taxation. Even with recent increases, arising especially out of the new constitution, Brazilian taxes on petroleum products are much lower than a wide range of developed and developing countries. 6.12 Finally, the adverse consequences of inefficient pricing policies are magnified when account is taken .of all the costs of the alcohol program, which are not fully ref lected in the fiscal deficit. Only about US$836 million appeared in the 1988 budget under costs related to sugar and alcohol. The figure does not capture the direct economic losses, which we estimate at US$1.8 billion (para. 6.6), or the indirect sacrifice of tax revenues, caused by the steady decline in gasoline sales. 6.13 The legal and institutional framework of the energy sector gave insufficient support to private sector partf aipation, thereby contributing to the fiscal iat-. Although foreign companies were permitted to explore for petroleum under risk contracts, private participation generally in the petroleum sector, both domestic and foreign, and in the electricity sector has been limited. As a result, the energy sector was deprived of a potentially significant source of investment financing, wbich could have provided needed relief to the Government. V1.2.3 Income DistribuJtin 6.14 The Government pursued broad regional development and political (including income distribution) objectives through uniform national energy prices, in electricity, petroleum products, alcohol and natural gas. To the extent that Government has a legitimate concern with regional income distribution, the issue is whether or not that concern is better pursued through energy pricing or %ome other vehicle. We conclude that energy pricing is the wrong vehicle to promote such policies, because it has potentially serious long- term consequences for resource allocation. Wasteful energy consumption is likely to ensue from uniform pricing, while regions are deprived of the incentive to exploit comparative advantage, in terms of local energy supply options. * 81 - 6.15 In the case of electricity, the direction of the economic transfers may be perverse, since the uniform tariff will be of greater advantage to the South and South-Bast regions. Their Lower LRXC of transmission and distribution would be more than offset by their higher LRMC of generation, if power tariffs reflected geographical variations An economic cost. The ftiancial transfers flowed in the intended direction, because the situation in the past was different. The higher financial coats of the power companies in the North and North-East reflect historical circumstances, when generation was comparatively cheaper in the South and South-East. The LRXC is forward lookings for the future, the more economic generation is in the North/North-East (paras. 1.9 and 3.21) and tariff policy needs to reflect changed circumstances. 6.16 It is widely accepted that it may be desirable to provide energy to the lower-income population :t a price below economic cost, to satisfy b&sic needs. The issue in Brazil is that subsidies have not been targeted to the poor and instead operate to the benefit of all residential consumers. Two fuels are involved: residential electricity and LPG. Brazil's indiscriminate subsidies to all residential consumers of electricity are excessive. They amounted to US$2.5 billion in 1989, nearly halt the total economic subsidies going to the power sector (para. 6.4). Additionally, as in the regional case, the personal income distribution effects may have been perverse, because most of the subsidies (75%) have gone to the largest consumers, who are likely to be the better-off. The indiscriminate subsidies to LPG (US$0.6 billion in 1989) also were of more benefit to higher-income families, since they consume the most LPG. At the tame time, the low price of LPG relative to other liquid fuels encouraged its clandestine substitution in motor vehicles and impeded the development of low- cost and environmentally-beneficial natural gas. VI.3 THE STRATG VI.3.1 Economic Efficiency 6.17 To address the economic efficiency issue, we recommend a determined move to energy prices based fully on economic costs, in level as well as structure. We include in this recommendation the dismantling of regional price uniformity, as economic costs should reflect regional differences in production, transport and distributior costs. Price adjustments must be made with sufficient frequency to safeguard the financial health of the sector enterprises. Second, to support energy conservatior efforts in the longer-term, actions are required to reduce market imperfections. 6.18 We estimate that, in the case of electricity, efficient pricing could cut consuw-tion by 10-15% by the year 2000, corresponding to a drop in system capacity requirements of 6,500-13,600 MW, depending on the system load factor. According to the assumed average cost per kW, the monetary savings could be US$1.0-2.5 billion in investment, which should be compared to the current power sector investment program of US$6-8 billion p.a. for 1990-2000. The impact on the total demand for petroleum products would be less, given that the price level has maintained a closer relationship with economic costs. We estimate that the demand for petroleum products could be cut by 5% by the year 2000, mainly through improvements in the price structure, which translates into US$0.3 billion p...., some 4% below the annual costs without efficient pricing. 6.19 The adjustment of the price structure, in parallel with the movement of the price level to economic cost, would foster more efficient inter-fuel 82 - substitution and eliminate economic subsidies, valued at around US$5.4 billion in electricity and US$1.1 billion in petroleum products (parad 6.4 and 6.5). Continued tAplementation of LRMC pricing in electricity could modulate the system load by a further 1500 MW, equivalent to investment savings of more than US$2 billion. Economic pricing of petroleum products will facilitate efficient gas pricing, based upon the cost of substitutes, which would provide the market incentives to develop natural gas usage, including imports if appropriate, given the high calculated netbacks. 6.20 The price of alcohol should be set equal to or above that of gasoline in terms of energy equivalence, including all taxes, which is likely to eliminate the sale of new alcohol cars. The implicit economic subsidies would decline parl pasu with the rundown of the alcohol fleet, from US$1.8 billion p.a., on the basis of 1988 consumption, to perhaps US$0.3 billion by the year 2000. Furthermore, the subsidies would no longer be hidden in the energy sectort they would become transparent. An alternative approach is to eliminate subsidies on alcohol as a fuel, by establishing a free market, consistent with the fact that alcohol production is in the hands of the private sector. The price of alcohol would rise in the short term, and to respond to equity arguments, some or all of the subsidies could be applied instead to the cost of converting alcohol cars to gasoline. We would expect the same long-run equilibrium under deregulation, i.e. price equivalence with gasoline, as conversion and replacement of alcohol vehicles to gasoline occurs over time and as competition drives down the LRMC of alcohol. Evidently, our proposals for alcohol have environmental implications, but the evaluation of such issues is beyond the scope of this study (see para. 6.28). By removing the burden of subsidies from the energy sector, the implicit tax on gasoline (equal to the existing excess of the price over economic cost) could be recovered as an explicit tax, depending on changes in energy taxation policy (para. 6.24). 6.21 To support the desired changes in pricing policy, we recommend institutional modifications. First, Government intervention in energy pricing policy, for reasons divorced from sectoral considerations, e.g. by agencies such as SMAP to implement anti-inflation policies, should be stopped. Par from keeping inflation down, the indirect effects of failing to raise energy prices to reflect economic costs probably exacerbated inflation, by magnifying the public sector deficit. Second, while Government intervention in sectoral pricing and investment policies was the main cause of resource misallocation, rather than public ownersh$j per so, the increased participation of the private sector, which we recommend primarily for fiscal reasons (see para. 6.25), would provide a competitive spur to the areas which will remain under public ownership. Further efficiency gains could accrue in the longer term. 6.22 The issue of departures of energy investment programs from least-cost solutions needs to be addressed in electricity supply by the intensified investigation of thermal alternatives to domestic hydroelectric power generation. We recommend in particular studies of imported coal and hydroelectricity and strongly encourage the ongoing efforts regarding, notably, the import of gas. Furthermore, over the next several years, efforts are required (and are being made) ..o redress the balance of investment between generation, transmission and distribution. In oil and gas, we conclude that further investment by Brazil in its domestic crude oil production is justified, provided that ongoing efforts to cut costs, through the upgrading of PETROBRAS' costing system, are vigorously pursued, given the very narrow margin of advantage enjoyed by domestic crude at existing and projected international price levels. VI.3.2 Fingal Problea 6.23 The actions we propose on energy pricing would make a substantial contribution to the resolution of the fiscal issues. An increase in the electricity tariff Level to Jag55/MWh In September 1989 prices (US$54/KWh in July 1989 prices), from the US-647/KWh (including taxes), operative in September, 1989, represents a minimum acceptable financial target, according to the recently negotiated Electricity Transmission and Distribution project. It will also raise the tariff level to the LIRMC of US$71/NWh, if taxes of at least 30% are added and if the average tariff received is brought into closer line with the average tariff billed. Consequently, the Government's financial contributions to the power sector of some US$6 billion would no longer be necessary, mirroring the elimination of US$S.4 billin a economic subsidies (para. 6.4). 6.24 In oil and gas, the economic pricing of petroleum products and alcohol would further strengthen sector finances, creating resources for the likely investment needs of US$2.0-3.5 billion p.a. through the year 2000; restore the sector's tax base; and thereby improve the fiscal situation. We recommend considering an increase in the tax rate on petroleum products, alcohol and natural gas. For example, the large existing implicit tax on gasoline, close to 100%, could be converted into an explicit tax, as it would no longer be necessary for the energy sector to cross-subsidise alcohol by US$1.8 billion and other petroleum products by US$1.1 billion (paras. 6.5 and 6.6). Furthermore, our proposals on alcohol pricing will reverse the decline in gasoline sales, which, for a given tax rate, will boost substantially total tax revenues, since alcohol sales exceed gasoline sales. However, we have refrained from making specific proposals on tax rates, as they would involve macroeconomic judgements, which cannot be made fvom the energy sector perspective. Hence, it is impossible to calculate the not effect of efficient pricing of petroleum products and alcohol, combined with wreasonable" taxation. 6.25 In conjunction with actions on energy pricing, we propose institutional changes, to support and encourage more active private sector participation in energy, which would reduce the investment burden on the public sector and alleviate the fiscal difficulties. In oil and gas, the Government shoulds (i) encourage a greater role for private capital in areas of PETROBMS which are not protected by its monopoly and in State-level gas companies; and (ii) review the scope for improving the petroleum legislation, to attract more private capital (foreign and domestic) into finding and developing oil and gas reserves. In electricity distribution, private participation at the State level also offers scope for tapping new sources of capital, as in gas; while a variety of models exist for private participation in generation. The NNE has a vital part to play in any initiatives for greater private sector participation in energy. VI.3.3 Income Distribution 6.26 With regard to regional income distribution, we argue (para. 6.14) that energy prices are not an appropriate vehicle to implement Government policy towards broader economic and social development questions. The full transition to economic energy prices would remove the distortions which arise when demand and supply decisions at the local level are based on uniform national prices. Energy prices are frequently used to address issues of personal income distribution. While this, too, has potential resource allocation problems, our concern is rather that the efficiency of the present system of subsidies can be -84 - improved, to achieve the desired goal. In the case of electricity, we recommend increasing the sise of the first block of residertial monthly consumpeion, from 30 to 50 kWh, which would continne to be sold at a subsidse "lifeline" rate. However, all subsequent residentiax consumption would be charged at least at full LRMC, requiring an increase in the residential tariff significantly above average for all tonausra, to eliminate subsidies of US$2.5 billion (para. 6.16). We are not persuaded by the social arguments for subsidizing LPG, given that it is difficult to target only the 1&0 consumption of the poor and recognizing the resource misallocation which results. The price of LPG should be based on its economic cost, as with other petroleum products. However, as natural gas penetration proceeds in residential markets, encouraged by efficient pricing of LPG, the basic eitergy needs of lower-income groups could be satisfied with the introduction of a "lifeline" rate for 5as, analogous to the case of electricity. VI.4 FUR ACION 6.27 The strategy which we recommond above represents a set of minimum steps which can be initiated immediately, given the large reforms which are necessary. To assist with its implementation in the longer term and to lay the basis for further change, we also propose a number of studies. 6.28 First, ZLNTROSAs' planning methodology should be strengthened, to deal in a more sophisticated way with alternative pricing policies, demand scenarios, risk and uncertainty. The analysis of generation alternatives should incorporate an updated and more extensive analysis of thermal and non-domestic options, notably natural gas and imported coal. Second, high priority must be given to PSTROBRAS' efforts to develop profit centers and upgrade its costing system, to ensure that domestic crude oil production remains competitive and to support the full economic development of natural gas. As the profit centers are established, PETROBRAS can base management accountability on profitability and thereby increase its operational efficiency. We would anticipate that the implementation of appropriate profit centers will call for relevant corporate restructuring. Third, to improve the analysis of different policy options for alcohol, updated information is needed on its LRMC; and investigations should be conducted into the "upstream" impacts on the sugar industry and the "downstrem" effects on the automobile manufacturers and the environment. ?ourth, the Government should review tte scope for improving the petroleam legislation to attract a larger contribution from the private sector in finding and developing nqw reserves. Finally, further work is needed on the fisca" potential for taxing petroleum products, alcohol and natural gas, recagnizing demand elasticities, environmental implications and the financial consequences for the energy sector. 6.29 Brazil has relatively abundant energy resources, which could be husbanded more effectively if they were exploited with a sounder energy strategy. This report, in conjunction with the Rnergy Matrix, is offered as the basis for an early discussion of such a strategy for Brazil. -85- ANNEXE AND MAPS BRAZIL ENERGY PRICING AND INVESTMENT STUDY - 87 - aRøstic InfjaM A12 Btes Page i of i 9.4k% 1 .4 9, .39% 9.% ??.w '191 4A% 0$7% 0,0 LA3% 94.7 7.8 198 L92 1.05% 0.05 #.85% 927 ,3 19#1 9Bas. 1.5 78 14. 77,0% .1 18 100,0% 100,0 6.20 10.0 100.0% 100.02 19 24.2% 24.25% 13.66 22AG 91.0% 99 t97 8.9% 79.1% 3.23 432.98% *0.4% a0.1% 198 419 58 49%2K ,45.0 6227.46% 62.2% 59.7% øes 38414.4% 1,37o 21$7251% 56.5% 562 jul-6 49462.4% 1,140 30.1 d6.a% 4.4 sep. ? 718.8 897.9 Ø2000 52438.79% 75.1% 60.3% t4,4900 7244.21%4 Wots + : Te tP wa intrdce i he i-90s. le figure pror to thet dat ar4 estimates Domestic Inflation & Exchange Rate Base 1985 140% 130% 120% 110% 100% 90% 80% 70% 60% 1974 1977 1980 1983 1986 1989 Frg Exch Index / General Price Index - 88 - ANEX L.Z Page 1 of 7 1. Brazil is endowed with substantial energy resources, both renewable (hydroelectric power and biomass) and non-renewable (hydrocarbons, coal and uranium). The situation is particularly favorable regarding renewable energy resources, which provided 60% of Brazil's total energy needs in 1987, a proportion which was much higher than that typically enjoyed by other countries in Latin America (around 25%). Taking into account her own indigenous energy resources, Brazil's dependence on foreign energy supplies has been less than 20% in recent years. According to the National Energy Balance, prepared by the Ministry of Mines and Energy, the proven energy resources as of December, 1987, were, in million tons of oil equivalent (toe): petroleum - 352; natural gas - 96; shale oil -8,884; shale gas - 2,304; coal - 6,504; hydroelectricity - 271 per year; uranium - 2,100; and peat - 153. Comparable data on biomass (non- commercial energy) is lacking, but Brazil undoubtedly has the potential to continue to exploit sugar cane and wood for energy purposes. EWdroelectric Energy 2. Currently, about 46,000 MW of hydroelectric generating capacity have been fully developed in Brazil, covering a reservoir-inundated area of about 20,140 km2. Rydroelectric generation represents about 94% of the total electricity produced in Brazil. The usable hydroelectric potential has been estimated at about 106,705 MW, of which 56,660 MW or 53% is located in the North/North-East and the balance (50,045 MW or 47%) in the South, South-East and Center-Westl. Presently, it is estimated that less than 90,000 MW could be economically developed2. Additionally, it is estimated that about 7,100 W capacity may be tapped through the construction of some 560 mini-hydro plants with annual firm energy of around 31 TWh. 3. The largest potential for development of hydroelectric schemes is located in the North of Brazil, in particular the Amazon basin on rivers such as Tapajos, Xing, Tocantins and Araguaia, where about 36,000 MW or 40% of Brazil's potential may be economically tapped. In contrast, the South, South-Bast and Center-West regions, where most of the price effective schemes have already been developed, may provide an additional hydroelectric capacity of 46,000 MW (51% of the total potential) mainly consisting of medium to small size schemes. The balance of this potential (9%) is in the North-Cast. 4. The Plano 2010 (the sector long-term development plan discussed in Chapter III) has set as an objective for the continued development of hydroelectric resources. However, this program is to be complemented with the development of thermal-based generation (heavy fuel/gas/coaL fired plants) to firm-up the availability of energy supply, especially during dry-season and peak-hour periods. The main concerns regarding the development of hydroelectric resources which are now being addressed by Brazil are related to: (i) the environmental 'National Energy Balance, 1988* (p. 68) and ELETROBRAS, Ptano 2010: Retat6rfo, Gerat (p.145). 