Loading...
HomeMy WebLinkAboutRailbelt Intertie Reconnaissance Study Vol. 4 Fuel Price Outlooks Crude Oil, Natural Gas & Fuel Oil 1988RAILBELT INTERTIE RECONNAISSANCE STUDY VOLUME 4 FUEL PRICE OUTLOOKS: CRUDE OIL, NATURAL GAS, AND FUEL OIL Prepared for Alaska Power Authority Anchorage, Alaska Prepared by ICF Incorporated 9300 Lee Highway Fairfax, Virginia 22031-1207 August 1988 RAILBELT INTERTIE RECONNAISSANCE STUDY VOLUME NUMBER 1 10 11 LIST OF VOLUMES VOLUME TITLE Economic and Demographic Projections for the Alaska Railbelt: 1988-2010 Forecast of Electricity Demand in the Alaska Railbelt Region: 1988-2010 Analysis of Electrical End Use Efficiency Programs for the Alaskan Railbelt Fuel Price Outlooks: Crude Oil, Natural Gas, and Fuel Oil Anchorage-Kenai Transmission Intertie Project Anchorage-Fairbanks Transmission Intertie Expansion and Upgrade Project Railbelt Stability Study Northeast Transmission Intertie Project Estimated Costs and Environmental Impacts of Coal-Fired Power Plants in the Alaska Railbelt Region Estimated Costs and Environmental Impacts of a Natural Gas Pipeline System Linking Fairbanks with the Cook Inlet Area Benefit/Cost Analysis PREFACE This report combines two reports prepared by the ICF-Lewin Energy Group for the Alaska Power Authority: an analysis of world oil price outlooks and a fuel price outlook for oil and natural gas in the Alaska Railbelt. These reports are part of a series of reports the Authority is funding to plan for Alaska electric power requirements. The former report, Section 1 of this volume, addresses world oil markets and should be useful to any energy analyst interested in comparative analysis of crude oil price projections. The latter report, Section 2, should prove useful to policy-makers and analysts interested in Alaska’s fossil fuel markets. Appendices to the latter report are in Section 3 of this volume. ICF-Lewin would like to thank the many persons who met with us to discuss the Alaska Railbelt energy markets. 20T00360 SECTION 1: Outlooks for World Oil Prices: Analysis of Alternative Schools of Thought SECTION 2: Outlook for the Alaska Railbelt Region: Oil and Natural Gas SECTION 3: Appendices 20T00360 SECTION 1 OUTLOOKS FOR WORLD OIL PRICES: ANALYSIS OF ALTERNATIVE SCHOOLS OF THOUGHT EXECUTIVE SUMMARY This report identifies the major schools of thought regarding the crude oil price outlook through the year 2010 and analyzes the reasoning and evidence put forward in support of each point of view. ICF collected oil market forecasts from organizations known for their expertise in this field. Subsequently, those forecasts judged complete were selected for use in this study. Table ES-1 lists these forecasts. A review of these forecasts reveals that they vary widely. Prices in the forecasts (expressed in 1987 dollars) range from $10 to $28 per barrel in 1990, from $10 to $47 per barrel in 2000, and from $10 to $56 per barrel in 2010. Nevertheless, the forecasts generally can be categorized into two schools of thought, a Consensus school and a Low Price school. The Consensus school has been called the "3:2:1" forecast by Texaco. This school’s assumptions for the free world economies are that: -- economic growth will proceed at about 3 percent annually; -- energy demand will grow at about 2 percent annually; and -- oil demand will grow at about 1 percent annually. In a typical Consensus forecast, OPEC regains control of the world oil market after 1990 and uses its market power to increase oil prices at about 3.5 percent annually in real terms through 2010. An alternative view is expressed by the Low Price school, in which forecasters believe that real oil prices will rise very little, if at all, over the 1990-2010 period. Forecasters in this school perceive a world of 20C0107 Page ES-1 Table ES-1 OIL PRICE FORECASTS DATED DECEMBER 1986 AND 1987 (1987 Dollars Per Barrel) (1) ECE 21.06 46.85 (2) INSEE 24.95 43.17 (3) ORIE 23.63 36.16 (4) DOE/NEPP 19.46 34.28 (S) DOE/EIA 13.38 18.52 22.66 27.78 = 33.95 42.19 (6) IEA 20.01 31.59 (7) TIASC 23.69 28.43 (8) CHVRN 18.95 28.43 (9) ICF 11.30 18.40 27.50 20.00 28.40 37.00 (10) GRI 20.53 27.69 (11) CERG 18.95 25.27 (12) ITAST 23.69 25.16 (13) PG&E 17.02 19.37 20.70 18.80 26.90 32.78 (16) WK 16.42 20.95 (15) ADOR 15.18 17.81 (16) ARTA 10.29 15.44 20.58 10.29 15.44 20.58 COUNT 1/ 16 COUNT 1/ ; 16 MEAN 1/ 19.77 MEAN 1/ 29.28 MEDIAN 1/ 19.46 MEDIAN 1/ 28.40 (17) CEC 18.51 KEY: FORECAST: ssse=sess (1) ECONOMIC COMMISSION FOR EUROPE (2) @) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (16) «15) (16) (17) NOTE: INSTITUT NATIONAL STATISTIQUE ET ETUDES ECONOMIQUE ORI, EUROPEAN ENERGY SERVICE DEPARTMENT OF ENERGY/WATIOWAL ENERGY POLICY PLAN - DRAFT PROJECTIONS DEPARTMENT OF ENERGY/ENERGY INFORMATION ADMINISTRATION ANNUAL ENERGY OUTLOOK INTERNATIONAL ENERGY AGENCY INTERNATIONAL INSTITUTE FOR APPLIED SYSTEMS ANALYSIS - CONVENTIONAL SCENARIO CHEVRON CORPORATION ICF INCORPORATED GAS RESEARCH INSTITUTE CAMBRIDGE ENERGY RESEARCH GROUP INTERNATIONAL INSTITUTE FOR APPLIED SYSTEMS ANALYSIS - TECHNOLOGICAL DEVELOPMENT PACIFIC GAS AND ELECTRIC COMPANY WORLD BANK ALASKA DEPARTMENT OF REVENUE ARLON TUSSING ASSOCIATES CALIFORNIA ENERGY COMMISSION 1/ - COUNT, MEAN, AND MEDIAN ARE CALCULATED USING BASE CASE FORECAST PRICES ONLY. 20€0107 24.30 20.76 10.29 COUNT 1/ MEAN 1/ MEDIAN 1/ ECE INSEE ORIE DOE/WEPP DOE/EIA IEA ITASC CHVRN ICF Gai CERG IIAST ARTA cec 34.75 38.10 42.01 35.80 31.69 32.38 49.70 51.89 15.44 20.58 35.78 35.80 41.89 DATE: 3/87 12/86 12/86 2/87 12/86 4/87 10/87 6/87 1987 11/87 4/87 3/87 9/87 9/87 7/87 4/87 Page ES-2 energy abundance in which technological development will ensure that substitutes always appear when any resource becomes relatively scarce and expensive. Table ES-2 presents the three crude oil price forecasts used to evaluate petroleum and natural gas markets in this study. In addition to the Consensus and Low Price schools, a third forecast is included which represents the lower range of the Consensus school of thought Table ES-2 ALTERNATIVE OIL PRICE SCHOOLS OF THOUGHT (1987 Dollars Per Barrel) 1990 2000 2010 Consensus $20 $30 $40 Consensus (low) 18 24 30 Low Price School 14 18 20 ASSUMPTIONS BEHIND THE TWO SCHOOLS OF THOUGHT The Consensus school is composed primarily of organizations that have developed detailed models of the world energy market, which they use to produce their forecasts. For this reason, their assumptions can be readily reviewed and analyzed. The Low Price school generally bases its forecast on a more general philosophical approach. The market assumptions behind the Low Price forecasts are, therefore, less explicit. A review of the reports from these two schools suggests the following implicit or explicit differences in their market assumptions: e The Low Price school believes energy supplies from the non-OPEC countries will be more abundant at low oil prices than assumed in the Consensus forecast. 20C0107 Page ES-3 e The Low Price school believes that OPEC will be willing to produce more oil at a low price than assumed in the Consensus forecast. ° The Low Price school believes energy efficiency gains will be greater than assumed in the Consensus forecast. e The Low Price school believes that substitute fuels are viable in the transportation sector at low oil prices. e The Low Price school believes that oil cannot compete in the boiler market if oil prices exceed $20 per barrel. ICF’s analysis of these assumptions indicates that energy supplies could be more abundant, OPEC production could be greater, and energy efficiency could be greater at low oil prices than assumed in the Consensus forecasts. Substitutes for petroleum fuels, however, are not viable in the transportation sector at low oil prices. Further, only gas is a good substitute for oil in the boiler market at oil prices below $20 per barrel. At this oil price level, coal cannot compete with oil in industrial boilers, and coal and nuclear energy cannot compete with combined cycle units in the electricity generating sector. CONCLUSION The assumptions behind the Consensus forecasts may be too conservative with respect to world energy supplies and energy efficiency gains at a low level of oil prices. The Low Price school, on the other hand, appears to have overestimated the capability of substitute fuels to compete with oil at low oil prices. Given the level of uncertainty associated with worldwide oil supply and demand relationships, however, neither the Consensus nor the Low Price schools of thought can be easily rejected. 20C0107 Page ES-4 TABLE OF CONTENTS EXECUTIVE SUMMARY INTRODUCTION Organization of the Report THE OIL PRICE FORECASTING PROBLEM The Forecasting Problem in Practice ALTERNATIVE OIL MARKET OUTLOOKS IN 1987 Forecasts Obtained For This Study Forecasts Obtained from International Energy Workshop Survey California Energy Commission Survey Forecasts Used to Identify Schools of Thought Classification of Forecasts ANALYSIS OF ALTERNATIVE SCHOOLS OF THOUGHT Consensus Forecast (1987) Critique of the Consensus Forecast Changes In Consensus Price Forecasts Low Price Forecast Critique of the Low Price Outlook Creation of a Detailed Low Price Market Overlook Review of Oil Demand Assumptions Economic Growth Rates Energy/GDP Ratios Oil Versus Its Substitutes Oil Demand in OECD Countries Transportation Sector (OECD) Industrial Sector (OECD) 20C€0107 II-1 II-3 III-1 Ti -1 III-2 III-4 III-5 III-8 Iv-1 Iv-2 IV-6 IV-7 Iv-9 Iv-10 Iv-12 Iv-13 Iv-13 Iv-14 Iv-15 IV-16 Iv-17 Iv-21 TABLE OF CONTENTS (continued) Electricity Generation Sector (OECD) Residential/Commercial Sector (OECD) The OECD Gas Market at $2.75 per MCF Non-OECD Oil Demand Assumptions Review of Oil Supply Assumptions U.S. Production CPE Exports Other Free World Production OPEC 1 Specification of a Low Price Forecast in 2000 Conclusions Appendix A -- Comparison of Saudi Light and Alaska North Slope Crude Value Appendix B -- ICF Forecast of Future Production Potential from Non-OPEC Free World Countries 20C0107 IV-24 Iv-24 Iv-25 IV-27 Iv-29 Iv-29 Iv-30 Iv-30 Iv-31 Iv-32 Iv-33 B-1 ES-1 ES-2 III-1 III-2 III-3 III-4 III-5 Iv-1 Iv-2 Iv-3 Iv-4 IV-5 II-1 III-1 Iv-1 Iv-2 IV-3 IV-4 IV-5 IV-6 IV-7 IV-8 200107 LIST 0 ‘S_AND FIGURES Oil Price Forecasts Dated December 1986 and 1987 Alternative Oil Price Schools of Thought Forecasts Sought via Direct Inquiry Forecasts Obtained from International Energy Workshop Survey Oil Price Forecasts Dated December 1986 and 1987 Consensus Economic Growth, Energy, and Oil Demand Forecasts Alternative Oil Price Forecasts 1985 Oil Consumption by Sector 1985 Gas Share of Oil/Gas Market Fuel Economy for Passenger Vehicles EIA Forecast of Industrial Oil Consumption Potential Free World Oil Market Outlooks Conceptual Model of the Oil Market Oil Price Forecasts Dated December 1986 and 1987 Illustrative Consensus Free World Demand Outlook Illustrative Consensus Free World Energy Consumption Outlook Illustrative Consensus Oil Supply Outlook World Oil Prices 1886-1986 Energy/GDP Ratios Over Time U.S. Passenger Car Fuel Efficiency Economics of High Efficiency Passenger Vehicles Lower-48 Onshore Drilling Costs over the 1970-86 Period ES-2 ES -3 III-3 III-4 III-7 III-9 III-10 Iv-17 Iv-17 Iv-20 Iv-21 IV-33 II-2 II1-6 Iv-3 Iv-4 IV-5 Iv-9 Iv-15 Iv-19 IvV-22 IV-26 CHAPTER I INTRODUCTION This report has two purposes. First, it identifies the major forecasts and classifies them into schools of thought regarding the crude oil price outlook through the year 2010. Next, it analyzes the reasoning and evidence put forward in support of each point of view. ICF’s approach for this work was to first collect the energy market forecasts or written reports prepared during late 1986 or 1987 by organizations known for their work in oil market forecasting. Next, forecasts initially appearing methodologically complete or important for other reasons were identified. Subsequently, these forecasts were sorted into categories which could be considered "schools of thought." The analysis also examined each school of thought to identify critical assumptions which cause the price outlook to differ from one school to another. Finally, the report reviews and assesses the empirical basis for these critical assumptions. The purpose of this report is not to present and document ICF’s forecasts, but rather to present forecasts principally prepared by other groups. While no informed forecaster can or should be completely impartial, ICF has made a conscious effort to keep this report from becoming a justification of ICF’s previous views on the future oil market outlook. Nevertheless, ICF’s most recent forecast was included in the group of forecasts examined. ORGANIZATION OF THE REPORT This report has four chapters. The next chapter provides background information on the oil price forecasting problem. Chapter III presents the forecasts collected for this study and categorizes them into schools of thought. Chapter IV presents an analysis of the assumptions adopted by these schools of thought. 20C€0107 CHAPTER II THE OIL PRICE FORECASTING PROBLEM The concept of a market price is fundamental to economic theory. Like any other price, the oil price at any point in time is simply the price at which the demand for oil and the supply of oil are in balance in the oil market. A forecast of oil prices is, therefore, based either implicitly or explicitly on a forecast of the supply and demand for oil. Conceptually, future oil prices are forecast by first developing a set of oil demand/price relationships and a set of oil supply/price relationships and then by calculating for each year the price at which oil demand and supply are in balance. This is the process that most forecasters use (sometimes implicitly and sometimes in extremely detailed mathematical models) to develop their oil price forecasts. The starting point for these forecasts is generally the most recent year for which complete energy consumption, supply, and price data are available. Since most of the data come from public sources, most forecasters begin the forecasting process with the same set of historical information. Differences in forecasts, therefore, are generally based on different views about future demand/price and supply/price relationships. The key factors in these relationships are shown in Figure II-1. The key factors affecting the future oil demand/price relationship are: ° future economic growth (worldwide) ; ° future structure of the world economy; ° future prices (and implicitly the supply) of energy substitutes for oil; and ° technological change. 20C0107 Page II - 1 FIGURE II-1 CONCEPTUAL MODEL OF THE OIL MARKET NON-OIL ENERGY SUPPLY | Y | onl Ol ECONOMIC | ENERGY | °EMANO | on aes Ol | | _GOVERNME Tf GROWTH DEMAND oa | MARKET on Ree eee | . POLICIES PRICES | PRICES BASE | | NEW OIL NEW TECHNOLOGY PRICES TECHNOLOGY ; In general, higher economic growth leads to higher energy consumption, but a shift from energy-intensive to service industries greatly limits this tendency. In addition, the ready availability of low-cost oil substitutes can effectively put a ceiling on oil price increases, and technological change can greatly reduce energy requirements even for the same level of economic activity. A forecast of the future oil demand/price relationship must implicitly or explicitly take these future changes into account. 20C0107 Page II - 2 The key factors relating to the future oil supply/price relationship are: ° the physical extent, nature, and location of the oil resource base; ° the technology available to find and produce oil; ° the cost of this finding and producing process; ° government policies covering leasing, taxation, royalties, and production controls; and ° anticipated military activity or civil unrest. Clearly, technological changes can greatly reduce the cost of finding and producing oil, but the actual geological resource base is a key determinant of cost. On the other hand, the potential supply available at a particular price can be greatly restricted by government controls, as it clearly is in the case of OPEC. Explicitly or implicitly all supply/price forecasts must take these factors into account. THE FORECASTING PROBLEM IN PRACTICE The many factors identified above reveal the complexity of the oil price forecasting problem in practice. The problem is particularly complex because the oil market is a world market, and the demand and supply relationships must implicitly or explicitly be examined for every country in the world. Three aspects of the problem, however, are particularly vexing: as Th ort- al : equired to balance the oil market are very different. Both the supply of oil and the demand for oil are significantly affected by investment over a long period of time. As a result, prices can change very significantly in the short-run without significantly changing oil supply or demand. Oversupply can severely depress prices, while a slight shortage can cause a major price increase. This means that short- run prices can be very unrepresentative of the long-run price level required to keep the oil market in balance. In other words to : o__inhe v. ty as_a_ st in int fo dev i -te ce ends 20C0107 Page II - 3 ae elatio i tween di factors ai oil demand complex that it Vv unpredictab over _a_lo iod_ of time, The demand for oil is related to the demand for energy, which in turn depends on economic growth, energy prices, and technological change. The demand for oil, in turn, depends on the price and availability of all other types of energy. While forecasters agree on the nature of these relationships, they do not agree and cannot predict precisely the quantitative interrelationships. e net re t_ is a wide 2.0 usible u e oi lemand timates at a) ve utu: ice 6 ti of low-cost _o s uw ers’ 2. bi oil market. In a competitive market low-cost resources are produced first and prices rise as necessary to cover the cost of increasingly marginal production sources. If demand falls, prices fall slightly as the highest cost producers shut down. This continuous supply cost relationship lends some stability to a competitive market. The current oil market, however, is not structured this way. Due to the activity of the OPEC production cartel, the world’s lowest cost oil producers have the marginal production capacity, and there is no relationship between the market price and the marginal supply cost. esult is that t oil vi be i arbi Despite these major uncertainties, forecasters do regularly create oil price forecasts which extend to the year 2000 or beyond. Each forecast attempts to tie together in an internally consistent way a view of how the oil market will develop over the forecast horizon. In this report we explain how forecasters do this and provide an analytic review which can guide a reader in assessing the likelihood of alternative forecasts. 20C0107 Page II - 4 CHAPTER III ALTERNATIVE OIL MARKET OUTLOOKS IN 1987 To identify the different schools of thought regarding crude oil price forecasts, oil market forecasts were obtained from many different sources. These sources included oil companies, trade associations, financial institutions, consulting firms, utility companies, international agencies, the U.S. Department of Energy, and the Alaska Department of Revenue. Forecasts were obtained in three ways: through direct inquiry, from the recent International Energy Workshop survey (by Dr. Alan Manne of Stanford University), and from a California Energy Commission survey. The two surveys served as secondary sources for this study. They were used to expand the list of forecasts obtained directly. FORECASTS OBTAINED FOR THIS STUDY Of the many firms and agencies contacted directly, some provided us with forecasts, some were unwilling to release forecasts, and some did not have up-to-date forecasts. This review includes only those forecasts dated December 1986 or later because earlier forecasts could not have taken into account recent developments in the oil market, such as the mid-1986 collapse in world oil prices. The first step in collecting the oil price forecasts was to determine a universe of potential forecasters to contact directly. ICF began by contacting 11 major oil companies. We were able to obtain oil market forecasts from four companies: Ashland Oil, Chevron, Conoco, and Texaco. However, only Chevron’s forecast met the acceptance criteria. The forecasts of Ashland and Conoco were outdated (August 1986 and September 1986, respectively), and Texaco’s oil market forecast (dated October 1987) did not include a forecast of oil prices. The seven other oil companies currently do not release their forecasts to the public. We next contacted trade associations whose members are very concerned about the oil market, including the American Gas Association, the American 20C0107 Page III - 1 Petroleum Institute, the National Petroleum Council; and the Gas Research Institute. Of these four, we were able to obtain oil price forecasts from the latter three. Of the three obtained, only the reports of the National Petroleum Council and the Gas Research Institute were dated 1987; the American Petroleum Institute's forecast was dated July 1986. The next contacts were to consulting firms, utilities, and international agencies. Some consulting firms and agencies did not have oil price forecasts. Others refused to release their forecasts. Some provided market forecasts but did not include a forecast of oil prices (such as the Lawrence Berkeley Laboratory). We obtained forecasts from the World Bank, Pacific Gas & Electric, and Arlon Tussing Associates (ARTA). Finally, we obtained forecasts from several government agencies. From the U.S. Department of Energy (DOE), we used the Energy Information Administration's Annual Energy Outlook, 1986 and the International Energy Ou 86. We also received forecasts from the National Energy Board of Canada and the Alaska Department of Revenue. Table III-1 lists organizations contacted, indicates the availability of oil market forecasts from those organizations, and indicates the relevance of these forecasts for this study. Those organizations whose forecasts were used to identify schools of thought are shown in bold type. Forecasts Obtained from International Energy Workshop Survey The next step we took was to obtain Dr. Alan Manne’s summary results of the International Energy Workshop Poll (dated January 1988), conducted at Stanford University. This workshop is a survey of 1985-87 oil price forecasts prepared by oil companies, government agencies, consulting firms, international agencies, and research centers. For our purposes, we considered relevant those forecasts dated December 1986 or later, which included approximately 20 forecasts. Of those twenty, we narrowed the sample to forecasts that were (1) produced by well-known organizations, (2) comprehensive and complete in their scope, and (3) not included in the forecasts already obtained directly. 