2 This assuming thermat generation costs below US $ 50 per NMh. ELETROBRAS, Piano 2010: Relat6rio ExecutIe (p.24). - 89 - ANNEX ILZ implications associations with developing the Amazon basin, for which measures are now being undertaken and for which comprehensive Environmental Protection Plane must be Implemented before starting construction; and iii) the economic feasibility of new hydroelectric schemes, for which alterntive sources of electricity generation (such as natural gas and coal) should be considered and/or conservation measures should be undertaken (Chapter III). TABLE Liquid Hydr2carbon Reserves (Million barrels) jit oth *pt a 4.7W i10 4.F5 1"17 A7 "*7 fa 293,13 44.68 46.96 *.ht 4,4A i Notheast 219.13 25.90 242 4.39 0.45 tiehas t7 286 0.30 Sothee t Setrt~ 48,52 17.06 O 3,77 to.&4* Cotty et 4 28Q,09 1996,56 10.08 19.33 t6.4 #+$4 5. Exploration for hydrocarbons in Brazil has been directed primarily toward oil-prone regions but discoveries of oil fields containing associated gas and non-assciated gas fields have steadily added to the proved reserves. The proved rserves of liquid hydrocarbons, crude oil and condensate, are estimated to be 2.8 billion barrels in 1988. Condensate reserves, the low gravity crude oil which is produced with natural gas in some fields, account for less than 1% of the liquid hydrocarbons and are estimated to total less than 20 million barrels. In the aggregate, liquid hydrocarbons account for approximately 804 of the energy content of proved oil and gas reserves while natural gas provides 20%. The estimates include estimates of proved, probable and possible reserves as well as estimates of potential reserves.3 As shown in Table 1, the proved, probable and possible reserve estimates for crude oil total 5 billion barrels. Over 70% of the proved reserves are located in offshore fields, almost entirely (95%) in the prolific South-East region offshore fields. The onshore reserves are divided more or less equally between Bahia state (10% of all reserves); Rio Grande do * In June, 1988, PETROBRAS adopted a new reserves estimation code, which classifieos reserves as proved or probable and possible. PETROBR.AS exploration officials believe the definitions for proved reserves are simiter to those used by the Society of PetroLean Engineers, but they are probably more restrictive. - 90 - P,age"o Norte (8%) t and Sergipe (7%). Very small quantities of proved reserves, 20 million barrels, have been identified in the North/North-East regions. The probable and possible reserves also are located almost entirely in the South-East offshore region. This region contains 90% of the total. 6. Since 1979, natural gas reserves have more than doubled. The average finding rate over the past decade has been approximately 10 BCM per year. The proved reserves, as of December 31, 1988 in onshore fields and offshore fields located in less than 400 meters of water, were estimated to be 109 billion m3 (BCM). The probable and possible reserves total 89 BCM in onshore fields and offshore fields inside the 400 meter water depth, primarily in the Campos and Amazon Basins. Additional quantities which are expected to be found in fields beyond the 1000 meter water depth are not counted as reserves, but they are estimated as 21.8 SCM. As of December 31, 1988, natural gas constituted 20% of the energy available as proved oil and gas reserves. 7. Half of the proved reserves are located in off-shore fields and approximately 60% are associated gas. As shown in Table 2, based on 1987 estimates, the reserves are concentrated in the offshore fields of the South-East region (37%); the onshore Bahia region fields (25); and Northeast region (19%), where most of the gas is located in onshore fields. 8. Natural gas liquids (NGL), which include ethane, propane, butane and heavier hydrocarbons, are extracted from natural gas and as production increases they will become an increasingly important source of fuel. Associated gas contains large amounts of extractable hydrocarbons and some of the newly- discovered offshore fields are also expected to be produced as condensate fields, yielding a relatively high ratio of NGLs. Based on the levels of NGLs to be recovered from natural gas projected for the PLANGAS study (see Chapter IV), it is estimated that the recoverable NGL ratio will be approximately 0.35 CM of NGLe per 1000 CM of natural gas processed. The total recoverable NGLe from the proved reserves is approximately 220 million barrels. This is equivalent, in energy content, to abont 5% of the proved crude oil reserves. 9. As in the past, PETROSRAS's exploration goal is to find oil, but recent discoveries have also proved up more natural gas. In the Campos basin, most of the natural gas is located in the Garoupa and Enchova fields, but there is some non-associated gas in the southern region. The hydrocarbon-bearing formations in the Santos basin are deeper than those in the Campos, but the water depth is only about 150 meters, so the drilling and production costs should be lower. The newly-discovered Tubarao field may yield 30 - 40 billion cubic meters of natural gas but the primary interest is in the condensate and other extractable liquids. Seismic data indicate that the fields in the Amazon basin will usually be small with an oil rim. . If these interpretations are confirmed, the fields will probably produce more gas than oil. - 91 - TMarx !.2 TABL 2 Natural Gas Proved Reserves. 1987 Worines~t 516.4 - 56.4 7140.5 71485 7664.9 QA ?666.9 ertit i4.8 * 10.8 St.4 991.0 t042,6 4#.4 9#1s 185, Northeast Centret 1149.0 6356.1 7S55. 182.2 1t92.9 1375.1 131,2 1549.0 8084. Vortheast 188.3 3356.9 5236.1 13307.1 1830.4 15137.5 15192.4 5181.3 20537 Bahia 11406.8 * 19406.8 14563.2 * 14363.2 25170, QA 2."70M > @fiteSano '190. 335.1 1345.8 8054$ * 005.5 1816.1 355.1 S1,3 Sothat$ 7702.4 $7798.4 - 174.4 1747.4 0.0 39649. 39449.8 Total 15979.0 477444 6372$,5 35858.1 561,7 4061948 Slasta 53506.2 1053434 Sources CNP. Anuario Estatistice 1986. 10. Brazil's estimated reserves of uranium have grown tenfold since 1'976, and now amount to about 300,000 tons of 11308. The breakdown of these reserves, according to the categories used by the Brazilian mining code (measured, indicated and inferred) and by location, is presented in Table 3.4 11. Allowing for losses in mining and beneficiation, these reserves are equivalent to 2,100 million toe. However, recoverable reserves are estimated at about 120,000 tone, half of which are located in Lagoa Real and Itatiaia. Exploration for uranium reserves has been relatively limited so far and additional reserves could be discovered. The only productive facility at present is at Pocos de Caldas, with a capacity of 100 tons of 11308 per year. In terms of electricity generation, it has been estimated that Brazil's uranium reserves could support a capacity of some 26,000 KW. E~mA3A 12. The two major sources of biomass energy are bagasse and fuelvood. The former is principally the by-product of the sugar industry, resulting from its sugar and alcohol operations. The latter derives from both natural forest and dedicated energy plantations. 13. Much of the bagasse is used as process heat and to generate electricity for internal purposes. About 3.*4 TWh of electricity was generated by the sugar industry in 1986. However, the sugar industry did not design its facilities to make the most efficient use of its bagasse for energy purposes; on the contrary, the objective was to minimize the bagasse surplus, which constituted an inconvenience. Furthermore, there are non-energy uses for the by-product, notably for paper manufacturi and cattle feed. Approximately 10% of the bagasse from sugar mills with annexed distilleries is available as surplus; while the figure is 20% for autonomous distilleries. With modernization and efficiency ' These definitions differ fron those used by the IAEA. Measured and indicated reserves correspond roughly to "reasonably assuredP, white inferred are possible additional reserves. - 92 - Pais5ig improvements, these proportions could probably be raised to 20% and 30% respectively. ELETROBRAS has cited data, apparently derived from ZAR, which show that electricity generation of some 13 TWh would be possible, over and above the sugar industry's own internal needs, equivalent to an installed capacity of 2,000 MW. A further discussion of the institutional and legal obstacles to the development of bagasse cogeneration is in Chapter 1I. Reserves Of Uranium (Thousand tons) Inferred Tital Qerrilteae . 10.0 is.0 Amierfopolis/GD 2.0 3.0 5.0 al rt/0050.5 1.0 aaC912 S.3 14.5 .a.. ke4$ 61.8 31.4 95.2 foT19.5109,0 301.5 tear4 oUL8A Paas BRm0rt~0 ot~c e ........ . . .... . 14. Datf on Brazil's fuelwood resources are uneven. A partial survey of forest cover carried out by the Brazilian Institute for Forestry Development (IBDF) in 1983 showed some 26 million hectares of native forest and 2.5 million hectares of reforested areas in the South and South-East, translating into a fuelwood stock of 4 billion m3 and 276 million m3 .espectively. These results showed that those regions had already lost a large part of their natural forest cover; while much of the reforested areas were not suitable for energy purposes. No national forest inventory has been carried out for the North-East and Center- West, but it is believed that forest cover is minimal or again unsuitable for use as energy. Bence, the major forestry resource in Brazil is the Amazon, where it is estimated that 300 to 350 million acres of forest exist. The fuelwood density varies considerably, according to the topology. Apart from the paucity of information on the energy potential of the Amazon, environmental considerations would severely restrict the extent to which it could be exploited for power generation. The disposition of Brazil's natural vegetation is shown in a map. 15. The huge size of Brazil and the variety of terrain suggest that there would be scope to develop energy plantations, for example through reforestation and appropriate management, without adverse effects on alternative agricultural uses of the land; and to employ those plantations in dendrothermal (power) generation. Although comprehensive studies have not been carried out in Brazil of such possibilities, the power company in the North-East (CHESF) has estimated, as a first approximation, that 50 million hectares -- corresponding to 33% of the region's land area -- could be reforested without prejudice to other uses of the - 93 - Pag 6s* f land.5 An annual production of 891 millionm3 of biomass would be theoretically sufficient to supply 62,000 MW of power capacity. While CHESP believes that the construction and operation of a pilot plant would be necessary to confirm its estimates, the economic justification is likely to be questionable, especially since production would proceed in relatively small units. coal 16. Brazil's coal reserves have been estimated at 32 hillion tons, equivalent to 6.6 billion toe These comprises 10 billion tons measured and indicated; and 2 billion tons inferred.6 They are concentrated in the South, in the States of Parana, Santa Catarina and Rio Grande do Sul (see map). About 80% of the coal reserves might be used for energy, the rest being more appropriate for metallurgical purposes. Brazil also has some peat reserves, but the economic potential is too limited and the available data too sparse to warrant inclusion in this discussion.7 17. The bulk of total coal reserves is In the State of Rio Grande do Sul (91%), with the main field known as Candiota (12 billion tons). However, the coal has a low calorific value (in the range 2500-4000 kcal/kg) and high ash content (50%). In Santa Catarina, the coal is bituminous, reaching higher calorific values (in the range 4,000-6,000 kcal/kg), but the ash and sulphur content are also high. The coal found in Parana is of better quality, but suffers from a medium ash and high sulphur contenti the calorific value is 6500 kcal/kg, which makes it suitable as a fuel. Shale oil and Gas 18. The possible reserves of shale oil and gas in Brazil are among the largest in the world. However, they have not been adequately evaluated and their potentiL role in energy production remains highly speculative. The National Energy Balance gives the data shown in Table 4, as of December, 1987. The reserves are mainly to be found in deposits in Bahia, the Vale do Paraiba (Sao Paulo), and Irati (in the South). See uEstudo do Florestamento do Semi-Arido NordestinoR. e See "Mational Energy Batance, 19880. The categories used are defined by the National Deportment of Mineral Production as fottows: the measured reserve is based on test drilling to 400 meters, with an associated surface area of 0.5 kof; the Indicated reserve is presumed to go beyond the measured reserve, to a depth of 1200 meters and with an associated surface area of 4.02 kWo; white the inferred reserve goes down to a maximum depth of 4,800 meters beyond the test dritting. 7 The National Energy BaLance cites an ainferred" potential for peat of 153 mtLion too. - 94 - Pa4 TABLA 4 Ngtimated Reserves og Shale oil and Gas (Million M3) 19. PETROBItS has carried out a more careful evaluation ....EA..0 J Z of the above-mentioned e deposits, concluding that the hW gas lt i 344 t,304 potential is at least in the $$4 i $ it,39 it# order of 2-3 billion barrels OureOW 198. of oil, most of which is at Irati. Of course, the economic viability of shale oil and gas production has yet to be established. Other Enera Resources 20. Brazil possesses a number of possibilities for non-conventional energy, notably solar, tidal and wind energy and urban wastes. Various bodies, including Universities and research centers, are conducting demonstration projects for solar energy, and CHESF is developing a 1000 kW unit with photovoltaic panels. Sao Paul* and CESP have concluded an agreement to prepare basic information on the scope for burning urban wastes in electric power plants; while ELETROSRAS has completed a preliminary inventory of possible tidal energy projects. What little information exists on wind energy is not promising, although research efforts are in progress: for example, CHESP has developed a 100 kW windmill, in collaboration with local research institutes and the financial assistance of West Germany. These sources of non-conventional energy are worth further investigation, but they could at best make only a marginal contribution to energy supplies before the next century, given the large size of the Brazilian energy market. TOE3 960BB 00METC FR0UCT10IN OP RD4RV 5NES7 80.U85 1970 1971 1972 1978 1974 1975 1978 1977 1978 1979 190 1961 192 198 1964 195 196 1987 tON0t00~e Ogn07 1056 1051 10461 10461 1187 11326 11408 11500 11016 12275 13530 15529 21464 25569 324~7 36930 37611 37826 Ptl#ete 009 183 6156 n29M 0646 O666M 644 . 02 8002 W262 9083 10675 ^2984 16598 2216 27493 28784 2403 Naturai 0as 1108 1074 1132 1076 1857 '142 1498 1649 1768 1732 2011 2257 2708 660 4471 49~6 5180 5272 Stea Cosi 599 641 644 50 615 729 984 1071 1224 1376 1463 1664 2155 2800 2605 2572 2443 233 etaIuroM Co~# 495 523 327 520 717 547 632 755 e29 905 973 713 716 737 021 M67 36 625 uranla 0 0 0 0 0 0 0 0 0 0 0 0 246 2269 1M74 992 360 1130 ReNAmL 8~0 7 47066 48840 51228 58175 86117 5165 60465 64896 66861 72494 7064 78762 U~3047 6976 96706 108954 103217 106559 Iedrasioset*Icl1 11542 12628 14697 167M8 19047 20968 24045 27109 29796 3~00 87363 87922 40928 43928 4312 51729 52902 53705 Ma3d 317M9 3109 32012 1520 32191 32739 31672 30409 29867 29014 30607 2N00 29127 29727 32677 32136 31772 30710 Suar Ca». Dcrlva6ive. 8566 S b.73 4221 4561 4585 4105 4662 6417 716 8067 9001 9955 11644 1506 16342 18569 1796 20221 Other R~1ewable. 221 280 296 306 344 858 406 463 552 005 998 1076 1146 1165 1377 1498 1747 148 TOTAL 57344 56691 61689 68656 67654 69468 71696 75896 76699 84169 91594 94291 104511 115445 181195 140884 140626 144362 114 s8MC 1970 1971 1972 1978 1974 1975 1976 1977 1978 1979 1980 1961 1962 19fl 1904 195 196 1987 #0-Båi m~ 8~ga 17.9 17.9 17.0 16.5 17.1 16.8 15.9 15.2 15.0 14.5 14.8 16.5 20.5 22.1 24.8 26.2 26.7 26.2 Potroleam 14.0 14.1 13.2 13.0 12.8 12.3 11.6 10.6 10.2 0.7 9.9 11.3 12.4 14.4 17.7 19.5 20.4 19.7 lcDral %9e 2.0 1.0 1.8 1.7 2.0 2.1 2.1 2.2 2.2 2.0 2.2 2.4 2.6 8.2 3.4 8.6 3.7 3.7 St~... ct 1.0 1.1 1.0 0.9 1.2 1.0 1.3 1.4 1.6 1.6 1.6 2.0 2.1 2.0 2,0 1.8 1.7 1.6 Matal urgient Q*$ 0.9 0.9 0.9 0.8 1.1 0.8 0.9 1.0 1.1 1.1 1.1 0.8 0.7 0.6 0.6 0.6 0.6 0.4 ~iraaluD 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.'0 0.0 0.0 2.7 2.0 1.0 0.7 0.3 0.6 R 0em E2.1 02.1 ab.0 .8 62.9 66.7 84.1 84.8 05.0 a5.g 05.2 83.5 79.5 77.9 75.2 73.6 73.3 73.8 fb#r~el9Icåty 20.1 21.8 28.6 26.4 28.2 30.2 63.4 85.7 37.9 39.9 40.8 40.2 Y9.2 3.1 36.6 36.7 87.6 7.3 Nod 88.4 64.0 51.9 49.5 47.6 47.1 48.6 40.1 37.3 35.2 83.4 81.6 27.9 25.7 24.9 22.8 22.6 21.3 Suger Ca* Derivatv~. 6.2 6.4 6.8 7.2 6.7 5.9 6.5 8.5 9.1 9.5 9.9 10.6 11.3 13.0 12.5 13.2 11.9 14.0 Other Rm~eab.e 0.4 0.4 0.5 0.5 0.6 0.5 0.6 0.6 0.7 0.9 1.1 1.1 1.1 1.0 1.0 1.1 1.2 1.3 7M7AL 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Source: KIllr7 of N.e~nd Era . ~*aimat Energy salanc. 1966. - 96 - 1 5 1 2:: :¡ 22 0 i *^ °° * E * ''v' 0.°@q°°° ° 1 -0 3 @048U e : wg°ijila o s!H0°0 00li@ fa,..°U U@fttq esg°I.igs §§§°jli1 ass llig så°illi - à i- - g w @q~o@~ qen 0 - . 1 a1 - 97 - Total Cnsutiono Dieset Oil (Nita J in m3) Tr~~~~~~,4 n... stoenxsanué aro éto - 1 54 0. 7 169 9 0 1 - 5.0 14 8 3 134$ 24 1.6 ø,4 2Z.4 . 16,3 - 3. 1.6 0.7 fre 23 4 171 36 1.7 0? SwC. N A seø orie lea, eot lca 4. tambusttve1 L a udos.AMutov Brstili, Fbr. 190. Autmotive Fuels in Transportation (Mo eion toS) 10e . Disl .oln lco Ott Cyt n ai ,,,,,,,,, ,,,,,.. Gas~.. •A A. Diea~sat0t 29. '47 7.9 . 88. 198 11,J 8 2,0 10.IL5. 1985 2.7 .9 5.6 11,. 19. 198 13.9 wi o.3 14.. 993 1911. 4 5 79. 940 39.1 -afmtios stt*, Febr -«Ot98C8.øb-(g AL Exnort and Consuut G~hasolne and Alcohol taBP.) ncn 192 18 the 1 e 56 Atth9 Cna 4 1/4 44 59 112 139 184 60 190 4400(¼Nporto 4 25 27 35 72 11 58 U8 '9 80W9: NS Ase**ri yøni¢a, PitU¢ø 4* C0b *t1 Liquid@* AtOmøtWOsÉ ANNEX I.8 Y.ar Total len lcohot Powered Vehicles 1980 980.3 240.7 24.6 27.? 8.6 198 ?Ø8.2 647.4 84.8 95.9 39.5 1986 , 866#9 699 0 80.6 92.0 37.8 1907 . 580.1 459.2 79.2 94.0 39,1 sources CENAL Nuay 1988 Notest a/ Percen~t -of total sales of Otto cycte vehi-ces (ALcdhot 6 Gale>. s/ of other vehfcles. - 98 - ANNEX 1.9 Page 1 of 8 ERY DEMAND FORECASTS AND METHODOLOGY back"rOund 1. Energy demand forecasts are central to the analysis of the energy sector, influencing investment, pricing and finances. Energy consumption is affected by income, prices and other socio-economic variables, The nature and. speed of economic growth determine to a large extent how energy demand grows. Energy consumption in Brazil is highly correlated with GDP growth, and GDP projections face considerable uncertainty. Total energy consumption grew at an average of 6.4% during 1970-79, while GDP grew at 7.6%. During 1980-88 energy consumption grew at 3% and GDP at 3.1%. The apparent income elasticity of demand from 1970 to 1979 was 0.85, increasing to one from 1980 to 1988. Given the uneven petformance in the past and the present macroeconomic problems, the outlook for the future continues to be uncertain. Thus, the energy forecasts should consider alternative macroeconomic scenarios and policies. 2. PETROBRAS' methodology prevents the simulation of demand according to alternative scenarios. The petroleum product models do not have the necessary variables to incorporate explicitly economic and policy variables. PETROBRAS' forecasts are in essence trend projections and educated guesses. Some of the models consider other variables besides time (such as population distribution, number of households, car fleets, etc.), but none of them is based on. a consistent analytical basis that would allow the direct consideration of alternative scenarios. The scenarios are taken into account indirectly, through educated guesses as to the likely future evolutior, of the independent variables. None of the models consider, for instance, price as an explanatory Mariable. 3. There are two problems with ELETROBRAS forecasts. First, there is only one demand scenario, which makes it difficult to deal with the extreme macroeconomic uncertainties confronting the Brazilian ecoromy. Secord, price is not considered, thus limiting the possibility of simulating .he market according to different tariff schedules, and different energy conservation goals. Forecast Aroach and Demand Model 4. In order to examine alternative macroeconomic and pricing policies, and supply constraints on a consistent methodological basis, we developed a series of demand models, predicated on the basis of a competitive market. For residential consumption, this means that energy demand depends on relative fuel prices and household income, while industrial demand is dependent on output and relative input prices. 5. We have selected the partial adjustment specification of demand, which states that demand in year t is explained by the value of all the explanatory variables in that year, plrs the lag-dependent variable (fuel consum-,tion in the previous year). The model formulation is double-log, so that the coefficients of the equation are the short-run elasticities of the exogenous variables. 6. Regression analysis was used to define the relationship between the dependent and independent variables. The base data correspond to time series at the national or system level for 1970'to 1988. The units for hydrocarbons - 99 - ANNEX 1.9 Page 2 of 8 and alcohol were in toe, GWh for )ower, and the income and price data in constant prices of 1988. Static and dynamic specifications were tested, selecting those that explained consumption better. In general, the chosen specifications were dynamic (i.e., the explanation of the level of consumption not only depends on current conditions but also on past decisions). Ordinary least-squares was generally employed along with two-stage least-squares for customers subject to tariff block pricing. 7. Given the existence of intez-fuel substitution, the forecast approach adopted went from the general to the specific: forecasts were made at different levels of aggregation, from global energy to specific fuels by sectors. Consistency was sought so as to avoid large deviations from the totals by specific fuels as compared to the global forecasts. Thus, the forecasts were developed in stages. In the first one, each fuel was forecasted separately. In the second phase, aggregate forecasts were developed, from total energy to total electricity consumption. In the third and final stage, both types of forecasts (disaggregated and aggregated) were compared and a final set of forecasts selected. This procedure permitted the incorporation of some of the inter-fuel substitution which is difficult to simulate. effectively in disaggregated models. 8. The adopted demand equations represent a compromise between econometric results and judgements. In the case of natural gas, where historical consumption is negligible and supply constrained, the regression results are meaningless. Thus, the adopted demand equations for natural gas are more judgmental, taking into account experience with elasticity values for other countries. In the case of fuel oil demand and industrial electricity demand, the income -measure used (industrial value added) was weighted by intensity factors of the different industries. Thus, demand equations were built for: o total commercial fuels; o total hydrocarbons and alcohol; o total petroleum products and alcohol; o total petroleum products; o total electricity; o specific fuels and sectors: - LPG; * gasoline; - kerosene; - diesel; - fuel oil; * alcohol; * natural gas; * residential electricity; * industrial electricity; .- commercial electricity, and - other uses of electricity. In the case of alcohol and natural gas, supply constraints were introduced. - 100 - ANNEX 1.2 Page 3 of 8 Exlanatory Variables 9. The income variables employed in the analysis were related to GDP and projacted accordingly, so as to simulate demand with different economic growth prospects. An average 5.2% annual growth rate of GDP is assumed from 1990 to 1995, growing 6% thereafter. Doriestic prices for petroleum products were pegged to the World Bank forecast of the international price of oil, assumed to increase (in 1988 prices) from US$15.33/bbl in 1988 to US$16.65 in 1995, and US$21.23 in the year 2000. Electricity tariffs are assumed to reach LRMC levels (in one scenario) by 1995. 10. As described in Chapters III and IV, two scenarios were developed on the basis of alternative pricing policies: the "Business as Usual" (BAU) and the "Efficient Pricing Scenario" (EPS). Both forecasts restrict the supply of alcohol to the present capacity of 16 million cubic meters per year, and of natural gas to 17 million e?/day, the estimated attainable level by year 2000, assuming no imports. Elasticity Results 11. An analysis of energy price and income elasticities was carried out. The short-run and long-run price and income elasticities of demand are presented in Table 1 below. The analysis showed inelastic demands for price changes for most fuels, except for gasoline, alcohol, fuel oil and industrial electricity. In spite of this, the long-run price elasticities are relatively significant for most fuels. 12. The results of the analysis suggest large inter-fuel substitution. The price elasticities decline with the level of aggregation: energy demand is less elastic as a whole than for any of the particular fuels because of substitution. The only exception is electricity, due to the large share of industrial consumption in the total and the use of different income measures for global electricity consumption, residential, industrial and commercial demands. The demands for kerosene, diesel, natural gas, and residential and commercial electricity are not very sensitive to price variations, indicating moderate effects on demand and significant revenue effects as prices are modified'. 13. Residential electricity demand is, as expected, more inelastic than the other demands. Once the effect of new users is taken into account explicitly, expressing residential demand in per user (household) terms, the long-run price elasticity becomes more significant and the income elasticity declines. Industrial power demand in Brazil is, as expected, relatively more elastic, indicating a rapid and effective response of industry to relative price changes. This elasticity is largely due to the ad hoc demand management programs implemented in the electricity sector. The two historic demand management programs were: the "tarifa hora-sazonal", and electroth9rmal fuel substitution. The first program, designed principally to reduce peak load, helped to achieve a system load factor higher than 70%. The second program, based on a concessional tariff (EGTD), expanded demand rapidly from 0.7 TWh in 1982 to 12 TWh in 1985 (i.e., 13% of total industrial demand). 1 These elasticities are preliminary. Further work is needed on cross-price elasticities. - 101 - ANNEX 1.9 Table 1: Page 4 of 8 t:ice and Income Elasticities of Demand for Fuels Price Income Adjusted Income enerqy Demand Form Short- run gr Short-run Loa-M Measure Total energy -0.04 -0.16 0.21 0.75 0.99 GOP Hydrocarbons & alcohol -0.12 *0.36 0.32 0.96 0.98 Disposable Income Total oil products -0.15 '0.82 0.15 0.79 0.97 Disposable Income LPG -0.10 -0.44 0.24 .1.06 0,98 Disposable Income asoln - Disposable Income Kerosene - 0.25 .3 0.92 0.94 Disposable Income DieseL -.09 .0.24 0.50 1.4 t10 Disposable Income Fue ofLri.eu, de -.0 -.7 .0 .29 _0.96 .,Vqfghtedn.vaeadd A3coho5 .13 0atural gas0.91 0posable Income tectricty 0.26 .08 1. MP sdential -0.15 t.6 1.00 Disposable Income esidentit per use-0.50 0.10 . 0.90 Disposable Income p rustrial.0 .00 Weighted d. value added C ommerefatl -0.03 -0.26 0.07 0.71 1.00 Comercial value added 01.. 0.96 Diposable Income MaXket Review and Forecasts 14. T h e Table 2 relative importance of Rates of Groyth and 622arent Income fuelwood. in energy Elasticities of gonsumption consumption declined from 41% ini 1970 to 11 i 187 TeRates of GrgMit b ) Elasticities 11 n 97. TeEnergy Sourc2l 1970-80 180- 1970 1980-87 share of electricity, Fuetwood -2.7 -2. -0.3 -0.9 on the other hand, .Charcoal 6 0.3 5.7 1.4 1.9 increased from 16% to SUgar cane bagasser 8.0 9.9 1.0' 3.3 34%. Te aprn Other ..bio-fucts . 18.0 9.4 2.4. 3.1 * h paetElectricity 12.4 6.6 1.6 2.2 income elasticity for O 05 products 8.6 0.2 1.1 0.1 firewood demand was 0c.hot 18.4 20.1 2.4 6.7 negative for 1970-80 Natural gas -28.8 18.5 3.8 6.2 an.1980-87 (-0.3 and other energy 11.3 9.9 1.5 3.3 4Total 6.4 3.6 0.8 1.2 -0.9 respectively). Total without fuetwood 10.1 4.7 1.3 1.6 For electricity, the values were 1.6 and 2.2, respectively. We expect these trends to continue in the future. 15. The share of sugar cane i, mainly due to the rapid growth in the demand forDalcohol. The high income elasticity of alcohol is explained by the subsidies and fiscal incentives to alcohol vehicles. Consumption-of petroleum products grew at 8.6% per annum from 1970 to 1980, but almost stopped growing during the 1980s. Although part of this decline was due to the slowdown in economic activity, the main reason was the policy of substituting imported petroleum with a: .ohol, hydroelectricity and natural gas. Thus, the apparent income elasticity of demand for petroleum products dropped from 1.1 during 1970- 80, to 0.1 in the 1980s. - 102 - ANN!X1. 9 Page 5 of 8 100% 90% 80% 0% 1970 1974 1979 19e2 198 M Power of products rF-Irewood * came Ingase M Other Fuels Charcoa AICOhOI Natural Gas FISure 1: Structure of Consumption by Energy Forms (1970-1987) 16. The evolution of the demands for petroleum products; petroleum products, natural gas and alcohol; and all commercial fuels are presented in Figs. 2 to 5. As the level of aggregation increases from (Fig. 2 to Figs. 3 and 4) the outer surface of the curve becomes smoother, due to inter-fuel substitution, i.e., the total demand for commercial fuels varies from year to year less than specific fuels, because of substitution. 17. The demand for petroleum products increased at an average rate of 0.3% per annum during 1980 to 1988, while total secondary commercial energy demand grew annually at 3.8%. Petroleum products demand grew at 8.6% p.a. from 1970 to 1980, and about 0.2% from 1980 to -1988. However, this performance was rather uneyen: there was rapid growth during the 1970s (19.2% in 1973), followed by sharp declines during the 1980s (for instance, the drop of 7.4% in 1983). This uneven- performance was largely due to the uneven performance of the economy2, and the policy of substituting imported with domestic fuels. The decline in gasoline and fuel oil consumption after the oil crisis and the ensuing substitution policies of the 1980s, was partly compensated by ineeases in alcohol and electricity demand. The correlation coeffIcient between petroteum products demand and GP is close to 0.9. - 103 - ANNEX 1.9 Page 6 of 8 90 Forecast 80 70Historical 60 S50- '30 20 1970 1975 1980 1985 1990 1995 2000 Efficient Pricing Scenario LPG Gasoline 9 Kerosene Diesel Fuel Oil ligure 2: Oil Products Demand (1970-2000) 18. The graphs also present forecasts up to the year 2000, assuming efficient fuel prices (see Chapters 3 and 4). The demand for oil products is forecasted to increase at 3% per anum from 1990 to 1995, accelerating to 4.7% if rational fuel pricing policies are in place, while total commercial fuel demand grows at 3.7% and 4.6% p.a., respectively. In 1988, petroleum products and electric power accounted for 43.8% and 48.7% of commercial secondary energy, respectively. These shares change by the year 2000 to 46% and 43%, respectively. Figure 5 shows the evolution of the structure of commercial fuel demand for different years. - 104 ANNEX 1.9 Page 7 of 8 120 Forecast 100 Historical /80 C 5 60- 40 20- 1970 1975 1980 1985 1990 1995 2000 Efficient Pricing Scenario LPG Gasoline + Alcohol [ Kerosene Diesel Fuel Oil @ Natural gas Figure 3: Petroleum Products, Natural Gas and Alcohol Demand (1970-2000) 250- 200- 1100- 1970 1975 1980 1985 1990 1995 2000 Erfadent Pricing scenaro 08 Products M Mcohol NM NG alericily Figure 4: Total Secondary Energy Demand (1970-2000) - 105 - ANNEX 1.9 Page 8 of 8 Box 709- 60%. 509- 30%- 1970 .1975 980 1985 1990 i9i 2000 Effcien* PrItng Senro * o Proucts g Acohot MG NeGtrtcity Figure 5: Structure of Commercial Secondary Energy Demand (1970-2000) -106- ;:4-~, 'i¶ ..~... . .. w& -D ølw m, . . . m g . -.h.l I 4N . . .. - g . . * . ø -~*L~tk9 £07- Å t '2 4 04 4 04*wa 'ea4Ø4<0~ .......j 4W ..... .....49~S$&.~ .44 4 . - - - -- * ~ 108 - POMER SYSTEM PLANNINIG, 1XPANION AND INVSTMNT 1. BLETROBRAS has the primary responsibility for the coordination of planning in the power sector, which includes the preparation of expansion plans for generation and transmission systems. Distribution planning is the responsibility of each individual utility but the participation of ELETROBRAS in the global assessment of the adequacy of the level of investments of the individual utilities ensures a consistent approach. BLETROBRAS is sbject to the supervision of DNER in the preparation of its plans and the plans themselves are subject to the formal approval of MMR. Since 1981, SIST within MOF has approved the sector's budget to ensure consistency with the level of public sector investment spending (see Chapter II). ELETROBRAS has developed a comprehensive set of planning models for the preparation of development plans to expand system generation and transmission capacity in the long-term (20 to 30 years), medium- term (10 years) and short-term (about five years). The long-term plans are revised every five years, while the medium- and short-term plans are revised annually. Given the size of the Brazilian power market, the large number of power utilities, the considerable time that it takes to implement large hydroelectric projects and their capital intensity, the power planning process in Brazil has evolved into a flexible mechanism that produces long-run plans which are compatible with frequent adjustments in the short and medium terms. The planning methodology is considered to be one of the best in Latin America, encompassing models that adequately represent and simulate the characteristics of the existing system, along with the candidates for expansion. 2. The sector ileneration expansion program is formulated mainly utilizing a linear-programming model, DESELP (Determination of the Long-Term Power System Expansion), developed by ELTROBRAS, utilizing an IBM package (UPSX/370, and complemented by various more detailed simulation models. The objective is to determine the optimal composition, sizing and sequence of power generating plaut additions to meet system demand requirements at the lowest economic cost (the present value of the associated investments as well as operational, maintenance and fuel costs are minimized) for given standards of reliability. The bulk of the generation expansion plan is based on the installation of hydroelectric power plants (mainly schemes with multi-annual regulation) complemented by a reduced amount of thermal-based generating plants. Several generating schemes -- for which information is available at different levels of precision concerning inventory, feasibility, basic and detailed design studies -- are analyzed through simulations using the above model and complementary models and the optimal sequence is selected from all feasible combinations of available generating schemes. The simulation models specifically consider individual power scheme characteristics, such as capital and O&M costs, capacity, firm/secondary energy, as well as power system data, such as system reliability levels and hourly system load curves for the different years under analysis. The level of system reliability is based on the determination of the probability of system power deficits and operational planning assumes that the probability of having an annual energy shortage is below 5% of projected demand in a given year. In fact, it should be noted here that Plano 2010 changes the risk of deficit from 3 to 5 percent, without identifying the consequences of this important change in one of the planning parameters. - 109 - 3. The sector transsission expansion program is also formulated in the context of long- , medium- and short-term scenarios, taking into consideration the market forecast, generation expansion programs, existing system characteristics, load dispatch and regional power interchanges. For the long- term, only basic characteristics -- such as voltage levels, capacities and tentative commissioning dates of large transmission links and substations -- are determined through system simulations, using load flow, short-circuit and automatic network expansion models. The results of these simulations provide adequate background expansion plans. For the medium-term, detailed analyses are undertaken in order to define the basic characteristics of all transmission lines and substations -- including conductors and power transformer sizing, reactive compensation and specific commissioning dates. In addition to the models utilized in long-term planning, system stability and economic optimization models are used. The short-term planning -- which provides the basis for decision making relating to the construction and commissioning schedules of works -- is determined using the above models, although the data are based on more detailed information. All the configuration parameters of the system are well defined and they provide the necessary elements associated with the physical sector expansion programs and the corresponding details of both the investment and financial plans. 4. Distribution planning is carried out by each of the sector utilities, following certain specific technical standards which are included under sector legislation. The distribution expansion programs are formulated for long-term (10 years), medium-term (5 years) and short-term (1-2 years) scenarios. Distribution planning using traditional methodologies is made by the utilities to determine the required high voltage facilities (69-34.5-24-13.8-6.9 kV) such as distribution substations and primary lines. Only three utilities, LIGHT, COPEL and CBMIG also carry out detailed planning for secondary networks (440, 220 and 110 V). In general, distribution system performance is monitored by the utilities through social aid topological parameters (population, income, development, education, location) and technical indicators (voltages, voltage drop, flicker, losses, reliability, loading, short-circuit levels). The frequency and accuracy of such indicators depend on the level of sophistication of each individual utility. Based on the utilities' load forecasts and on the above parameters, different expansion alternatives are technically and economically analyzed. The least-cost schemes are then selected, provided that they meet the prescribed standards regarding load and voltage profiles, as well as system losses, voltage drops and reliability levels. 5. The methodologies utilized by the sector for planning the expansion of its generation facilities are in line with economically accepted concepts of long-run least cost power system expansion plans which optimize sizing and commissioning dates of future plant additions. The nuclear and coal-based power generating programs, which we consider a deviation from the least-cost expansion plan, are discussed further below, under the sector investment plan. However, the treatment of reliability could be improved by evaluating alternative reliability levels in terms of the value to the economy of unserved energy. 6. The analyses made for planning transmission works are comprehensive and the methodologies are satisfactory, although an attempt should be made to utili-e probabilistic methodologies and due consideration needs to be given to the evaluation of the cost of power outages. In general, the selection of the transmission line configurations (voltage level, structures, conductor sizes, - 110 - span lengths and routing) is based on economic comparison with other possible configurations. In each particular case, the alternative is selected with the lowest present value at a discount rate of 11%, assumed to be the opportunity cost of capital in Brazil. Furthermore, substation capacities (power transformers, capacitors, reactors) and their lay-outs are established to minimize investment and op6cational costs over the facility's useful life, taking due consideration of both substation load requirements and applicable technical standards currently being used in the sector. 