200107 Page III - 2 TABLE III-1 FORECASTS SOUGHT VIA DIRECT INQUIRY Reason Not Obtained| Reason Not Relevant Forecast No Don’t Forecast No Price Forecaster Obtained? Forecast Release evant? _ Info. Outdated | AGA NO x Alaska Department of Revenue YES YES API YES NO X ARTA YES YES AMOCO NO xX Ashland Oil YES NO xX ARCO NO xX Cambridge Energy Resrch Group NO xX | Chevron YES YES Conoco YES NO x DOE National Energy Policy PlanYES NO DOE/EIA Annual Outlook YES YES x Exxon NO Xx | GRI YES YES IEA NO Lawrence Berkeley Laboratory YES NO xX Mobil NO xX National Energy Board, Canada YES NO xX NPC YES NO Xx Norwegian Royal Ministry of NO xX Petroleum and Energy Pacific Gas and Electric YES YES PlanEcon, Inc. NO x Shell NO xX Sohio NO xX Sun Co. NO x Texaco YES NO xX World Bank . YES | YES ote: Organizations whose forecasts were used to identify schools of thought are shown in bold type. 20C0107 Page III - 3 International Energy Workshop Survey forecasts from the Manne survey from the following organizations were added to the original forecasts collected: the International Energy Agency (IEA), Data Resources Inc. - Europe, the Economic Commission for Europe, the Institut National Statistique et Etudes Economique, the International Institute for Applied Systems Analysis, and the Cambridge Energy Research Group (CERG). A draft of the DOE National Energy Policy Plan forecast was also included in this group.t TABLE III-2 FORECASTS OBTAINED FROM INTERNATIONAL ENERGY WORKSHOP SURVEY Forecaster Data Resources, Incorporated - Europe DOE, National Energy Policy Plan (draft) Institut National Statistique et Etudes Economique International Energy Agency International Institute for Applied Systems Analysis Economic Commission for Europe Cambridge Energy Research Group California Energy Commission Survey The California Energy Commission (CEC) periodically undertakes a survey of oil market forecasters to obtain a baseline forecast for energy policy development purposes in California. The Commission's most recent survey was carried out in March and April of 1987 and included 33 forecasts from academic, financial, government, industry, and private consultants. The CEC does not identify their sources, so their survey results can only be used to confirm that forecasts collected by ICF are a representative sample of available forecasts. The Commission has reported that private consultants and financial institutions have had lower-than-average price forecasts, and oil companies have had higher-than-average price forecasts. Academic and la comprehensive analysis of the 20 forecasts presented in the survey was beyond the scope of this study. Consequently, the exclusion of certain forecasts from ICF’s lists does not necessarily imply that these forecasts are invalid. It may only mean that ICF did not have enough information to evaluate them. 20C0107 Page III - 4 government institutions have had forecasts slightly above the average obtained in their surveys. 2 Forecasts Used to Identify Schools of Thought The following organizations provided the forecasts used to identify schools of thought on oil price forecasts. Economic Commission for Europe Institut National Statistique et Etudes Economique Pacific Gas and Electric Chevron Corporation International Institute for Applied Systems Analysis Cambridge Energy Research Group Gas Research Institute ICF Incorporated DOE (Draft National Energy Policy Plan) DOE/EIA (Annual Energy Outlook) Data Resources, Inc. (European Energy Service) Alaska Department of Revenue International Energy Agency World Bank ARTA California Energy Commission eoooo0o000000000000 These oil price forecasts were plotted in a graph to show the projected prices for 1990, 2000, and 2010. (Only about half of the forecasts projected oil prices in 2010; the others stopped at 2000.) Three forecasters -- ICF, PG&E, and DOE/EIA (Annual Energy Outlook) -- projected Base, High, and Low Cases in their forecasts. The International Institute for Applied Systems Analysis provided two scenarios, conventional and one of technological development. Arlon Tussing Associates (ARTA) provided only a range of prices; ICF selected the midpoint of the range as ARTA’s Base Case forecast. These price forecasts are shown in Figure III-1. The numbers plotted in the figure are shown in Table III-3. Prices (in 1987 dollars per barrel) ranged from about $10 to $28 per barrel in 1990, from about $10 to $47 per barrel in 2000, and from about $10 to $56 per barrel in 2010.3 2california Energy Commission, LET of 0 Forecasts Summary of Round II (July/August 1986) Results, March 1987. 3unless otherwise stated all prices in this report are in 1987 dollars. 20C0107 Page III - 5 60 ' 4 $0 4 ” 1 1 " 1 40 + 1 we Lay 4 ‘ ' 30 4 it i] 1987 J aS 1 $/BBL L 'n L 4 20 4 L L : ' i) 10 4 L L (—_——_—___Ssa ee _ _ eae __avxXxvovrCrrTlOOOOoooOOCCCCCCC 1990 2000 2010 NOTE: All points represent base case forecasts, except those designated by “L” or “H", which denote iow and high oll price scenarios, respectively. 20€0107 Page III - 6 FIGURE III-1 OIL PRICE FORECASTS DATED DECEMBER 1986 AND 1987 TABLE III-3 OIL PRICE FORECASTS DATED DECEMBER 1986 AND 1987 (1987 Dollars Per Barrel) 1990 2000 2010 LOW BASE HIGH LOW BASE HIGH LOW BASE HIGH (1) ECE 21.06 46.85 (2) INSEE 246.95 43.17. (3) DRIE 23.63 36.16 (4) DOE/NEPP 19.46 34.28 56.06 (5) DOE/EIA 13.38 18.52 22.64 27.78 33.95 42.19 (6) IEA 20.01 31.59 (7) LIASC 23.69 28.43 34.75 (8) CHVRN 18.95 28.43 (9) ICF 11.30 18.40 27.50 20.00 28.40 37.00 24.30 38.10 49.70 (10) GRI 20.53 27.69 42.01 (11) CERG 18.95 25.27 35.80 (12) TIAST 23.69 25.16 31.69 (13) PG&E 17.02 19.37 20.70 18.80 24.90 32.78 20.76 32.38 51.89 (14) WBK 14.42 20.95 (15) ADOR 15.18 17.81 (16) ARTA 10.29 15.44 20.58 10.29 15.44 20.58 10.29 15.44 20.58 COUNT 1/ 16 ? COUNT 1/ ’ 16 COUNT 1/ 8 MEAN 1/ 19.77 MEAN 1/ 29.28 MEAN 1/ 35.78 MEDIAN 1/ 19.46 MEDIAN 1/ 28.40 MEDIAN 1/ 35.80 (17) CEC 18.51 27.80 41.89 KEY: FORECAST: ACRONYM: DATE: seesss=s2 szsszesza =eezsz=s2 (1) ECONOMIC COMMISSION FOR EUROPE ECE 3/87 (2) INSTITUT NATIONAL STATISTIQUE ET ETUDES ECONOMIQUE INSEE 2/87 (3) DRI, EUROPEAN ENERGY SERVICE ORIE 12/86 (4) DEPARTMENT OF ENERGY/NATIONAL ENERGY POLICY PLAN - DRAFT PROJECTIONS DOE/NEPP 12/86 (5) DEPARTMENT OF ENERGY/ENERGY INFORMATION ADMINISTRATION ANNUAL ENERGY OUTLOOK DOE/EIA 2/87 (6) INTERNATIONAL ENERGY AGENCY IEA 12/86 (7) INTERNATIONAL INSTITUTE FOR APPLIED SYSTEMS ANALYSIS - CONVENTIONAL SCENARIO T1asc 4/87 (8) CHEVRON CORPORATION : CHVRN 10/87 (9) ICF INCORPORATED ICF 6/87 (10) GAS RESEARCH INSTITUTE GRI 1987 (11) CAMBRIDGE ENERGY RESEARCH GROUP CERG 11/87 (12) INTERNATIONAL INSTITUTE FOR APPLIED SYSTEMS ANALYSIS - TECHNOLOGICAL DEVELOPMENT IIAST 4/87 (13) PACIFIC GAS AND ELECTRIC COMPANY PG&E 3/87 (14) WORLD BANK WBK 9/87 (15) ALASKA DEPARTMENT OF REVENUE ADOR 9/87 (16) ARLON TUSSING ASSOCIATES ARTA 7/87 (17) CALIFORNIA ENERGY COMMISSION cEC 4/87 NOTE: 1/ - COUNT, MEAN, AND MEDIAN ARE CALCULATED USING BASE CASE FORECAST PRICES ONLY. 20C0107 Page III - 7 The average Base Case price in the forecasts is $20 per barrel in 1990, $29 in 2000, and $36 in 2010. The California Energy Commission March 1987 survey obtained similar prices for 1990 and 2000. That survey only extended to 2007, but an extension of the price trend to 2010 yields an average price of $42 per barrel. In most cases forecasters do not identify the type of crude used in the forecasts, but generally it is an average quality oil or Saudi light. The U.S. forecasts generally refer to the price of crude delivered to the U.S. Gulf. Alaskan North Slope (ANS) crude oil would be expected to have a similar price. A comparison of ANS and Saudi light crude values in Appendix A indicates that ANS crude has been worth on average about $0.28/barrel less than Saudi crude light over the last three years. CLASSIFICATION OF FORECASTS A review of the price forecasts reveals quite a divergence of opinion about the future oil market. Nevertheless, these forecasts can be divided into two very different schools of thought. One group is composed of organizations with large world energy models who develop detailed forecasts of energy supply, demand, and prices on a world-wide basis. These organizations generally have a similar view on the world economic growth (GDP), energy demand, and oil demand outlook. These views are summarized in Table III-4. 20C0107 Page III - 8 TABLE III-4 CONSENSUS ECONOMIC GROWTH, ENERGY, AND OIL DEMAND FORECASTS (1985-2000 Annual Average Rate of Growth %) oEcD} Countries Free World Countries Economic Energy Oil Economic Energy Oil FORECAST Growth Demand Demand FORECAST Growth Demand Demand CERG 206 1.3 0.5 CERG Qed 116 Tok DOE/EIA 26 1.3 0.4 DOE/EIA 228 Lis? 057 DOE/NEPP 2.1 1.1 0.1 DOE/NEPP 2.4 1.5 0.7 ICF 2.9 1.2 0.8 ICF 3.4 ead 1.2 IEA 2.5 4.2 0.3 TEXACO 2.8 1.9 1.1 TEXACO 2.6 1.5 0.8 Average i 1.3 0.5 Average 2.8 i? 0.9 1 The OECD is the Organization for Economic Cooperation and Development; it includes 24 industrial countries, including the United States, Canada, Japan, Australia, New Zealand, and the Western European countries. Texaco has dubbed the mainstream Consensus view in 1987 as the "3:2:1" forecast. In this forecast world GNP rises at about 3 percent annually, world energy consumption rises at about 2 percent annually, and world oil consumption rises at about 1 percent annually. The oil price associated with this 3:2:1 forecast generally rises in real terms at 2.5-4.5 percent annually over the 1987-2000 period. Post-2000 there are few complete forecasts, but this price trend generally continues in those forecasts which extend out that far. The other school of thought is composed of organizations who do not have large computerized models of world energy markets, but who believe that oil prices will not rise much in the future. These organizations include the World Bank, the Alaska Department of Revenue, and ARTA, Inc. Table III-5 presents three alternative crude oil price forecasts. The Low Price and Consensus forecasts represent the views of the Low Price and Consensus schools of thought in 1987. The third forecast represents the lower range of the Consensus school of thought. 20C0107 Page III 9 TABLE III-5 ALTERNATIVE OIL PRICE FORECASTS (1987 Dollars Per Barrel) 1990 2000 2010 Low Price $14 $18 $20 Consensus (low) 18 24 30 Consensus 20 30 40 20C0107 Page III - 10 CHAPTER IV ANALYSIS OF ALTERNATIVE SCHOOLS OF THOUGHT The review of current oil price forecasts in Chapter III identified two distinct schools of thought: the Consensus school, in which prices in constant dollars rise over time to a level around $40 per barrel (1987 dollars) in the year 2010, and the Low Price school, in which prices remain indefinitely in the $10 to $20 per barrel range (1987 dollars). Most of the Consensus forecasts have been developed using detailed models of the world energy market, which contain specific regional forecasts of economic growth, energy use per unit of GDP (gross domestic product), and energy consumption by fuel type. Many of these forecasts are based on analyses of fuel-switching economics and the supply cost and availability of alternative energy sources. Such disaggregation provides sufficient documentation for the Consensus forecasts to give them credibility. Nevertheless, this credibility has diminished because market analyses with the same level of disaggregation and documentation were providing very different price forecasts just a few years ago. In marked contrast to the Consensus forecast, the Low Price outlook is not based on the same level of disaggregated market analysis. Rather, this outlook is based on the view that energy is abundant, that oil is not unique, and that technology will advance at a sufficient rate to prevent price increases as the lowest cost resources are depleted. The Low Price forecast’s credibility rests heavily on the fact that oil prices have returned to the $10-20 per barrel range after ten years in which the Consensus forecasters were predicting that real prices would continue to increase. The weakness of this view is that it provides no fundamental reason why any commodity’s price should remain at the same level over a very long period of time. 20C0107 Page IV - 1 CONSENSUS FORECAST (1987) The Consensus "3:2:1" forecast in 1987 is based on a market outlook in which OPEC regains control of the world oil market after 1990 and uses its market position to increase oil prices in real terms through 2010. In this outlook increasing world demand for oil and stagnant non-OPEC production slowly reduces OPEC’s overcapacity, such that by 2000 OPEC is producing about 28 million barrels per day and has raised prices about $10 per barrel to $28-30 per barrel (1987 dollars). The U.S. Department of Energy/Energy Information Administration (DOE/EIA) Annual Energy Outlook 1986 is representative of the Consensus forecasts. In addition, it is extremely well-documented and publicly available. Although the DOE/EIA price forecast for the year 2000 ($33 per barrel) is slightly higher than the median in the consensus group, it follows the "3:2:1" assumptions used by most Consensus forecasters. At this time, the 1986 DOE/EIA Annual Outlook is about a year out-of-date, but reportedly the 1987 Outlook will be similar. The critical DOE/EIA world oil market assumptions on the demand side are shown in Figure IV-1. The Free World economies (the DOE/EIA forecast includes the centrally planned economies only as a net oil supply contributor) are assumed to grow at a rate of 2.8 percent annually over the 1985-2000 period. Energy consumption grows at 1.7 percent annually and oil consumption grows at 0.7 percent annually. These aggregate growth estimates are the weighted sum of lower OECD growth estimates and higher growth estimates for the OPEC and oil-importing developing countries (OIDC). Since EIA's demand estimates are a bit lower than the Consensus "3:2:1" estimates, they are slightly conservative as a representation of Consensus thinking, at least in 1987. 20C0107 Page IV - 2 FIGURE IV-1 ILLUSTRATIVE CONSENSUS FREE WORLD DEMAND OUTLOOK (Percent Annual Growth) 2.5% 2.6% | | | | | | | | | 1.4% | gop | 1-2% Gop s | 0.7% ENERGY ENERGY | oll , 0.2% U.S. OTHER OECD 3.6% 2.8% | | | 1.7% | Gop 7 | ENERGY | aor | : | | | enenay. 0.7% | | | on OPEC/OIDC FREE WORLD Source: DOE/EIA, 1986 Annual Energy Outlook. 200107 Page IV - 3 The energy consumption growth rate is projected to be less than the economic growth rate due to structural change in the world’s economy and technological developments. For example, over the 1985-2000 period, DOE/EIA projects a decline in the U.S. energy/GDP ratio of 1.4 percent annually.4 This rate falls between the 0.7 percent decline over the 1926-66 period and the 2.6 percent decline over the 1966-85 period. The world oil consumption growth rate is lower than the energy consumption growth rate because coal, gas, nuclear, and hydro power are projected to increase their market share. Figure IV-2 shows how energy consumption is projected to change between 1985 and 2000 in the DOE/EIA forecast. FIGURE IV-2 ILLUSTRATIVE CONSENSUS FREE WORLD ENERGY CONSUMPTION OUTLOOK (Quadrillion BTUs) 120 i 103.9 40 Source: DOE/EIA, 19 ua 4 DOE/EIA, Annual Energy Outlook 1986, February 18, 1987, p. 15. 20C0107 Page IV - 4 The critical assumptions in the DOE/EIA oil supply outlook for the 1985-2000 period are shown in Figure IV-3. U.S. production declines by 31 percent, while OPEC production rises by 65 percent. ‘Other free world production declines slightly. Net exports from the centrally planned economies fall to insignificance. OPEC’s overall share of free-world supply rises from 38 percent in 1985 to 55 percent in 2000. Other Consensus forecasts show some minor variations on these demand and supply trends, but the general story is the same. The mean oil price associated with these forecasts in 2000 is $28-30 per barrel. FIGURE IV-3 ILLUSTRATIVE CONSENSUS OIL SUPPLY OUTLOOK (Million Barrels/Day) a SS Sy QS RN WN » LAAN WU, ®. SFI TiN vee, x wie _ 0.8 U.S. OPEC OTHER CPE EXPORTS EXPORTS 1986 2000 Source: DOE/EIA, 1986 Annual Energy Outlook. 20C0107 Page IV - 5 Critique of the Consensus Forecast The methodology used to develop the Consensus forecasts is rigorous, and the results are credible, but the Consensus forecast suffers from three potential weaknesses: Ls It may miss the forest by focusing on the trees. By examining the known information -- which is based on existing institutional structures and technology -- the detailed forecast may miss fundamental changes which could significantly change future energy markets. An approach based on what is known is conservative and may overestimate the scarcity of any resource in the future. Ra The Consensus forecast may implicitly assume that oil is in some ways a unique energy resource for which other substitutes are imperfect. In reality, all resources are unique, but on the margin there are good substitutes for all of them. i The Consensus forecast expects that OPEC will continue to limit production to support increasing prices and will be unwilling to sell increasing amounts of $2 per barrel oil at $18 per barrel. Given that OPEC, and Saudi Arabia in particular, have a vast amount of low- cost oil resources, this assumption is not necessarily correct. The Consensus forecasters have a poor track record. The fact that a detailed set of assumptions supports the Consensus forecast is no longer sufficient to give it credibility. As has been shown repeatedly, these detailed assumptions can all be wrong. Forecasters have made major errors in the past in their forecasts of economic growth rates, energy/GDP ratios, non-oil energy supply growth, non-OPEC oil supply growth, and OPEC’s willingness to produce oil at low prices. Nevertheless, they offer the most rigorous attempts to tie together the latest understanding of world oil market forces in analytic, internally consistent models. 20C0107 Page IV - 6 Changes In Consensus Price Forecasts For the purpose of gaining some perspective, it is worthwhile to remember that $10 per barrel forever (in 1987 dollars) was the consensus forecast up until the OPEC oil embargo of late 1973. At that time oil prices were not a subject that concerned many people outside the oil industry. After the 1973 OPEC embargo the consensus forecast of future oil prices was raised continuously through 1980. During the 1974-87 period the consensus oil forecasts were prepared by a growing number of interested parties using increasingly sophisticated models. Since 1980 the consensus forecast was reduced each year until 1986. It has been relatively stable since 1986. Consensus thinking during the last twenty years can be summarized as follows :° ° Vintage I (Pre-1974): In the late 1960s the major oil companies were managing the Middle Eastern oil fields to match supply to demand at a price (in 1987 dollars) of about $10 per barrel. Their expectation at the time was that oil production could be increased as necessary at that price. By the early 1970s, however, demand was growing so fast that officials in some oil companies thought future prices would have to rise. ° Vintage II (1974-77): After the first OPEC price increase, views varied between those who thought OPEC could not hold the increased price (about $24 per barrel in 1987 dollars) and those who thought prices would These historical forecasts were divided into five vintages by Arthur Anderson & Company in The Future of 0 ices: The Perils of Prophecy, 1984, p. 10. The fifth vintage was adjusted and the sixth vintage was estimated by ICF. 20C0107 Page IV - 7 climb in the future. The Federal Energy Agency (FEA) projected constant real prices into the future. ° Vintage III (1978-79): By 1978-79 OPEC had held together, non-OPEC supplies had not _ responded significantly to the higher prices, and demand had not slackened substantially. As a result, the consensus forecast moved to one of rising prices in the future. ° Vintage IV_ (1980-81): The 1979 oil supply disruption shocked forecasters because it showed OPEC to be less dependable than expected and because short-run oil demand was relatively unaffected by the significant price increases in 1979. As a result, price forecasts were raised again. ° Vint Vv 982-1985): By 1982 world oil demand was falling rapidly and non-OPEC oil supplies were increasing rapidly. Forecasters realized that they had let short-run events influence their thinking to an exorbitant degree. Prices were forecast to decline during the remainder of the 1980s before rising again. ° Vintage V 986-87): In 1986 prices collapsed to $10 per barrel. Forecasters further reduced their estimates of future demand growth and future oil prices, but the consensus was that prices would rise again in the post- 1990 period. This review indicates that the use of a detailed disaggregated forecasting methodology does not necessarily ensure a more accurate outcome than from a general theoretical approach such as that used by the Low Price forecasters. 20C0107 Page IV - 8 LOW PRICE FORECAST The Low Price forecast is not based on an explicit worldwide energy market forecast, but the argument that oil prices will not rise does rest on some specific assumptions that can be reviewed. The Low Price school notes that resource scarcity fears have recurred throughout history but have never materialized due to technological development. They reject the assumptions that oil is a unique commodity whose price must rise. Instead they see a world of energy abundance in which substitutes always appear when any resource becomes relatively scarce and expensive. In the specific case of oil, they refer to its 100-year history from 1886 to 1986, shown in Figure Iv-4, which can be interpreted to support their assertion that the "normal" price of oil is low. FIGURE IV-4 WORLD OIL PRICES 1886 - 1986 so 40 1966 Dollars ., Per Barre on 1900 1900 1010 19020 1930 19460 1960 1960 1070 1000 | 1606 eee Sources: 1886-1981 Arlon R. Tussing, “Reflections on the End of the OPEC Era," view i onomic Conditions, XIX, No. 4 (December 1982), pp. 1-16. 1982-1986 DOE/EIA, Monthly Energy Review, October, 1987, p. 93. 20C€0107 Page IV - 9 Proponents of the Low Price outlook refer to the following specific assertions about the energy market as a basis to support the continuation of a price below $20 per barrel: ° Energy efficiency gains in oil-using equipment are sufficient to prevent a significant increase in future oil demand. ° At a price above $20 per barrel, oil is priced out of the boiler fuel market because gas, coal, and nuclear fuels can compete with oil at that price. ° Numerous "viable" alternative fuels such as_ LPG, compressed natural gas, ethanol, and methanol can replace petroleum fuels in the transportation market. They do not indicate the price at which these fuels are "viable". ° Extensive supplies of oil and natural gas are available outside OPEC at a price below $20 per barrel. ° Saudi Arabia alone could increase its oil production level to 20 million barrels per day at very low cost. Critique of the Low Price Outlook The principal criticism of the Low Price forecast is that the proponents have not gone through the exercise of showing how it could happen by examining the detailed oil supply and demand economics. The fact that there is a lot of low-cost oil and gas in the world does not ensure that enough will be produced to meet oil demand at a price below $20 per barrel. In addition, 100 years of low oil prices, or rather 100 years in which prices were low for 90 years, do not by themselves mean that oil prices will be low forever. Substitutes are imperfect and their existence cannot guarantee that oil prices will not rise. Even if energy is abundant, this does not ensure that one type of energy will not become relatively scarce over time and have its price rise relative to other energy types. 20C0107 Page IV - 10 The principal problems with the assertions backing the Low Price outlook are the claims made for the boiler market and the transportation fuel market. Economic studies made to date show that alternatives to petroleum are not attractive in the transportation sector when crude oil is priced below $20 per barrel. ° Using current methanol production technology, methanol is not competitive with oil products until crude oil prices rise to at least $30 per barrel, even if natural gas is available at no charge in isolated locations. ° Compressed natural gas and LPG fuels are not attractive when oil is available at a price below $20 per barrel because the potential savings from switching are too low. Even the use of low-cost natural gas in vehicles involves incremental out-of-pocket and other costs relative to petroleum fuels, which cannot be offset at such a low oil price. ° Ethanol production, even in Brazil, requires enormous subsidies to compete with petroleum fuels at an oil price of $20 per barrel.