7. Planning of the utilities' subtransmission and distribution systems is based on the application of reasonable optimization methodologies, although again consideration ought to be given to the use of probabilistic approaches. When reasonable alternatives exist for power system expansion (e.g. new supply points for the distribution network), the utilities seek to select the least-cost alternative at a discount rate of 11%. Generally, however, no reasonable alternatives exist, because the works involve expansion of existing installations, which should use predefined sites and rights of way. In these cases, the sizing and other characteristics of the works are defined in accordance with standard configurations which represent an economic solution. Nevertheless, there may be cases where reasonable alternatives do exist which are not considered, such as local generation, supply from other primary circuits, phased coastruction, or deferring investment. Such alternatives may not be considered because they involve different reliability levelsa the current distribution planning methodology predetermines the reliability level for the alternatives to be considered, rather than allowing reliability to be another variable in the optimization process. At the aggregate level, presently neither BLETROBRAS nor the Government have satisfactory ways to assess the adequacy of expansion plans for power distribution. At the power utility level, the distribution companies need to revise their planning and design criteria in the light of economic considerations and financial constraints, and should improve their methodologies to define priorities in power distribution investments. To address the aforementioned problems, distribution planning studies at the sector aggregate level as well as studies aimed at improving the methodology for expansion planning and design of the individual utilities power systems are being carried out with Bank support (Loan 2565-BR). 8. Since the mid-1970s, the Brazilian power authorities have made significant progress towards implementing environmental and social safeguards in the power sector. Notably, they have taken initiatives to mitigate the harmful side effects of large hydroelectric projects on river basin ecosystems and human populations. By 1985, ELETROBRAS and various regional utilities had created environmental units, carried out studies, and in 1986 ELETROBRs prepared a "Manual for Studies of Environmental Effects of Electric Systems" (the Manual), which has since served as the basis for environmental studies in the various stages of planning and implementation of electric systems. The studies proposed in this Nanual also provided the basis for preparing Environmental Impact Reports (RIMAs), which have been, since 1982, a fundamental legal requirement for licensing the construction and operation of electric systems (power generation plants and transmission lines), by relevant Federal and State environmental agencies. While this RM&I became an effective first step for sector-wide progress in environmental management, LETROBRAS and the regional utilities soon recognized the need to follow it up in more depth, including social questions. - 111 - 9. Since 1986, ELETROBRAS has prepared the Environmental Master Plan (SMP) for the Power Sector in connection with Loan 2720-BR. The EMP consists of sections ons (i) national environmental policy, legislation, regulations and guidelines; (ii) specific guidelines to improve environmental planning and operation on the part of the power sector, including criteria to assess environmental costs and benefits rulated to the construction of power projects and the operation of the system; (iii) preparation of environmental and social action plans on a project-by-project basis; and (iv) measures to strengthen the institutional capabilities of the sector for implementing the EMP. Together with the gna and the general Federal guidelines for the preparation of RIMAs, the EMP provides guidelines for the treatment of environmental, resettlement, and tribal matters and specific project-by-project social and environmental action programs. The EMP is a dynamic performance evaluation of the power sector and is periodically updated and re-issued. Results of the implementation of the RMP can be measured primarily in terms oft (i) the environmental, resettlement and Indian assistance programs, which have been carried out since 1986 by ELETROBRAS and the power sector utilities in compliance with its guidelines; (ii) improvements in the investment selection process; and (iii) institution building. 10. Recently, ELETROBRAS started to develop a methodology to integrate environmental effects, both ecological and socio-economic, more directly into the basic project analysis through an Environmental Index (BI)I. The El attempts to take into account, for each alternative, the physical and biological effects, e.g erosion, impact on local fauna, and tropical forest coverage; and the socio- economic effects, e.g. population displacement, infrastructure requirements caused by the project, and changes in local economic activity. The index increases as the environmental costs increase. The El is then considered jointly with the standard cost-benefit measures, so that ELETROBRAS can at least eliminate projects which, for little or no gain in standard cost-benefit terms, involve significant increases in I. TN POWER SECTOR XPANSION PROGRAM Plano 2010 11. The qector's current expansion program for 1990-1995 was prepared on the basis of the long-term expansion plan, Plano 2010, a comprehensive planning study completed in early 1988 and elaborated in accordance with the planning methodologies described above. Before considering the current investment program, it is therefore necessary to analyse briefly the principal features of PlUa 2010. 12. Plano.2010 is essentially based on developing hydroelectric power in the expansion of generating capacity. As seen in Annex 1.2, Brazil is relatively rich in hydroelectric resources. Five basins which are being considered ares o the Parana, which is the dominant southeastern basin (with more than half of the existing hydro capacity) and which includes Itaipu; a the southern basins, the most important being the Uruguay and Iguacu; a the Sao Francisco basin, the main basin to supply the northeast; 'See R. A. Coo, aReview of Nethodo. *i used for Expansion Plaming of the Brazilian System Generation and Transagasion, Deceber 1989. - 112 - o the Tocantins-Araguaia, on the border with Amazonia and sharing Lto potential with the northeast and the central-west systems; and o the Atlantic eastern basins, comprised of a series of relatively small rivers (the most important being the Paraiba do Sul, Doce, and Jequitinonha in the eoutheast-centerweet power pool). With the development of the hydroelectric resources of the Amazon, three more important basins could be added -- the Xingu, Tapajos, and Madeira -- roughly as followss o development of Volta Grande in the lower Xingu, which includes a hydroelectric complex of 17 GW (Cararao/Selo Monte and Babaquara); o a large 9 GW project on the Tapajoe-Itaituba basinj 0 various hydroelectric developments on the Madeira, for a total 12 GW capacity; and o other small projects. 13. The hydroelectric share of projected installed capacity in the reference case of Plano 2010 (see Fig 1) stays above 90% until the year 2000, after which it declines only slowly, to reach 88.6% of total capacity by the year 2010. The additional hydroslectric capacity after the turn of the century comes principally from the Amazon region, with the developed share of economic hydro- potential rising sharply from 17% to 78% in just ten years, implying a rate of growth of 16.6% p.a. (see Table 1). Installed Capacity 1985 - 2010 .. .. .. . .. . . . . . EM&!v EMas Ma aw Mo 14. A number of comments can be offered on the hydroelectric solution embedded in Plano 2010. The transmission of large blocks of power from the northern to the southern part of the country, representing a magnitude similar to Itaipu every five years, would be a major technological challenge and will constrain LETROBRAS' ability to develop the Amazon, aside from environmental - 113 - considerations (see below). There are other hydroelectric resources that should be studied which might offer a more manageable transition, slowing the rapid pace in the development of the resources of the Amazon region, among which the hydro potential of the Araguaia-Tocantins basin has been noted in the literature as promising2. In this context, there is a need to update Brazil's catalogue of hydroelectric projects. The only available comprehensive hydroelectric study is still the CANANBRA study, financed by UNDP in the 1970s. The new study needs to examine in depth the resources of the Amazonia, to update the available technical, economic and environmental knowledge. Tals A pevlopment of Economic Hydro Potential: 1995-2010 (GW of Firm Energy) Uth 54 6.1 17.1 28.2 KrEst 7.8 4~ 7.3 7.4 7.7 tadth. spht/Cento 5.t 0.9 21. 25.6 26.7 28.1 foath 149 6. 9.6 11.6 13.1 15. The development of large-scale projects in the Amazon presents serious environmental risks and ELSTROBRAS will not find it easy to reduce the environmental costs to an acceptable level, despite the significant strides forward that it has made in environmental analysis. The adopted hydroelectric solution is also weak in terms of the limited flexibility that it gives to the power system compared with other alternatives. The construction of smaller thermal projects, for example, would provide flexibility: (i) to respond to unexpected changes in the rate of growth in demand; (ii) to take advantage of the possible emergence of new technologies or the improvement of existing onee and (iii) to gain more experience with the environmental effects of hydroelectric construction in the Amazon. The inclusion of additional criteria in long-term system planning, to reflect environmental effects, to credit flexibility and to penalize risk, might postpone developments in the Xingu and Tapajos basins until the Tocantins basin has been fully developed. In this way, development would be consolidated in an area which has already experienced growth in the last twenty years rather than developing a whole new region. 16. Brazil's main alternatives to the development of the Amazonian hydroelectric potential ares nuclear powerl gas (including imports, e.g. from Argentina and Bolivia); oil (including shale oil); domestic and imported coal; international hydroelectric schemes (e.g with Argentina); renewable energy sources, such as bagasse, solar, wind, tidal and mini-hydro; and energy 2 Alan 0. Poolo and Jos6 Roberto Moreira, "Atds do atternativa miclear so piano 20100 (Sao PauLo, grazit: Instituto do ELetrotdnica e Energia do USP, 1988). - 114 - conservation (especially through pricing) .3 Plano 2010 found these alternatives to be generally unattractive. Nevertheless, there are options which are worth exploringi and the thermal options which Plag IflQ did include, namely nuclear power (for strategic reasons) and a limited expansion of plant using domestic coal, do not appear to be economically attractive. 17. Plang 2Q10 gives very limited attention to the use of natural gas. It considers steam turbine plants with high capital costs operating in base mode, concluding that 4.5 million cubic meters of natural gas per day would be needed in 1995 to supply 1,000 MR. However, considering combined cycle, it would seem that this limit could be boosted to at least 1,500 NW.4 Similarly, bagasse co- generation is described favorably in the plan, but none is projected, due partly to institutional problems (see Chapter 11) and because the estimated potential is too small to change the strategic panorama. While this may be true, the additional use of bagasse to "firm up" some hydroelectric power should be explored further. 18. Other hydroelectric options are available to Brazil through international co-operation. These could be attractive, even adjusting for the risks associated with more external energy dependence.5 For example, the Roncador project, a large binational project of 5,400 DN, is not presently considered in the Plano 2010. This project should be examined to determine its operational and cost characteristics. It is located on the border between Brazil and Argentina, in the Uruguay basin, and is interesting to both countries. The hydrology is quite distinct from that of the Parana basin, which dominates Brazil's hydroelectric capacity. A second interesting binational project is on the border with Bolivia, the 11,000 MR plant on the Madeira river which is even less studied than Roncador. A third binational project that has been studied on and off by Argentina and Brazil for several years is the interconnection of their systems. The characteristics of both systems are complementary. Argentina has a power system about 25% that of Brazil, where thermal generation is proportionately larger than that of Brazil and their load curves are compatible. Perhaps a two-way exchange of power of about 3,000 MW could take place, the direction of which would depend on the season and the hydrological conditions of the year. 19. Finally, energy conservation and demand management could be integrated more effectively as an alternative to additions to generating capacity. The criterion in Plano 2010 is that energy conservation options are considered only if their cost is below one-third the cost of marginal generation expansion cost, which seems to be restrictive.6 Energy conservation offers not 3The potential from domestic supply resources is discussed in Annex 1.2. uibject to the available sUPPly and LNC of natural gas. See Chapter IV. Additional studies could be made within Brsit to ssign weights explicitty to risk and energy Indepwdence, since experience in these techniques exists elseshere. 0See It. A. Com, Op.c..Sa - 115 - only a low-cost and environmentally-sound alternative to investment in generating plant, it reduces the risks related to demand growth. The Sector's Current xoansion Proaram 20. Although the sector'a current investment program for the period 1990- 1995 is broadly based on the precepts of Plano 2010, in the sense that it preserves the optimized construction sequence of power generating plants and associated transmission grids established in that document, it has been adjusted in some key respects by LETRDBRAS and by the sector utilities. The latest version takes into account the sector's financial capability to execute the investment program, based on actual and expected sources of funding; prioritizes the implementation of transmission and distribution projects, so as to re- establish the reliability levels of power supply and transfer the energy generated by new power plants to the centers of consumption; and restrine expansion plans for large electricity-intensive industrial users under subsidized tariffa.8 The underlying demand forecasts are those of ELTROBSM shown in Annex 111.1 21. The commissioning dates of several power generating schemes (such as Porto Primavera and Taquaracu of CBSP, Segredo of COPEL, Nova Ponte of CEMIG, Serra da Mesa of FURNS, Ita of ELTROSUL, Candiota of CEE) have been delayed for periods ranging between two and four years as a result of both the lower electricity consumption patterns which are expected during the coming years and the financial constraints imposed upon the power sector by the adverse macroeconomic situation. Furthermore, no plans exist to construct further nuclear stations beyond Angra. Details of the power generating program are included in Annex III.4. The sector expansion program is adequate, except for the program of coal-fired plants --which is limited to the installation of less than 5% of the total capacity being considered, to provide system peaking and base capability as well as energy for expected shortages during dry season periode--has also been delayed. We continue to urge the Brazilian authorities to analyze in more detail alternative sources of generation, such as natural gas and biomass (asgar cane) within the least-cost expansion context. 22. During the late 1970s and early 1980s, the power sector allocated a large portion of its financial resources to deelop power generating projects, which accounted for as much as 65% of the total-investment. Consequently, the financing required to implement associated transmission and distribution works was curtailed beyond prudent levels. The decline in real investment in the mid- 1980s was due to the completion of the major works in Itaipu and the financial crisis of the sector. Consequently, transmission and distribution projects suffered, and the quality of supply was reduced. As a result, the risk of energy shortages -- estimated to have a 15% to 20% probability of resulting in energy shortages equivalent to 10% of projected demand by 1992 and 1993 - exceeded the 7 The screening criterion for energy conservation smasures should of course, take into account the fact that effective demnd managemnt (DN) would probably reduce LANC throu; the contractian in demd. in other words, for ON to mke sense econmaicat ty, the investor has to recoup its investmnt in conservation, and this can onty be done whean the ex-post LANC is higher than (or at least equal to) the average kWh saved with conservation. a Currently, these consumrs *- mainty the aluinm and steet industries ** account for about S of the total consumptin and benefit fram subsidies equivalent to about 40% of the marginal cost tariff tevel. - 116 - level of St traditionally used for sector planning. Under the Power Sector Loan (Loan 2720-8R), the sector undertook a commitment to balance the investment program and the proposed 1990-1994 program reflects more realistically the sector requirements through the following six% generation, 53; transmission, 23%; distribution, 16%; and general investments, 8. 23. Nevertheless, the risk of energy shortages, mainly within the South and South-Bast systems, will continue to rise in the coming years and exceed the 5% target level, because of the delays in commissioning several power schemes and the historical imbalance in transmission and distribution investment. Under the Power Sector Loan (2720-BR), BLBTROSRAS agreed to implement measures to minimise the risk of energy shortages and improve the quality of power supply. Some measures to restrain demand growth have already been implemented and includes (1) discontinuing, as of December 31, 1986 subsidized industrial tariffs for the promotion of oil substitution ("Bletrotermia loads"); and (ii) implementing the long-run marginal cost (LRMC) tariff structure for large industrial consumers (currently representing about 50% of the sector electricity consumption). Quality of power supply is partially being improved through rehabilitation of distribution networks, installation of capacitor/reactor banks and improved public lighting fixtures. in addition to measures already taken, the regional power utilities will need to: (i) produce additional electricity using existing oil-fired thermal plants during dry periods; (ii) accelerate construction of transmission and distribution networks; and (iii) implement the National 3lectricity Conservation Program (PROCEL) in a timely fashion. Further corrective action may include selective curtailments of electricity supply. The levels of reliability in the North/North-East systems are within acceptable limits. Detailed risk analyses for the different regions are included in Annex MU.S. Phsical Works 24. In line with the assumptions and premises described earlier, the power sector expansion program necessary to meet the energy requirements expected by ELETRORAS through 2000, as shown in Annex III.1, includes the commissioning of about 34,600 MW of power generating capacity, of which about 30,850 or 89% is hydroelectric; 1,260 or 4% is coal firedl and 2,490 MW or 7% is nuclear. The transmission and substation expansion programs through 1998 and the distribution program through 1991 are detailed in Annexes 11.5 and 111.6 respectively. Invesaten 25. The estimated cost of the power sector investment program during 1990-1994, corresponding to the physical works described above, amounts to about US$39 billion, excluding investments required to complete Itaipu (US$840 million) (see Table 3) but including Angra i and III (US$3.3 billion), which are being constructed by FURNAS. The financial liability of the sector for Angra II and III is limited to an amount equal to the cost of the equivalent hydroelectric plant which would be next in sequence under Plano 2010: the balance is financed by the Federal Government. It is estimated that the sector will continue to invest after 1994 amounts equivalent to US$6.0 billion/year to keep pace with demand requirements. Details of these estimates are included in Annex III7. - 117 - Planned Tranamission and Distribution Works Tenitaion 45980408 3 AB d247 ircuttm 230 kV 13,468 efrcu-10 1G kV 19,409 o$rouitiae Retorf an~d~ Casp.citora 35,000 NVAR Lemth9=mit ~1939.93» NWIN 107,57 km 11,398 Its Power Sector Investment Proaram. 199.0-1294 ... 0n enerat... ..21 .. 53, otnmtf--4 «¾- 26. The above estimates allow for the acceleration of construction of projets underway which have been delayed for the past four to tive years because of Inadequate funding. The estimates are optimistic, as the sector's ability to generate internal funds is impeded by the low level of tariffa. They were prepared by the sector utilities' planning and construction staff and they are based both on preliminary and detailed design and comprehensive bill of materials and equipment as well as detalled evaluation of the corresponding civil works and construction costa. - 118 - a¶j~~ i,r~ars PLANnEC . S 4 ott1954 4.5 ros CEN MCUP t8> 2 . tt13258 84 5300 10 Nydro ILhTNTS sua .0f~ 1 '1. anN Øt. 3MW01 169 Ti* pant to be constructed try gentInW and - rtls . The tyt y, KW.tT 電 H ド りD 季 ら 120 L.V.LIMØ OW No øf units C~Ity (KVA) .... . . .... tri ø . . . . . . . . . . . . . . . . . . . . . . . . 4 .... .... . . . ~A W Ik ram re No af units 6430 MS løg C~ Ity 19 øf units U4 254 C~ Ity (HVA) 117 li? 113 - 121 - 199 1991 12 1995 1994 . 2m2.9 3?45,7 4522 49l 4094, 4tibtinf07.24 1*3. 1289.0 1294.3 15M, toA1 ecs6tss.s 7559.4 es24.7 as.4 *sE56,o . .82. 24. .9. ' .....-............................. u30.0 516.2 3922.5 421.2 420S.1 Tom ¿Ecic 4$58, ?5$94 $524.7 ø84# ~ø450. rwp 48.7 41.8 46,49. 94. of Jae 1989. 122 - _ __.' 's a S.... .~ .* •mw.regene can - 3,7 3,9 4,340 4,63 4,945 5,23 5,568 .f'sk LOWI m2) o 0.0 0.4 0.9 2.1 5.8 11.0 9.9 6>8 M 0.0 0.5 1.6 3.7 11.1 21.1 20.7 .e, .. - 16,60 17,.0 16,58 19,447 30,463 14.274 3.4 tot Lam.t: m. 0.1 0.4 1.2 2.8 7.0 12.4 10.5 9) nitvty quIrement aO (> " 16,996 18,068 19,215 20,191 21,440 22,584 2,j17 Risak < 2 0.2 0.8 2.2 5.4 13.2 23.2 *1.2 -a~ cmW ta o> 1.274 1,405 1,614 1,763 1,961 2,198 2,48 R41k-*UV@l <2> L O 0.0 0.0 0.1 1.1 0.? 1.1 8> uenitlvity seöisnwta SIW>" 1,338 1,604 2,17 2,331 R,646 2,rO L.09 * -fia elwe t.) 0,0 0.0 0.8 3.9 13.3 7.9 11.5 * 4 lnUts aU> * 3,185 3,413 3,631 3,979 4,344 4,70 5,00: .4Uk L;nt > O1 0,0 0.0 0.2 1.7 1.0 .0 - 8> senttlvty * -1gW.Wdg, (g> " 3,372 3,701 4,06 4,397 4,719 -5,049 5,381: . fkLeet ,> 0.1· 0.2 0.8 4.0 12.5 7.6 1l * O as corr*esponds to nry equrents frlhd ln PIUr 2010 (Jne 1M). f bq-o*¶MUmdnl taysrly Iverge d 9f1 1 % 19~ M 198 191$ 194 = 19 18 I981 47 3et StW 1s1 SI 4M 420 4 4 982 ner d 144m 12 12 M 7 1719 u 1240 s 144 10 <,n«v n ~ctr U% 405 26 S 221 i 4 25 3 1 2 486 5M . tota nowherr.ø nd% sfl1Ø 134 9F22 285 .519 '3CØ -d9ø -40 .41 2.5f h. borrosinl 2424 aM 3815 -4.52 1 . -9.6 124. -32. -19. 2167 -U2G Tot'at Bsrc 5616· 6385 6767 6149 6523 6197 4227 6131 3791 6775 4186 .anment.. . I u¥M 4195 450 483 4077 412 4555 366 348 3781 392 5083 b. Loa to il p 59 739- 698 783 97 1106 706 950 887 698 780 c. Total Inves.t 4754 3300 5251 4842 3111 5661 4402 4198 4d8 4190 583 6. OUher 932 1073 1486 1286 1413 537 -175 19534 -878 588 -1208 Totut eplicatdmna 5686 4386 6745 6148 1524 6198 .4227 6132 3790 4778 4455 Debt-service soverug ) 2.09- 1.67 1.49 0.92 1.11 0.4 o.5 .37 0.34 0.41 0.44 Self-financin rati. €> 23.5 17.7 21.5 .4.5. 5.9. -9.6 -22.3 -77.4 -111.5 -101.1 .88.0 nsunr-basd finano <> 3.9 46.8 48.8 23.8 39 17.6 11.2 -36. . -80.2 -71.5 -64.4 0errutnlinàstent51.0 55.7 -75.6 84.8 72. 65,0 25.~9 150.7 89.9 56.5. .31.1 o:. -Val M h: h unrtud. to S win pri e Iee . tnd- hachags r.te of 197. nteril, cak~ Gnraefcm• Total tevenm mnu ~ toalf opeimtwi (b>. . Total intørest numrtfaatfor at tomns bf dometic and:forle ~gii BatumuLlbcsLs C> b> te» i (e) te) (d). .fau ,uimu .l9øst f hersin Fud -pous Ap or ln (the ter consthe poser Bstor 5 OCntribution dias to the very Md*64MSWUB CndtIón f 1 Ii N¶I,nttrnis>. plus..Sol. Tox, receivd by the.. poer ue¢toi witltles Ø4%nt tots hitoe ahe reuaining.481 conmtg. (of -t ~-da f direØt titW the h-tIe. th) . . . from amastic idu forignp sme This total 'AIes- hBuse..de Srmil - •HP IIICIh are sham as part ftme hoversmet 43ø .:#tavtsftatI uensmn< l.e. enorptfwn, - ransefsetmn. .*eg ·Ø . ..eteaatuwq · t 7datig apta rgareene ndalnexOea  a~t 9 Bq , g9 us; au3 *m JO U p -11os q talk I qS m> NUI&S .usts uo>fla sim4 älä*uuoo voaaJsaa lai ro unjon atp AiMa1,3a Aun 0 * s-. l al p-epu e u . 43gi u3el ogajwun 0,p44. auspib aSaS eqa st tien se unS3dansuo tlmpIli MuOm~ ~ ~ i*l! ~1 wOlM WIMM niq4s 1'41A ~ -IaMW 61k $9364 MIMSJ AI.4C Auo k5D0 ta*pp .sn pS ' 03040 .t aw.iuup %bs ek~ mujl" anuioe owe gacussc3axiän Jaid at asi siag lo =aas - su sU*fJ3 lCui6Jeans1fl ao %"vi~a nutejud sA lo 343 Ut hat..~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~ma ~uan mik mjq~ WIO h ~ ~m>jmh-~w ~. J ··~ fuglA Buyteles Mulff,~In IRIN me851p -kul41u)e a e."pno Folk jwfflåkJ-Jazul boa o3 quolan3u~ 50ux UMM u ase u~saepu; n.simq Wuba e aah *fd •eml~wwq isusiba a Ounrs^a sim -a uofss qs ai s;djn pm oz sasd '9 .gP 4314c pndla PUReau~ 1~en uuo up P *q3 o uDa namiano ~pa pu vauuus ::0fgIqSI P UIW Mb 01 sufIWtqI349W M S ptm Aup #o Un3ado ut j e*s u»saapluom eau; a:p; oi SMOi,å $a PU SIllvnk i Agd- N IWEu t44 a i~ ~ m sv soui~d -ag ps Måilant3n Da Aq 8 loUu~ 19301 -:.. . pn.o st c m ai ja 39=~p.s e jo mag ams ' b Dun ujlop s u n Aq =sopetn t pm sophhnuI impg6n K'Asm k ~i ,j~c iq csapmn3 smgc io u*¥~3a fw.>ja~ AJ p..· dm ea- JMdusM » on eva- sud Ie ' queiw o mG:3 pap-A M - -li ftu.•it.i,-», CqUdiO-i tenAs n u. sv44 sj0i ni 06-må6 p>;ad mapt Ginp 903|935 ru W96-J66 uSis PUgR pungeAIæn 1s-gog9 tbs ~hk -8p1^s Aolmd MIt Ik[~e w~4 PO961.41£61 gut~p CM> pu'å auuwiffl IMP11S 95 01% 996 1 tatt 999t ·019492 tin9 £96u u9st9E 90mm I 0969 68601 1091 £m119 999 sidpeä 31% 09961 lSEgI .499E 01919E £1693 £1596E. 191W 9001Z 09 6601 £01911 £l 9 11999 53 "~d LOt- 6665- UUI• 900- 065%- 19559- u691.- 9%I- 90W 9611- 6649- 950E- ff61- wu an 921 Sit 6E902 lOpt 1o9er929'10E 69992 909s2 690,1 61% 919 0269 £216 56602 Uadpeml M5St 1U v=S om- o95 m g~» ^0 £1M 011s man MS a u«9;d~ 4 -41A nun 8 94.98 90M£ om9 550 M9 0519 1865 1449E ew6 aum fao. L98 609- 9116<- *99- 9695- 691£- 90105- ^ito!- 91161- 041£- U361- 99LEL- 9~u8 am fl- 99· · £U£ 69609 110E0 14065 9S99 999t i0o9» 99969 66E59 £8021 £911? S3dsmis1 12-1- 519 981 ·1 2EB 0% IA! 6E0 95O9 M 1 om1L 295% 62911 1211 61011 $9% 69991 29601 6*0911 9S0*% *119e 099E 99100 56E091 U1as 59059 mEit 06151 69015 uWU Sm 15 - -09% OMI6 UP9 tU! 9M net 6611 am 66 4559 f1im Sut 1909! 901 061 68605, 10451 01111 1992 659191 29E1! 292991 1519 99611 9462 91265 £5619 saus4d 0^11~)US lix 4U- 00E9- 9om£~ 9099- iE9St 1990- 061- 69uE- 159- ull- EOG- 951m- 1L- 9EM6- aeq 3n 9w6 00 EES 999E m£91q. 51t 91%9 ELS 6119 Ml9 £5161 0091E 9949 1EM11 sad ;nu 61; 002 9011 9ME 6 Ptt f%%• 1a SU 9001 992£ I99EL 99051 £665 569£ Sue~ 6 D9OMt - 91991- 2~1E1- z99501- 90090- l155- I5r 96*2- 156- £E111- U1S agn ft 901 9626 .m9>91 . 9185 1111% 0Z9 1m64 f 095Z U5161 00 11511 sadsCo IS, 0 0 9 59 0 0 f1 Owl 1911 3191 99 18 L V1 951 1 SSSd urW r c f-W -or Nr r I-m -9r Mw -nw u w w053 (~21, >0 ffi111MIN OlYMIM OW81I dog-ft #am839 SOmTI d0 miaiVlNoID3LU333y:jn Anv~ ;111flX3O - 127 - PATROLEUM SECTOR DEVELOPMENT PLANS, INVSTUMNT AND STRATEGY PTROLBUM SECTOR ACTION PLAN IF PASP I 1. In November, 1988, PETROSRAS issued an Action Plan, the "Plano do Acao do Setor Petroleo" (PASP), which laid out the broad lines of a tentative development program to the year 1997. The strategic targets for the company were to reach a domestic production of crude oil of about 1.48 million bpd by 1997 to meet a consumption forecast of 1.54 million bpd, i.e. to attain virtual national self-sufficiency in oil. In parallel, natural gas production would rise to 70 million m3/day. Much of the production increase, for both oil and gas, would come from the Campos basin, with the Marlim oil field playing a particularly important role. These highly ambitious production target. would be accompanied by a program to reduce the costs of petroleum transport, notably by more extensive use of shipping and pipelines to substitute for road and rail. PASP required an investmerit program which would increase from US$ 2.5 billion in 1988 to US$ 4.1 billion in 1991 (an average of US$ 3.4 billion per year). Within a few months of its publication, PASP had become obsolete as the Government introduced yet another of its anti-inflationary programs, the "Summer Plan". PETROBRAS suffered drastic cutbacks in its investment and operating budget and was forced to rethink its short- to medium-term strategy. PLANGAS 2. At the same time that PETROBRAS issued PASP, an extensive analysis and development plan for the natural gas subsector was completed, under the auspices of the CNE. The plan, known as PLANGAS, had two development scenarios. Uftder the Basic scenario, natural gas would be produced from known fields and delivered through an expanded pipeline network to the principal markets. Production would increase by 8.5% per year and reach a level of 36.8 million m3/day in 1997. Under the High Growth scenario, production from known fields would be expanded and gas would also be produced from fields which have not yet been discovered. Production would increase I5% annually and reach 71.0 million s3/day in 1997. In both cases, some of the gas would be reinjected; some used for field operations; and a declining portion would be lost. Heavier hydrocarbons, such as ethane, LPG and condersate, would also be extracted. Some of the natural gas would replace town gas but would go to new industrial and domestic consumers. 3. To meet the goals of the Basic Scenario, the expanded system would requires o the attachment of 600,000 new domestic, commercial and industrial consumers by 1991 and 1,500,000 new customers by 1995; o installation of natural gas processing plants to produce 800,000 tons/year of LPG and 200,000 tons/year of ethane in 1995 and 1,500,000 tons/year of LPG and 700,000 tons/year of ethane by 1995; o conversion of 10,000 vehicles, such as urban and intercity buses, to operate on compressed natural gas rather than diesel fuel or gasoline. - 128 - Natural gas consumption for power generation would increase to 3.9 million O/day by 1995. If the gas is used for baseload thermal plants, this quantity of gas could fuel 700 - 900 MW. Use in steel manufacturing would be 2 million O/day and consumption as a raw material for fertilizer and petrochemical manufacturing would increase modestly. These are ambitious goals and would require very heavy capital investment in an already strained capital market. The same problems affecting PASP have rendered PLANGAS obsolete, although, even if the resources for investment were available, the High Production Scenario seems unrealistic. 4. Consequently, PETROBRAS has for all practical purposes been without a clearly-articulated investment program since early 1988. Hence, we have considered tentative alternative scenarios of petroleum development and estimated the associated investment requirements. Since a number of the basic parameters and assumptions employed are drawn from an evaluation of recent experience in the subsector, it is appropriate to start with a review of past efforts. PAst xf forts 5. Brazil's domestic production of crude oil has grown at a rate of about 13% per year for 1980-88 (see Annex IV.3). Furthermore, the cost per barrel of finding and developing oil reserves (including the oil equivalent of non-associated gas reserves) has declined steadily over the last 10 years (Annex IV.3). However, there was a marked slowdown in the growth of domestic crude production after 1985. While refinery capacity has been sufficient to handle the increasing supply of crude (Annex IV.4), there was a need to implement projects to improve the yield of diesel oil relative to fuel oil (Annex IV.5). In terms of expenditures, exploration, development and refining constituted about 90 to 96% of PETROBRAS' total investment program in the period 1980-1988 (Annex IV.6). 6. The increase in natural gas reserves has varied annually between 7,200 million mS and 15,300 million m3 from 1980 to 1988. Associated gas reserves are larger than non-associated and associated gas production has grown faster than non-associated gas production in absolute terms, due to the rapid development of the Campos Basin oilfields (Annex IV.7). Non-associated gas production increased at an annual rate of about 20% after 1980. By the end of 1986, gas reserves were almost equally divided between offshore and onshore. Gas production reached about 6.7 billionm3 for the year 1988, out of which 84% was associated with oil production. While the reserves-to-production ratio for associated gas reached twelve yeers by the end of 1986, closely following that of oil production, the same ratio for non- associated gas was still at a high of 44 years. This means that there are non-associated gas reserves which could be developed to supply an additional 2,000 million m3 a year over and above the 1,057 million m* produced in 1i88. Such a possibility is taken into account in 'the gas production forecast prepared below. On the other hand, the gas-to-oil ratio for associated gas has been steadily declining over the period 1980-1987, indicating that the Campos Basin oilfields (which weigh heavily in the overall production) have generally lower ratios than other fields. This trend has proved to be useful in forecasting associated gas production in the period 1989-2000. The pattern of natural gas utilization is discussed in Chapter IV. - 129 - 7. In light of the analysis of these past efforts we have projected the domestic supply of crude oil and natural gas to the year 2000, using the following parameters: o rate of growth of domestic crude supply, 13%/year; o minimum reserves-to-production ratio (for oil), 10 years; o gas-to-oil ratio for associated gas, 150 m /mPg o growth of non-associated gas production, 20% a yearl o minimum reserves-to-production ratio (for non-associated gas), 20 years; o availability of gas to non-PETROBRAS users, 41.5% of total supply; o refinery investments, as required to process total crude supply; and o investments in exploration, development and refining, 90% of total PETROBRAS investments. 8. Three scenarios were developed to represent poasible levels of investment in exploration/development and the corresponding investments in refining and infrastructures o Scenario AA, keeping exploration and development investments at same level as in 1988 (about US$ 1.5 billion per year); o Scenario B, raising exploration and development investments to 150% of the level achieved in 1988 (i.e. US$ 2.3 billion/year); and o Scenario CC, increasing exploration and development investments to 200% of the level achieved in 1988 (i.e. to about US$ 3 billion/year). The resulting projections of crude oil and natural gas supply are summarized in Table 1 and further details appear in Annexes IV.8 and IV.9. C-ouarison with PAS? 9. While PASP is already outdated, it is instructive to compare its results with Table 1. A direct comparison of all the main projections is not, in fact, possible because PASP is not articulated in sufficient detail, but some general remarks can be made. PZTROBRAS was anticipating an average annual investment of about US$3.4 billion through 1991 (para. 1). If PVTROBRAS continued that level of investment over 1992-2000, it would be roughly analogous to Scenario CC. Using 1997 for comparison -- the last year for which projections are available from PASP -- gives the comparative results shown in Table 2. - 130 - 10. In broad termu, the projection of doestic crude oll supply under Scenario CC is quite close to that of PASP. The principal dif ferences arias in the estimnte for the total production of natural gas and in the ratio of available supply to production. The supply figure in the CC Scenario (42 million m3/d) may be less than that of PETROIMS because PASP may have assumed that larger amounta of gas would be discovered by PETROEhS and possibly that gas reserves would be developed at a faster rate than in the past. The figure in the CC Scenario (22 million m3/d) for available oupply appears to be consistent with PASP if 5 million m3/d per day are added to the latter to account for the Shell- Pecten field. However, for deand forecasting purposes, it would be prudent to anticipate a maximum available gas supply to non-PETROBRAS users of 17 million 3/day by 2000, corresponding to Scenario Aa (i.e. with investments maintainad at the 1988 level). ................ . . . . . ...... ;~ASSMa.Y (NIL. Cal> 19 27 27 v,PAV IOPMY (i 8 20 23 amd.y turet omh Iquids bG tua ttl-Peoten gas l (0) Zretds 5ge 000 /dby natura gas iquids (0 "Avuftable g s spty" is faturet es 6upp(y mvaitet to an-PETR~BS Uwes. - 131 - ..44 katurvt IOs Seply Ave able (Miiion fod> 22 16 Availability of S~ply (Z Pr~Oductin) 52 23 iffF the EP(UO 9'cmnrifo, edjusted to Includk nenenergy petrotam pro6 (b) Correspnds to PETRORAS A.t Scmenr~o for 1997, inoiuding oeergy petrotwu. procs: onsut ian. (c) This snbr has been rprted by PiTCBRAS Undmr fte A.2 Scenr o for the yw 200 n aw y not inctude Shelt-Petten production. Inveatment Strateoles 11. Evidently, the investmente which will be required in the development of domestic crude ol and natural gas and in refining will be a function of the demand for petroleum products and, notably, the pricing and taxation policies followed by the Government. These policies are considered in Chapter IV. For the present, tour broad strategic options are considered, according to the demand management policies in effect and according to the quantum of investment resources allowed to PETROBRAS to develop crude ol and natural gas supply. For this analysis, the EPS(U) demand scenario was selected rather than EPS(L), to distinguish the changes due solely to petroleum product pricing and to abstract from the effect of the alcohol program. The four options are as follows: Option #1. Highest demand and minimum inveetmento, i.e. a combination of Scenario AA and the BAU demand Scenario. Option #2. Lowest demand and maximum investments, i.e. Scenario CC and the EPS(U) demand Scenario. Option #3. Highest demand and maximum investments, i.e. Scenario CC and the BAU demand Scenario. Option #4. Lowest demand and minimum investmente, i.e. Scenario Ah and the EPS(U> demand Scenario. Annex IV.9 draws togetlier the projections of domestic crude oil supply under the three development scenarios considered (AA,BB and CC) and the high and low projections of demand for petroleum products under the tvo demand scenarios formulated, i.e., BAU and EPS (U). The results obtained from the analysis of the four options for 1989-2000, assuming the World Bank projections of crude oul prices will hold in the future (Table IV-3), are shown in detail in Annexes IV. 10 to IV.13 and summarized in Table 3, which leads to two important lessons. -132- 12. Iesson 1 is that the policy option of importing crude oil rather than relying on PETR08MAS to carry out the required investments to develop domestic petroleum resources would lead to some resource savings if PETROBRAS could continue to import crude oil at the prices projected by the World Bank to the year 2000. The relevant comparisons ares a for the high-demand growth pathu, between option #1 (minimum investment and hence greater emphasis on imports) and Option #3 (maximum investments and hence greater emphasis on self-sufficiency in crude oil production); and, o For the low-demand growth paths, between option #4 (minimum investment and hence greater emphasis on imports) and Option #2 (maximum investments and hence greater - emphasis on self- sufficiency). 