§ The weakness in the boiler fuel market argument relates to the economics of oil/gas substitutes in small boilers. While it is correct that coal and nuclear energy can substitute for oil in large baseload utility steam boilers at $20 per barrel, the economic advantage of doing so at that price is low. More importantly, coal at $2 per million BTU cannot compete with combined cycle turbines for power generation unless crude oil or equivalent gas prices exceed $22 per barrel. If the coal units require flue gas desulfurization, the coal break-even price rises to $28 per barrel. Brazil has discovered it cannot afford its ethanol production program and is trying to blend gasoline into ethanol to reduce the ethanol required for its ethanol-powered vehicle fleet. 20C0107 Page IV - 11 In smaller industrial boilers coal is not attractive as a boiler fuel when oil costs $20 per barrel. Even when oil prices were much higher and expected to remain so a few years ago, industrial boiler operators were not interested in taking on all the handling and other problems associated with coal. On the other hand, the Low Price oil and gas supply argument has a stronger basis. The amount of oil and gas available worldwide at $20 per barrel is enormous. Saudi Arabia alone probably could produce up to 20 million barrels of crude oil per day. The OPEC supply limitation is political, not technical or economic. Of course, although it is true that Saudi Arabia could produce more oil, it might choose not to do so. For example, at the moment Saudi Arabia is unwilling to produce even 5 million barrels per day. CREATION OF A DETAILED LOW PRICE MARKET OVERLOOK The Low Price forecasters do not have a disaggregated energy market forecast to support their outlook. If their price outlook is credible, however, it should be possible to create a corresponding energy market outlook for the year 2000. Although changing technology could cause 2010 to be far different from the past, the year 2000 is only 12 years away, and technology will not have transformed the world we know by that time. In this section we attempt to find plausible changes in the demand and supply assumptions behind the Consensus forecast which could cause the future oil price to remain at $18 per barrel or below until the year 2000. The starting point is an identification of the points of agreement between the Consensus and Low Price outlooks. Both outlooks see $18 per barrel as an important threshold because above that price oil is not competitive for power generation in large utility steam boilers. Both outlooks also agree that extensive oil and gas resources can be produced at an oil price of $18 per barrel. Both outlooks also agree that new technology will significantly reduce the energy/GDP ratio over time even at $18 per barrel. The fundamental differences between the outlooks are the following: 20C0107 Page IV - 12 ° The Consensus forecasters believe that at $18 per barrel energy demand would rise significantly and sufficient non-oil substitutes would either be unavailable or too unattractive to energy consumers to prevent a substantial increase in oil demand. Even at $30 per barrel the Consensus forecasters believe oil demand will be higher in 2000 than it was in 1985. ° The Consensus forecasters do not believe that a sufficient increase in world oil supply would be forthcoming at $18 per barrel to meet demand at that price. Even at $30 per barrel the Consensus forecasters believe that OPEC will have to increase production by 10 million barrels per day over 1985 levels to balance the market. These forecasters believe that OPEC will be able to use its market power to both raise prices and greatly increase its production rate. Clearly then, to support the $18 per barrel forecast for 2000, plausible changes in the assumptions behind the Consensus forecast must be identified which could reduce oil demand and/or increase oil supplies to keep the market in balance at this price. Review of Oil Demand Assumptions An overall review of the Consensus oil demand assumptions requires an analysis of economic growth rates, energy/GDP ratios, and the economics of oil versus its substitutes. To reduce the oil demand forecast, economic growth rates must be revised downward, energy/GDP ratios must be revised downward, or oil’s market share must be further reduced. The starting point for this effort is the EIA estimates of 2.8 percent annual GDP growth, 1.7 percent annual energy growth, and 0.7 percent annual oil consumption growth. Economic Growth Rates EIA's assumed free world annual economic growth rate of 2.8 percent is already quite low. Six years ago predictions of future annual growth were 20C0107 Page IV - 13 about 3.7 percent. The higher rates were based on the economic experience of the 1960s and 1970s, which had strong growth. The free world market economies grew at a 3.6 percent annual rate during the 1970s and at a 2.0 percent rate during the 1980-85 period. Current assumed rates reflect the lackluster performance of 1980-85, and in that respect they may be too conservative. In any event, EIA's projected free world economic growth rate, although not the lowest possible rate, should not be further reduced for projecting Base Case energy demand. Energy/GDP Ratios The relationship between economic growth and energy demand is tenuous. This relationship can be projected using aggregate statistical techniques or engineering analyses of energy-using devices, but neither can hope to capture with any precision the complex changes which occur as economics and technologies change over time. Substantial reductions in energy use per unit of GDP are technologically possible and can be identified. The rate at which technological developments will be adopted commercially is much harder to project. Normally, this rate is determined in part by energy prices, but it can.also be forced by government regulation. In addition, technological advances which make a product more marketable may have greater energy efficiency as a side benefit. Needless to say, the rate of improvement is heavily affected by the rate of investment in new technologies, which also varies considerably among countries. Changes in energy/GDP ratios also reflect structural shifts in the economy. A dollar of output generated in the service sector can require much less energy that a dollar of output from heavy industry. Figure IV-5 compares changes in energy/GDP ratios over time in the U.S., Western Europe, and Japan. Over the 1972-87 period the energy/GDP ratio declined at 2.7 percent annually in Japan, 1.9 percent annually in the U.S., and 1.2 percent annually in Europe. The higher capital investment rates in Japan may explain the greater decline in the energy/GDP ratio in that country. 20C0107 Page IV - 14 FIGURE IV-5 ENERGY/GDP RATIOS OVER TIME Source: Ed., "A World Awash with Oil," The Economist, November 14, 1987, Ds) 76). In the EIA forecast the free world energy/GDP ratio drops by 1.1 percent per year over the 1985-2000 period. While this pattern is obviously quite feasible it clearly does not represent the maximum feasible decline in the energy/GDP ratio over the forecast period. This ratio dropped much faster over the 1979 to 1987 period, after dropping at a slightly lower rate over the 1972 to 1979 period. Clearly, this component of the forecast requires analysis in greater detail at the sectoral level. Oil Versus Its Substitutes Oil has numerous substitutes, but many are not as attractive as oil when oil costs $18 per barrel. The principal, potentially available substitute for oil at this price is natural gas. Nuclear and hydroelectric power can also compete under some circumstances at this price. Coal is viable in existing steam boilers but not in new boilers at that price. The 200107 Page IV - 15 degree of potential substitution for oil by these energy resources requires a sector-specific and country (or region) specific analysis. As mentioned earlier, there are currently no energy substitutes which can replace oil in transportation at $18 per barrel. The substitutes being used currently, including ethanol and natural gas, can compete only if oil prices rise significantly above $18 per barrel or if the substitutes are subsidized or required by government regulation. Petroleum fuel prices are generally maintained above world market prices in most transportation markets. Particularly in Europe, high taxes on gasoline have been applied to reduce petroleum fuel use. Nevertheless, replacing $18 per barrel oil with higher cost substitutes, even if they appear attractive due to tax policy, is not a practice most countries can afford to any extensive degree for long. The best way to evaluate the potential for reducing energy consumption per unit of economic output and the potential energy substitutes for oil at an oil price of $18 per barrel is to examine oil use by sector in the OECD and non-OECD countries. This approach permits an analysis of the availability of economic substitutes for oil in developed and developing regions. The non-OECD countries generally differ from the OECD countries in terms of climate, technological development, and availability of infrastructure to transport natural gas. For this reason, there are some significant differences in energy economics between the OECD and non-OECD countries. Oil Demand In OECD Countries There are great differences in how oil is used even among the OECD countries, as shown in Table IV-1. Over half of all oil consumption occurs in the transportation sector in the OECD countries, but the proportion varies from only 27 percent in Japan to 63 percent in the U.S. Japan uses a lot of oil outside the transportation sector because its geographical isolation and lack of low-cost domestic energy resources have limited the substitution of other energy resources, particularly natural gas, for oil in 20C0107 Page IV - 16 these sectors. Europe has natural gas resources, but the amount of gas is less than in the U.S., so that gas has been a smaller share of total energy consumption than in the U.S. TABLE IV-1 1985 OIL CONSUMPTION BY SECTOR (Share) Transportation Industrial Electric Residential/Commercial US: .63 425 04 .08 Japan 27 -36 ee aL) Europe awh a2) LO) poe) Average 05! 129) 3} «a3 Table IV-2 shows the gas share of those markets in the OECD countries which can utilize either gas or oil. While natural gas has half the oil/gas industrial market in the United States and three quarters of the oil/gas residential/commercial market, it has a much smaller share in Japan and Europe. Japan uses no natural gas in its industrial sector and has only recently begun importing LNG in substantial quantities to back out oil in its electric generation sector. Europe uses proportionally less gas for electric generation than Japan, in part because government legislation was passed to limit the use of what were previously perceived to be scarce gas supplies for generation purposes. Recent changes in the future outlook for European natural gas supplies are leading to changes in government policies which limit gas use. Thus, future gas substitution potential must be examined on a sector-specific basis. TABLE IV-2 1985 GAS SHARE OF OIL/GAS MARKET Industrial Electric Residential /Commercial U8. -47 .74 aa Japan .00 oo .24 Europe ro 34 45 Transportation Sector (OECD) Given the high proportion of oil use in the transportation sector, future demand in this sector is a critical component of world oil demand. The EIA Outlook includes only a 0.3 percent annual oil consumption growth 20C0107 Page IV - 17 rate in the transportation sector of the OECD countries during the 1985-2000 period. The forecast of fuel use in passenger vehicles is a 0.9 percent annual decline. This growth rate is the net change resulting from a 2 percent annual growth in total miles driven and a 2.9 percent annual improvement in vehicle fuel efficiency. In contrast, consumption grows in the other parts of the sector. Alternative fuels are not projected to play a major role in the transpor- tation sector even though EIA’s transportation consumption estimates are based on a projected oil price of $33 per barrel in 2000. Ethanol and methanol cannot be produced and delivered (without subsidies) at a price which will compete with petroleum fuels in vehicles unless crude oil prices exceed $40 per barrel. The economics of natural gas use in vehicles is very complex because a gas supply system must be created. Nevertheless, studies show that gas can be economic in high-mileage fleet vehicles (e.g., taxis) in urban areas at a crude price which is less than $40 per barrel. The extent of transport market penetration at this oil price is related to local gas market conditions. Because gas is more desirable as a fuel in stationary than in mobile applications, gas use in vehicles will occur only in those areas where low-cost gas supply exceeds the demand in stationary applications locally or within pipeline distance (e.g., in New Zealand). Given current gas supply projections, this limitation means minimal use of gas worldwide in vehicles at oil prices under $40 per barrel. Oil consumption will be determined in part by the rate of turnover in the vehicle fleet, which delays the effects of the improving fuel efficiency in new vehicles. Nevertheless, by 2000 most of the vehicles on the road will be composed of the relatively fuel-efficient vehicles produced since 1985. As shown in Figure IV-6, the average new 1985 U.S. vehicle could travel 26.8 miles per gallon, which was over 50 percent farther than the average vehicle in the 1985 existing fleet. In the EIA forecasts the new fleet’s fuel efficiency rises to 35 miles per gallon in 2000, while the existing fleet's efficiency improves to the level of the new 1985 fleet. This efficiency gain seems potentially conservative, particularly since many recent technological developments seem to have considerable promise for reducing vehicle fuel consumption. In addition, the high tax rates on transportation fuels in many countries will maintain the commercial pressure to convert these developments into commercial practice. 200107 Page IV - 18 Miles per Gallon Source: 20C0107 FIGURE IV-6 U.S. PASSENGER CAR FUEL EFFICIENCY 40 ~-- New Vehicles 30 + | L 20 _ - 10 — Existing Fleet | | 0 | 1960 1966 1970 1975 1980 1965 1990 1995 2000 Historical - American Petroleum Institute, Basic Petroleum Data Book, January, 1987 Vol. VII, No. 1, Sec. XII, Table 8 and the Motor Vehicle Manufacturers Association. Forecast - DOE/EIA. Page IV - 19 The World Resources Institute recently published a report promoting fuel efficiency. The report documents some of the technological developments which make possible large vehicle fuel efficiency gains. Table IV-3 presents the fuel economy estimates associated with some existing and prototype passenger vehicles. TABLE IV-3 FUEL ECONOMY FOR PASSENGER VEHICLES Car Fuel Fuel Economy Curb Weight Passenger (miles per (kilograms) Capacity gallon) (persons) Commercial 1985 VW diesel 47 1,029 5 Jetta 1985 Nissan diesel 55 850 5) Sentra 1985 Ford diesel 55) 945 5 Escort 1986 Chevrolet gasoline 57 676 4 Suzuki Sprint Prototype VW Auto 2000 diesel 66 780 4-5 Renault EVE+ diesel 70 855 4-5 Toyota Ltwght diesel 98 650 4-5 Compact Source: World Resources Institute, Energy for a Sustainable World, 1987, p. 64. Most of these design efforts rely on diesel engines for heavy vehicles and gasoline engines for lighter vehicles. These estimates indicate that the new vehicles entering the U.S. fleet in 2000 could have an average fuel efficiency above 35 miles per gallon. The problem from a market penetration standpoint is that the economic incentive to raise fuel efficiency above 30 miles per gallon drops quite rapidly at low fuel prices because fuel costs become a smaller and smaller proportion of total vehicle operating costs, and the initial cost of the vehicle rises somewhat as fuel efficiency is improved. Figure IV-7 shows how total vehicle operating costs change as fuel efficiency is improved using existing technology. Ultimately, fuel 20C0107 Page IV - 20 efficiency is likely to improve as a byproduct of the effort to reduce vehicle emissions and improve vehicle operation, but the speed at which this will occur is highly uncertain, and the lag between technological improvement and fleet turnover will delay much of this change until after 2000. This discussion highlights the difficulty of predicting fuel efficiency improvements. Nevertheless, the technological developments underway lend support to the argument that oil use in the OECD transportation sector could actually decline over the 1985-2000 period, even with crude oil prices of $18 per barrel. Industrial Sector (OECD) The industrial sector is the second most important for OECD oil use, accounting for 29 percent of total OECD oil consumption in 1985. The EIA projection for oil consumption growth in the OECD industrial sector between 1985 and 2000 is also quite small, as shown in Table IV-4. An analysis of the fuel consumption projections reveals that changing gas use explains the different growth rates for the United States, Japan, and Europe. In the United States a tighter gas market is forecast to increase the oil share of industrial fuel use, while increasing gas supplies in Europe cause the oil share to decline. In Japan the falling oil share in the industrial sector is due only to structural change in the economy, efficiency improvements, and the increasing substitution of electricity for oil. No gas is used in Japan's industrial sector in the EIA forecast for 2000. TABLE IV-4 EIA FORECAST OF INDUSTRIAL OIL CONSUMPTION (Quadrillion BTUs) 1985 2000 -2000 Annua owt! U.S. 7.7 8.5 0.7% Japan 3.2 3.3 0.2% Europe 7.4 75 0.1% Total 18.3 193) 0.4% 20C0107 Page IV - 21 FIGURE IV-7 ECONOMICS OF HIGH EFFICIENCY PASSENGER VEHICLES Continuously — Vasiable Transmission Continuously Variable ~~ Tranemission S28 Weight — Reduction At Idle And Coast 3 33 : a Garaging. Parking, and Tolls Repairs. Parts. and Maintenance 0. en 20 10 as 6 $s 4 3.5 3 2.7 Liters Per 100 Kilometers The indicated energy performance is based on computer simulations of an automobile having various fuel economy improvements added in the sequence shown at the top of the graph. The base car is a 1981 Volkswagen Rabbit (gasoline version). The figure shows that the reduced operating costs associated with various fuel economy improvements are roughly offset by the increased capital costs of these improvements over a wide range of fuel economy. Source: F. von Hippel and 8.G. Levi. “Automotive Fuel Efficiency: The Opportunity and Weakness of Existing Market Incan- tives, Resources and Conservation (1963): 103-124. 20C0107 Page IV - 22 The keys to future industrial consumption of oil are technological improvement, gas availability, and the rate of substitution of electricity for fossil fuels within the sector. To both reduce the Consensus oil price outlook in 2000 from $30 per barrel to $18 per barrel and simultaneously improve energy efficiency and maintain gas market share in the sector is a tough proposition. At $18 per barrel for oil, gas prices would be about $2.75 per MCF. At this price gas supply availability could become quite constrained by 1995. The overall reduction in energy use per unit of industrial output is about 1.2 percent per year in the Consensus forecast, which is not particularly high considering that electricity is simultaneously increasing its market share from 39.5 percent in 1985 to 46.8 percent in 2000. Electricity is more expensive and must be used more efficiently than fossil fuels to justify substitution in the industrial sector. Greater efficiency gains than forecast are possible in the industrial sector, even at an oil price of $18 per barrel. Nevertheless, both electricity and coal would probably lose some market share growth if oil prices do not increase. The growing use of coal projected for 2000 in Europe's industrial sector undoubtedly would not occur. This loss would have to be offset by higher oil or gas consumption unless efficiency gains are underestimated. Natural gas is a direct substitute for oil in oil/gas-fired boilers, which is the marginal use of gas in the industrial sector in EIA’s forecast for 2000. If oil prices were to remain at $18 per barrel, could the EIA forecast level of gas be obtained? Since the industrial sector must compete with the electric generation and residential/commercial sectors for gas, this question cannot be answered without examining the total gas supply and demand situation at $2.75 per MCF. If the gas supply turned out to be quite limited at an oil price of $18 per barrel, then the demand for oil in the industrial sector could surge. 20C0107 Page IV - 23 Electricity Generation Sector (OECD) As a result of the oil price increases since 1973, the electricity generation sector no longer is a major oil consumer in the OECD countries. With rising oil and gas prices EIA forecasts significantly rising demand for electricity over the 1985-2000 period. The electricity consumption growth rate is projected to equal the average GDP growth rate of 2.8 percent per year. This rate is consistent with end use efficiency gains as well as a significant increase in energy market share over time. With rising oil and gas prices in the Consensus forecast, there is no incremental boiler demand for oil/gas over the 1985-2000 period. There is, however, increased demand for gas in combined cycle units in the mid-1990s because these units are less capital-intensive than coal units and can be built more rapidly, which is an advantage when electricity demand is uncertain. Further, combined cycle plants can generate electricity at lower cost than a large coal plant if oil prices or equivalent gas prices are less than $22 per barrel. This increased use of combined cycle units causes total oil/gas demand to rise in the U.S. electricity generation sector over the period. With stagnant projected gas supplies, this increased oil/gas demand ends up causing increased U.S. demand for residual fuel. This pattern is not repeated in Europe or Japan in the forecast, but it could occur in Europe if reduced gas prices and altered legislation permit the construction of combined cycle generation units. As in the industrial sector, gas availability is a key determinant of oil demand in the electricity generation sector. Residential/Commercial Sector (OECD) Natural gas and electricity have substantially reduced the use of oil in the residential/commercial sector, especially in the United States. Oil use is not projected to increase much in this sector in the Consensus forecast over the 1985-2000 period, and increases should be limited even if oil (and gas) prices are much lower than forecast in the Consensus outlook. Nevertheless, if oil prices were expected to remain at $18 per barrel, 20C0107 Page IV - 24 residential/commercial oil consumption would probably increase in Europe and Japan as new households are built and incomes rise. The OECD Gas Market at $2.75 per MCF Supplying gas at $2.75 per MCF requires that both transportation and production costs be covered at that price. Gas transport has a lot of up- front capital costs, but once pipelines and LNG transport equipment are built, variable costs are low. For this reason, a supply analysis must examine the location of potential supply sources and existing transport infrastructure to evaluate supply costs. The EIA gas supply forecast of 18 tcf for the U.S. in 2000 is similar to the 1985 level. At this supply level, off-peak pipeline capacity is in surplus and transport costs would be low. Canada has significant gas resources which can be produced for under $2 per MCF. EIA’s forecast does not include the maximum possible level of exports from Canada. Consequently, at $2.75 per MCF U.S. production would probably be lower, but Canadian gas supplies could exceed the level in the Consensus forecast. ICF/Lewin has previously estimated that up to 15 tcf could be produced in the lower 48 states in the year 2000 at $3.75/MCF, but no production cost reductions were included in this analysis. As shown in Figure IV-8, drilling costs in mature production regions have been declining at 3 percent annually or more over the past 15 years. If these production cost declines were to continue, 15 tcf of gas could be available at a price below $3.75 per MCF. ICF/Lewin concludes that total U.S. gas supplies (U.S. and Canadian exports) could conceivably be 18 tcf in 2000 even at $2.75 per MCF. In the European market gas supplies are projected to grow at about 2 percent annually over the 1985-2000 period, but much of the necessary transport infrastructure is already in place to move this gas. Incremental gas will be obtained from Russia, Norway, the Netherlands, or Algeria. Again, it appears that production costs could be low enough so that supplies could increase at 2 percent annually even at a gas price of $2.75 per MCF, and some of these supplies are already under contract. Nevertheless, all of these European supplies have important political elements, so the 2060107 Page IV - 25 FIGURE IV-8 LOWER-48 ONSHORE DRILLING COSTS OVER THE 1970-86 PERIOD (Average Cost of Wells Below 5000 Feet in Thousands of 1987 Dollars) 180 Cost Care > Cece Trene 160 @ Preactes 140 120 100 80 1970 1972 1974 1976 1978 1980 1982 1984 1986 Source: ICF/Lewin willingness of these potential suppliers to deliver at $2.75 per MCF is unknown. To obtain supplies at what is a low price relative to recent prices would require a realization on the part of suppliers that the prospects for future big price increases are poor which, in turn, depends on the oil market price outlook. The incremental LNG market is all but dead at $2.75 per MCF. Even for very short distances and with free gas, the cost of expanding LNG production and transport is about $2.00 to $2.50 per MCF. Consequently, at $2.75 per MCF LNG is not a financially attractive prospect for potential suppliers. In addition, much of the world’s potential source of LNG is in OPEC. If these countries expanded their LNG sales, they would cut into their own oil sales, and their profits on LNG sales would be much less than their profits an oil sales. Since LNG trade is not substantial in EIA’s forecast for 2000, however, the gas price reduction associated with an oil price decline to $18 per barrel would have little impact on forecast LNG sales. 20C0107 Page IV - 26 ICF’s conclusion, surprisingly, is that much of EIA’s forecast of year 2000 gas supplies at $5 per MCF ($33 per barrel of oil) could conceivably be available, grudgingly, at $2.75 per MCF. This is an important finding for the feasibility of the $18 per barrel forecast, particularly since (as discussed below) it also holds up for natural gas supply from the non-OECD countries. The demand for gas in the OECD countries would undoubtedly be higher than forecast by EIA if gas were available at $2.75 per MCF. At that price gas would replace coal in the European industrial sector and would look very attractive for electricity generation in combined cycle units, where it could replace baseload coal units in both Europe and the U.S. If additional combined cycle units were built instead of coal, they would reduce the gas supplies available for replacing residual fuel in utility steam boilers. Nevertheless, if electricity growth rates were lower than forecast in the 1995-2000 period, gas demand could be lower than forecast in 2000 even with $2.75 per MCF gas. The EIA forecast projects OECD electricity demand growth at 2.5 percent per year. If this growth slowed and generation capacity were overbuilt in 2000, then oil/gas demand would be lower than forecast and gas supplies might be sufficient to prevent a large increase in residual fuel consumption over the 1995-2000 period at an oil price of $18 per barrel. Non-OECD Oil Demand Assumptions Until recently, the Consensus forecasters thought the non-OECD countries would contribute to a continuing major increase in world oil demand, as these countries’ economies slowly caught up with the OECD countries. In the 1986 outlook, however, EIA projects an oil consumption growth of only 1.3 percent per year for these countries over the 1985-2000 period. The lack of data on the non-OECD countries and a lack of knowledge about their energy situation caused forecasters to underestimate the potential for fuel substitution in these countries. While it is true that many small countries do mot have the alternative resources, the technological capability, or the capital to shift away from oil, these countries are not the major non-OECD energy consumers. In addition, these 20C0107 Page IV - 27 small countries are in such a poor economic state that their overall energy demand is not growing very much. The situation is quite different in the larger countries where cheap hydroelectric power and natural gas are still plentiful. In addition, for numerous Asian countries, such as India and Thailand, domestic coal resources are sufficient to permit diversification away from oil. Moreover, the increasing sophistication of technical personnel now permits some non- OECD countries to operate nuclear power plants but this remains the exception rather than the rule. In the EIA Outlook, over the 1985-2000 period non-OECD gas consumption is projected to grow at 6.5 percent per year, coal at 3.1 percent per year, and other energy sources at 2.8 percent per year. For this reason, oil consumption growth is limited to 1.3 percent per year. At issue is whether these relative growth rates would hold up if oil prices were to remain at $18 per barrel. In part, the answer would depend on perceptions of future prices. If oil prices are expected to go up, then other energy sources look attractive and significant substitution out of oil would continue. If oil prices were expected to remain at $18 per barrel, then the rate of growth in consumption of coal would probably be reduced over the period. It does appear that the switch to gas will continue in the non-OECD countries even at oil prices of $18 per barrel. This price is high enough to justify exploration for oil in many countries, and drilling for oil often yields gas. Once this gas is found, it is often economical to produce and transport it at $3 per MCF or less. This price is too low to justify LNG sales among countries, but it is often adequate to support the creation of domestic pipelines to transport large supplies to industrial and population centers, typically to replace oil use in boilers. Even though the demand for oil in the transportation sector will increase, the vehicles used in non-OECD countries will have the improved efficiencies developed in the OECD countries. Also, since transportation fuels have been subsidized in recent years in the non-OECD countries, lower world oil prices will not have much effect on domestic transportation fuel prices. Substitution of gas for oil in the industrial sector and the 20C0107 Page IV - 28 limited use of oil for generation would limit oil consumption growth even at $18 per barrel. REVIEW OF OIL SUPPLY ASSUMPTIONS Free-world petroleum supply can be divided into four categories for analysis. ° UES. ° Exports from Centrally Planned Economies (CPE) ° Other free world ° OPEC As discussed earlier, OPEC is the most important supplier and is not supply- constrained to any meaningful degree in a physical sense. For this reason, ICF’s approach is to analyze what the other suppliers could reasonably make available at $18 per barrel and then discuss whether OPEC might be expected to make up the remaining demand/supply gap at that price. U.S. Production The U.S. oil production outlook has been very extensively analyzed. U.S. crude oil and NGL production is declining and is expected to decline significantly even if oil prices rise. At $18 per barrel U.S. production would decline more rapidly. The major unknown relates to Alaska oil production. With the TAPS pipeline tariff declining, the cost of transporting Alaskan crude to the lower 48 states is declining. By 2000 this cost will be very low. The major hope for significant incremental supplies are potential resources in the Arctic National Wildlife Refuge (ANWR). If this area is made available for leasing, and if the resources turned out to be significant, then incremental Alaskan production could keep U.S. supplies almost at the EIA forecast level even at a 2000 price of $18 per barrel, especially if improving oil production technology lowers production costs. 20C0107 Page IV - 29 CPE Exports The USSR is the world’s largest single oil producer, and its success at finding new fields will be a critical determinant of CPE export potential in 2000. Russian geologists believe that additional giant fields exist in Siberia, but so far most exploration efforts mostly have resulted in gas resource discoveries. The CIA has previously forecast the decline of Russian production and the elimination of all Russian oil exports. So far, this has not happened, and the CIA has changed its forecast. The Russians have managed to maintain oil production levels and replace declining Eastern European oil supplies with natural gas. As the rapid gas substitution policy continues, the CPE countries may be able to maintain net oil exports at current levels; it is a feasible outcome of Russia’s desperate battle to maintain a source of Western currencies for their import purchasing requirements. Obviously, an oil price of $18 per barrel provides a smaller incentive for Russian production efforts, and lower prices would also be expected to increase CPE oil consumption. Nevertheless, continuing CPE oil exports at or above EIA’s forecast for 2000 are potentially feasible even at a price of $18 per barrel. Other Free World Production Other free world production includes supplies from over 100 countries. Many of these countries have carried out quite limited exploration efforts to date, and data about what they have discovered are often unreported or unreliable. In this uncertain environment EIA has been conservative in its forecast of future production outside the U.S. Even with oil prices rising to $33 per barrel in 2000, EIA forecasts production from this region to remain about the same. ICF has also developed production forecasts for this region using optimistic geological resource estimates and more optimistic assumptions about future production behavior. A recent article on the results of this analysis is included as Appendix B to this report. ICF’s analysis suggests 20C0107 Page IV - 30 that oil production in this region could rise by 4-5 million barrels per day by 2000 at the Consensus forecast price of $28 per barrel. At $18 per barrel ICF has shown that production at current levels potentially could be maintained until after 2000. OPEC Ultimately, the biggest unknown in the world outlook is OPEC production. OPEC’s oil resources are enormous, and a large proportion of these resources are very low cost. Saudi Arabia and several other Middle Eastern countries currently control the rate of production from these resources. When the major oil companies made their plans to expand Middle Eastern production in the early 1970s, they expected oil prices to remain about $10 per barrel (1987 dollars). Expanded production was expected to be and still would be quite profitable at that price. What has changed since the early 1970s is control over production. In the early 1970s production from Saudi Arabia was expected to increase to about 20 million barrels per day and production from all the OPEC countries to 40 million barrels per day or more. In fact, production peaked at 31.3 million barrels per day in 1977 and has since declined. The oil is still there, but for political and economic reasons it has not been produced. Currently, OPEC is controlling production at a low level to prevent a decline in the market price below the $16-18 per barrel level. How much OPEC might be willing to expand production in the future at the $18 per barrel price is unknown. EIA uses a “price reaction curve" to forecast the OPEC supply/price relationship. In this method the historical relationship between OPEC capacity utilization and price is used as the basis for future pricing behavior. Capacity utilization is affected by production capacity, however, and production capacity can change significantly over time. In its forecast for 2000 EIA assumes this capacity will be about 33 million barrels per day. With this methodology EIA has forecast that OPEC would be willing to produce 28 million barrels per day at the $33 per barrel price, which is 85 percent 20C0107 Page IV - 31 capacity utilization. The same methodology would yield different results if OPEC's assumed capacity were greater. The only thing required to change this capacity or the historical production/capacity relationship would be an OPEC or a Saudi Arabian policy decision. SPECIFICATION OF A LOW PRICE FORECAST SCENARIO IN 2000 Table IV-5 presents a potentially feasible oil market outlook for the year 2000 at an oil price of $18 per barrel. In this forecast total free- world oil consumption in the year 2000 is only 500 thousand barrels per day above the EIA Outlook level of 51.3 million barrels per day. Slightly higher oil supplies from the U.S., CPE countries, and the other non-OPEC producers would still require 27.6 million barrels per day from OPEC. OPEC limited oil production in the early 1980s under the mistaken impression that the world would continue to buy large amounts of OPEC oil at the high prices set by OPEC at that time. Similarly, OPEC will restrict production in 2000 to obtain higher prices if it appears that this practice will maximize OPEC revenues. If world oil demand grows slowly at $18 per barrel, however, and non-OPEC supplies remain ample, then OPEC will rethink its pricing policy. It is quite possible that OPEC would be willing to produce 27.6 million barrels per day at the $18 per barrel price in 2000 if the other world oil supply and demand elements turn out as shown in Table Iv-5. 20C0107 Page IV - 32 TABLE IV-5 POTENTIAL FREE WORLD OIL MARKET OUTLOOKS (Million Barrels/Day) 2000 Consensu: Low Price Consumption 1985 ($30/Bb1) ($18/Bb1) U.S. L537) Thee Tile Other OECD 18.8 19.1. 19.1 OPEC/OIDC 122. 14.8 S53) Total 46.7 S153) 51.8 Supply e URS®, bln Tish: 7.4 CPE Exports 1.8 0.5 Ly Other non-OPEC 18 \c7, 14.9 L5i.7 OPEC EV a2 28.3 27.6 Stocks 0.8 <07.1> <0.1> 46.7 51-3) 51.8 CONCLUSIONS A critical review of the 1987 Consensus forecast and the Low Price forecast not surprisingly has revealed limitations in both. In ICF’s view the Consensus forecast is conservative with respect to oil and gas supply availability at the level of prices projected. In addition, energy efficiency gains could be greater than projected. If these two factors were to shift, the supply surplus would continue and there would be continuing pressure in the marketplace to keep oil prices from rising significantly over the 1985-2000 period. ICF has also determined that certain assertions made by the Low Price forecasters are not correct. Specifically, none of the alternative fuels proposed for the transportation sector appear economical when crude oil is available for $18 per barrel. In addition, ICF believes that combined cycle units burning oil or gas would substitute for new baseload coal generating units at this oil price. Despite this finding, ICF has still confirmed that $18 per barrel crude oil in 2000 is potentially feasible if (1) non-OPEC oil and gas supplies reach maximum feasible levels, (2) OPEC decides to produce 20C0107 Page IV - 33 an additional 10 million barrels per day at the $18 per barrel price, and (3) energy efficiency gains are slightly greater at $18 per barrel than the Consensus forecasters anticipate under rising oil prices. 20C0107 Page IV - 34 Appendix A Comparison of Saudi Light and Alaska North Slope Crude Value 20C0107 Page A- 1 ABSOLUTE REFINING VALUES IN HOUSTON BASED ON CRACKING YIELD (NOMINAL $/BBL) ALASKA ARAB NORTH ISSUE LIGHT SLOPE OI FF sasasze22 SSSSeeezsss222225s5525222=2 1985 1/4/85 26.47 26.45 0.02 1/11/85 26.12 26.11 0.01 1/18/85 eres 27.16 (0.01) 1/25/85 27.39 hear 0.02 1/31/85 26.46 27.65 (1.21) 2/8/85 27.45 27.41 0.06 2/15/85 27.62 27.60 0.02 2/22/85 28.11 28.05 0.06 3/1/85 27.72 27.65 0.07 3/8/85 27.82 27.65 0.17 3/15/85 28.31 27.96 0.35 3/22/85 29.13 28.61 0.52 3/29/85 28.81 28.35 0.46 4/12/85 29.34 28.86 0.48 4/19/85 28.47 28.08 0.39 4/26/85 28.71 28.22 0.49 5/3/85 28.31 27.80 0.51 5/10/85 27.82 27.21 0.61 5/17/85 27.54 26.77 0.77 5/26/85 26.68 25.88 0.80 5/31/85 26.53 25.71 0.82 6/7/85 25.98 25.25 0.73 6/14/85 3.77 25.46 0.33 6/21/85 25.91 25.66 0.25 6/28/85 26.00 25.62 0.38 7/12/85 26.12 25.66 0.46 7/19/85 26.17 25.63 0.546 7/26/85 25.88 25.26 0.62 8/2/85 25.89 25.246 0.65 8/9/85 25.87 25.20 0.67 8/16/85 26.41 25.67 0.74 8/23/85 26.65 26.00 0.65 8/30/85 27.07 26.56 0.53 9/6/85 27.60 27.10 0.50 9/13/85 27.42 27.02 0.40 9/20/85 er St 26.98 0.33 9/27/85 27.76 27.39 0.37 10/4/85 28.21 27.76 0.45 10/11/85 28.22 27.63 0.59 10/18/85 28.50 27.66 0.84 10/25/85 28.75 27.90 0.85 11/1/85 28.84 27.95 0.89 11/8/85 28.76 27.85 0.89 11/15/85 29.23 28.42 0.81 11/22/85 29.22 28.49 0.73 12/2/85 29.40 28.63 0.77 12/13/85 26.22 25.81 0.41 12/20/85 25.50 25.09 0.41 12/27/85 25.59 25.09 0.50 1986 ABSOLUTE REFINING VALUES IN HOUSTON BASED ON CRACKING YIELD (NOMINAL $/8BL) ALASKA ARAB NORTH ISSUE LIGHT SLOPE OFF szzszesse 33 332222222222222252S522==2 1/3/86 26.08 25.59 0.49 1/10/86 25.22 24.95 0.27 1/17/86 23.98 23.73 0.25 1/26/86 21.32 20.96 0.36 1/31/86 20.79 20.38 0.41 2/7/86 17.83 17.43 0.40 2/14/86 17.33 16.86 0.47 2/21/86 17.06 16.59 0.45 2/28/86 16.25 16.07 0.18 3/7/86 14.81 16.84 (0.03) 3/14/86 16.02 15.83 0.19 3/21/86 16.45 16.25 0.20 4/6/86 16.36 16.36 0.00 4/11/86 14.79 14.70 0.09 4/18/86 14.08 13.96 0.14 4/25/86 15.45 15.26 0.21 5/2/86 16.78 16.66 0.12 5/9/86 15.55 15.70 (0.15) 5/16/86 15.38 18.62 (0.26) 5/23/86 15.22 15.467 (0.25) 5/30/86 15.27 15.54 (0.27) 6/6/86 12.89 13.30 (0.41) 6/13/86 13.61 13.41 0.20 6/20/86 13.21 13.14 0.07 6/27/86 13.45 13.22 0.23 7/3/86 12.75 12.467 0.28 7/11/86 , 11.36 11.16 0.22 7/18/86 11.76 11.35 0.41 7/25/86 11.02 10.72 0.30 8/1/86 11.467 11.21 0.26 8/8/86 13.48 13.37 0.11 8/15/86 16.02 14.08 (0.06) 8/22/86 16.62 14.52 0.10 8/29/86 16.76 16.61 0.13 9/5/86 15.08 16.91 0.17 9/12/86 13.92 13.89 0.03 9/19/86 13.56 13.43 0.13 9/26/86 13.61 13.38 0.23 10/03/86 16.39 16.00 0.39 10/10/86 16.37 14.06 0.31 10/17/86 13.78 13.55 0.23 10/27/86 14.26 13.82 0.46 10/31/86 13.38 13.06 0.32 11/7/86 16.26 13.93 0.33 11/16/86 16.61 16.22 0.39 11/21/86 16.88 16.82 0.06 12/5/86 16.62 16.60 0.02 12/12/86 16.52 16.46 0.06 12/19/86 14.78 14.79 (0.01) 1987 ABSOLUTE REFINING VALUES IN HOUSTON BASED ON CRACKING YIELD (NOMINAL $/BBL) ISSUE ssssss22 1/2/87 1/9/87 1/16/87 1/23/87 1/30/87 2/6/87 2/13/87 2/20/87 2/27/87 3/6/87 3/13/87 3/20/87 3/27/87 4/3/87 4/10/87 4/16/87 =/26/87 3/1/87 5/8/87 5/15/87 5/22/87 5/29/87 6/5/87 6/12/87 6/19/87 6/26/87 7/9/87 7/17/87 7/26/87 7/31/87 8/6/87 8/14/87 8/21/87 8/28/87 9/4/87 9/11/87 9/18/87 > 25/87 *2/2/87 10/9/87 10/23/87 11/6/87 11/13/87 11/20/87 11/25/87 12/4/87 12/11/87 12/18/87 12/24/87 AVERAGES: 1985 - 1987 ALASKA ARAB NORTH LIGHT SLOPE OLFF 233232a22222222222222222225 16.44 16.51 (0.07) 17.05 17.16 (0.09) 19.02 18.91 0.11 18.51 18.42 0.09 18.08 17.93 0.15 17.90 17.69 0.21 Wat 16.90 0.27 16.67 16.26 0.43 15.99 15.70 0.29 16.95 16.47 0.48 17.80 17.40 0.40 17.82 17.50 0.32 18.04 17.68 0.36 18.18 17.86 0.32 17.96 17.7% 0.20 17.42 7251 0.11 17.59 17.38 0.04 18.02 17.88 0.16 18.89 18.56 0.33 19.10 18.84 0.26 19.17 18.89 0.28 19.22 18.92 0.30 19.21 18.89 0.32 19.30 19.06 0.26 19.29 19.03 0.26 18.93 18.75 0.18 19.56 19.39 0.17 20.12 19.98 0.16 20.25 20.11 0.16 20.28 20.06 0.22 20.37 20.10 0.27 19.80 19.51 0.29 18.93 18.61 0.32 18.65 18.29 0.36 18.59 18.23 0.36 18.38 18.18 0.20 18.26 18.13 0.13 18.39 18.26 0.15 18.57 18.23 0.34 18.87 18.48 0.39 19.48 19.03 0.45 18.88 18.44 0.46 19.08 18.59 0.49 18.49 18.11 0.38 18.50 18.18 0.32 17.91 17.73 0.18 17.33 17.20 0.13 15.43 15.47 (0.04) 15.88 15.82 0.06 27.39 26.95 0.44 15.31 15.15 0.17 18.36 18.12 0.24 20.36 20.07 0.28 SOURCE: PLATT’S OILGRAM PRICE REPORT Appendix B ICF Forecast of Future Production Potential from Non-OPEC Free World Countries 200107 Page B - 1 Reprinted from Petroleum Economist, Oct. 1987 .by Theodore R Breton and John C Blaney Outlook for OPEC’s competitors The non-OPEC free-world countries have greatly increased their ot production ance 1973. They have a production cost structure which 1s much higher than OPEC's, but if OPEC 1s walling to support a $20 per barrel or higher price, these countmes could connnue to increase their producnon over the next 20 years. the other non-CPE i1.