13. The total and annual costs are about 5% higher; and the average ex- refinery costs per barrel of incremental oil about 6% higher for the Option representing greater self-sufficiency. These are not large differentials in costs given the uncertainty inherent in estimates of oil reserves and production. We have also estimated the ex-refinery opportunity cost of a barrel of incremental oil, assuming it is imported instead of produced locally (Table 3). While slightly lower (3-6%) than the cost of a barrel produced locally, the difference is not significant, for the same reasons. However, the following two qualifications should be noted. First, crude oil prices have increased in real terams since mid-1988, which has a greater impact on Options #1 and #4 than on Options #2 and #3. Under Option #2, a 20% increase in the import price of crude leads to only a 5.5% increase in the total cost, due to the relative importance of domestic supply. A much larger change in the cost (10.8%) occurs under Scenario #1, where domestic supply has a smaller role in the total supply. second, the larger ex-refinery costs per barrel for Options #2 and #3 reflect the fact that a large volume of production which would be developed in the 1989-2000 period would not yield benefits until after the year 2000. To illustrate the order of magnitude of this effect, the average cost per barrel of domestic crude oil under Options #3 and #4 falls by 10% if the forecast is extended through to the year 2009 and no additional investment occurs after 2000. 14. Taking these qualifications into account, we conclude that, given the uncertainty in the cost and production estimates, the policy of producing domestic crude oil is as economic as importing; while it has the advantage of greater self sufficiency and provides some protection against higher oil price movements than projected. However, we must emphasize strongly that our estimates of the cost of domestic oil production are preliminary and need to be refined; and that PETROBRAS must make every effort to reduce the cost of domestic crude to be able to remain competitive in the face of wide variations in the international oil price. The upgrading of PVTROBRAS' costing system, currently under way, will provide valuable additional information on the cost of domestic crude production and on the ways and means to achieve and retain competitiveness against imported crude. 15. Lesson Toe is that resource savings can be obtained from more effective demand management, through efficient pricing. The comparisons in this case are: - 133 - ..ÉL . .I L .L - .il ...... iI DEN/LO IV. LO DEN/NI INV. #1DNN N. -O-EA (V S8,369 100 58984 t00 41,024 t00 56,510 100 -ddrae10 F259 18 10 259 f? 1029 ? 1,29 1 t159 28 32,537 5 2,50 53K AM .... . .. . . ANNUA C'T.94$952 45 ,9 - agtdQ 11 X-,E 11.Må #ote: Totat east are in present vetu terms, with a 42 dicount factor nd factid fwstments fu reffnorfe and associated Infrastructure, as welt as the Invstument cost of developing dmtilc crude aspty; ~n the operating costs associatud with ofitfilds, Infrastructure, refineries- and the purchase of iffortud crudo. Annual coste are expressed in annuftles, o For the minimum Investment scenarios, between Option #1 (higheot demand and no change in petroleum product pricing policy) and Option #4 (lowest demand and efficient pricing); and o For the maximum investment scenarios, between Option #3 (highest demand and no change in petroleum product pricing policy) and Option #2 (lowest demand and efficient pricing). Total and annual costa would fall by 4%; and the annual import bill by 7t-13% with the implementation of an efficient structure and level for petroleum product prices. As stated In para. 11, the EPS(U) dumand scenario was used In order to distlnguish the efficiency galns due sotely to petroteu product pricing. Additionat resource savings would accrue from reducIng atcohot cons~uption, as under EPS(L). - 134 - bN IV.2i PETAOLEM MLA OIUE £NLL&CM AfhELUt Pcsete CM.a 31.4 36.5 s86.1 M. 46.3 390.4 43.1 434.4 445.3 .tqprts -26.2 10.1 5.3 11.2 s . 13.1 15.9. 18.9 3.4 . .EpRts .3 29.8 39.5 43.7 67.2 61.7 48 54 56.9 kt :ortPrmt C3 -t190 .2 32. 61.9, 48.6 S5.$.. cam~wten 40ä 37 373 .3s 350 365 403 .41 41a Covsintiu Ö.lpsM) lom 1030 1022 9m 3 959 .1000 li 1137 113.9 *4~nst PuTiluuA 196~AmmS RIost. (Years 1980·19, and 9UTROGME 198I Almnui Rep,rt. 8of. de M dIM Aotnutt-.::#aøs actiade de paioSfSa Ce* 9 .ss. AE IJ.3: RESERVS EVWLUTICS MND COST av SBL FCUIlD AND DEVELOPED 016 163l1s m3 . 198 198 . g98 193 94. 18 9& ‡7 18 off UUU8rve» VS • 321 s $m, 40 .4M -IIUionLt * 132$ •148 1717 169 2019 2170 34. 2516 2818 *fÉ$rU$'66. 7 95 M3 ¶49 m-9 2M 201- 20, MUio -185 ¶34 - 15.. 9 . 8 316 3.2 . 27 1.. - onsreProd. . 4 4.6 49. 7. 84.6 9 94 9 S RIo buo9. 15 i2 11: '11-14 au.r MI .I.t9 50- 9 · 34 .356 c~ v^:otg Ag fl 247 40W 1715 14 g.2 stes ~ ~ ~ ~ ~~~., Y-4s »sh oa ta Rtos Yas (s #~ettM - 135 - ANMIV.4Lt REINtNO CAPACITY ØsrstIsgDy - 1446 l46S 1544 1565. 1546 1510 1409 1411 MÅa Mtlfan ml 40.9 40.3 415 39.7 42.6 42.3 432 44.3 44.3 mittan j4a 2M, 2w 21 25- 26B ad 2n 279 2 i int 2t. 30.9 I8. 18.7 19.6 19.9 2. 21.5 Mu 4 M..I L 42.t UJ. 6r 68.6 s n %t i 14 la *M MuuBs .,..? SS3 380 $71 3 400 424 432 må Isouc.; PEltaMS Atql Import 19iBt MEX IV.I: YIELD OF SFINIES (X) . A . . . . . x . . ..... ..... . .. ..~½ * x '~'~ ' W k zii « ccf.nd 8æ(W £4 37 25 2r 80 DA m a A lt asrve Groth R1tto 11.6 .3.9 13. 17.8 A mA HA HA WÆO Rue'm F;M,t Peo¢utIn - 2 2 2.67T 3,02 4,013 9.08 . 466 5,686 5,786 6,869 % s t r: 7a. 8 M9 8 -- 1,15 offshore) .7980 ·t,2† 1,40 1,999 2,804 3,336 3,586 3,869 wA srva qra*ntr tit ts,05 i,$85 7,¶188 14,30m a,78m -tS,29 10,196 igalimnC 67 1 e *. 48 95 58.- 10 68 0M Asu~f~tad las (Vuf gas fOfl . 16 1 176 169- 169 151 148 146 148 MA R/P Affnthed 17 - 18 13· 12 12 M NÅ 14P HC-AIl:tWe. .. 75 7 53 39 37 45 44 -A A motet RiP v O~wves-tPtøntmnttatio In r fl - Sms'te-oi ratto. a 'tImted fraP#fie tsetit J980 an:PÉMBSA AnnutSspott of 1988. -- HaBarrl o: ofi euØnat. (100* t.f:·ga. u d.5?8oø.. .. -137- &MAUxL. CUE OIL -A ^ MTURAL.AS UMY £~ ?Q§CTtO scnr A Øpidetin/ %weopen ,twostUents sm as in 1988 .. abtRste ø1 #0LNa f Ott -n 2700 U*e 3/Vear et N4on-se. Gau utAG) .MB 121:...92..1U99L .. . 1997. 9. 19L MI 31 289 M 25 294 2980 2997 3014 3031 304& 3065 308 $10 12? 2t1 2 269 30 9 300 304 30 300 300 00 301 $4a KM. S "0 74 T2 820 20 20 82 82 82 2 20 li R/ 513 i 1 0 lo 1 10- 10 1o 10 to 10 1.14 lM. 21.3 24.7 28.4 31.3 35. 33.2 33. 31V.1 35.8 364 %.6 Aser/" 18,3 19,1 . 19.74 22.4 24.4 2J 2, 25. 27.1 271 27,1 27.6 r80C Mo$ 156 1J 1. 9 930 10.7 193 19J0 19J 195 19 94 198 1989 1990 1991 1992 1993 1994 1999 1996 1997 1998 1999 2000 2 88 301 322133% 3528 3633 3738 3f,3 3938 4013 4078 4133 418 P 21!' 238 -269 302 341 370 370 380 380 400 410 420 420' 1190 8 640 700 ?70 850 930 1029 1100 1100 1100 1100 RI'13 13 12 11 11 10 10 10 10 10 10 10 10 BA8. P. INla/D 18.1 19.1 21.3 24.7 28.4 31.3 33.2 33.2 33.5 34.8 31.8 36.4 366 ASOc. 15.3 1-.6 17.7 19.7 22.4 24.4 25 25 25 26.3 27 27.6 276 f~ Thø~t a db~* 58 pe f Otw 9..9L...t9 99 9 199 .9 . 99 . 9 9 21 50 29 302 34 38 435" 48 490 500. 51; 520 ~530 MiPD -580. 640 700 7?0 850 930 1020 1130 1240 1360 1500 - - m/p· 13 16 14 14 14 13 13 10 10 10 10 to 10 Bps im. Estu/0 · .. 18.1 19.1 21.9 24. 28..4 -32.6 37.4 41.1 41.6 42.6 42.1 46.1 .4 ABsso. 15.3 15.6 17.7 197 22.4 25.4 28.6 31.5 32.2 32.7 $3 34 34.6 9' u Production 0tBS ilPD- u Thousani barrtls per day mso/ iI t m3 per day. - 138 - ANNEflJV.9: PETROLEIM BLANCE AND GAS SUPPLY FORECAST ~onario M 238 269 300 300 300 300 300 300 30 300 300 300 Smnrio fl 138 29 302 341 390 370 380 - 400 410 430 4M0 MriCC 238 269 302 341 3M 435 480 490 500 510 320 30 .Pd (Nes> 343 5 37$ 389 406 424 442. 465 490 516 543 571 76 79 0 86 90 94 99 103 108 113 110 124 TötI 419 47 456 47 416 518 341 568 598 69 60 -0 9~atr(oA) -18 148 156 175 196 218 241 268 29M 39 341 35 ~usfro 103 103 148 144 i81 199 210 44 27 301 331 3. . 6on1o =3 181 14 146 123 111 i9 13 164 1i n 91 208 2 4$ nr CC 191 163 146 123 96 64 38 53 72 91 108 135 1.uvli LA» 1&. 19.1 21.9 34.7 25.7 26.9 27.1 27.1 27.1 27.1 27.1 27.1 ~ 9nori m ( ,S> 18.1 19.1 219 4.7 28.4 31.3 U.2 33.2 33.5 34.8 35.8 34.4 Seurt. .18.3 19.? 21.9 24.7 28.4 2.6 37.4 41.1 41.8. 4,6 43.1 44,1 - 139 - ANNEXJf¥.10: guaim At iA < . f ø.'c . ..~ ~:< t6 <>4W.. %~.............. 4 SLM z,Wtøøa>2 9n en4t mo 1a91, -It Udo IW4*1*:**a (TotaL 8a1*u ø/986 '»1> CMat atm o Tta ($ISNPUt' PUBU Cak 7793114 NIS (Ia . .i 0 . tit oQ,dtjc - -n d 397 UDI Nililmn7 OtilÈS'\C O. 29 -05 16 1 177 166 143.9 -13 43 4 .ota$fa(lgl' s~ u 6 US N d ( Ison OW.l momt uS 102 * tu is o mqbt.*arud 53291 US i ti 10813 8 $8 flM r (g .~g . EIIr- a . 1 or;i t EUI aet 25Purflsgu.tBIwsOt. i . ra%sshfds.adItfmntuspilæe in paoacin cas.efstto attasinm a ie MftISdflq I4 th0t tOIfllWo TI PET11RA9 toPnaste -A ~.tek.kuSa øta abut10 9 xpoe* y,bwflØU.- ¿ . -340- AL.IE ZJ1±' PEf0LEttI BALANE PIRECAST AUB C0ST (OPTIOhI 2l et 1960 c 9 d fiten åtteßt 2t 'N Y Ied . 4 åS *~~~ ...... .~ .........~3~#~s~~ . . . . .... ....................... . .......... . Ekft$evK··427 -2 4<.' hrogho tj2e to2004 tota j920 34248¾~f'~*. 2» 30 M0 550 4W0 35 3 2 200 201 M. Ottrk 31 3.5 346 35 341 3C5 3I, 315 3M 305 -' 3;.29 M8 a totot (19890 20) t 41934.USS MiCiotn (u~0 MILt ear) 0 jot (1989-M00O Rattig Costa: 66124.00 Uft wilan ut of. *fob kont Cu 19133 U #i Utton Atotdwth ader a to~ ^ Uss. KLfn -O Cr~ om129Nite;anåtc T.ith ~....5.......ittn tgp0g acmtd f4n7 2812%.lwaqhute laro 40 oas8/3:34 Ie imry1al dtd CrIdNI. -10 trg enet 4r Mde: 2009 WoSm a Tiai 3 134 354 54B340 .55'30 14 04-5 05 93 .15 0 al(199 l ~ ~'af b 4193mt4 of .N * Ttal(199400>Cpra t rqostLeaty 6612 UD .66/88 ,ry v ats aeW prad oma r e.tonga puf9d-sinna d~ -141- 611EX 3V1L2' PETROLEIII SALANCE FCRMCAST AllS CT 199 1990 1991 1992 1993 194 1995 1996 1997 1998 1999 2000 ....... . ...... .e . ... ... n .a .......... lmprts 151 16 154 u 111 81 61 77 98 119 141 168 o Total (1989 to 2000> investmnts : 41875 »SBBMNtll.n (U43589 U(tlimnTar> o Totat <19~92000) ~ur l-cost3 69470 uS ttion Gat.f Ldttch.,aæI Cn 4700t Us Milion o Olmmngtad CÒftgt (-13 ftaaMit) . Least Crues'.4F.USt it(øf (Old.Cnldu. Uls10259 Mfillo ianorranuul Ciré: .g 5Was0 wiuøø>e 'amortt 186 (U ~lt jin . o Disautd. CasØa:i2IlOtscosati- leust Cnidé .34 MBBLI (014 Criu: 169? Nøptip lIMMsat CnkI 1369 0186 > Toteh aI 17. WMLu¡. * Disouted EnV WefIPy Gst-o e u arr oft -*,nflqØ%.4 CrisiiPraile44qQy tys. Al85.41/im. lncreSuuV Cfuds IqpCred< : B it. 2/. .. - 142 - S0EU'IV.13 9IO~=L BAtC 95CM CMS (£PMnM 43 . * .. W Imm etutnmstiLp>$(El *maurfe & .SmrrMl ... .. lm i. Am- I..M m- .,ag t c . 7. . ~w . 4f5i0 4i0 m80 .55 o 400 350 u sam i a~ ffib 20- rser - 45 3 205 a13 19 1s in r7n i. t- i 151 143' .TestiL <1999 to <0*0 léwetmet rf22494 lUNlImitn (U0$3074NfllHan/Tber> o Total (190P2i00> *weratinggC's.s 745f8 dSS Hill le. Cue of ddk .Lat Cruden..52967 OSS alt tIan ·Aonal raImoi 2 x.48.95 mtun <4 rud: UlS 10259 c1~.1n; Incromnut Credm Unc e~0 ci- a Mn tat: 4uSi.MSIO Hittlon .. . . , . o $¶%40 CIIaeu Cqalnt 12 GiCcut> --, " .. - --- ... *:- .':,ý ,:-:ý': ý -': ý. ... -......,... -. - - -, --, --- -::----::,ý . .- : --::. . ' .. ---- ----- -: .... ý:, ý .... ý._ý~ý~ - . ý ...... ' - > .. . --;:- ,,...lý .ý~~..'-- - -.' ., - --- ,.: ----- .ý ... - ,"-zý::-:ý,,l.":.:.11.1---I.I: -:ý:;-ý'.I.-----.:. , :....:,:ý:,::ýý, -;' .... '-.:' ,ý.,:::ý:'ý:,:,ý","ý,ýý.'ý.", » -: ->ý-'.'- -.........."ý-~ ~:... -.., -, ... -ýý---., .... .. ''::. w ;ý;.ý ... ,.: .. .: . - I . . .. - 11 :. ,;: -: ': ... .. *:.,....:,.,.ý,:,.""..::..":*..,...,"." - ." ', ...,, .., - - ':. -..... -', ... ' .. ' .: .. ' : 11.1. .:" , - ,- :- I.,.", . :ý- ,.ý: - I ' .. .. --,- ,..: -.ý -.-. .. .... ..... :. ...... -::. -..." .:,:.. I.. -::- .. ý,- -:,ý-.",ý< - , -ýý . - ý -- - :. ,,,. -----." .....,:,. -,,. I :. " . , - . . ' - ' - .. .:,- - .... - - ' -::.:ý:ý.., ..., ý:-: --ý...:,,. .:.. ,,:.ý t ' ':.-.., ...,. .. ' .. ý ,....., .. . -ý6ia.:":-,,-' ,,... Z.,...,,...:;' : ,.,:,.,.:.-.:"" : _.. ' .. . .;::- . .:.:,:.,.,;.. ,,...ý '.ý. ý1.1 ,,:",' <,, . - ' 'I - ' - -"ý.:: ....... I .. ýi- ~ .........:,. I . ý,.... ... - --- - . - -- : --,,-----.' ý. ---.- .. I . ... .. -".^,.:,;:ýý:*:,ý".*", :ý...:, -- ' ' ' ' ý." - - : :- -"':::ý - ".^, ý::'ý-...- ~1. .. ~ ...'<, ... -ý::ý, ' -.., ' ' . - ." --...:", - -, .",:: ": 1ý-" . - . :,,, -ý:: ý '., .. - ~1 1, '- ------... .. , , kýý,:. .. : ý, - - ", ' I . . . . - - ' " -:" .«;"..:.-., -:ý:-. .; . - - -,ý: .. ý -': -..:- - ..,.. -... . ... ' ::- ,-:-ý.:-7 I -- . .ý ...5...ý ... I.... I.I. - --.. -- :-.'ý,:,, .,:. "I:,ý..,:"ý,,,., ".-.ý..:,.:., ý - . . . I:,,...,ý .:7::. .." ..... . .:ý:. ' . -:-ý,".,;:" .-.., : - ":::.,;.*..:":...ý,:""..".:::. ' I. . .' .. . .:.,..--..:,..,...ý--, .. ::"'.::ý.1:"l': -ý1,1:- ...: ---.:«"...::::. . . ' .;*.": -: ,". ... -, ' : " ,',' , ý.. - ý ". ý " ,!.. .. - - ... ': . . ..... ,",:......,...,."'. ý1. . .ý ,ý -': ..,. ' ý- - :,:,.:., . , ,,.. ý. --- ";- ...,:"'.ý,,,,,,. .",", - ý.1,1 - -, --ý .,.... ..,..:...:" :*. -----,.'- ..,ý.....:.,j'ýý:--.".:- --...,...1 .. ..:.,..:ý,:,*:., -;'. ' - - -':'ý': ..: _ ....::, . . . :":.:.:.... ..:,.ý":::.:ý.,ý ...:, .9. .,*.",,,..... ..- A. '-tti« ý _N. ,,..."~,i. . ý *.,*.",*,.": «.....".,..^-".,ý-":-,.. --,-- -,,,, .....,..'.« ý ": - ..,::..,:..,,' ý-::, - ..ý,..., .: ,:-,. , .., I - :l:".E... -,,Y:ý:-:2.-:ý:.. .. ... . ::. - - ý. .. . . . ' I ': -:':'.. z:.ý,:""::,ý":,:::::: ýý'.:-::-:- : .::. .' --...,:: ... ý. .. ... . :.... ...... :.'-:,-;ý,:":.....,.,,,ý-ý,ý.ý.,,.....-:"..",' ... -!..ý-,,,..'- ,: .. . ý I. . ':.ý~....: .. -;.,,,..ý::,ýý, ...ý::" - ::»:.,...,.ý-,;.,.. : ý',-,-ý', ,":::..:: ,,:,:: - . --ý:'"::ý:ý.ý,>:- . ,:,:.,.,-:,.."' ' - ' - ... - - .,:.:. ' :::'.,:.'..:«. :. .:..:..: ' - .. - ,,,,ý:", :,....,",:..." ,,.. -. - 1,*,..,,....,..ý, :,:.::",.' .. .... , -- - . .. ,.: --.2:ý ,:. ' .", -- - -, .:. ,ý .ý -::.. I ,:,,,"':--,,..ý~,ý-:. *ý:, ý.ý~~ ,"- I- - ..... - - . ... . .:ý . -."ý..... --- ý... ,,. . ..:,:,:. ...:~ " - - ,!:ý "' W ý:ýZ.:.z.* ,:;.,.".'-.z.".::». - . . . ..ý : ....,...:ý: .:...,ý:. "'' ' '. ý ýý. ' * . ,ý:ý. ' ..,*::' .ý..ý::"".,.. . ", Yýý1 , ,- ý1:, - »' ý,-,-:-,.ý...-. 2 .' ,:::-ý,''ý.,:.~,:-: -ý, -:ý ý, -;Y . ' - ":.:,.:,:.., " .., 1.ý ..... ;*.W : ":.."ý.::.. . . - ' ý - - - ,. . - .. » . lwi k. .» -:'-.--,-,ýý .-----: .,,ý.::ý: -." ': -... ý-..- .ý.. - I. .. ... ý -- - ..... .. .. ..:1..,, ,.. - ý . : ,,,,,. '. - ýý .ýý. ,,, . - - -: . .:% . . :,,ý ,:.... - - - -:1...... i . ..........-.., * - .. . .... n ' . ' - ' ... ,:" ::".- ,....ý ...... . ::ý:.:: ý: .. .. ---- " ý ' . . ý . ... - I ' ' . " ""' . ., : ... I "' . ' ,,.: ý. . : - I.: .. jW * - . ý:.: . . . $ i* "' 40 . m ' I ... . ý:ý';ý."::, "' - """' .- - .~~ 1....ý,' - -- ,.."- I ' . ' 2 0'" ' .... . . I. ... " ," -, ý - ý 11. :, .,....." ýý~ 1" * " 11 .11 . : --..:ý :- ",''---,,,,,.. .. . "" :": ýW ,,ý': ýý:-. ý:_ -,:,::' ý * :ý- ý ý:... . ý . ' - ;.....;ý,.ý ýý.:,'" ýý .:: -: . I:....:-ý-:A''.ý.."" :"- - 1.11 - ' ' :: .4 :ý - - -.ý:.,:,:.:. .... ,::ý-."'. il..."ý,ýý,.,,.::ý::,,ý, W ":::.,;:,. -.1ý,---J ... I .ý - I ' - ' 1.1.' :, . : " - .. - . .:,:-,:,,ý-"---I -"' --.",. "".,.".., '.ý5 ...,,. ,,.. -,,Jý .ýý* --- 1, ... .: :,-,-,:ý-ýý . . . . . . - -» . I 'ý,":,:,:;:W . !! ... 4Ø .: ý:ýý,ý W ý I. ,-.. . * ' '''' ' . . -, '*,ý'* ' - .,,.. " ' ' ' ,:: .,j --.- . . -:.., .:......ý:. ...,.. : - - ý - ' .,, - :-, :", :.... .. - . .:,... ý ,# - ' .. .2 . «, - , ,...W1t0.:. . . :: '' - ::-----k ... :' .:. '"'",.::."",.,.,-.-.",::.: ' -, ' ' - ' ' . .. ý ý::.*:---:.,.:-ý'461t :,,..:ý....., .. .. . ' ;-ý:ýý ,,4 . - -:-: ýøe, - ' . .. . .1 .., -,- .,.. 2a . -1 - . .» ' . - - .. - - - .. ---- :. ý ,ý:,,...., .:", - ! --- ,ý ' ... . ý. ... -- - ' ý 0.: : - . ý.:..:.: .:.:.:,... .. ... 5 . ,.., . .."," .:.. .. . ,:: I ' . . -I ., ý:".,:;ý,,: - .. ... - . ...: -: -:-.:ý":- ,, »... >:ý," ". -, .':.ý ,ý;:",:- ' . .:. .ý... -: ' .: ,.ýý. .., .... -;:..ý - -..;: : --- - ---:, -,ýý l- , - , . lý. ' ,,.. ý . ý - . :-. . - ...... . - .-,.::.ý.-'-ý.:.:.::. .... -,, -''::, ,,,jý, --: -:-.. - ,,>:::!:::,ý.- - . . - - I ' - .. ý .. - - .,,:........ ---.".:.ý. : ,- ý ,,,..ý~*:ý ý:- ý,..,..ý..,.'ý1.- ' -.ý . -. .:ýý:::,, ' ..::ý::,-,:::ýý:::,,..ý.ý -.":,:ý::ý, - . u tk ý:,ý-:-:,:;-ý-ýý::.:-I.M ,,.. -.:ýý.ý":»..::%% ý,.,:::,.= - I ,...":11-1 .. '" ' . ;ý,åsb.:: - ' 11 I:, -ý,-, .ý.:-ý'.:ý,:,:-.ý-.,:ý-.ý-.:. . -- -, ø ...". "ý,.ýl 'ýý: ,..: -".:Z :-:::ý:ý..$>..:ý:,<. ".. -ý:, ý::''.:I:.::.:.::...:----~-i ,". - .1 ý- - I .:.': '. - - ý;-: ,.ý. - .:. --,: i - * . ' ' -ý-,, . ~ - -ý I...,...... .. -., .....- ýý :,,- .":..,,2:"::,,.ý:",;:ý^:*,......ý.:.N., ...,.":-:ý'; . .' 'l',,. ýi'ýý->::ýýi.,-,.ý:....-..,.-...-.....,..ý': . . ,:":"ýJ,":". ý - ' . . - ' ' . ..., .:ýý:.....:. ýý,.-.::,ý::ý::z ýý - - ..- ,".:::,,ý, ".ý'... . ----.........i:,,. - .; -- - --.---,,..,ý --: --: ,-------- .ý::, ý, - ý ' . --,, ... ý 1.ý~ ~ l:,ý ,-ý -.,,::".ý*. ,--. , ,..,.,:ý::,:,.,:,.",.*,'':"':-<- , ' ;Z.Z,:,ý...-: "' ý . . ,ý_ ~ :ý:>;, . -, -,- - -- - .. ,.... ,,,.;ý.. - - ----. .. ... ý_, ......,.....,,,.:-, -, -".":,:,:,:"ý-".ý-;,---;-'-j:.:,,."-- -, - - - - - -- ... ý .. ý.. --: '.I:ýý' . fl it,".ý., l -...,,Ik- <%:-..ýý ,:~ ".,:".:. -- .ý.1----.... .,.-::.:::-::ý ... ... :.":,,:ý-.::.ýý-.---ý,,.. ..,...,..,.,....ý~. .. . ... ' -ý.,ý.* .-'ý..,:ý. . -w i . 1ý z..:,-ý.......,..,. - - ...... .. -... ,:ý",,::"ý- .:.:.:, -. ,ýl:. 11 ....:.:,ý -."::-:: . . ' .. ".:.:", - , - . .. . .........ý.:......:,,,:,ý .ý::-- .':..... i i~ .. -11. - . ' . . - :,:_ ,ýý,:., :. ,:ý ý>.. "".". '.:::^:. : .. - - - - .ý - --.","-...,ý,.....:"::- ....., - .". .. ý . ,;..: -ýý:.:,ý': .. "-:ý,:?;- ....:- - ..... ..': ... .. .... : ... ' :,... .".:.::.::."",. -'. ''::.",:., -, ,. . . :::.- -:.::,::ýý,:ýý:,»:z --."..":",;.I.,:::-,:ý. ..:,.:..:,ý.' -.",,,,. * ' ' : .,. .. ý « - . ....: -ý-. -:.":ý,:::ý. .: ,., ' ý:,::'. --, "::,:..:::: ...:::... ý . ..,:..: -.':" ::.::..,:,:,. ..": 1-- - ::,ý,:.-:,ý,::.,.::?ý,.:^ýý ' . - ... - ,"-l":::.:::..:..::::.::.*;''.-.,ý">:,....ý,.",::ý.I.:..,z.;...,.,. - . . . .. .... :,,ýý.:, . .. -: .:; .. . ",:.::,.,:;",:, ., -,'.'.'::., - -: ...::.,::.: "'.":.;.": 'ý:: . ;ý>,"-W Ilillý,.-,ý, ::ý:ý- . -,,.. . ..,:. ' --.._ , - :: ... :,,,, ' . --:, ":- -~:.ý ' :;..!-,,...,...-.,...:,,.........:..ý~ ': ...... -ý-,,»:ý,:,ý:ý:-ý.,...., " " ... - <:" :" .,.....,ý : - ,,ý .::,.,: ..ýý,,,..ýý....ý.. . ' I:>",..::....:..,..:...:,:,..:::::..,.,.*"": - *, -::-.':,:::--,,.:,,.ý,;":->,...,:,,:..ý. , ý-l- ...ý . "" ' ... ,:."::"...: 'ký ýý ......":-:,... - -,:".ý,.::- :- - ------:": . ' ", .. -: :: W:,,:ý,;,...,::ý:::,:",,."--..-:*::,.",, :. .>::.....,.. . .. . -:ý:.:,..::. .,.,..,;::.j:.,., --....... ':.....ý,..,:-:::.- .............-~...:".ý:- _ý - 'ý'-- ----,:ý:.,ý::-:,: * ".ý,ý, -0 .":,-:--i:ý: ý.ý::":":-ý----, . I . . .:,.. ?.,-,:::, ý:,,ý:,-:-----: "".''.:,::":ý-...::,:::-ý,ý"....:ý-'..,...'...,.:.ý......:..,..»ýý.ý.ý:j-.ý:ýýý' --,.'::--ý:ýý ' ,,ý ..-,,;-:< -: .... w ,- ý!."- :"-.'::,..,...:.:-..:"::-.'- .. .:;. ' ' :::::-2.:,::;ý:ý:>::: ,i , ,.:.,:.:*.",*:.: .,::.-,. ....ý.::-ý.:.:,:ý...,.". '-I - I -. -::-, " .;,-, --- :.,-..:.....:.''.:;.:..,.,.::.. ..«...'- ,:::,;-ý"ý::- -*::,-ý:;ý,ý:, _, -." ,.:.:,* .- - ,,...- - ,- - -:, -ý,,:,::,,:::ý:-ý::,,,...ýý:^-,:,-,:" - ", I ..- - - ..ýý ý - ,L -- - - - ,, -. . -- .i..0 ... * . -.11.14.,,.......M ..,i* .,,,ý....-ý:":,,..,ý, -ý:ý,: :" . . .. . ----- .. . - - s:ý-",:ý---ýý:ý,:. ý.:ý,zj...:ýý--..,,..,:-::---."-:,:;:ý:ý ,:2M .F.- ' '..:..-.'. ': ' ":- -ý:,...;.. :,,..... ., m ,,. ; k. - 'lý, - ' . .ýý.ý ,ý- l,ý,-,ýý.*ý -.'.-.-., -.'--,:ý.ý-;:ý ý, - ,, --ý: - ,. -;:. -, ... .. . ,.. -, .: ýý .... ~: . ' .. ý. _ , ' -ý,,ý. ----.--- .. ..ý---ý..Z.ý'. " ,...