¢. “free-world”) countnes pro- duced 15 mb/d. Through its management of the world ol price over the 1973 - 86 period OPEC managed to reverse its market posiuon. In 1986 OPEC produced 17.6 mbvd while the other non-CPE countnes produced 23 mbvd. Figure | shows that the market share changes were even more dramauc than the above stausucs indicate. While US producton declined slighuy. the remaining non-CPE coun- unies increased their producuon from 5.3 to 14.1 mbvd, an increase of over 140%. I: 1973 OPEC produced 31 million barrels per day and FIGURE 1 FREE WORLD CRUDE OIL PRODUCTION 38 , a a ban 110NS OF BARRELS PER DAY ‘ees “ers sors 1960 ‘90s OPEC is again debaung whether to increase producuon or keep output low to dmve up near-term prices. ICF's anaivsis undicates that the decision (o uncrease prices could be a cosuv one for OPEC because the non-CPE producing countnes have the potenual to increase producuon again if they are given the mgnt incenuves. This arucie presents the resuits of ICF’s anaivsis of oul production potenual outside OPEC. the Centrally Planned Economues. and the United States. These producers are denoted as the “Rest-of-World” group io cus arucie. There are about 100 producing countnes outside OPEC and the Centrally Planned Economues. Argenuna. Brazu. Canada. India. Mexico. Peru. and the North Sea countnes are the largest or potenually largest producers. bur the total producuon from all the others is substanual. Prepamng a producuon oudook for ail these countnes 1s difficult because geologic condiuons vary enormousiy, informauon is lacking, reported data are someumes inaccurate, and the insuruuonal environment is frequendy uncertaun. The aiternauve to a country-specific approach 1s to exam- une the whole group together as a stausucal unt. ICF’s renew of the Rest-of-Worid group's aggregate hustoncal stausucs reveaied a surpmsing consistency in the behaviour of kev wndicators of ou resource development acuvirv. This Gnding Theodore R. Breton is a Vice President of ICF Lacor- porated, a Washington DC-based consulting firm. Joha C Blaney is a Senior Associate at [CF. provided empuincai support for the validity of an aggregate approach. OU production is the end result of a process in which she economuc poruon of ulumately recoverable ou-in-piace Grs< becomes proven reserves and is then produced. [CF’s asaiv- sis of the furure producuon potenual in the Rest-of-Worid group of countnes is based on a simulation of this process. The extent of the Rest-of- World petroleum resource base and the cost of producing these resources are unknown. Nevertheless. geologists have made estumates of the worid resource base. and there are data on actuai producuon costs in many countnes. Based on a review of geologists’ published forecasts, reported production costs, the hustoncal response to msing prices. and an assumpuon that the cost distnbuuoa to produce these resources 1s conunuous. [CF has developed an oil supply cost curve for the Rest-of-Worid group of countries. Table | shows ICF’s esumate of the amount of oul resources orginally in place which were economuc to produce for up to $50 per barrei 1986 US dollars: in the Rest-of-Worid coun- tunes. These esumates are based on actual tustomcai produc- uod costs. whuch in countnes with stare ou companies may oe much hugher than secessary. [f proaucuon costs can Se reduced. then :he amount of economuc resources at eacn pics would be fugher. Table [ REST-OF-WORLD ULTLMATELY RECOVERABLE ECONOMIC OIL RESOURCES 1986 DoUars Barret Buon Barreis 30 000 0 360 30 °30 <0 $60 10 M40 This amalvsis undicates chat the Rest-o1-World countnes resources are much more cosuy to produce than those un ne Maddie East. Nevertneiess. uo OPEC is willing to support an ou price of $20 per oarrei or more. :ne Rest-o1-Worid coun- tnes profitablv can produce a consideradie amount of ou. even at reiauveiv low levels of progucuvity Resources must be proven before snev can be producec Proven reserves are reported bv most <sountnes. out me validuev of these esumates vanes gready. paruculariv since state petroleum compames are somemmes uncer pressure ‘0 exaggerate they reserves. [n add::on. wnea prices fall. some formerly “proven” resc. ves become uneconomuc. Sut ofdcus estumates mav not de reduced. According to reported esumates. (he Rest-o1-World group had 116 billion barrels of proven reserves on | January 1987. The associated annual producuon-to-reserves (P/R) rauo was 0.045, whuch is consistent with 22 years of proven reserves. Under an opumai development schedule, proven reserves are produced in 12 - 15 years, which leads to a P/R rauo of .067- .083 in a mature industry. In a declining industry the PR rato is higher (¢.g. it is .10S in the US) because proven reserves are being depleted. Similarly, the P’R rauo is lower in an expanding industry. If the Rest-of-Worid group's esu- mate of proven reserves is accurate, its .045 P’R ratio is quite low. even for an expanding industry. ICF believes that the group's proven reserves are over- stated, bur that adjusted P’R rauos stull would be under .067 due to unsuruuonal inefficiency in many countnes within the group. As a result, there is potential for increasing produc- uon through umprovements in P/R ratios without proving up addiuonai reserves. There 1s also potenual for reducing pro- ducuon costs, paruculariv if more production acuvity is moved from state to pmvate oil companies. Furure producuon in each country will depend on geology, local polucs. and the charactemstics of the ou-producing unsutuuons in each countrv. Nevertheless. fucure world oil prices will be a cmucal determunant of future production acuvity. An assumpuon about the future price path is a cmucal starang point for any forecast of furure production. Forecasts of future oul prices are almost as voiatiie as oil prices themselves. and the uncertainty reiated co future prices exacerbates the problem of forecasung future produc- uon. In the 1970s producuon activity responded very forcefully to the increase in prices. in part because prices were expected to keep msing. Given the perspecuve of the last ten years, another price huke probabiy would not engender as large a response. Producers. after all. have no guarantee that prices wul sull be high when the dniling activites have finallv resulted in sales of oul. For the purpose of this exercise. we evaluated the porenual effect of two price paths on Rest-of-World production: a constant dolar Saud: light price: fob: of $18 per barrel indeni- mutely, and a current ‘consensus’ type forecast in which the Saudi light price: fob: increases from $18 per darrei in 1987 at arate of 3%: annually. Many current price forecasts follow this second path. althougn growth is often siower during 1987 -90 ana more rapid dung 1990-95. The actual price path is likeiv co De consideraoiv more volauie. but these two paths are reasonaoie aiternatives for production planning purposes. Future production in the Rest-of-Wortd region could con- unue at a ieve: consistent with histomcal experience if current FIGURE 2 REST-OF-WORLD PROOCUCTION sm PmCE $10 9ER BaneeL) momoveo seoouCTIVITy tebe ee Mu LIONS OF BARMELS FEM DAY seco oes e°3 ars 900 ves 980 vee = 2900 3238 insutuuonal practices conunue unchanged. Aiternauveiy, if Private sector pracuces are adopted in more piaces ana pro- ducuon costs decline, the group's furure producuon might be sigmficandy higher than historical expenence at the same price level. Figure 2 presents ICF’s esumate of the Rest-of-Worid group's 1987 - 2005 producton with current and umproved levels of productivity at the $18 per barre! Saudi Light pre. At dus price. Rest-of-Worid producers in the aggregate may uncrease or decrease production. depending on insuruuonal developments in the group. FIGURE 3 REST-OF-WORLD PRODUCTION J PERCENT D1 PCE CROWTm bau L10MS OF BARAELS PER DAY 1908 erasers veo ee SL Figure 3 presents [CF’s producuon esumates for the “con- sensus” oil price forecast. At the higher consensus price forecast. Rest-of- World producers could increase producaon quite signuficantiv bv 2005. pernaps of 4-5 mullon barrels per dav. This analysis indicates that producers in the Rest- of-Worid region are responsive to ou pnce changes and wul be significant competitors for OPEC. parucuiariv uf. as expected bv most forecasters. OPEC decides to increase real prices over the next 20 vears. FIGURE 4 NON-OPEC FREE WORLD PRODUCTION 19EOTEN’ 22028 Cw te as = pe wen ewe bres 999 3 vs vez vas a00 a9 Figure 4 presents the Rest-of-Worid producuon forecasts combined witn ICF’s prorection of US proaucuon over the 1987 = 2005 period for the increasing ou price scenario. This analvsis indicates that non-OPEC oui suppiy outside the CPE countnes will remain relauiveiv constant uf ou prices mse at 3% per vear. [f OPEC wishes to achieve tus price growtn. nen OPEC production increases impuciuy will have to be umured tO Match the increase in worid ou aemand. 7 For more information. contact Theodore Breton. [CF Lacorporated. (703)934-3000 ' \ | | | SECTION 2 FUEL PRICE OUTLOOK FOR THE ALASKA RAILBELT REGION: OIL AND NATURAL GAS TABLE OF CONTENTS Page EXECUTIVE SUMMARY. foc (is. 6.) SF teeth | st fet as! exes, rol ooh ella ek al els) oe) | ESe OD ME TT, PE ea ee cet ee ee ae ae ei | alae ie ee oe |e | oe | | ee METHODOLOGY! 55 4s se ie este pera ete erode oe Ae ae oh RL FUEL (OTL (MARKET 6. osc fcc +) eis csshle cettteitattnn | af sl teoteyPes” |) Sell) «, Mais oi hate gh ee NATURAL, GAS: MARKED «i023 jl eto tse) Shee) c| ie |e) abies oe oe oes? 4] ue Oe WSS CONTINUED :MONT TORING °) 5-50 see Re a te ie a ee ee er a oh SF uO. | MA PRODUG LION oo) | en to!" baid| ct ibe encsl wit) ePt| col 23h] Path cola at || aii euees ahs t el fre | ealiteay (at |) foals 1-1 PURPOSE) oss jin! iol islets Prezscae Geyer, ot | fat ce! | let | |e [te | oe) miter arse) waste lee. Worm mel cel [ee tees i-1 SCOPE (OF WORK. icc aha cis ce sete | toe) art Shia) | Pata oes oy italy ad cietor as 8s 1-2 WORLD. O81, PRICE RESULTS. 0) ) slo |e | he) |e) ot ll fo | pth tel as [wo ated) |e [te 1-3 REMAINDER OF* THIS) REPORT. 226.035 ee ase os os ME se G1 ee 1-7 te Ri POM ect |e | ee oe ome fs em alae ll a | a de 2-1 BUEEJOLL APPROACH TS Gite eels = we soot te fe a aye Po we geet tenon a 2-3 NATURAL GAS ANALYTIC APPROACH .*%. . 2. 1 6 0%. dhe ee eld wee we 2-6 Wremand,: Approachs: 5/5 se es ee we ge te 2-7 Supply Approach. . . alse | las | hil food isc (sopeay! |i] teiggteytton’ 15\| Hag, de ist ts f ppeuie a Market Balance Approach. eee lincetaiecc te | hdl fal eet ae an Manta Pat | el apts St) | Pa 3.0. FUEL: OIL MARKET ANALYSIS. 00s fi) 60 6 se uh ele Biggie igs Meee 4 BEL CURRENT RAILBELT OIL SUPPLY & DEMAND. ....... 2... 2. eee 3-3 Alaska Refinery Capacity . Major Railbelt 0il Flows ‘ : Residential Fuel Oil Disposal 7 Light Product Flows . “ . Current Consumption of Fuel Oil ‘by Fairbanks ‘ Electric Utilities WWWwWw ' OonnNnunw 20T00360 TABLE OF CONTENTS (cont’d) Page Current Residential/Commercial Fuel Oil Consumption and Distribution in Fairbanks. . 2. 6-0) me ee oe wk te ew te er LO Heating O11. Supply Costs... 2. 2 2 2 wee te te we el | SLL Heating Oil Prices in Fairbanks . . 2... ss % ww « +. 3014 RAILBELT PETROLEUM DEMAND OUTLOOK ............... =. 3-15 Electric Utility Demand. i) 0) ie) oi) leis | et eis lit ie | |) i ote” lo ae Home Heating Oil Demand. . . | jam) [tol bo | lok ble | ja lel st atl | et | | fe] (9!) lay] SRT Other Petroleum Product Demand ay! (at, |e) |e | | l | fot |e | /s| || lel Set ale | cel |e | op | [Steak RAILBELT PETORLEUM SUPPLY OUTLOOK ................ 3-19 FUEL OIL PRICE OUTLOOK. .. . owe oe be ee ol Fuel Oil Price Projections for GVEA. 6 iH, s/t | las | | fot | eo] lee len) fier” | SPR Fuel Oil Price Projections for the Fairbanks tel | i [et || fot al el oe Heating Oil Market @.G NATURAL GAS DEMAND ANALYSIS .. 6 sek ww KA ke es 4-1 REY (MARKET, SEGMENTS. |.0-)/5, ie) | ethic) ie! | aj ie alles) fe | atte be | lay bie ot | 8 | oe! |e | ee |e 4-1 OVERVIEW?OF. DEMAND TRENDS 3: «oss soe swage 8 @ se we ee ew 4-4 MARKET OUTLOOK BY SEGMENT -- DEMAND AND VALUE. ......... 4-6 RESIDENTIAL AND COMMERCIAL DEMAND... .........0.2.2604 4-6 BLECTRIC. GRILIT® (OMAND . wt ce ee ee ee ee 4-9 Mimi | | | fk) | | | | || at || oe) tl |e? | fel | defo) || | | at || tn | ke fe | | Background . . mel fom okt ae | tel or et fee | ome] [oll fe | (et | || | en Estimate of LNG “Net- “back Value swe ew ee ol lll | ee AMMONIA/UREA EXPORT . . . BTN Inet fel Vet |e |e Vit | dee | gel el fo | foel ay | | e | | lle Fertilizer Market Outlook. S| is ss | La! | ay tol | a) | lol fe! | Jt [te |e || | | eee Estimates of Fertilizer Net- back Value oie et em ee a et OES METHANOL FUEL MARKET ASSESSMENT ................ +. 4-24 Background .. . as | ol oat | |e | | || et | | Methanol Market and California Price Outlook ee ee ee | SUMMARY OF DEMAND ANALYSIS. ........-..-62 5656+ 2 + + 4-30 5.0 GAS SUPPLY AND DEMAND INTEGRATION. ...........-2 5.884 5-1 SUPPLY ANALYSTS SUMMARY. «© ss. /s se ws awe we we es me 5-1 20T00360 TABLE OF CONTENTS (cont’d) Page TRANSMISSION DIFFERENTIAL ISSUES... ........-......- 5-7 SUPPLY/DEMAND INTEGRATION. .... 2.2... 2-2-0. ee ee) 529 NATURAL GAS PRICE. OUTLOOK 30)... SG we ee ge oe ee. S210 PRIGESDEFRERENTIALS 02.0 ide Se ete ge ogee St Wee, gine BO SLS 0" ECONCLUSION O88. 3 5028) 0. Sate.) ye ghee eg ees Soe Ge APPENDIX A. Cook Inlet Supply Analysis: Proved Reserves APPENDIX B. Cook Inlet Supply Analysis: New Supplies APPENDIX C. Fuel Oil Price Projections for Fairbanks APPENDIX D. Comments on the Draft Report and Responses 20T00360 LIST OF TABLES EXECUTIVE SUMMARY ES.1 Summary of Electric Generation Fuel Price Outlooks. CHAPTER 1. INTRODUCTION 1.1 OIL PRICE FORECASTS DATED DECEMBER 1986 AND 1987. 1.2 ALTERNATIVE OIL PRICE SCHOOLS OF THOUGHT. CHAPTER 3. FUEL OIL MARKET ANALYSIS WWWWWW DUNFwne ALASKAN CRUDE OIL REFINERIES. . . 1987 CRUDE OIL ACQUISITION COSTS FOR. ALASKA REFINERIES . ere PROJECTED PETROLEUM PRODUCT DEMAND FOR THE ALASKA RAILBELT. FUEL OIL PRICE PROJECTIONS FOR FAIRBANKS ELECTRIC UTILITIES RESIDENTIAL HEATING OIL PRICE PROJECTIONS FOR FAIRBANKS . . LARGE COMMERCIAL FUEL OIL PRICE PROJECTIONS FOR FAIRBANKS . CHAPTER 4. NATURAL GAS DEMAND ANALYSIS fF nor on -10 «hl -12 CONSUMPTION OF COOK INLET PRODUCTION: 1986... FUEL OIL PRICE PROJECTIONS FOR SOUTH CENTRAL ALASKA: RESIDENTIAL/COMMERCIAL MARKET . . . : VARIABLE COSTS OF TRANSPORTING GAS FROM WELLHEAD | TO BURNERTIP . ESTIMATES OF NATURAL GAS WELLHEAD VALUE FOR THE RESIDENTIAL/ COMMERCIAL MARKET DEMAND. . . FUEL OIL PRICE PROJECTIONS FOR SOUTH CENTRAL ALASKA ELECTRIC UTILITY MARKET. . . ESTIMATES OF NATURAL. GAS " WELLHEAD VALUE FOR THE ELECTIRC UTILITY MARKET DEMAND . . ALASKAN LNG EXPORTS TO JAPAN. BY YEAR. AND. PRICE LANDED CRUDE PRICES IN JAPAN. ESTIMATED NATURAL GAS WELLHEAD VALUE. FOR. LONG RUN (NEW FACILITY) LNG DEMAND. . . ESTIMATED NATURAL GAS WELLHEAD “VALUE FOR. SHORT RUN. (EXISTING FACILITY) LNG DEMAND. . . ESTIMATED NATURAL GAS WELLHEAD “VALUE. FOR. EXISTING AMMONIA/UREA FACILITY DEMAND . . SUMMARY OF VALUES AND QUANTITIES “FOR POTENTIAL GAS. DEMAND BY. SECTOR AND OIL PRICE FORECAST . 20T00360 ES-4 PR 1 uf 3-4 3-13, 3-16 (3-23 3-26 3-26 LIST OF TABLES (cont ‘d) CHAPTER 5. GAS SUPPLY AND DEMAND INTEGRATION 5.1 ESTIMATES OF RESERVES AND PRODUCTION DECLINE COEFFICIENTS FOR COOK INLET FIELDS . mn -3 PRICE RANGES FOR NEW COOK INLET NATURAL GAS (ISOLATED CASE) 5.4 PRICE RANGES FOR NEW COOK INLET NATURAL GAS (NEW LINKS TO WORLD MARKET) . . . -5A SUPPLY/DEMAND DATA WITHOUT NEW EXPORT PROJECTS. os p -5B SUPPLY/DEMAND DATA WITH POTENTIAL NEW EXPORT PROJECTS - -6 LOCATION SPECIFIC PRICE DATA. uw Ad 20T00360 -2 GAS PRODUCTION POSSIBILITIES IN COOK INLET BY PRICE INTERVAL. LIST OF FIGURES CHAPTER 1. INTRODUCTION 1.1 OIL PRICE FORECASTS DATED DECEMBER 1986 AND 1987. 1.2 ALTERNATIVE OIL PRICE SCHOOLS OF THOUGHT. CHAPTER 2. ANALYTIC APPROACH 2.1 OVERVIEW OF ANALYTIC APPROACH... ....... 2 ee ee eee 2-2 2.2 ILLUSTRATIVE DEMAND FOR NATURAL GAS ............... 2-11 CHAPTER 3. FUEL OIL MARKET ANALYSIS 3.1 MAJOR FLOWS OF CRUDE OIL AND PETROLEUM PRODUCTS IN THE ALASKA RAILBELT ... . oes 3-6 3.2 REFINERY MARKET SHARES IN THE FAIRBANKS AND OTHER INTERIOR HOME HEATING OIL MARKET . . . oe 3.3. PROJECTED PETROLEUM PRODUCT DEMAND FOR. THE. ALASKA RAILBELT. +s 2 « S916 CHAPTER 4. NATURAL GAS DEMAND ANALYSIS GROSS MARKET SEGMENTS FOR COOK INLET PRODUCTION: 1986. HISTORICAL RAILBELT GAS CONSUMPTION BY SECTOR... . PROJECTED RAILBELT DEMAND FOR NATURAL GAS BY SECTOR . JAPANESE LNG IMPORTS BY COUNTRY OF ORIGIN . PROJECTED LNG DEMAND BY COUNTRY . . 4 COMPARISON OF JAPANESE LNG DEMAND AND CONTRACTED SUPPLIES aS fas CONSESUS (LOW) DEMAND ESTIMATES FOR COOK INLET NATURAL GAS, 2000. OVERVIEW OF COOK INLET NATURAL GAS DEMAND OUTLOOKS. PP ' WNAUEF AWN PRP EEE CONDUFWNHE FREE t WWHrHH I CHAPTER 5. GAS SUPPLY AND DEMAND INTEGRATION 5.1 FORECAST PRODUCTION DECLINE FROM EXISTING COOK INLET RESERVES . . 5-3 5.2 SCHEMATIC DIAGRAM OF METHOD USED TO INTEGRATE GAS SUPPLY WITH DEMAND . . 5-6 5.3 POTENTIAL SUPPLY OF COOK INLET GAS. 5-8 20T00360 EXECUTIVE SUMMARY In this report, the ICF-Lewin Energy Group analyzes the price outlook for fuel oil and natural gas in the Alaska Railbelt region. Fuel oil and matural gas are two key fuels for Alaska electric generation. The purpose of this work is to assist the Alaska Power Authority in its decisionmaking regarding the need for transmission upgrades in the Railbelt. The fuel price information will become an input to the assessment of the comparative economics of alternative configurations of future Railbelt generation and transmission capacity. Special attention was devoted to the question of whether and how the Railbelt gas market would be linked to world energy markets. WORLD OIL PRICE This report incorporates three world oil price scenarios presented in ICF-Lewin’s January 1988 report to the Alaska Power Authority. Briefly, those three scenarios follow: Crude Oil Price (1987 Dollars per Barrel) Price Scenario 1990 2000 2010 Consensus $20 $30 $40 Consensus (Low) 18 24 30 Low Price 14 18 20 METHODOLOGY The ICF-Lewin analysis began with the world oil projections. The research team then developed estimates for natural gas supply and analyzed the fuel oil supply situation. The study next incorporated the latest 20T00360 ES-1 available demand information for Alaska-oriented users (residential commercial, electric) and developed independent assessments of export- oriented demand, including liquefied natural gas (LNG), fertilizer (ammonia and urea), and fuel methanol. The final analytic stage integrates the demand and supply projections to provide an assessment of what resource economics can conclude about price behavior in the Railbelt fuel oil and natural gas markets. The report also provides information on economic based cost/price differential factors (e.g., variable costs of gas transmission) and potentially important sunk cost factors (e.g., full gas rates and contracts). Thus, the report provides the information necessary to calculate price paths either based upon fundamental economics or also incorporating other business/regulatory factors. FUEL OIL MARKET This report provides data on Railbelt fuel oil supply and demand and analyzes the Fairbanks market. The analysis of the Fairbanks fuel oil market focused on two key fuels: e No. 4 distillate fuel oil ° No. 2 home heating fuel oil The former fuel’s importance relates to the fuel costs for electric generation at Golden Valley Electric Association (GVEA). The latter fuel’s importance arises in the context of the economics of constructing a natural gas transmission line from Anchorage to Fairbanks. The attractiveness of such a pipeline " system would depend on the savings to the residential, commercial, and electric sectors. 20T00360 ES-2 The Fairbanks fuel oil market has a number of unusual features that make it unique. These features include: . The unusual slate of petroleum products required, e The role of Trans Alaska Pipeline System (TAPS) crude oil and transport capacity, . The special relationship between GVEA and a nearby refinery, and . The exchange of petroleum products between Fairbanks and Anchorage. The isolated and cold conditions in Fairbanks make for a special set of product requirements, dominated by jet fuel and Numbers 2 and 4 fuel oil. All crude processed by Fairbanks area refineries is from TAPS. GVEA owns and operates the 2.5 miles of pipeline connecting the refineries to TAPS. The refineries return the bottoms of the refinery runs to TAPS. The report provides price figures for each crude oil scenario by sector. NATURAL GAS MARKET The Railbelt market for natural gas is concentrated in the Cook Inlet region. Cook Inlet has five major gas use markets. The disposition of production for 1986 follows: LNG (Export) 62 billion cubic feet (Bcf) Electric Power 44 Ammonia/Urea (Export) 36 Residential/Commercial 23 Field 18 Producer 15 Unaccounted for ¢ 5) 193 Bef 20T00360 ES-3 TABLE ES.1 Summary of Electric Generation Fuel Price Outlooks (1987 $ per million Btu) Natural Gas Price Range b/ Fuel Oil Price a/ Cost of Producing Marginal Scenario ear No. 4 New erve Value Consensus 1990 $3.54 -- cf 2.28 d/ 1995 4.23 0.73 2522 2000 Seto 1.65 3.07 2010 6.80 59 4.62 Consensus (Low) 1990 3.19 -- of 2.03 1995 3.68 0.47 2565 2000 4.16 0.86 3.09 2010 S413) Bhai S105 Low Price 1990 2.30 -- cf p+ 1995 2.84 0.47 1.96 2000 SeL9 0.86 23260) 2010 3.54 Sig La! 2.54 Fairbanks Cook Inlet No reserve additions required before 1991. The 1990 marginal value is for a lower total without new exports (which are assumed to begin in 1995). gece The LNG export market represents the key market for the future of natural gas pricing in the Cook Inlet. The ICF-Lewin analysis has found that LNG export offers the highest potential netback value among all new export opportunities for natural gas suppliers in Cook Inlet. This option would prove more financially attractive than additional fertilizer production in all the cases examined. The Alaska Power Authority also asked ICF-Lewin to examine the economics of potential fuel methanol production. Alaskan interest in this concept has grown as California air quality problems have pushed that state towards serious consideration of significant fuel methanol consumption. The ICF-Lewin analysis found that Cook Inlet gas cannot compete economically for 20T00360 ES-4 the California fuel methanol market because of the combination of fuel methanol’s estimated value in California (as a substitute for gasoline) and the estimated costs of production and transportation from Cook Inlet. LNG demand is likely to increase from about 61 Bcf to 69 Bef annually. This increase is based on current Philips-Marathon plans and requires approval by the U.S. Economic Regulatory Administration. The outlook for additional export activity depends on the world oil price scenario. This subject is discussed later. The electric power demand for natural gas will dip slightly in the early 1990s when the Bradley Lake hydro facility displaces some gas-fired generation. Gas consumption is expected to return to 1990 levels. ICF-Lewin expect the current ammonia/urea facility to remain at relatively high utilization factors. As noted above, however, expansion requiring significant new plant investment is not expected. Residential and commercial demand is expected to remain approximately level. The supply analysis consists of two major elements. The first stage projects production profiles from known reserves in active fields. The decline coefficients range from 2 percent (e.g., Kenai Field) to 15 percent (e.g., Beaver Creek). This analysis of baseline production capacity showed that expected demand cannot be met without new supply investments. The second stage developed hypothetical supply curves by applying well- established statistical theory to estimate the economics of remaining resources. ICF-Lewin examined three categories of potential resources: ° Developmental (including extension of existing fields) . New field exploration (including new exploration in known productive formations) ° Speculative (including exploration in formations not known to be productive) 20T00360 ES-5 This supply analysis yielded the following estimated production possibilities: Marginal Production Bcf Available Cost (Wellhead) ‘ve lopmenta New Field Speculative (1987 $/Mcf) < $2.00 800 650 1500 $2.00 -- $3.00 50 200 250 $3.00 -- $5.00 55 300 400 Total < $5.00 905 1150 2150 Total Resource 1070 2100 3400 The technical details of this supply analysis are presented in two technical appendices to this report. As a final step before proceeding with the integrative analysis of the natural gas market, ICF-Lewin examined the economic basis for cost differentials at key locations in the Cook Inlet area. The real cost differential identified is the variable cost of transmission, which ICF- Lewin estimated as having an upper bound of $0.03 per million Btu. Just as the Fairbanks fuel oil market has unique and important features, the Cook Inlet gas market has important differences from a classic competitive market. There are few buyers and few sellers. The current major export projects are served through gas supplies from affiliated producers. ICF-Lewin has found that without significant new demand from new export-oriented facilities, the Cook Inlet natural gas market is likely to remain in a situation of supply surplus. (The potential linkage to world energy markets through new 1NG exports will be addressed shortly.) Service to Fairbanks could represent another important new gas load that could potentially have effects similar to new exports. This issue was, however, beyond the scope of this study. 20T00360 ES-6 In a competitive situation, such a surplus would lead to prices close to marginal costs of production. There is no evidence at this time that Cook Inlet gas prices are approaching full production costs, which ICF- Lewin estimates will be ‘about $0.33 per million Btu in 1990 for new reserves. In this market with few buyers and sellers, the price can vary between a high of the value to the marginal user, which is LNG for the period examined, and a low of the marginal production costs. The price outcome will depend on expectations about future prices, demand, and costs, as well as the negotiating skills of the parties trying to arrive at a price. Clearly, a quantitative analytic solution to these qualitative matters was not possible. Nevertheless, ICF-Lewin believes that the best available indicator of the net effect of the expectation and skill factors is the current contract price. The price emerging from contemporary contract negotiations and renegotiations (of existing contracts) is a highly confidential subject. Table ES.1 shows how the potential gas price range varies with oil price scenario. The wellhead price for new supplies is the key factor for considering future prices. Other factors, such as cost differentials and existing contracts, can also play a role. This report provides data to assist in incorporating such information into the fuel price analysis. CONTINUED MONITORING Because of the considerable uncertainty and potential for volatility in the fuel prices relevant to the Alaska Power Authority, ICF-Lewin recommends careful monitoring of new price developments. To the extent the price paths presented here are to be applied a year or more in the future, such monitoring and appropriate adjustments should play an important part in any comparative assessment of Railbelt electric generation and transmission options. 20T00360 ES-7 1. INTRODUCTION This report presents the results of ICF-Lewin’s analysis of the price outlook for natural gas and fuel oil in the Alaska Railbelt region. This is one of a series of studies that the Alaska Power Authority has funded to analyze the alternatives for electric interties between the Kenai Peninsula and Fairbanks, Alaska. In an earlier volume, ICF-Lewin presented an overview of the world oil market and the various "schools of thought" that have evolved in the analytic community of energy economics. This volume presents the analysis of the Railbelt natural gas and fuel oil markets and assesses the extent to which these Railbelt markets will likely be linked to world markets. A third volume provides technical appendices with additional detail in areas including fuel oil prices in Fairbanks and the natural gas supply outlook in Cook Inlet, where ICF-Lewin has completed significant original research in the course of this project. PURPOSE This study develops price projections for two key potential fuels for electric generation in the Alaska Railbelt region, natural gas and fuel oil. The work provides an important analytic component of the Alaska Power Authority's assessment of the need to upgrade the Kenai-Fairbanks electric intertie by providing fuel price projections necessary to assess various candidate sites for the construction of new electric generation capacity. Only by examining the economics of generation at specific geographic locations can the Power Authority assess the relative economics of major alternative configurations of new electric generation and transmission capacity. This report presents price projections through the year 2010. Among the major issues examined is the outlook for new natural gas export-oriented projects. The extent to which export-oriented activity could grow has 20T00320 Page 1-1 important implications for the extent to which natural gas prices in the Cook Inlet region is likely to be linked to world market activity, or alternatively, to evolve as a more isolated marketplace. SCOPE OF WORK This analysis examines the Cook Inlet natural gas market as well as the Anchorage and Fairbanks fuel oil markets. The analysis involved the review of available literature and other documents (especially gas contracts), the incorporation of related ICF-Lewin work related to Cook Inlet natural gas resources, and extensive interviewing of participants in the Railbelt energy market. The analysis proceeded in parallel with two other related Power Authority studies on (1) employment and population forecasts and (2) electricity demand forecast. This work, however, did not include input from these other studies. With regard to natural gas, this work focused on the supply outlook and the outlook for export-oriented activity (in the demand analysis). In-state demand forecasts are from the most recently published Alaska Department of Natural Resources (ADNR) projections. The fuel oil demand analysis used ADNR growth rates and applied these growth rates to the ICF-Lewin base year (1987) demand estimates. The new analytic results concerning fuel oil emerged from the supply side examination of the fuel oil markets. 20T00320 Page 1-2 WORLD OIL PRICE RESULTS! Before proceeding with the Railbelt fuel market analyses, a review of the world oil price outlook report could help the reader. ICF-Lewin presented the results of the world oil price analysis in an earlier report. The ICF-Lewin analysis of world oil prices involved several steps. First, we collected oil market forecasts and reports prepared in 1987 and late 1986 by organizations known for their work in oil forecasting. The analyses collected were screened so that only those appearing methodologically complete or important for other reasons (e.g., some low price opinions) were considered further. ICF-Lewin then sorted these screened forecasts into "schools of thought." Next, the analytic underpinnings of the two major "schools" identified were critically reviewed. Finally, the analysis presented three world oil price scenarios representative of the groups identified. Price Forecasts A review of the oil market forecasts from organizations known for their expertise in this field reveals that these forecasts vary widely Table 1.1). Prices in the forecasts (expressed in 1987 dollars) range from $10 to $28 per barrel in 1990, from $10 to $47 per barrel in 2000, and from $10 to $56 per barrel in 2010. Nevertheless, the forecasts generally can be categorized into two schools of thought, a Consensus school and a Low Price school. The Consensus school has been called the "3:2:1" forecast by Texaco. This school’s assumptions for the free world economies are that: Readers who have reviewed Volume 1 of this report need not read this section. 2 ICF-Lewin Energy Group, Qutlooks for World 0 rices: Analysis 2 ti ve Sch ought, report to Alaska Power Authority, Fairfax, VA, June 1988. 20T00320 Page 1-3 TABLE 1.1 OIL PRICE FORECASTS DATED DECEMBER 1986 AND 1987 (1987 Dollars Per Barrel) 1990 2000 2010 LOW BASE HIGH Low BASE HIGH LOW BASE HIGH (1) © ECE 21.06 46.85 (2) INSEE 24.95 43.17 (3) ORIE 23.63 36.146 (4) DOE/NEPP 19.46 34.28 56.06 (5) DOE/EIA 13.38 18.52 22.64 27.78 33.95 42.19 (6) IEA 20.01 31.59 (7) LIASC 23.69 28.43 34.75 (8) CHVRN 18.95 28.43 (9) ICF 11.30 18.40 27.50 20.00 28.40 37.00 24.30 38.10 49.70 (10) GRI 20.53 27.69 42.01 (11) CERG 18.95 25.27 35.80 (12) TIAST 23.69 25.16 31.69 (13) PG&E 17.02 19.37 20.70 18.80 24.90 32.78 20.76 32.38 51.89 (16) WBK 14.42 20.95 (15) ADOR 15.18 17.81 (16) ARTA 10.29 15.44 20.58 10.29 15.44 20.58 10.29 15.44 20.58 COUNT 1/ 16 COUNT 1/ 16 COUNT 1/ 8 MEAN 1/ 19.77 MEAN 1/ 29.28 MEAN 1/ 35.78 MEDIAN 1/ 19.46 MEDIAN 1/ 28.40 MEDIAN 1/ 35.80 (17) CEC 18.51 27.80 41.89 KEY: FORECAST: ACRONYM: DATE: essssess2 szeses=22 gezezszs= (1) ECONOMIC COMMISSION FOR EUROPE ECE 3/87 (2) INSTITUT NATIONAL STATISTIQUE ET ETUDES ECONOMIQUE INSEE 2/87 (3) DRI, EUROPEAN ENERGY SERVICE DRIE 12/86 (4) DEPARTMENT OF ENERGY/NATIONAL ENERGY POLICY PLAN - DRAFT PROJECTIONS DOE/NEPP 12/86 (5) DEPARTMENT OF ENERGY/ENERGY INFORMATION ADMINISTRATION ANNUAL ENERGY OUTLOOK OOE/EIA 2/87 (6) INTERNATIONAL ENERGY AGENCY IEA 12/86 (7) INTERNATIONAL INSTITUTE FOR APPLIED SYSTEMS ANALYSIS - CONVENTIONAL SCENARIO ITASC 4/87 (8) CHEVRON CORPORATION : CHVRN 10/87 (9) ICF INCORPORATED ICF 6/87 (10) GAS RESEARCH INSTITUTE GRI 1987 (11) CAMBRIDGE ENERGY RESEARCH GROUP : CERG 11/87 (12) INTERNATIONAL INSTITUTE FOR APPLIED SYSTEMS ANALYSIS - TECHNOLOGICAL DEVELOPMENT IIAST 4/87 (13) PACIFIC GAS AND ELECTRIC COMPANY PGE 3/87 (14) WORLD BANK weK 9/87 (15) ALASKA DEPARTMENT OF REVENUE ADOR 9/87 (16) ARLON TUSSING ASSOCIATES ARTA 7/87 (17) CALIFORNIA ENERGY COMMISSION CEC 4/87 NOTE: 1/ - COUNT, MEAN, AND MEDIAN ARE CALCULATED USING BASE CASE FORECAST PRICES ONLY. 20T00320 Page 1-4 ° Economic growth will proceed at about 3 percent annually; ° Energy demand will grow at about 2 percent annually; and ° Oil demand will grow at about 1 percent annually.In a typical Consensus forecast, OPEC regains control of the world oil market after 1990 and uses its market power to increase oil prices at about 3.5 percent annually in real terms through 2010. An alternative view is expressed by the Low Price school, in which forecasters believe that real oil prices will rise very little, if at all, over the 1990-2010 period. Forecasters in this school perceive a world of energy abundance in which technological development will ensure that substitutes always appear when any resource becomes relatively scarce and expensive. Table 1.2 presents the three crude oil price forecasts used to evaluate petroleum and natural gas markets in this study. In addition to the Consensus and Low Price schools, a third forecast is included which represents the lower range of the Consensus school of thought. Table 1.2 ALTERNATIVE OIL PRICE SCHOOLS OF THOUGHT (1987 Dollars Per Barrel) 1990 2000 2010 Consensus $20 $30 $40 Consensus (low) 18 24 30 Low Price School 14 18 20 20T00320 Page 1-5 Assumptions Behind the Two Schools of Thought The Consensus school is composed primarily of organizations that have developed detailed models of the world energy market, which they use to produce their forecasts. For this reason, their assumptions can be readily reviewed and analyzed. The Low Price school generally bases its forecast on a more general philosophical approach. The market assumptions behind the Low Price forecasts are, therefore, less explicit. A review of the reports from these two schools suggests the following implicit or explicit differences in their market assumptions: e The Low Price school believes energy supplies from the non-OPEC countries will be more abundant at low oil prices than assumed in the Consensus forecast. ° The Low Price school believes that OPEC will be willing to produce more oil at a low price than assumed in the Consensus forecast. . The Low Price school believes energy efficiency gains will be greater than assumed in the Consensus forecast. . The Low Price school believes that substitute fuels are viable in the transportation sector at low oil prices. . The Low Price school believes that oil cannot compete in the boiler market if oil prices exceed $20 per barrel. The ICF-Lewin analysis of these assumptions indicates that energy supplies could be more abundant, OPEC production could be greater, and energy efficiency could be greater at low oil prices than assumed in the Consensus forecasts. Substitutes for petroleum fuels, however, are not viable in the transportation sector at low oil process. Gas is a good 20T00320 Page 1-6 substitute for oil in the boiler market at oil prices below $20 per barrel. At this oil price level, coal cannot compete with oil in the industrial boiler market, and coal and nuclear energy cannot compete with combined cycle units in the electricity generating sector. The assumptions behind the Consensus forecasts may be relatively low with respect to world energy supplies and energy efficiency gains at a low level of oil prices. The Low Price school, on the other hand, appears to have overestimated the capability of substitute fuels to compete with oil at low oil prices. Given the level of uncertainty associated with worldwide oil supply and demand relationships, however, neither the Consensus nor the Low Price schools of thought can be easily rejected. REMAINDER OF THIS REPORT The next chapter of this report describes the method of analysis applied by ICF-Lewin to develop the reported price outlook. Chapter 3 presents the fuel oil price outlook analysis. Chapter 4 presents the natural gas demand analysis. This is followed by the gas price outlook analysis, which integrates the demand and supply analyses, in Chapter 5. Chapter 6 presents the study’s conclusions. 20T00320 Page 1-7 2. ANALYTIC APPROACH This chapter summarizes the analytic approach employed by ICF-Lewin to perform this analysis. This approach included analyses of world oil price projections, the outlook for natural gas supply and demand, and the outlook for fuel oil supply and demand. The period of interest is 1990 through 2010. As Figure 2.1 illustrates, the three major elements of this analysis have significant connections. World oil prices affect fuel oil prices directly. These prices also directly affect certain market segments on the natural gas side; in particular, Japanese importers of LNG generally tie prices paid to crude oil prices. Fuel oil prices, in turn, affect the value of natural gas for certain major Cook Inlet market segments because fuel oils represent the best available alternative in these market segments. World oil prices can also have an effect on natural gas and fuel oil demand in the Railbelt. Because the Alaska economy depends heavily on the energy supply industry, high oil prices would generate general economic growth in Alaska and in turn, might lead to higher demand than the projections employed here. As noted earlier, however, this analysis is one of a set of studies commissioned by the Alaska Power Authority. The comprehensive analysis of domestic demand in the Railbelt Region (residential/commercial and electric) is being undertaken in another effort and was beyond the scope of this analysis. The Authority’s directive to ICF-Lewin for these demand segments was to rely on demand projections published by the Alaska Department of Natural Resources to the maximum extent feasible. Alaska Department of Natural Resources, Division of Oil and Gas, 20T00303 Page 2-1 Forecasts a Identify Key Fuels FIGURE 2.1 OVERVIEW OF ANALYTIC APPROACH Categorize by "Schools of Thought” Price Scenarios Develop FUEL OIL PRICE ANALYSIS Adapt ADNR Demand Projections Identify Major ' { Representative [—; 1 { ' ' Segment Markets Analyze Costs and Economics by Market Segments End Uses identify Segment Suppiles . Suppiles 20T00303 NATURAL GAS PRICE ANALYSIS Identify Major End Uses Adapt ADNR Demand Projections Characterize Known Fuels 1 Establish Statistical Distribution of Resource Estimate Remaining Fieids Assess NG Wellhead Value by End Use Estimate Marginal Costs integrate 0&s to Forecast Market Balance Page 2-2 FUEL OIL APPROACH This study developed price projections for the following fuels: e No. 4 distillate fuel oil prices for electric utilities in Fairbanks e No. 2 distillate fuel oil prices for residential and commercial users in Fairbanks ° No. 2 fuel oil prices for the Southern Railbelt In a competitive market, the price of distillate fuel oil is determined by the point where the quantity demanded equals the quantity supplied. Such a competitive market typically consists of many suppliers. No one supplier or group of suppliers has market power, and there are no barriers among market segments. Unlike the theoretical competitive market, the Alaska fuel oil market consists of a few suppliers, and the different geographic markets are separated by significant transportation costs. For the two major Railbelt fuel oil markets, Cook Inlet and Fairbanks, the supply curve can be divided into segments: ° Low-cost local refiners . Higher-cost local refiners e Outside suppliers -- West coast, foreign, or (for Fairbanks) Anchorage suppliers In such a market, the future price of distillate fuel oil will be bounded on the lower side by the variable costs of the marginal supplier and bounded on the upper side by the full costs to the user of the next best alternative source of supply. The marginal supplier is determined by the level of demand and refiner capacity. 20T00303 Page 2-3 While focusing primarily on the Alaska distillate fuel oil market, this analysis also had to examine the effects of potential changes in the supply and demand for other fuels in Alaska. The cost of producing distillate fuel oil is a joint cost that refineries must recover from the sale of all petroleum products produced. Thus, factors that affect the supply and demand for gasoline and jet fuels can have an effect on a refiner’s ability to recover the cost of production from the sale of these products and therefore affect the price at which he is willing to sell distillate fuel oil. In addition, the Alaska petroleum product market is affected by developments in the world market. On the supply side, the world oil market influences the price of distillate fuel oil and other petroleum products in the Alaska Railbelt. The wellhead price of Alaska North Slope crude oil, which is the primary crude used in Alaskan refineries, is determined as a net-back price from the U. S. West coast and Gulf coast. Thus, factors that affect the value of Alaska North Slope crude oil on the U. S. West coast or in the Gulf coast will have a direct effect on what Alaska refiners will have to pay for crude oil and therefore on the price of distillate fuel oils in the Alaska Railbelt. On the demand side, the world oil market affects the price of distillate fuel oil in the Railbelt because the price that Alaskan refineries receive for exported residual fuel oil is determined by the value of these products on the U. S. West coast, the Caribbean and the Pacific region. 2 A decline in the world residual fuel oil price would tend to force Alaskan refineries to increase the price of their other petroleum products in order to cover the costs of production. In addition, the price that Alaskan refineries can charge for petroleum products is limited by the price of these products on the U. S. West coast plus transportation costs to 2 Although residual fuel oil is the highest volume product of Alaskan refineries, little or no residual fuel oil is consumed in Alaska primarily due to its poor flow characteristics in cold temperatures. 20T00303 Page 2-4 Alaska. The price of petroleum products on the West coast in turn are determined in the world oil market. In summary, the price of Alaskan distillate fuel oil in the Railbelt will be determined by the marginal supplier. The price set will be at least sufficient to cover the variable costs of production. The variable costs of production are a joint cost that must be recovered from the sale of all the petroleum products produced by the refinery. Therefore, the price at which the marginal supplier will be willing to sell distillate fuel oil will be affected by the supply and demand for other petroleum products in Alaska and in the world oil market. Alternatively, the upper bound price of distillate fuel oil will be set by the cost of obtaining it from the next best supplier. Working within this theoretical framework, the starting point for this analysis was an examination of current Alaskan petroleum product demand, supply and prices. Data on 1987 refinery capacity, production, fuel sales and prices were collected and analyzed. ICF-Lewin conducted interviews with representatives of each component of the petroleum product market in the Railbelt including refinery executives and marketing representatives, fuel oil distributors, and electric utility fuel procurement personnel. The collected data and insights gained from these interviews were used to disaggregate the price of distillate fuel oil into the following components: e Delivered cost (to refinery) e° Refiner gross margin (including processing costs) e Transportation cost to market . Distribution costs and margin Summing the first two components yields the refiners rack price. Adding the transportation costs and the distribution costs and margin to the rack price yields the current delivered price of distillate fuel oil. These price components were calculated for each of the key refineries in the Alaskan market. 20T00303 Page 2-5 The next step in the analysis was to project future supply and demand for petroleum products in the Alaskan Railbelt. The focus of this analysis was to examine those factors that could cause the supply/demand balance to change dramatically from the 1987 baseline. Assumptions about future world crude oil prices and market conditions were derived from earlier ICF work done for this Railbelt project.? Petroleum product demand estimates for the Alaska Railbelt region were taken from projections made by the Alaska Department of Natural Resources (ADNR) .4 The primary determinants of demand in the ADNR methodology are population and economic growth. Additional analysis was conducted to determine how these ADNR estimates might be affected by factors not directly related to population or economic growth. In particular, potential changes in the airline industry and the Alaska fishery market were examined to determine what effect they might have on the demand for jet fuel and marine diesel fuel. These demand projections were then compared with in-state refining capacity to determine the source of marginal supply, and to assess their potential impact on distillate fuel oil prices. Based on this supply/demand market assessment, fuel oil price projections were developed for the three key distillate fuel oils of interest. Price projections were developed for each of the three alternative world oil price scenarios. NATURAL GAS ANALYTIC APPROACH The analysis of the natural gas markets in the Railbelt focused on the Cook Inlet region. The analysis had three major elements: 3 ICF-Lewin Energy Group, oks for W ices; sis ools t 19 : , also in this volume. 4 Alaska Department of Natural Resources, d ected Qil_and Gas Consumption. 20T00303 Page 2-6 . Demand analysis. ° Supply Analysis. ® Integrative/balancing analysis. As noted earlier, the demand analysis relies extensively on Alaska Department of Natural Resources published projections for domestic uses. The export-oriented demand components incorporate significant original work. The supply analysis incorporates major new analysis of Cook Inlet natural gas resources. The integrative analysis examines the demand and supply curves developed to assess the prices expected in the resulting supply/demand balance. A key aspect of this step is to determine whether the market will clear on Cook Inlet-generated demand or export-oriented demand. Demand Approach The demand side of this natural gas market analysis develops an approximation of the demand curve for the Railbelt natural gas marketplace. The demand curve is the relationship between price and quantity of a good bought in the marketplace. For natural gas, this demand occurs in the Cook Inlet Region. 5 The analysis estimates price-quantity pairs for each of the three world oil price scenarios in 1990, 1995, 2000, 2005, and 2010. The demand curve is approximated by segmenting the market into several distinct categories. Each category has similar fuel alternatives or other important similarities. These segments are: ° Residential/Commercial. ° Electric utility. e LNG (existing plant) Chapter 3 includes a brief discussion of a proposal to supply natural gas to Fairbanks. 20T00303 Page 2-7 e LNG (new plant) e Ammonia/urea (new plant) . Methanol (new plant) The first two segments cover "domestic" (i.e., Railbelt) consumption. The quantity figures are taken from the ADNR projections and are not varied by world oil price scenario. Under most expectations, these segments are relatively stable. The associated price levels are derived on the basis of alternative fuels: Segment Alte tive Fue Residential/Commercial Retail distillate fuel oil Electric Utility Wholesale distillate fuel oil The other segments relate to export-oriented consumption. There are currently two major export activities: liquefied natural gas (LNG) production and ammonia/urea (fertilizer) production. Both processes use significant volumes of natural gas as feedstock. LNG production growth at the existing LNG facility is based on recent projections by the facility operators. © The effective price for LNG production at the existing facility is estimated through the following steps: Specify world oil price scenario Adjust world oil price to crude FOB price (Tokyo) Relate crude price (Tokyo) to LNG price (Tokyo) Net estimated full costs of LNG transport (Kenai to Tokyo)’ 8 uF WN BF Net estimated marginal costs of LNG production at Kenai. 