- . -ý-".-ý:,,..:.-1: -- ' - - ý... ...",ý'.*-k", ..<ý,::-:,:- .::,.",."«:,j ..,..x .. _ý . ý - ' - - ,ý," ..,tó ;:ý 14-,ý.-."ý,.........'- . -- ..... ,,1.- ---... " -- - ý1:1:-"'ý-ý5:-ý,::::-. . ' ' W ' ' .-0 .ý:",":...,:- ý::,;'-ý,".:.ý. ..., ... ---.. --- - : -- -::::::;,-,::ý.'ý:ý- -:--.:ýý :.:-.:-ý, ... .-. -: - ---- .. ýý " . .ý ,..,......... .,.. ..., . ,:,...:-ý.:.,:ký>......:.: -ý::,..ý..:,:ý::;ý:ý,:'':::.":::ý,"',,,k:-5:. - . "" "' ' .. - - .,.. . , .. ' .,.. .. ý .., - ý..., ,- ,,... . - - - ...::. -..,ýý:ý-ý;:ý:ý:ý:ýH ý';: . - .. , .ýk:ý::: .ýý::::: . .. - .: - I- .. .. . . " - ' -ý.,..":ý. -;:ý.,ý- . - --- ... ,,.......,ý."':" .:ý,ý:ý -- ,, - ----- ---,. ý-, -- -1 .....l." - ----- ---- ' .ý - ---- ..:ý-.2ý:., : * ý, - -, .. -:ýý: . ' -,",...ý~.., - - - -: ,: ' :..,..,..._-,,....." -ý',-ý:ý .,..., --, -ý'--------,i -1---.---ýI.,,,...:.ý-----." ' ----*-~ " i,K ý",",...:.......ý:-,_ýA -,:" ' " - - I I . ' .... -... _,..ý... -: ,,.:, -: - .. -ý:'. --,, . - ...... .... . ..... "',",", - - - - - - . -:"ý: .,:"- ---:: ý ......... - - .. 11. ' - ". ,..........;- --- -- -*,- ---.'ý..... .... ; ,,zý,":ý'ýý.,,;. - :, 'ýý " - - -- - -- - - ' . .... ý ý ý -----,,,ý"---------- 'I * -..ý- - ý,.,...::...,::,..,-." . ... .. ::;....:, ...:." ",:.:,.,." ". ý.. . " ý-:-:.z,- :ý:....!::ý::.;:,,- .. . - -----.....,..,..,,,.-: ...... ... . - - - - - ....... ý ',,.". -,,..,,ý,:-ý,ý-,----:- - .. - - ý, -----ý--,,1---- ... .. .....::: ý ....- : -,..., I ýi, " ý.ýý:ý::*-ýý,:::ý:.ý:ýý;:.»Ifl." '-".",'' 11 ,,,, -- ýýý ,,, .,.,:.",.", ý,:. - - ': '., :"- '.: :. ," -,,;' -:, ,,...ý:.:::..,.., -:- ,:,::".: '. . - ... - ..ýý . .. ......"':.: .ý~~la." -- , . . . ..,N ....''. - .:---. ' . :.. ....,.. -:- , ... .... : ' -, ' .: ..... - -::... - ý: . . .... -:..... :: ". . . '. .: I - - -:-- ,... ý- ý::. - :ý:,,1-."'. .:":..":""ý ........ .,",:.:, . I . -:.. -..:.ý. :., , . -ý"-ý;.,... .. - ' ' ' ' I .. - ,: .,<,ý-:.:-,,. .- . .. : ýl,::Jø iýW lik:ýz,'ý:-ý,:.- ::, ;.": : . ' :..,- - ... ý - ,,,.. . ., .... . I " . .. " - "'' ý': -. . ý -::.... I .. - : . . - . . . >: -: ,....,;ý::,, .:, .ýý.-ý.ýý,* .ý<.ý,.">",-,:.;.ý<.<*-"":ýýýýý.,-ýý>,ýý-ý.,",ý... " " ' :,,.."'.ý:x ' ,....".". . ': ' ' ,... -... . . :, - ' ' - .:ý::::- ý -:.:: -, " .... -, -, ' - - - -:,:': - ... .:..,.:,.,..:-....ý.":::.:. ý- :. ,. zý. :,..."::::--,,...:::ý:-. ---, ' * : "" ' ~ " - :,." ,ý .ýý . . . . . . . :::: -,,. -.'-- ý . ' ". ,., ": ' ,,w :":--;-ý:,ý-.",,,:,,.... .'- ,-- . - ,,ý:.:ý» . . . . . . . . . . . . . ý, ---,ý . . . ..... ' --- -:::ý,.. . ,,,::, . . ý1 < -:ý:. ' .ý, ..... .. ..,::: ..: . - .. . -ý-,,;,,,,.... ..." . ..ý -..-.*.,..:,:.,.,..:":.,ý,::,...:ý:::::".:.*..:..:ý,:,:":.::.. - -- 1: , 1.4 11 ý .. ; ........ . . . . . . . . . . . . :,::. ,* i iff.""Vý.":::ý.ýwýn,,,,..,,ýý........ ý '.;.... .. ý:... ':,- --,ý---- ---:" -': . . . - . '. . ý . ..ø ... .. .... ::. - ... .,.......- ..., .:--, - ". ,.ý...;$::. .. .: .,,::, .. ýlý,,....".: -----,-----ý, .:.::.ý.",:.....,.. ....: ' ,:,-:,,...:,ýý:.,,.ýý:ý, -::- ,'...ýý: -;--.., :::...ý..,:": ,:.::.., - -: ---,, .. . ::. :::::: . , ' . . : ' .. . - ... 11 . :. : .: -- * ' ' : , . . I 14W -::":-:.ý1. - - .,: -,-- -- " . .... . .. . ý- ý . ý . ':ý.ý..' ' . ' : :ýý,W & . . ...:- Ic ..... - .:",", ..:..,:-:: I * ý0 '"'., - - ý . . . . . . . . . . .. I ,::,:". .1 . .. . .... ': ,,. -.:. ,,.. ,- - ;,., . ., .. . .- - " . ... - - .. ... ,ý.. -: .....,..-,.,: ..ý,-, -....::.., :?ý- ' :. ::.': ý:.. -. ' .:ý' * Q :ýý:::::":ý; .:-::ýý1.... . . ..:",." _. . I - - - . .:- :ý - ."ý. ' . . ,:; ;-:... ý...., . - ý .ý , -". ..: .:::Z; A :::::,:..ý:.:::::..:,. ,... -,:-.' .ý.:';..ý. - -ý' --ý. '-.-'... ----- -:,,..,.'.>-ý,...",ý:- : ý:- .. .- -:,:ý- .. , -. .. .,:.::...."":."ý:....ý:.ý,:..ý";,.,:ý,ý,..:.ý.,.:.",.::,.ý.:ý...:"':ý,<:ý:.'. ý..'-""<-.ý,:....ý,..-.".,.:.ý:','.,ý'...:ý.>,,-'ý:<.."....,lý''-ý:*,-,.:,..,,'.ý"::,ý",.'':: - .. . .. . . -.. '- . .:. .ý2-,,".,..'ý.... -,,ý -,- - ý,....., ...., .. - ,,, ': . . . . -ý,,......: - - .,...:::,ý,.".:;..". ..... . ..,,,..:,...:,.*.:.,.-:..-,:" ý.-,'-.-.'-'-"ý''.,.:::,.:"ý..ýý.-'-:.,:,.,... . ý ý ' I " - - - . ' - .." ' ' " .. .- .. ' ':"..ý 'ý'...'ý.:..,.:ý:,..:...,..ý.:...,::....'ý',,:,.: ..."' ' ... " ..... .,,..::ý:" .. . ' ý. .ýå . - :ý - .. ' - - ,,ý*:,,:, ý . ý, - . .. -, ý.:: ' :: --- , .. . ' . .. :I - - - -':.ý.:::ý.' ý, . ' ' . . . :,-:. I. - - - ... .-- - .: .,..- ,,.. ' ' - ,- ::.: - - ---ý"::ý- . - .., .. :. . . .. - .. .: . " - . 111-.' -. -,,, :, -- ..... -"', ,,,,, ý:; "-». .z,:" ." ,... . ý..,..!."..*^..:.: ý - - :' - ",","." .. . . .:> .. ::.. ::.- - ..:,::". ý '':':ý:; - ' « .: ' -----:-----.'-----..- -ý. ~ - - ý .. - " - ::::: : - ' -", -': ' .: . ,': ».... .. . :. .*:: ".:.,.., - . - ' ,- I. . ...,.. .ý::,;ýý - ý:,- :....;:-- -:- ,,. ...,ý.:ý.«,.::.:. ý . .111.1. . . ., - ýl. .! ... ,.... M :ý....,,ý...ý'ýý:ýý-".:: ý ..:.: . - -:'..,':.':.:.,:,::''. . I . ' ..., ,,, " ' :ý;.'ý ..'- - ý - ":, -:-.'. "-:. : . - - ,". .... .. . .. ý.1. ý.:ý' ý ., ý::..... - -: .".,:.:, .ý - ý - - 11--. - ....... ,ýý-,,,..ýw ,ý,,,,,.:.- ~ý',* ,,,,, . . , .. ý.. ' ---. ..... ' . ý - .. im . - :.:.::..".: . ....- - -- . -ýý - kýý........,..,:.,:ý.:-...,,.z.... -ý----,.'ý,.., ,:", :-... -:;:,:":,::,,:;ý::,:- -........ - ýý:,:.....j , ,.,:.:,:.-.,.:, . ''::.,::,::,:"::.:,ýý,:,.:.. ."ý... -..ý:- -,. ' ' - --. I . .. . ",ý."',ý- ~ :.:,:...:-:ý,.::;".-::":- -,..,ý:ý...... ..,:ý.:ý,,..... :,:",ý:,,»..ýý,....*ý:,:,::,,,.."::,-:_ ::.,ý ... ý.ý. - - --:.- . . - . ' . , ,,. ,,..I.- .: -I 1::.I:.",ý;- ' :" I.- - ý.ý-,: - 11 ---;, '. 'ýý~::;.....,.:: f t f "..," -ý.n.:'ý,-::;.:,ý'-j-.::: . -ý:.,.. . ' , I...,.. '<:,::ý ý .:":..... - ,:%;.::ý,....ý:ýI:.:ý:-ý::-ý,$;:::.:,.:-:ý. ý, I- . :.ý~:.ý..:.:, j : . - - ----- -:-ýiý.., ,ý:,Aý: --z ' ,ýýa ' . - ..: ':---:.--..ý-----...... . ,1-...,. --, .::ýý::, ý::..- . ý ,"," .ý . . ý . .;. '. ý. - --.. ý.,- - . . .1 " ......... .. ............. ...... . ' - ,- ý - I ý,-: -ý;ýý: ý. - --:" . . a- W ' ý -;: " ' " ' * ý ...... - . ... . ...:.:ý,....,....:- ...:,<",^:.:..ý:-:ý.:.. *"- 'I ý - --: .... . . ...,.... . ... 4 : . ý '- --- -- -' ....... ' .,...I:,ý ' .. ---. .......:ý'-k:::ý,ý,-,:::2:",; ýý ý::::ýý.:L i .,.."ýý..-,, < . u- - ý -. 0 ; .... . . -, d. -::.- - " ý:-;«..;.... - - ...:..:.!.ý',: "' ý...:. ........ " . - . . . ... '. .. ...l. ' .. - - - - ...:: ::: . - --..?,..... . . . ' I.:. .'. - ,,,.. "- :. ,,ý,, : - ' ---: . ...... I, :. ". ,":-ý:ý."':":....ýý%,..:: " -.' .ý -:, --:,ý,-,,:.:-ý, .:.ý ::,.. ýý:ýý:5:: , ...ý ------------,,,_..,;"., , ",.:,-I... - 1: ..:.': . ..::,::.ý.1-..-..ý,ý., - --- - - .. ',> ...... :,,... ,. '. I .. . . ....9, ."- ------- ---- - - .." ;: - --."':".ý:.::,:....:,... .... ..- . , . . - : - ,.., -.-ý: ..., -- - , -----ý-t -ý----,ý-----.'--,--,.'--,--w ,-------.------.., . ý-,- " ---:-':-ý-ý, ,:-::".:::,ý-.':":f":.ý'«.,. ý ý ...... . .ýw,..,.. . ý;...l.ýI..m _ ý1ý--........ . . ' ý .. - :. : . ' ý . . -- . .: ... :-ý,-:;:"-.' ".:" J, .- .. -:,ý:.. .. . . . - . ' . . ' * ... -.; ... :... -ý. - ý,..l.: ... . I.,ý*..,.,,..l. ":.,:, ý:: ".I - .' ' ' . . ' - ,, " ' .... ' .,....ý.. -:-..,ý;, -.ý.ý~-::-:-ý:;_..:,,.. ....:,:-:.-":ý-:.,:':.,:.'-ý,ý.". : ýý11 - I...lý, ýý,:' ..ý . . .. - 4 .- . .. .. .. ý ..., ..,.. - q ý - ......,-: '. ý- ... :.- I --- .. . ý.. ý .:. . , . ' -::" -'.'..."..: '" ,- ý-j :" :1-.:,::"ý.ý:-:.,ý.:ýý: -: 4 .ý:j-:-.".'-:,... :...::. . II:... -, . . .:::;,ý..... ,:.:.,.., . .. ...,., :,,:_,ý:.:, -: ' ,', ý ,.:-:ý-.,. ..4 ': .... - ':0 -,,ý, .... I . :. .. 4 : ......,.-:ý .... :- , ..,.:ý,-"::,.. : ý - ' - - - -."' ... ' . . ý': : . '4ý .....,ý . - - - -- --- '. . ... ,:,:-«.ý ..,::,:,:.- ' -- -.'''. .ø :", ' .::--:,.: ;...:.-.-.- ýý-tt ..:... :: .".".::.-:-:,..", ' " 'lý,:': 4.::-, .,...:::.".':.: ..... ý....". -.-.-....,I..- .-1:-.'----,....,...------. --- ' ýý . -- ý:ý- .ý,ý. ý:..,. ...., .... I - :.:: . : ý ':. - ........ *ýý ,:: ":--- . . ..:.. ' -::.-* ý:...I.: ý: .. ',.ýn .ý:.:',ý::'..'. ..ý~:. - .. ,, - : ,'- "," .. . ....ý:ý .ý.ø .. - -- ..:: -i ý:: ý:":ý::ý, I. .... , : :... .:.:- .- -. .,:..ý. .. - - ' - ,ý-,.ý, .... I...X:.. .. ,,.... ' " " ....,...,,,...- '. ' .1 ..ll:,- . .« .: ,,W ý:"-I' ....,. . ' . .: ,,. ... I jit'- i , ' -" - ' " ýý ' ' ' i . . ... . » . . . - - -ý;:.-.,:-.*. -, * : ý." , . I . ý...ýý,.... -ý-:,,. ..>,:---ý......-,---ý.:,*-......:- ý .. .:" " .. . :- ....,ft",,.., - ' -:-ý ,ýý":.-- - -----.':,, - "- - -- - -- -1, -- - - " -ý"" :::::.: -:: ,- :,-""-..:-. . :ý . - -.:1...5ý ' ' ý % -, ý. : ., - ..., - -w - " ' - *1 1,1 ' . . - - -" - - "t -, .'- - - '' - 't. ' - ' ' ' . " "' ' ýQ ý 1; 1* - - -- . .ý.:. I., . * ' . . ,.ý:-...:ý: .....,:",:::-.,.*"": -- -- - - ., -', -" :. .... -,ý......ý...'- .." -: --- - ..:.ý." ., .l'. - :: ,. ý :,- --- - .. ,..-.ý.::,;::.:ý::.:...-.,.",:.., - "I ---- ---- - .. -- -- ý I ýý ..." :, ý ' . . --. .. .ý -- , ----:ý.,.ý.- - .., I .ý*,,:.:,::":ý:-,:,-;- :- - -.' - - .....,,..., - .. - ' . ... .. ' . . . . ... - ---- .. . - . . . ' 'lý, . - . ý:.. .' - ...:.,. .. ,,. --: - - ýý::,:;*ý ý ý'.-- - .. ::-,:-ý.....,,ý...,....,::-"-" ... .:" - . -, -. ý åý ..,.. :,,...," .- . , . -,,,. i-,,.."i..,.:..::ý---:,,.':,,.....,..,,.. .. .. ":-.. '.'- ý " - .. . . . . ,- . .. ,,. " :.. ~ - ' 4. - - ' - . ' ,ý ,». ' .. .. :. ..ýý.,.... -, -.ý~ - ,-: .::,,...:.ý- ' - ... - ..ý..d,": ,å . .. : ý .:.,::.... ... '.- -: ,,.. ... . ' -: ,.,..;, - , : . --- .:.-".Iýl...4 .1 ....:;.ý~...,,..:,.,......":-:: :_ " * . ...: . . ..... .. . . .. .:::X....ý- ,"'-ý:.:ý-:-'ýý-:,:,,ý, -& :;:::,*ý ,: -- ' å ."ý::::ý.. -ýý-,,,,,ý.,..,....": . - - :.,,.ý.:,..,ý" : ý4, :: ' ,,,.... ,...l. , ....ý,,.. . ' " " .. ., . ý ' - ' - - - - - ----",",''ý;-,,ý..":::ý:-:,,:-ý::.ý::"ý::,,:: ý. - , -:, ý:,,k .11 " - -: .:'ý.' - ': - - -- -ý .:,,;l . -: ,.:,.- ., -": - :::. -, "". - . .. .~ ý. . "-.."".,". ..::1ý . - . - ... , .....:ý, .. -'r-,...,ý-, . .:> I ' " .... . -"," ýý~~~ .... -;,:,:,..ý,..,ý ' :. ..:" :, . .. .7 ..: -- - ' - .. I .. _ -,- -. .- :.-ý;: :,:_, ...I., -:ý,.., . - - - - - i ý: ý, . -.'- ...,.. ý - . ,... . ... ......, ....,... W .,,.....k:", ,:.: . - . . . * ý ,,5:--, -'' .:.,...'::. ' . .ý ,- I ,.:ým .,:ý. . :,.... --- :---.10 .: ,,......:. . ý ' . . . ......,:,ý-,,... .. - ý " " I I - . lý.: ' . :,:l; ' ''. - ..ý:'n .: .: ,,.,..W .,.."., -: . . . - - ' . . .. . . . .:- ,'* .: . - " - ø -", """' . .. 11. # - . :.::: -In - . -:ýý-.' ^.*. " . ::Z".... '11... - . : . .,:,::,.:,:..:::"., 1,*,ýW .,,IýIt - :-ý:::ý,:: :., . ,.... - 11 I ... *,. -.. .,....,...: ....,:..,ý,::ý- - -- ....."., ::.. 1ý ... ' « . ' . . .. -, .. . -..I...,. . ' - " - -" . ... - ,:-, --,,...,.ý* ...... . . . ... ý :.... . - : .,:- :- - - - --- - --- :'. :,.,. :.- . ::ý:::ý,"::,-ý:. '. ' .., .. v . . -, " ... .., ý8 ' .. . , . : -,:, ý . --: ., - .. - --: .. . . . . > - . . . ' ý. Ik . - -- ---- -- - ,ý-.<... .... ýIk: ::ý:,:, .... . . ' I . . : ,,-. . '. ý: -.,.,ý-,,»."".'-", # - ' : . ^ 1: , .. : - .. -ý ý,:: " .. .. - ... ... . ..:,.. . ' .. . 'O- " "'"" - .. ,l, . ý ... .... I ý .. ' ' . . . .-.. . " , 'ý- .." -,,, ::, -:- , ..,v.'- ",;----ý..,..,..,:. . - - -- ' ' ý- I ý . . .. .. . I ý -.... . I ..... . .... . I ...., :"'. . : ý:'-". -: I - ýý ~ I - . - .. - . .. . '. I . . . . .:.. -: ... - . - ' "" ' :.4 ~ ý1 . . I.., I ý . . ... . . . .:., - -: ..:.: I. - - I l.. .., .-:......-. .. ,.:-:.-.. -.,:,.:-..:;ý" ,, . . . . .. I ý ... . '. ' ' ' - . *. 'l' ý, .,.:..-..:::.:.""...'.-, .. - I ' lý... . . . . . - . . . .." -:-:.: --- - ' ---: ' ~ *.4 . - . . ,,."': ,- .:ý.....> .....: ...... . ... ,... .. : . .... ý- . .. ... :.... . ......... - . ...ý , .... -:: :" ,-:::----:ý::,.:: ,".ý - .:.»,» -::.,;..,. . .-.. ý . . . . . . :, - -:: -. ., :::-. .. .. .,..,... - ' - - : -: .. - --..., I... - .. - : . -- - ' ..,ý:.. . ., . 111. .: ' -, --k," . " . . " ý. . ' I . . . . . .. - - ' ' * ' ' . : -:ý,...,. .,,:,:ý ' ' ' . '. ý... I .... ' . ' I ....., . . . .- . . : :- ,. - ý . ...,. ","-: ' .ý . . :.ý. - :: 'ý->:":,.: ': ý, -.. -- -- ..ý .. ¥, ..: ,":ý . . ý . . - , . . - ý : ... :,:. . .. ..0 4 31* t," -.---ý,",,,:It ,,.:, , : ý_-,:---: --,ý ,,... -." . ' ' ' ' ' . .. .. .. ý:. - ..:.: .: .": - . I - : ý::,:ý::,ý ..".:ý:ý:ýý.. .,.:::...,.,i* .. .:. I . I ": '.- . ' ..... ..,. '. : ,,.. ' ,,, . ' ' :.:.. ,,. ,,,.....ýý:-ý,ý:,,,,5~ ý. -ý'., -.. .. .,:.: :,7,t.t.:., ,-, -, - -:- ,: - ' .- ' ' . " . .. . .. ... - -- '. .m ~ :.. iý» ' - . ,:,....:-,,.*..ýý.:.;ý..:w , -,:"., ' l" - .. . ::.. ,..:: -, . - --, ' ý :ý. ' " .. -- - - - ý.1 ... - - ,,..,,, ' ' . - :, 11» ý:.:,::.:-:*..,:,.:,..ý.:... ,:,.::""ý:,:'' .... l- ý... ' .ý . ' ' . ý ' ý . . . - , --- . ' - ' - - - . :Iý- .-. -, ..::',::'"ý:ý":. .,:, - ' ': - -,;,:.'ý:. ...., . ... .:,,,.,. ý .*,,::----ý.!w . . - . :..,.------.» -- , ' -,... - ý.,< .. .. : -- -ý - . . -: ý.:",,.. .:.::.:..:,;.::«,. :ý -1,ý ...:. ' :ý..>..-..::.: ":,."::. ' ':.,.:, - : : : . . ' ý ----- - . . , :l ::ý': -:ýýj:,ý:::- .::* . . . . . . ' ý ,;-ý:,,:,,.:-, :: .ý:-ý-:: *,",.,"ý..: ý .., . . - .. -..;.. ::ý..:,...ý, . -l I:,- - : - ; - . - . . . . ------------ ---- -J.. .:: ,,... ":-,ý . " ý,'ý:, ."::.: .,.... 'n.,... . . . . . . l- ý.,..."', ,.,:::> !-'>-..:. . ... . I - -------. ..!. . . ,..,,., ..,:,. .. . . - : :. ...... -."'. . . I m j - .. - " ":..... - ..... : -.' - - .* -: ;,ý- .: '. . . I .. . : .ý:.ý . ..ý, :.. ., ...:,,..':,ý: W . fý-gø . ý . ;,. . I.: -.'-'.' ~- $ýý.... ' ' .. .. , " ' ,, "' ý...Ik . . ' . ý, ' . ..,.ý:: ý:-: - , -:- -:,:- . . . .. .. . .:"- - :- ... . ' ý . ' - ---,,,ý:," -".- :'-M -.'", *.., ' ' -: ,- ,.... '. .. - .w, . - 'ý:...:-ý:;ýý-:- :- . 11 : : :: - - -": .. .... ::: . "" .. ,.!, - ..... .ý,-- - - - .:I . ... .: - ,,..ý.....,...... : .. ...,....,,ý,,,:.:ý.;. . - - .,. : - ....... 'ýZ.:, - ý ý:-.":: ..., . .. . ,;:::,:ý: ,,.ý,:.:.., . ,.,.,:, ,:.ýý:, ý,:-,:.::.ýý, <ý".,...,,," ,". :.ý...ý. ,",...: ý,..:,ý'.:,:..."',".'"..",;:...."':-". -."', .. .,.",., -.,-::: , ' - -, - - :-,ý:- .. . .,..... -- - :....,,":-ý:........":<-ý...- ... ::.- .,.,ý...,.,:,..... . :- .,., ..,. - .. ... .. I - -;::ýý,..::ý..,..... .:, . ... - .: ; -l:,ý,.. ' ' ,z ' .. ---: I ý " ' .. - -ý:,- ....:- ... . ... .. - :...:.>: k :.:. . - . .".,:., ..,..... ý:.'. . ý.:ý': ý.", ::. -ý;:.,,.:-::- -:..ý,.'...,..,:,.,,.,.:.,ýý:.*.....,::;.,::.:ý.,*ýý:..::."..,..,.- .,.........."%ý,'ý,,.:-ý..,.,.,. - .:ý .:.::.::,:,....:: ,: ::; ' ý:...ý,":ýý::. .. - .- ..::- ý.',- -:,:--. . . ý. .ý,ýý.. ' s': - ý .,ý., ,:-.::,:.:ýM ,,..ýý::,,.::'" . :ý. . ý..: - ......, .. ' . I :. ... -ý... - ...4 -ý;- .., --- - .". ,::ý - ... :ý. -:ýý:"ý -. -.,:ý-ý ,:::: - .;, ý ,,ý:-,..:-:-.'zý.ý:: . .. .ø .l. :.,:ý -ý:ý:ý:ý -: :....:.ý..:, .,.ý, ,,... ..,ý_ ~..ý~ ~:_:,,....1-......- .11... . . - .,: ... -::.:ý:::,:. - .. .:- .. -.- . .. .. ......,iý:-: . : . .ý..» .ý:,»,..:* .. . ..".:..... .:.:-;. L- i -:::ý,:ý: -- ... ,,:. ,.....,,I.- . 1 -1 ' "".,::.::.:, .. . ý:'':-- l-,,, :::ý,ý:ýý :i<,ý...,:.:....,::,,ý:,ý:,:"-:,,,.--,,, :.. . - . . .. - ' .ý. 4 ý::.-ý:,:::.: ,;. ..:.! ..,.....:ý-:.m..::.. ý::ý:ý.":ý:ý: -, ý - .U .. - ,-: : . ..!:"::: .., "::::<,ý:ý,ý -::" , . I ::: ý. " .:" . ,v.,ý..ý= ..,:-...I:,,::,ý.:-:ý I *- ý:: ... x ..ý . -ý'111* . .. . ý..., --. -.. " - -, -ý-.- .",: I . - I ., ..,. .:'.- ,,..., " ":-ý:,:7::: ,:,,;ý,"'::-...,9ý ... - " ' - ý:.:"::,..... .. ' - - ...::....": . .. I ' 6: - ' ,.. ",",", ...l.' .. .. . :-. ..ý;* .... :..-.....:....:,. ::".:, . .."-'>'".-" øl .... ,,k ý,,."":::,,:::: :14,..I.: '. ' -.ý: . :, ' - . ...: - ' :: -".,....,,. -- -,,...- - - - ý . ý. . - " . - 146 - ANSSE -iV.16 -_PART..il: STINATIS OF EgER=? C0U8NPTWN my UnA vEICLE ft AMING mo FRTIER DIMSEL2ATIM9 OF FLEET AFTER 1974 1975.79 1750 250 TU 180 70 40 3560 fl-n 1200 200 10 50 60 50 20 59 19508 3m50 600 120 110 70 110 50 4610 total 10000 1250 130 250 400 350 120 12400 *. DITRIUTM BY PACUMT (x of total> upto-31 - Zps, 100 100 55 48 % diest 100 45 52 100 -100 % åt66%. 1975-79 Xs 100 100 50 50 tdesel 100 50 50 100 100 % Mech % ens. 69 55 % diset 100 sa 50 96 100 %Moo 31 45 50 50 4 1985-sa Sos: 7 , 30 1aet 100 50 50 96 100 9 ae 93 70 50 50 4 Totat t % gas 63 59 0 18 33 0 0 a disl 0 0 100 50 50 97 100 $·slo 37 41 0 32 16 3 0 0. A 11ML ISM! as 11 -1$ 20 25 3 : 50 diset 10 20 15 18 28 40 t1bil 14 16 25 30 45 55 1.. TOTAL h1TIKATlO MIN M cuhI~PTfON bfifon Iors p.a., I9O0): 39.1 gasoUm 6.? 1.4 0.0 0.3 1.6 0.0 4.0 10.0 diest 0.0 - .O 1.6 0.7 2.2 5.5 4.j 14.7 aole 9.9 1.8 0.0 0,9 1.6 0.2 0.0 14.4 ~hurossr .NS, smpol, Fransporte Nod~rn., saan U nergetfoo Uslcimnm 198. 4 ItuliMn th unhydrto (in gsohot. 2M03 ond hyo anht. *taffApulassl leprt, NIghKsys Nangsn sd iR~ aftun Project, 060adr 29, 1989 w iort go. 3000-pRJ, Annx 4. - 147 - TIM ECONOIC COST AND NETBACK VALUE OF NATURAL GAS 1. The Average Incremental cost of finding, developing and producing non-associated gas offshore is difficult to estimate, because of the lack of experience in developing such fields. The cost estimates have ranged from US$2.30/million BTU to $1.54/million BTU. The former is a World Bank estimate made in 1983 but may reflect an unrealistically high estimate of finding costs.1 The latter value is from CNP, based on gas from the Santon Basin, and may be more indicative of future costs. The AIC of associated gas is much lower, because it includes no finding costs and only a portion of the development and production costs. Based on COP's estimates of the investment and operating costs of producing offshore associated gas in the Campos basin and delivering it onshore, the economic cost, at a 12% discount factor, is estimated to be $0.75/million BTU. The AIC of gas in each region will be the weighted average of the economic cost of producing associated and non-associated gas. The ratio will vary and the best basis for projecting the production ratios is the ratio of associated gas reserves to non-associated reserves. The proportion of associated gas will vary from 25.7% in the Northeast; to 44.3% in the Bahia region; to 95.6% in the South- East. Based on the estimated economic cost of associated gas, $0.75 per million BTU and $1.54 per million BTU for non-associated gas, the AIC for the North-Bast, Bahia and South-East regions is shown in Table 1. These estimates may be on the high side, because some portion of the gas will come from onshore sources, which are less expensive to develop. 2. The depletion allowance is the foregone future value nf the gas if it is consumed today. The future value will be the discounted cost of the replacement fuel, which could be either petroleum products or new natural gas discoveries. In view of the large unexplored potential in Brazil, it is reasonable to assume that non-associated natural gas will replace the existing supplies when they are depleted. At the projected rate of consumption, proved reserves will be depleted in 18 years, so the depletion allowance is the present value of the cost of developing new sources of gas in 18 years. It is estimated that the historic finding cost of non-associated gas was $0.67/million BTU 2. The development and production cost add about $0.30/million BTU 3, making the AIC for onshore non-associated gas US$0.97/million BTU. See "DiU and Gas Sector Review," Report No. 4816-BR. PLMGAS Sufgroup No. 1 Report, 1988. a See Vortd Sank, OfI and Gas Sector Review," 1985, Report No. 4816-8R. - 148 - ,... .... ~TABLE 1 ...4...139 ..139..9 City Gat 2.066 2.009 1.20 LA? ldsty 0.14 0.14 0.14 St rdutry 0.50 0.50 0.50 mesidentis 2.78 2.78 2.78 Trmuport 2.62 2.62 2.62 Large Industry 2.21 2.15 1.34 Offit Indtry 2.57 2.51 1.70 Reientfal 4.0 4.19 3.9 Tragport 4.69 4.63 3.82 3. Most of the gas would come from onshore fields but assuming 15% was from offshore field* at an average alIC of US$1.54/million sTU, the overall average AIC for both onshore and offshore production would be approximately $1.07/million BTU. The depletion allowance, calculated at a 12% discount factor, would be $0.139/million 8TU, as shown in Table 1. The development costs are subject to significant variation, depending on the location of as yet undiscovered resources, field conditions, etc. However, even if the more accurately estimated cost of developing offshore non-associated gas were used, the depletion allowance would still only be $0.20/MMtu. If additional natural gas were not discovered, the replacement fuel would be heavy fuel oil. Based on a 1989 price of $15 per barrel and assuming the real price would increase 14 per year, the depletion allowance vould be $0.39 per million STU if fuel oil wre the replacement fuel. 4. The long-run marginal transportation cost will vary according to the capacity of the pipeline system, the distance gas is transported and the terrain the pipeline must traverse. For example, the PLANGAS analysts estimated that pipeline coste per kilometer would be tvo to three times higher in the Amason region than in the South-East. Based on typical international costs for pipeline construction, the estimated LRMC of transporting gas to markets in the North- East, Dahia and South-East regions is in Table 1. It is difficult to estimate the cost of transporting gas in the Amason region. The infrastructure is very limited and construction conditions are rigorous. As part of the PLANGAS study, the alternativen for transporting gas via pipeline, as CHG, or as liquefied - 149 - PW! Y ar- natural gas (LNG), were revieWed in a preliminary way. However, a much more detailed analysis would be required before either of these alternatives to a conventional pipeline could be recommended. 