6 See Phillips-Marathon, Application to Amend Authorization to Export Liquefied Natural Gas, EPA Docket No. 88-22-LNG, 4-7-88. 7 ICF Incorporated, aska Natural Ga velopment, An Economic Assessment of Marine Systems, for U.S. Department of Transportation, Maritime Administration, Sept. 1982. 8 Ibid. 20T00303 Page 2-8 The full cost of LNG transport assumes the operators would require additional vessels to handle the incremental volumes from the existing plant.? The marginal costs of production are netted out but no capital- related costs are netted-out because no additional investment would be 10 required for this production. This yields a natural gas price that the plant owner would be willing to pay for feedstock to the plant. There is also an existing ammonia-urea plant. No incremental additions are planned at this time, however, and currently this activity is essentially severed from the Cook Inlet market because the gas supply is provided completely by the owners of the plant. The estimated economics of this demand segment are included, however, to take account of the possibility that this plant could seek other natural gas supplies in the future. The remaining three segments relate to potential new export projects. For LNG, the price calculation is identical to the steps noted above, except that step 5 nets out the fully-loaded (average) costs of LNG production. For a new ammonia/urea facility, the following steps yield the gas price estimate: I. Specify World Bank price projection! Adjust to price on the U.S. West Coast. 9 Phillips-Marathon, Application. Under the new contract, Phillips- Marathon will increase exports to Japan from about 50 trillion Btus/year to about 58 trillon Btus/year. 10 Phillips-Marathon, Application. One might further segment this demand into current and incremental production. This further refinement would imply greater precision than is present in these figures. ‘ 11 William F. Sheldick, World Nitrogen Survey, World Bank Technical Paper Number 59, April 1987. 20T00303 Page 2-9 Sis Net out estimated full costs of ammonia/urea transport}? 4. Net out estimated full costs of production. 13 This provides the maximum price the plant owner would be willing to pay for feedstock to the plant. For a new methanol facility, the following steps yield the gas price estimate: i Price and quantity estimates for methanol and gasoline fuel in California 2. Price adjustments (distribution and transportation) to Anchorage B, Net out fully loaded costs of methanol production. This yields the maximum price a methanol fuel plant owner would be willing to pay for feedstock to produce methanol for sale as a fuel in California. The price/quantity pairs generated through this exercise provide the segments of a demand schedule, as illustrated in Figure 2.2. The analysis required fifteen demand schedules (for 1990, 1995, 2000, 2005, 2010 and for each of the three world oil price scenarios). Supply Approach The supply analysis represents major new work in appraising the marginal economics of future discoveries from the Cook Inlet oil and gas resource base. The supply analysis begins with an evaluation of the production potential from existing fields. Assuming no additional investment in developmental wells, the analysis shows the decline in production that would take place from each of the producing gas fields. 12 ICF Incorporated, Alaska Natural Gas Development. 13 qpid. 20T00303 Page 2-10 FIGURE 2.2 ILLUSTRATIVE DEMAND FOR NATURAL GAS R/C ELECTRIC UTILITIES - LNG (EXISTING) (NEW) Q In order to produce enough gas to satisfy future levels of demand, additional investments will have to be made. Using well accepted reserves estimates and statistical techniques, the analysis identifies the costs of adding reserves and production to meet the anticipated demand. To do this, the study classified the undiscovered resource base into three categories based on increasing levels of uncertainty associated with the estimates and costs: . Developmental resources. ° New field exploration. e Speculative resources. : The analysis looked both at onshore and offshore resources, and associated and non-associated gas fields. Using economics, investment and statistical criteria, it estimates the marginal costs of adding incremental supplies of gas. These marginal costs are then used in the market balancing effort. 20T00303 " Page 2-11 3. FUEL OIL MARKET ANALYSIS This chapter of the study presents the ICF-Lewin analysis of the Fairbanks fuel oil market. Two components of this market are of particular interest: ° No. 4 distillate fuel oil for the Golden Valley Electric Association (GVEA), and ° No. 2 fuel oil for residential and commercial customers. The first of these two components is important because the economic attractiveness of upgrading the electricity intertie between Anchorage and Fairbanks will in large part depend on the relationship between the delivered price of natural gas to electric utilities in South Central Alaska and fuel oil prices’ that GVEA would have to pay in Fairbanks. The second component, home heating oil consumption, is important because the economic attractiveness of building a natural gas pipeline from Anchorage to Fairbanks will depend in large part on the difference between the delivered price of natural gas and fuel oil to residential and commercial customers in Fairbanks. These two components of the Fairbanks fuel oil market have fundamentally different structures. Golden Valley purchases royalty crude oil from the state, sells it to the nearby Mapco refinery, which then processes the crude oil and sells back No. 4 fuel oil to GVEA for a price slightly above its costs. This is a unique but relatively simple and isolated supply/demand structure. In contrast, the Fairbanks residential and commercial fuel oil market is much more complex. Heating oil is supplied to the Fairbanks market by two local refineries, Mapco and Petro Star, and the Tesoro refinery in Nikiski. The price of heating oil in Fairbanks is set by the marginal supplier, which is the Tesoro refinery. 20T00284 Page 3-1 Because the costs of refining crude oil are joint costs that must be recovered from the sale of all petroleum products, the price at which Tesoro is willing to sell heating oil in Fairbanks is affected by the price that it can obtain from the sale of the other petroleum products that it produces. Thus, for example, a decline in the price that Tesoro receives for gasoline sold in Anchorage, or the price it receives for residual fuel oil in Japan may lead Tesoro to increase the price of No. 2 fuel oil that it sells in Fairbanks. Because the electric utility fuel oil market and the residential/ commercial fuel oil market have fundamentally different structures, the analysis performed for each market was very different. The projection of No. 4 fuel oil prices for GVEA focused on estimating the components of Mapco’s crude oil acquisition costs: wellhead price plus transportation costs. Projecting fuel oil prices for the Fairbanks residential and commercial market was more complex. In addition to estimates of acquisition costs, it required a review of the entire Alaska petroleum product market. The focus of this analysis was on those factors that might lead Tesoro to change its Fairbanks fuel oil price. The next section of this chapter discusses the current supply and demand situation for these two components of the Fairbanks fuel oil market. It also presents a brief description of the current supply and demand situation for the other major petroleum products in the Alaska Railbelt. Working from this description of the current situation, the next section of the chapter presents projections of future demand for petroleum products in the Railbelt. This is followed by a discussion of the adequacy of in-state refinery capacity to meet projected future demand in the Railbelt region. The chapter concludes with projections of future prices of No. 4 fuel oil for GVEA and residential/commercial fuel oil for the Fairbanks market. 20T00284 Page 3-2 CURRENT RAILBELT OIL SUPPLY & DEMAND Total Railbelt petroleum product consumption in 1987 was approximately 75 thousand barrels per day (mbd). Railbelt consumption accounted for about 67 percent of total Alaska sales. The largest component of Railbelt consumption was jet fuel, which amounted to 38 mbd, or roughly half of the total petroleum product use. Jet fuel is used in the Railbelt for both domestic and international flights and by the military. The next largest category of consumption is distillate fuel oil, which amounted to nearly 25 mbd in 1987. In addition to heating oil and fuel oil for electric utilities, this category includes marine and highway diesel fuels. Gasoline made up the bulk of the remaining Railbelt petroleum product consumption. Total gasoline use in 1987 was 12 mbd.1 Most of the petroleum products consumed in the Railbelt were supplied by in-state refineries. However, roughly 10 percent of the jet fuel, all of the aviation gasoline, and much of the marine diesel consumed in the Railbelt is imported, primarily from the West Coast. Alaska Refinery Capacity There are a total of six refineries in Alaska with total crude oil distillation capacity of 225,000 barrels per day. (See Table 3.1). The two ARCO refineries are simple units dedicated to serving local requirements on the North Slope and, therefore, are not discussed further in this report. ICF estimates based on Alaska Department of Natural Resources, "Historical and Projected Oil and Gas Consumption," January 1988. ADNR published actual 1987 data for the state as a whole. Actual Railbelt vs. Non-Railbelt data was not available. But, ADNR projected Railbelt and Non-Railbelt consumption from 1988 to 2001. ICF developed 1987 Railbelt petroleum consumption estimates based on: (1) Actual State totals; (2) projected Railbelt consumption for 1988; and (3) information gained from interviews with industry and government personnel. 20T00284 Page 3-3 Current refining capacity available to serve Railbelt requirements is 191 thousand barrels per day (mbd). The largest refinery in Alaska, with crude oil distillation capacity of 90 mbd, is the Mapco Petroleum Inc. refinery located in North Pole, about 13 TABLE 3.1 ALASKAN CRUDE OIL REFINERIES CRUDE REFINERY c ITY LOCATION (bbls per day) ATLANTIC RICHFIELD CO. 12,000 KUPARUK ATLANTIC RICHFIELD CO. 22,000 PRUDHOE BAY CHEVRON U. S. A. INC. 22,000 KENAI Mapco PETROLEUM INC. 90,000 NORTH POLE PETRO STAR INC. 7,000 NORTH POLE TESORO PETROLEUM CORP. 72,000 KENAI ALASKA PACIFIC REFINERY INC.* 100,000 VALDEZ TOTAL STATE 325,000 Source: OQil & Gas Journal, March 30, 1987. * Proposed, not yet constructed. miles southeast of Fairbanks. The much smaller Petro Star refinery is also located in North Pole. The remaining two refineries are located in Nikiski, Alaska: the Tesoro Petroleum Corp. refinery with a capacity of 72 mbd and the Chevron refinery with a capacity of 22 mbd.2 With the exception of the The Chevron refinery does not compete in the Fairbanks fuel oil market and is not likely to because it will likely be at a cost disadvantage relative to Tesoro, therefore it is not discussed further in this report. 20T00284 Page 3-4 Tesoro refinery, these Alaska refineries are simple units with little downstream conversion or upgrading capacity. A seventh refinery has been proposed for the Valdez area by Alaska Pacific Refinery Inc. (APRI). This refinery would be an efficient state-of- the art operation. Although it would be primarily focused on the export market, it could also sell some products within the Railbelt.3 Major Railbelt Oil Flows Figure 3.1 depicts the major flows of crude oil and petroleum products in the Railbelt region. All of the crude oil processed by Alaskan refineries comes from in-state production. The two Fairbanks area refineries, Mapco and Petro Star, process only ANS crude oil, which travels from Pump Station No. 1 on the North Slope through the TransAlaska Pipeline System (TAPS) to the Fairbanks area. In Fairbanks, the crude oil travels the 2.5 miles from TAPS to the refineries through six and eight inch pipelines, which the Golden Valley Electric Association owns and operates as a common carrier. The two Nikiski area refineries, Tesoro and Chevron, process primarily ANS crude, but also refine smaller amounts of Cook Inlet crude oil. ANS crude destined for Nikiski first travels from Pump Station No. 1 through the TAPS to Valdez. The crude oil moves from Valdez to the Nikiski docks via 70,000 - 80,000 ton tankers. Thus, these two refineries must pay several dollars more than Mapco and Petro Star for transporting crude oil from the North Slope. The Tesoro and Chevron refineries also obtain crude oil from Cook Inlet. Crude oil from the West Side of Cook Inlet travels to Nikiski via tanker or barge. Oil from the east side of Cook Inlet travels via the Kenai Pipeline Company to the refineries. The potential effect of this proposed refinery on the Railbelt is discussed in more detail below. 20T00284 Page 3-5 FIGURE 3.1 MAJOR FLOWS OF CRUDE OIL AND PETROLEUM PRODUCTS IN THE ALASKA RAILBELT ° ,— ' = \ TAPS =e | ! ! eT) | | ee | ., FAIRBANKS | | ~ A} BB ans To marco | im & PETRO STAR ‘|| weve GASOLINE = as _* > oa & a o ee a: eee ANS TO TESORO & CHEVRON ae ea ere | ical JET FUEL & MARINE C ) ie IS FROM WEST COAST amin RES TO LOWER 48 ANS TO LOWER 48 & BUNKERS ae eae 20T00284 Page 3-6 -- Residential Fuel Oil Disposal Over half of total Alaskan refinery production is residual fuel oil. However, due in large part to the cold climate, little or no residual fuel oil is consumed in Alaska. As a result, substantial amounts of residual fuel oil must be exported, primarily to Pacific Rim countries. Some resid is also used as bunker fuel for the ships carrying ANS crude to the Lower- 48. Tesoro and Chevron particularly suffer from the lack of a local market for their largest product, residual fuel oil. Netback prices from the export of residual fuel are currently not sufficient to cover average refinery acquisition and processing costs. In contrast, Mapco and Petro Star, the two Fairbanks refineries, do not export their bottom of the barrel production as residual fuel. Their close proximity to TAPS affords them the unique opportunity to return the bottom of the barrel residual fuel oil to TAPS, where it is commingled with ANS and ultimately sold in Valdez as crude oil. They must pay a quality penalty for this returned oil of about $1.26 per barrel returned.“ Nonetheless, given the relatively low value of residual fuel oil and the transportation costs of shipping it to distant markets on the West Coast or in the Pacific, Tesoro and Chevron are at a disadvantage relative to Mapco and Petro Star, who effectively receive a higher price for their bottom of the barrel products. -- Light Product Flows The major flows of petroleum products into Alaska are jet fuel, aviation gasoline, and marine diesel fuel oil. About 10 percent of the commercial jet fuel used at the Anchorage International Airport is imported, primarily from the West Coast. Smaller amounts of military jet fuel are A quality bank has been set up for the TransAlaska Pipeline. Those who put oil into the pipeline with lower than the average quality as measured by a lower API gravity pay a quality penalty. And those who put oil into the pipeline with higher quality as measured by a higher than average API gravity receive a quality payment. 20T00284 Page 3-7 also imported. Because no Alaskan refineries can produce aviation gasoline, it must also be imported. Some marine diesel fuel for the domestic and international fishing fleet that operates off of the coast of Alaska is also imported, primarily from the West Coast. Additional amounts are sold over the side by the international fleet of ships. Within the Alaska Railbelt, the major flows of petroleum products are directed towards Anchorage. Smaller flows travel north to Fairbanks. The Nikiski area refineries, Tesoro and Chevron, send jet fuel, gasoline, and smaller amounts of other products to Anchorage via a 70 mile, 10 inch pipeline, which is owned by Tesoro Alaska Pipeline Co. Mapco, located in Fairbanks, sends jet fuel and gasoline to Anchorage in tank cars via the Alaska Railroad (ARR). The small Petro Star refinery in Fairbanks serves only the local interior market. Petroleum products for the Fairbanks market come primarily from the local Mapco and Petro Star refineries. However, Tesoro also sells fuel oil and gasoline in Fairbanks. To get petroleum products to Fairbanks, Tesoro first sends it through its own pipeline to Anchorage, where it is transferred to tank cars and hauled to Fairbanks on the Alaska Railroad. Current Consumption of Fuel Oil By Fairbanks Electric Utilities In 1986, the total use of fuel oil by Railbelt utilities was 1.5 mbd.° The Golden Valley Electric Association (GVEA) in Fairbanks accounted for over 98 percent of this total Railbelt electric utility use in 1986. The next largest user of oil by Railbelt utilities was the Fairbanks Municipal Utility System, which accounted for slightly over 1 percent of the total. Complete information on oil use by Railbelt electric utilities is not available for 1987. However, GVEA data shows that its oil use declined by over 60 percent in 1987 from 1986 levels to 0.6 mbd. This decline occurred 5 Alaska Power Authority, "Alaska Electric Power Statistics 1986: Summary Supplement on Railbelt Utilities," November 1987, pp. 4-5. 20T00284 5 Page 3-8 because GVEA shut down its combustion turbines at its North Pole No. 1 and No. 2 units from April 1987 until December 1987. GVEA displaced this oil- fired generation with cheaper power purchased from the Chugach Electric Association and the Anchorage Municipal Light & Power Company. This purchased power was transmitted from the Anchorage area to GVEA through the existing Railbelt electricity transmission intertie. Golden Valley purchases all of the No. 4 fuel oil used at its North Pole units from Mapco. The terms of sale for this fuel oil are unusual because Golden Valley is also a crude oil supplier. GVEA purchases royalty crude oil from the state. It then sells the crude oil to the nearby Mapco refinery. Mapco processes the crude oil and sells back No. 4 distillate fuel oil to GVEA. Mapco sells oil to GVEA under two contracts. The first contract is associated with the royalty oil that GVEA sells to Mapco. A second contract covers volumes over and above the royalty oil sales. Golden Valley's royalty oil contract with the state stipulates that the crude oil purchased thereunder is assigned to Mapco, which will process the crude oil and sell back fuel oil for generation at a reduced refinery charge. 6 The maximum price under this contract for 1987 was $0.44 per gallon. But, the actual price may be as little as $0.40 per gallon.7 Golden Valley has at times also bought fuel oil from Mapco under spot sales agreements, but is currently not doing so. The purchase price of $0.40 to $0.44 per gallon indicates that Mapco is selling No. 4 fuel oil to GVEA at only slightly above its supply costs. The Fairbanks Municipal Utility System purchases much smaller amounts of No. 2 distillate fuel oil. Total purchases in 1986 were only 264,000 gallons, or the equivalent of 17 barrels a day. MUS purchases its No. 2 fuel oil from Mapco at a reduced price, relative to the price of No. 2 fuel Contract between the state of Alaska and Golden Valley Electric Association, signed February 8, 1985, Sec. 2.11. Robert Hansen, Manager of Finance & Administration, Golden Valley Electric Association, February 1988. 20T00284 Page 3-9 oil sold to residential users. However, the FMUS price is substantially above the GVEA price for No. 4 fuel oil. Current Residential/Commercial Fuel Oil Consumption and Distribution in Fairbanks Total consumption of No. 1 and No. 2 fuel oil in Fairbanks in 1987 was approximately 2.9 mbd.8 An additional 1.5 mbd of heating oil was consumed in other interior areas of Alaska. Three different grades of distillate fuel oil are consumed in Alaska: No.1, No.2, and Blended No. 2. These fuels are primarily distinguished by their flow characteristics in cold weather. Blended No. 2 oil is often substituted for No. 2 fuel oil in the winter months in the Interior parts of Alaska. In Fairbanks, No. 2 fuel oil, including Blended No. 2, accounted for about 70 percent of total home heating oil consumption. Prior to the construction of the Earth Resources refinery in 1977, which was subsequently purchased by Mapco, all heating oil and gasoline consumed in the Fairbanks area was imported by five or six branded dealers including Texaco, Chevron and Tesoro. Each of these major oil companies with branded dealers maintained bulk storage facilities in Fairbanks. There were no independent distributors at that time. The market structure changed with the addition of the local Earth Resources (Mapco) refinery. Earth Resources began selling to existing large distributors and to other new independent distributors. The market was further opened to small independent distributors as a result of antitrust litigation that forced Mapco to sell all heating oil at the same rack price. Independent distributors were able to enter the market and essentially use Mapco as their bulk storage facility. As a result, the number of distributors increased from roughly 5 to around 20. Conversation with Steven T. Lewis, President, Petro Star Inc., February, 1988. 20T00284 Page 3-10 The market structure changed again when the new Petro Star refinery came on line in 1985 and proceeded to buy two distributors, including Sourdough Fuel, the oldest and largest distributor in the Fairbanks area. Both Mapco and Tesoro have followed suit by buying their own distributors. Currently, Mapco, Tesoro, and Petro Star supply all of the heating oil used in Fairbanks and the surrounding interior region. The major oil companies, with the exception of Tesoro, no longer maintain bulk fuel storage facilities in Fairbanks. Instead, they purchase their heating oil from the two local refineries and Tesoro. As shown in Figure 3.2, Mapco has about 44 percent of the Fairbanks heating oil market. Mapco has a larger share of the heating oil market in other interior areas surrounding Fairbanks. Petro Star has about one-third of the combined Fairbanks and other interior market. It has a slightly larger share of the in-town market than of the surrounding interior areas. The remainder of the market is supplied by Tesoro. It's sales are concentrated in the city of Fairbanks, where it has about a 22 percent. market share in 1987. Tesoro supplies under 10 percent of the other interior areas. -- Heating Oil Supply Costs Tesoro is the highest cost supplier of home heating oil in the Fairbanks market. As shown in Table 3.2, Tesoro paid about $2.20 per barrel more for transporting ANS crude oil from the wellhead to the refinery gate than did Mapco and Petro Star. It must also pay about $4.00 per barrel to transport fuel oil from its Nikiski refinery to Fairbanks via pipeline and 20T00284 ‘ Page 3-11 FIGURE 3.2 REFINERY MARKET SHARES IN THE FAIRBANKS AND OTHER INTERIOR HOME HEATING OIL MARKET (PERCENT) 100 80 + 60 40 5 20 4 ° A. FAIRBANKS OTHER INTERIOR FAIRBANKS & OTHER INTERIOR TESORO PETRO STAR 7 MAPCO Source: Petro Star Inc. railroad tank cars.? As a result, Mapco and Petro Star have a significant location advantage in the Fairbanks market. As noted above, Tesoro is also at a disadvantage because there is no market for residual fuel oil in Alaska, which accounts for between 40 and 50 percent of Tesoro’s total production. Tesoro must export the residual fuel 9 $0.80 per barrel pipeline cost from Nikiski to Anchorage, and $3.20 per barrel railroad tariff from Anchorage to Fairbanks. The pipeline from Nikiski to Anchorage is owned by Tesoro. Although Tesoro could cut this pipeline tariff substantially, it has little incentive to do so because the resulting drop in wholesale heating oil prices would not be enough to offset the other sellers’ price advantage. 