5. The distribution cost of natural gas in urban networks is a function of the investment in facilities, the quantity of gas consumed and the system operating costs, such as customer billing. CUE estimates a typical residential consumer will use 19 m3/month, which is somewhat less than consumption in other Latin American countries with similar climates. As gas use grows in Brazil, consumption per customer can be expected to increase. Assuming an investment cost of US$ 250/customer and consumption of 1 a /day, the economic cost of gas distribution is estimated to be US$2.78/million BTU (see Table 1). The cost of distributing gas to industrial consumers would be much less, because the consumption per dollar investment is much higher than for residential consumers. Based on the actual investment cost of an industrial network in the Paraiba region, as estimated by CNE, the cost of distributing gas to large industrial consumers is US$0.14/million BTU. The cost of distributing gas to smaller industrial consumers, such as artisans, would be higher than for large industrial users, say US$0.50/million BTU. These figures are also included in Table 1. 6. The economic cost of supplying gas to each class of end user includes the AIC of production, the regional transportation cost, distribution costs and the depletion allowance. Table 1 reveals that, in the North-East, where non- associated reserves predominate, the AIC would be higher than in the South-East, where essentially all of the reserves are associated gas. Transport distances are longer in the North-East than the South-Bast, so transportation costs are higher. The cost of distribution and the depletion allowance are considered to be the same in all regions. The delivered cost of natural gas to large industrial consumers ranges from US$1.34/million BTU in the South-East to US$2.21/million BTU in the North-East. Similarly, the economic coat of gas to residential consumers is US$3.98/million BTU in the South-East and US$4.85/million BTU in the North-East. NETaCK VALUES 7. In 1987, Brazil was a net exporter of heavy fuel oil but in 1988 exports and imports were nearly in balance. If natural gas replaces heavy fuel oil in the industrial sector, Brazil may once again become a net exporter, in which case the opportunity value of natural gas should be calculated as the export value, minus the cost of transportation from the refinery to the port and port charges, plus the cost of transportation to the point of use. We calculate here the opportunity value on the basis of the c.i.f cost of imports plus inland transportation, which is valid for the situation where fuel oil is imported. From Table IV.5, the estimated price of imported fuel oil was US$14.60/barrel, to which we add the average transportation charge of US$4.23/barrel for bulk supply. Thus the delivered cost of fuel oil for large industrial users would be US$18.83 or US$3.09/million BTU. The cost of fuel oil delivered to smaller industrial consumers would be higher, due partly to the higher costs of transport and handling. Smaller industry is also less able to burn lower quality fuel oil and will tend to consume distillate (kerosene) at a higher price. The netback value of gas to smaller industry is estimated based on a mixture of heavy fuel oil and distillate. For the latter, the delivered cost from Table IV.5 is the import cost (US$24.09/barrel) plus transport and distribution costs - 150 - (US$4.93/barrel), i.,e., US$29.02/barrel. Assuming a mix of one-third heavy fuel oil and two-thirds distillate, the average netback to smaller industry would be about US$ 25.63 /barrel (or US$ 4.21/million BTU). The netback values for natural gas are understated to the extent that they do not include a credit for the lower cost of utilisation, higher efficiency and premium value of natural gas. 8. Brasil is a net importer of LPG, so that the opportunity value of natural gas which replaces LPG in the domestic sector is the c. i. f cost of imports plus inland transportation, distribution and marketing costs. Table IV. 5 estimates that cost at US$26.89/barrel, exclusive of taxes, or US$7.31/million BTU. In relation to low-voltage electricity supply to the domestic market, based on the data in Table 111.6, the LRMC for residential customers is close to US$100/NWh or US$ 29.30/million BTU. The very high netback in the residential electricity market reflects the high capital cost in meeting a short-duration load, essentially for cooking and lighting. 9. It is technically possible to use natural gas as a fuel in the transportation sector, thereby replacing the highest value liquid fuels, gasoline and diesel oil. Gasoline is exported but in recent years diesel oil imports and exports have been nearly balanced. If large quantities of diesel oil are replaced with CNG, Brazil may become a net exporter again. Therefore, the opportunity value of natural gas when it replaces gasoline or diesel oil is the f.o.b export price minus the cost of transportation to the port and port charges plus the inland transportation, distribution and sales margin. Based on 1987 f.o.b export prices, the economic value of gasoline "at the pump" was US$29.31/barrel and the value of diesel fuel was US$31.15/barrel, which corresponds to the netback values of US$5.58 and US$5.34/million BTU. The cost of transforming natural gas to a useful form, such as CNG, is quite high and users must also pay a conversion cost to operate their vehicles on CNG. The two principal components of the CNG cost, in addition to the economic cost of the gas delivered to the CNG station site, are the cost of the fueling station and the cost of compression energy. We estimate the cost of producing CNG, at an operating load factor of 75%, to be US$2.62/million BTU, not including the cost of the gas. - 151 - SOM= SOCIAL CONSIDERATIONS IN GAS PRICING 1. In the case of natural gas, the costs to be recovered via the tariff include thtee primary componentas a the cost of gas supply, which is the bulk supply or city-gate cost; o the cost of attaching a new customer to the network, which may include a portion of the general network facilities, as well as the facilities specific to that individual customer, although it will not normally include the customer's cost to convert appliances or equipment; and a the cost of operating the network, metering the gas, preparing bills and similar "customer-related" costs. These costs could be billed to the residential customer as a single cost per unit of gas consumed. Such volumetric or "commodity" tariffs have the virtue of samplicity but the tariff is an average of the cost to all customers within the class and if the tariff is subsidized, all consumers would benefit regardless of their income level. 2. A more common method is to establish a two-part tariff, which requires the customer to pay a fixed monthly charge, to recover the fixed customer-related costs, and a variable charge, to recover the cost of gas supplied at the city gate. The connection charge would be amortized as part of the fixed charge. The two-part tariff is relatively simple and easy to administer. A typical residential consumer might use 19 ae per month. The subsistence needs of a lower-income consumer would be less than the average consumption. Based on natural gas subsistence requirements of 12 m3 per month and a LRNC of US$4-5/million BTU (see Annex IV.17), the average monthly cost could be US$1.69-2.11. Adding in financial costs which the distribution company is likely to wish to covar could cause the lower-income consumer s bill for subsistence quantities of natural gas to exceed the desirable ratio of the minimum wage, say 2-3%. Therefore, it appears that some type of "lifeline" tariff would be desirable. 3. One of the principal drawbacks of "lifeline" tariffs is the manner in which they are applied. In many instances, the subsidy is provided through the tariff structure. That is, the cost of amortizing all investments, including the customer connection costa, is included in the tariff. An increasing block tariff is established in which the first block, the subsistence requirements, is priced well below the financial cost. However, the first-block price is usually available to all residential customers, so that the higher-income consumer is also subsidized. This is exactly the situation in the electricity supply subsector. In natural gas, it may be less appropriate, as it is essentially a new and growing market, so that it will be necessary to build up and extend non- marginally the distribution networks. To reflect the different burden of distribution investment costa, an alternative is to require the customer to pay a connection charge to cover, inter alla, the cost of the service entrance, meter and other customer-related facilities. The tariff then includes only the operating costs and the cost of amortizing the common facilities such as city - 152 - gate *tatons, fder mains, tc. The fixed costs could be *tiltad" to subsidis the lo»er-Incom custoers anu the comection charge could be subldised by a diect cash payment from the government to the distribution company. This system is much asaler to control, because the qualifications of lower-incom. consumers can be confirmed when the gas meter is installed. vW£' m%g' m9' gig' MO gL lsgl 9Rg 89 888~ W~ gngWlift ... ....... ........ .. ... ... ......i.. ...... . . w. . ..Wi . :-. . . . . 90E m 5 2 5 019 2 dut o5m 4~ggasealo la ee a 400 (9) ( c9f(z 2 (620 ( 40 (9089= 41*sseau Aib.3Ass M9 692 E Ou 90 s nis m - -aW u ff6' in' m9' DuL' 98'I £0p0 ZIL' in' =öf 9~isw AOlyp Wa0 G9 , 5L &t 59 585 S 06 2OL 56 sanam J~Io 5 pWpl 4o u2 jaa OM9O EEL mi 694 016 98' as-'L ez'mee Sam 31%s g jg o 95e m1, gI 2 92' mo mo' vo l99t snidAi ~- 9~A 3d M • 95, MOL' 9m»& mo0z ca'L (05 964 00U 40 ~ lidsad¥ .al9 6 6 ' mo-$ om* ' 5: 9£o ' gE0'1 mo ' flc e« 9 e~1agpc 1' m69% p9'd1 ~6 2 6 w 5 528' 90'£ Aff ieul E9 s19'a 42' MI 059 & 11 68'4 L .169 e*U os go 9 9 9 a2 62 aU M 9, 5 SL I.'S 9m9 9it 6p m 5. 94' 5 , s Mu ns 095 m'9 MN sMo'9 oa' 5 onth implementation period for his project, a 12-month grace period and a four-year repayment period. Large projects would continue to be handled, as before, directly by BNDES. 6. In December, 1985, through "Portaria Interministerial" No. 1877/85, the Ministries of Mines and Energy and of Industry and Trade launched the National Program for the Conservation of Electric Energy (PROCEL). ELETROBRAS has the responsibility for the overall management of PROCEL, through its Coordinating Group for the Conservation of Electric Energy (GCCE). The main objectives of the Program are far-reaching and ambitious: to rationalize the use of electric energy, so that the same product or service can be provided with less consumption, eliminating losses and securing an overall reduction in costs and in investment in new installations for the power sector. Such rationalization will be achieved through a more balanced relationship between electricity consumption and supply -- by enhancing efficiency in usage, on the one hand, and improving the allocation of resources on the other. These objectives are to be achieved through the implementation of adequate policies and legislation, the availability of necessary financing and the provision of a suitable institutional framework to develop the demand management and conservation programs. Although, as we have seen, implementation of a tariff structure which would reflect more appropriately the costs of supplying different consumer categories was initiated long before the creation of PROCEL, the use of tariff structures to induce a more rational use of electricity was incorporated as a formal objective of the Program. However, item 5.3 of the "Portaria" significantly excludes mention ' using the tariff level (as opposed to the structure) to reflect costs and promote effective energy conservation. 7. The wide-ranging goals of PROCEL (including a proposed World Bank loan now under consideration, as a "special project") may be summarized under the following headings: o Load Management, via (i) the consolidation of LRMC-based electricity tariffs and improvement of the average system power factor (reduction of the reactive power component); and (ii) the implementation of pilot programs for load management, through centralized and decentralized teleconmanded load control schemes and devices especially designed to restrict demand. o Energy Efficiency, via a set of measures (including legislation and rvles to establish norms, standards, codes etc., the formalization of energy saving targets, the carrying out of demonstration projects, studies and energy audits, the launching of publicity campaigns and educational programs, and the dissemination of The timit refers onty to iaprovements within the coopany itself. The purchase of damesticatty- anufactured equipment can be financed up to an aditionat NCM300 mittion by FINAME, a subsidiary of BNDES. - 156 - information) which would lead to an improvement in the end-use of electricity by residential, industrial and commercial consumers, especially with regard to (i) electrical furnaces; (ii) public sector and commercial buildings; and (iii) public lighting. o Loss Reduction, via the installation of capacitor banks in distribution networks belonging to different power utilities. o Cogeneration, via the undertaking of studies on cogeneration, using residual biomass fuels such as bagasse from the sugar and alcohol production processes. 8. All of the above programs provide significant potential for savings in peak capacity, energy or both. In aggregate, the Bank has estimated that the actions included in PROCEL will result in electric energy savings of about 3,940 GWh per year after 1992, equivalent to installing about 900 MW of additional generating capacity, valued at about US$1.3 billion. BLETROBRAS has offered higher projections, reaching a figure of 39,000 GWh by the year 2000, apparently assuming that action is also taken on tariff levels. 9. Because of the inadequate support available in the past to implement the policy reforms and measures included in PROCEL, as well as the financial constraints imposed on the power sector as a result of the current macroeconomic difficulties, which have restricted the availability of funds for electricity conservation purposes, PROCEL has been implemented at a slower pace than originally scheduled. At the same time, significant progress has been made on demand management in the complementary area of electricity pricing and through the implementation of basic low-cost conservation measures; and the Government also has been pursuing a comprehensive distribution network rehabilitation and lose reduction program, with the Bank's support, through Loans 2138-BR, 2364-BR and 2565-BR. Based on the initial success of these actions, the Government and the power sector are placing increased priority on demand management and conservation programs to complement supply planning options. 10. Given this increased priority for energy conservation and the plethora of programs in existence, CNE has proposed an umbrella Chamber of Energy Conservation -- which would include consumer representation -- to bring about some degree of consistency and coordination. The Chamber would function under the auspices of CNE and propose: national energy conservation goals; actions which would result in a more rational production and use of energy; measures to stimulate energy conservation; and minimum norms and standards of efficiency. Under CNEs proposal, a Program of Conservation and Rationalization in the Production and Use of Petroleum Products (CONPOT) would be created to complement PROCEL and subsume other programs such as PRODEL. Hence, the Chamber would assume responsibility to coordinate all existing energy conservation programs, notably CONPET and PROCEL, under the immediate management of ELETIOBRAS and PETROBRAS respectively; carry out and promote studies of rational energy use; and disseminate results. IBRD 22230R 50ý- 700 40 BRAZIL SURANAME ENERGY PRICING AND COLOMBIAVENEZUELA A ic INVESTMENT STUDY 0 cen NATURAL VEGETATION Savanina 00° 0-- - Steppe Steppe savanna -J Swamps and sand farmations Dense Amazon / Atlantic forest - ¾ Natal Pine foret - João Pesso - Semideciduovs forest a d4 Noi' Deciduous forest 10 1 Piaoer formation of marinu influence \[Z-Pianeer formation of river influence E Rr Areas of ecalagical tension High plains BOAI-I- Legal Amazonio boundary Rivers ) State capitals * National capital international boundaries 2rVENEZUA \. 20*° F ARAGUAY- ancir æ a ri t; BRAZIL PERU ARGENTINA j 50 24006W MØ I URUGUA0wT|5R - - OcTOBER 1991 IBRD 22231R BRAZIL ENERGY PRICING AND INVESTMENT STUDY NORTH-NORTHEAST TRANSMISSION NETWORK 500 kV WA DO rNDE 230 kV 138 kV 69 kV SAO"s Pia A Hydmelecric a Therml * substations Fr~ State Capitals FORIALEZA Rivers Reservoirs nl- Stle Boundaries CER intem~1ina Boundanies MARAUTRA RIO GRANDE ARANHAO CU* DO NORTE ONosa C«T P A R s oomos u su~ uz vEDR 0 F* PAkAfBA ®JooPeoa C LAG RE OM GRAN muSmURm SERG0MNOME___EG_P - s.1.O PERNA BUCO -gPMACU da Nort 20A 00 20 @MWAi6 N o -- .-AUL LARG o . S A . A O R A E J ø PENED0 TOCANTINS ENf SDO SERGIPE wwEC OUNDIN-Arcj B0 R aZI BAHIAo % BOMJESU~ APAA L-ido -GOIAÁS 2122'' 4C5BE 1991 IBRD 222321t Nqi UESA ama BAHIA A GROQ GGRÁS C~< .. BRAZIt ENERGY PRICING AND INVESTMENT STUDY SOUTH-SOUTHEAST TRANSMISSION NETWORK INAST e... D .C. ±6oo kv 0 ----- -- -- 50k mo om AE A A T-G GERAIS Cot Hydroelectric MT us - - n sookV \ ~statiarn PARAGUAY œP $f@g S- stateøounldaries i ~ P N~ nternaiona Boundaris ) - D 0 SUL ASANTA, CA~ CATARN Paw ~ a~ JANICAt JAtic O ARGENTINAA RYG> Y -0 0 20 300p TO moMmm ' Ian D URUGUAY CA~ O Rømrvoin oco 1991 IBI山0 22233R ㅣ숟 一一江 甲‘方一■■〕計一。｝ IØRD 22235R BRAZIL ENERCY PRICING AND INVESTMENT STUDY . . . . . . . . . . Power Generation Sources In Northern Region B R A Z IL Distance From load Centers i MARADA AREA QUA CU1ABA AREA tro VWH0 Wl LYADOR SELO ORIZONTE Ria DE tAI4EULO SAO PAUL CURIT tT00 CANOIO ALEGRE "MIN M1 IBRO 22236f BRAZIL ENERGY PRICING AND INVESTMENT STUDY Principal Regional Electrical - BRA z Interconnections NORTH LEFT SIDE MARABA NORTH - EAST NORTH\ CU [ABA\ SOUTHEAST CENTER-WEST TRANSMISSION LINES EXISTINGSOT POTENTIAL SIPTWSII IH IBRO 222371 BRAZIL ENERGY PRICING AND INVESTMENT STUDY Principal Coal Deposits -R A IIt MATO GROSSO DO SUL * SÅO PAULO PARAGUAY PARANA SANTA CATARINA ARGENTINA 9 RtO CRANDE DO SU L s 6 7 r 23 * COAL RESERVES: URUGUAY i Candlota COASTAL PLAIN 2 SAo Sepé 3 Iruf SH IED 4 Le8o-ButIå S Charquedas 6 Morungava 7 Chico Loma Š Santa Tresinha 9 Sul de Santa Cataina BRAZIL ENERGY PRICING AND INVESTMENT STUDY Interregional Power Flow IBRDA22238L Firm Energy (MW) u * 2 0 3 0 4 0 NORHT (INCLUDING MARANHAO) 1 2 4? 3MI 4 .300 to 6.030 NORHT EAST (LESS MARANHAO) 1 0 2 9 SOUTH - EAST 3 33 CENTRE - WEST 417 90(LESS MATO O01 GROSSO 0DO SU L) INTERREGIONAL FLOW 2 1 NGIONAL 3 430 1995 2 2000 0 3 2005 4 2010 3 0 410 SOUTH (INCLUDING MATO GROSSO DO SUL)