20T00284 . Page 3-12 TABLE 3.2 1987 CRUDE OIL ACQUISITION COSTS FOR ALASKA REFINERIES TESORO Mapco PETRO_S' ANS cI AVERAGE WELLHEAD! $11.05 $16.00 $11.26 $9.96 PRICE PIPELINE TARIFF2 $ 3.93 N/A $ 2.50 $ 2.58 TANKER RATE? $0.75 §$ 0.27 N/A N/A ACQUISITION 915.73 $16.27 $15.86 $13.76 $12.54 COST 1 Average price of Royalty Crude Oil purchases obtained from Alaska Department of Natural Resources. Assumes non-royalty crude oil purchased for same price as royalty crude oil. 2 Actual tariff obtained from Alaska Department of Natural Resources. 3 Tesoro. Note: Variations in wellhead prices are due to the fact that Mapco and Tesoro purchase Prudhoe Bay crude oil and Petro Star purchases Kuparak crude oil. In addition, variations in price reflect differences in the terms of royalty oil contracts. Variations in pipeline tariffs reflect differences in points of origin and destination. oil, primarily to the Pacific Rim. The netback price from these sales is below Tesoro’s average acquisition and processing costs. In contrast, Mapco is able to return the bottom of the barrel products to TAPS. Although it pays a penalty of about $1.26 per barrel on returned oil, Mapco’s costs associated with this returned oil are not as large as the costs incurred by Tesoro on its export sales of residual fuel oil. 20T00284 Page 3-13 -- Heating Oil Prices in Fairbanks The average price of No.2 heating oil at Tesoro’s terminal at the Fairbanks Airport in 1987 through November was $0.67 per gallon. 10 The annual average wholesale prices for No. 2 fuel oil sold by Mapco and Petro Star were not available. However, according to Fairbanks distributors, Mapco and Petro Star wholesale prices for fuel oil tend to be a few cents per gallon below the Tesoro terminal price. Subtracting the estimated cost of transporting fuel oil from Anchorage to the Fairbanks terminal yields a No. 2 wholesale price in Anchorage of about $0.59 per gallon. In contrast, the average refinery sale for resale price in Washington state was lower in 1987 by about $0.06 per gallon, or $2.50 per barrel.11 his difference is sufficient to cover the cost of transporting fuel oil from the West Coast to Anchorage. In 1987, West Coast refineries could potentially have delivered fuel oil to the Fairbanks market at a lower cost than Tesoro, but they did not do so probably because of the relatively small volumes and the in-state refineries ability to cut their prices. Complete data on the annual average retail price of fuel oil in Fairbanks are not available. However, each fall the Fairbanks Community Research Center publishes the September price of No. 1 and No. 2 fuel oil for each of the major distributors in the Fairbanks area. In September 1987, retail No. 2 fuel oil prices ranged from a low of $0.80 per gallon to a high of $0.93 per gallon, with an average price of $0.85 per gallon. This relatively wide range of prices reflects for the most part varying levels of service, from simple delivery to complete service including maintenance. Comparing this average price to Tesoro’s September, 1987 Fairbanks terminal 10 Heating oil prices supplied by Enstar. The price reported here is a combination of No.2 fuel oil prices for summer months and Blended No.2 fuel oil prices for winter months. 11 U. S$. DOE/EIA. Petroleum Marketing Monthly. 20700284 Page 3-14 price of $0.68 per gallon yields a gross distribution margin of $0.17 per gallon. As noted above, Mapco and Petro Star generally have lower wholesale prices. Distribution margins would be higher if calculated using these two Fairbanks refineries’ wholesale prices. However, regardless of which refinery’s wholesale price is used, Fairbanks distribution margins were below the U. S. average of $0.27 per gallon. 12 RAILBELT PETROLEUM DEMAND OUTLOOK Figure 3.3 summarizes the Railbelt petroleum product demand projections used for this analysis. The more detailed product demand projections are presented in Table 3.3. The rates of growth in petroleum product consumption for each demand category were taken from the Alaska Department of Natural Resources (ADNR) report: "Historical and Projected Oil and Gas Consumption," which was published in January, 1988. The 1987 petroleum product consumption data were developed using ADNR estimates and data provided by the Alaska Department of Revenue (ADOR). The ADOR data showed actual total state demand. The ADNR estimates were used to disaggregate this total into Railbelt and Non-Railbelt estimates. The ADNR estimates that total petroleum product demand in the Railbelt will increase at 1.2 percent per year, rising from 75 mbd in 1987 to 100 mbd in 2010.43 Total Non-Railbelt petroleum product demand is projected to increase from 37 mbd in 1987 to 46 mbd in 2010, an increase of about 1 percent per year. These demand projections are based on assumed relatively low rates of growth in economic activity and population levels. The 12 U.S. DOE/EIA, Petroleum Marketing Monthly. 13 The ADNR projected petroleum product demand to the year 2001. Estimates for the year 2010 were developed assuming that demand would continue to grow at the same average annual rate of growth that was projected for the period from 1995 to 2001. 20T00284 Page 3-15 FIGURE 3.3 PROJECTED PETROLEUM PRODUCT DEMAND FOR THE ALASKA RAILBELT (MBD) 120 — —— 100 80 = [ asOLINE sco fF OISTRLLATE 40 — JET FURL 20 aj thew 1990 1998 2000 2008 2010 TABLE 3.3 PROJECTED PETROLEUM PRODUCT DEMAND ’ FOR THE ALASKA RAILBELT (MBD) RAILBLT 1987 1990 1995 2000 2010 JET FUEL 7.8 39.3 42.9 48.1 0.9 OomesTIC 17.8 19.0 21.8 3.9 36.3 INTERMATIONAL © 20.0 2.3 21.1 22.2 6.6 MILITARY GASOL Ime 12.2 WA7 Wo 103 11.2 AVIATION 0.9 0.9 0.9 0.9 0.9 NIGMMAY 10.8 10.3 10.0 9.9 9.8 waning 0.5 0.5 0.5 0.5 0.5 DISTILLATE 2.7 2.3 6.6 3.7 2.2 SPACE HEAT 3.9 3.7 3.7 3.8 4.1 UTILITY GEWERATION 0.7 0.7 0.7 0.7 0.7 MARINE DIESEL 10.2 10.2 10.6 Wee 13.2 MIGMMAY OLESEL 9.9 9.7 9.6 9.8 10.2 InDusTRY 0.0 0.0 0.0 0.0 0.0 TOTAL RAILSELT 7.7 - 73.3 78.9 s.1 100.3 Sources: Alaska Department of Revenue; Alaska Department of Natural Resources, with ICF adjustments. * International includes military purchases. 20T00284 Page 3-16 Railbelt demand projections for each end-use category are summarized below. Non-Railbelt projections are discussed where they are relevant. Electric Utility Demand As noted above in the Current Supply & Demand section, the Golden Valley Electric Association (GVEA) accounted for over 98 percent of Railbelt electric utility fuel oil consumption in 1987. However, GVEA cut their 1986 demand by more than 60 percent in 1987. The combustion turbines at their North Pole generation plant were shut down from April until December. GVEA shut down these units because it could obtain purchased power from Chugach Electric Association and Anchorage Municipal Light & Power more cheaply. The ADNR projects that electric utility demand for fuel oil will remain constant at the 1987 level of 0.7 mbd. This is consistent with a scenario in which GVEA continues to purchase much of its power requirements and run its oil-fired combustion turbines at minimum levels. This forecast also implicitly assumes no new oil-fired generation and no switching from oil to gas in the electric utility sector. Home Heating Oil Demand Railbelt heating oil demand is projected to remain essentially unchanged over the forecast period. It declines slightly between 1987 and 1990 from 3.9 mbd to 3.7 mbd. It remains unchanged between 1990 and 1995 before rising to 4.1 mbd by 2010. Non-railbelt heating oil demand is projected to increase at a faster rate, rising from 6.3 mbd in 1987 to 7.2 mbd in 2010. Assuming that the projected changes occur uniformly in all areas of the Railbelt, Fairbanks heating oil demand would increase very slightly from 2.9 mbd in 1987 to 3.0 mbd by 2010. 20T00284 Page 3-17 Other Petroleum Product Demand The largest component of Railbelt petroleum product demand is commercial and military jet fuel. In 1987, jet fuel accounted for nearly 50 percent of total Railbelt product demand. The ADNR projects that total jet fuel consumption in the Railbelt will increase at an annual average rate of growth of 2.1 percent over the period from 1987 to 2010, rising from 34.7 mbd to 55.7 mbd by 2010. However, several developments in the international airline industry could cause jet fuel consumption to be substantially below these projected levels. First, Boeing has developed a new more efficient jet engine for use in its new Series 747 jets. These more efficient engines may permit intercontinental passenger flights that currently refuel in Anchorage or Fairbanks to fly over Alaska. Freight traffic would not be affected by this development because of the economic disincentives of trading fuel for cargo. Secondly, Russia has recently opened its airspace to commercial air traffic, which again may permit international traffic to bypass Alaska. 14 Because of these two developments, some analysts have begun to revise their jet fuel demand projections downward. Mapco, the largest supplier of jet fuel in Alaska, currently projects that the rate of growth in jet fuel demand will slow down over the next several years, and in the long term demand may decline by as much as 10 percent. 15 The rate of growth in marine diesel demand could also be substantially different from the ADNR projections. Marine diesel is currently the largest component of distillate fuel oil demand (See Table 3.3). ADNR projects that marine diesel consumption will remain constant between 1987 and 1990, then grow at a rate of less than one percent per year through 1995, and 14 Conversation with Heinz W. Noonan, Airport Development Planner, Department of Transportation and Public Facilities, February 1988. 15 Conversation with R. Jack Turner, General Marketing Manager, Mapco Alaska Petroleum Inc., February 1988. 20T00284 Page 3-18 subsequently grow at a rate of 1.5 percent per year. In contrast, state marine diesel consumption more than doubled in 1987. This large increase in demand was due to the growing size of the domestic shipping fleet and the more aggressive efforts by Alaskan refineries to supply this market. Increased rates of growth are likely to continue due to the expansion of the Alaskan fishing fleet in the near future. Railbelt gasoline consumption in the ADNR demand projections decline at an average annual rate of -0.4 percent over the 1987 to 2010 period with most of the decline occurring before 1990. Highway diesel demand remains essentially constant over the forecast period. RAILBELT PETROLEUM SUPPLY OUTLOOK Given the low rates of growth discussed above, current refinery capacity should probably be ample to meet projected requirements for most fuels. However, current in-state refinery capacity is not sufficient to supply total Railbelt jet fuel demand. Roughly 10 percent of total Jet A fuel oil is currently imported from the lower-48 states. An increase in jet fuel or marine diesel sales would lead to a need for additional in-state refinery capacity or additional imports from the West Coast or Pacific Rim. Alaska Pacific Refinery Inc. (APRI) has proposed to build a new very efficient refinery in the Valdez area. This refinery is reportedly intended 16 to sell distillate products into the export markets. However, it could also supply in-state requirements of distillate fuel oil, replacing imports from the West Coast. FUEL OIL PRICE OUTLOOK This Railbelt Study projects prices for the following fuels: 16 Oil & Gas Journal, September 1, 1986, p. 31. 20T00284 Page 3-19 ° No. 4 distillate fuel oil prices for the Golden Valley Electric Association in Fairbanks . No. 2 distillate fuel oil prices for residential/ commercial users in Fairbanks. The following section presents projected prices for these fuels for three different world oil price forecasts: Consensus; Consensus (Low); and Low Price School. These world oil price forecasts are taken from previous work done by ICF for this Railbelt Study. 17 The projections reported below illustrate that fuel oil prices for the Fairbanks electric utilities are likely to track changes in world crude oil prices fairly closely. Residential/commercial prices contain a higher proportion of transportation and distribution costs. As a result, residential/commercial prices will not follow changes in world crude oil prices as closely. As discussed below, electric utility fuel oil prices in Fairbanks are projected to be set by Mapco, a local refinery. Mapco sells fuel oil directly to GVEA with only a small mark-up, thus avoiding additional distribution costs. In contrast, marginal wholesale residential/commercial fuel oil prices are projected to be set by Tesoro. Tesoro must pay several more dollars per barrel in transportation costs for North Slope crude oil. It must also pay an additional $4.00 per barrel to transport fuel oil back to Fairbanks. Distribution margins must be added onto the Tesoro wholesale price to derive the retail heating oil prices for Fairbanks residential customers. Because of these higher transportation and distribution costs, there is less of a link between world crude oil prices and changes in delivered prices for Fairbanks residential/commercial customers. 18 17 ICF-Lewin, Qutlook for World 0 ces: Ana s_ of Alternative Schools of Thought, June 1988. 18 In 1987, the North Slope wellhead price represented over 60 percent of the delivered price of fuel oil to GVEA. In contrast, the North Slope wellhead price represented about 30 percent of the delivered price of fuel oil to residential customers in Fairbanks. 20T00284 Page 3-20 Fuel Oil Price Projections for GVEA This chapter has shown that the electric utility and home heating oil components of the Fairbanks fuel oil market have different structures. The Golden Valley Electric Association has a unique situation for acquiring fuel oil. It essentially buys back its own oil that has been processed at the nearby Mapco refinery. GVEA’s current contract with the state stipulates that Mapco is to process the royalty crude oil and sell it back to the GVEA at a reduced margin. 19 Thus, the primary determinants of the cost of fuel oil to GVEA are the wellhead price of ANS and the TAPS tariff. ANS wellhead prices are determined by developments in the world oil market. The TAPS tariff is set by the Federal Energy Regulatory Commission. As a result, the price GVEA pays for fuel oil is to a large extent isolated from developments in other segments of the petroleum product market in Alaska. The future price of fuel oil for GVEA is likely to be set in much the same way. GVEA’s current contract for state royalty erude oil extends through 1995, which gives GVEA leverage in negotiating with Mapco. Even without this, Mapco’s transparent cost structure for supplying GVEA with fuel oil is likely to limit Mapco’s ability to raise its margins much above current levels on sales to GVEA. No. 4 fuel oil price projections for GVEA were calculated by first estimating Mapco’s refiner acquisition costs. Then, gross margins, which include Mapco’s processing costs and return on investment, were added to get a delivered price. The gross margin was assumed to remain at its current level. Refiner crude oil acquisition costs consistent with the alternative 19 The GVEA royalty oil contract states: "Purchaser also receives from MAPCO a lower refining charge or processing fee which Purchaser passes directly through to its consumers in the form of reduced electric rates. If Purchaser's arrangement with MAPCO ever fails to yield these benefits, Purchaser has a commitment from the Rural Electrification Administration for mortgage funds to be made available to pay for the prompt conversion of Purchaser’s generating units so that the Oil could be burned as fuel by Purchaser without first being processed." pg. 7. 20T00284 Page 3-21 world. oil price forecasts were calculated by first calculating the ANS netback price and then adding on the transportation costs to move the crude oil from the North Slope to the refinery. Table 3.4 presents the resulting projections of No. 4 fuel oil prices to GVEA. Appendix C presents the detailed components of these price projections. Fuel Oil Price Projections for the Fairbanks Heating Oil Market The heating oil segment is more complex than the electric utility component of the Fairbanks fuel oil market. Wholesale heating oil prices in the Fairbanks market are currently set by the marginal supplier, which is the Tesoro refinery in Nikiski. Because of its location, Tesoro must pay several dollars more for ANS crude oil than the two local Fairbanks utilities, Mapco and Petro Star. In addition, Tesoro must pay an additional $4.00 per barrel to transport heating oil to Fairbanks. The two Fairbanks refineries follow Tesoro’s lead in setting heating oil prices for the Fairbanks market. However, Mapco and Petro Star sell heating oil for somewhat less than Tesoro. Given Mapco’s and Petro Star’s lower costs, there is substantial room for further declines in wholesale margins, but there is little incentive for them to reduce these margins. If a natural gas pipeline were built to Fairbanks, heating oil sales would decline substantially. The remaining heating oil market would consist of those more remote areas beyond the limits of a gas distribution system. In these circumstances Tesoro, which currently has the smallest share of the "Other Interior" market (See Figure 3.2), would likely stop selling heating oil in the Interior. Nonetheless, heating oil prices are likely to continue to be set by the cost of transporting fuel oil from the Nikiski area. Mapco and Petro Star would have little incentive to reduce their prices below this cost of marginal supply. This assumes that natural gas service would be initiated only if it could be assured a significant price advantage vis-a- vis fuel oil. Such an advantage could require a substantial subsidy (e.g., state provided transmission system). The economics of this issue were beyond the scope of the ICF-Lewin project. 20T00284 Page 3-22 TABLE 3.4 FUEL OIL PRICE PROJECTIONS FOR FAIRBANKS ELECTRIC UTILITIES (1987 $/GALLON) 1987 1990 1995 2000 2010 WORLD OIL PRICE SCENARIO CONSENSUS $0.44 $0.51 $0.61 $0.74 $0.98 CONSENSUS (LOW) $0.44 $0.46 $0.53 $0.60 $0.74 LOW PRICE SCHOOL $0.44 $0.36 $0.41 $0.46 $0.51 See Appendix C for the detailed components of these projections. Within the Fairbanks area, Mapco is the price leader and principal supplier (Mapco’s refinery has 13 times Petro Star's capacity). Petro Star has nothing to gain by undercutting Mapco’s prices. Therefore, Mapco should be expected to continue to charge somewhat less than the cost of bringing up fuel oil from the Nikiski area, and Petro Star should follow suit. Thus, the future price of fuel oil in Fairbanks will be determined by the factors that affect Tesoro’s costs and margins. Tesoro’s .crude oil acquisition costs will be a function of the wellhead price of ANS and Cook Inlet crude and the Taps tariff. As noted above, ANS and Cook Inlet wellhead prices are determined in the world oil market; and the TAPS tariff is set by the FERC. Absent additional investment, Tesoro’s processing costs should remain relatively constant. Because a refiner’s costs must be recovered from the sale of its full slate of petroleum products, the margin that Tesoro will seek to obtain on the sale of heating oil in Fairbanks will be affected by the price it receives on the sale of its other petroleum products. A decline in the price that it receives on the sale of its other products will tend to increase its desired margins on the sale of heating oil in Fairbanks. 20700284 Page 3-23 The above sections of this chapter noted several factors that could adversely affect Tesoro’s margins. First, a potential decline in jet fuel sales could lead to increased competition and lower margins. Secondly, the proposed Valdez area refinery would probably have a cost advantage relative to Tesoro and Chevron because of its additional processing capacity that will eliminate residual fuel oil production. Although this refinery is reportedly intended to serve ‘the export market, aggressive marketing by the new refinery could lead to more competition among in-state refineries for jet fuel and other petroleum product sales and a decline in refiner margins and prices. However, pricing practices more consistent with historical pricing behavior would leave the cost of imported jet fuel and marine diesel fuel from the West Coast as the price setter. Tesoro currently is making a reasonable return on its refinery investment in Alaska. 2° Consequently, Tesoro could continue to cover its product supply costs even if increased competition causes a decline in Anchorage product prices. Therefore, for this analysis, we have assumed that the gross margins, including processing costs, on Tesoro’s sales of fuel oil in Fairbanks over its crude oil cost will remain at the current level. The distribution component of the Fairbanks market is currently competitive, with more than 15 distributors selling fuel oil. As a result, distribution margins in Fairbanks were below the national average in 1987. (See the Current Supply & Demand section above.) For this analysis, gross distribution margins were assumed to remain at their current levels. The net effect of these assumptions is to provide a conservative (low) forecast of Fairbanks heating oil prices. A further reduction in these margins is unlikely even if a natural gas pipeline were constructed. The methodology used to estimate fuel oil prices for residential and commercial sales of fuel oil in Fairbanks was based on the assumption that 20 This assessment is based on ICF-Lewin’s best estimate of Tesoro’s costs, product slate, and netbacks prices. 20T00284 Page 3-24 the price would continue to be set by the Tesoro refinery. The first step was to estimate the wellhead price of ANS and Cook Inlet crude oils. The ANS wellhead price was estimated as a netback price from the U.S. West Coast because it was assumed to be the clearing market for North Slope crude oil. The Cook Inlet wellhead price was calculated based on the estimated price of ANS in Valdez adjusted for quality differences. Tesoro’s crude oil acquisition costs were then calculated as the wellhead price of these crudes plus the transportation cost to the refinery. The wholesale price of No. 2 fuel oil was then calculated as the crude oil acquisition cost plus the transportation cost from Nikiski to Fairbanks and plus Tesoro’s current gross fuel oil margin, which included its processing costs. Table 3.5 presents the resulting price projections for No. 2 fuel oil for residential and small commercial customers in Fairbanks, including the retail distribution margin. Appendix C presents the detailed components of these price projections. Price projections for large commercial customers are presented in Table 3.6. These prices are calculated as Tesoro’s wholesale price at the Fairbanks airport. 20T00284 Page 3-25 TABLE 3.5 RESIDENTIAL HEATING OIL PRICE PROJECTIONS FOR FAIRBANKS* (1987 $/GALLON) 1987 1990 1995 2000 2010 WOR. RICE SC CONSENSUS $0.85 $0.89 $1.00 $1.12 $1.36 CONSENSUS (LOW) $0.85 $0.84 $0.91 $0.98 $1512 LOW PRICE SCHOOL $0.85 $0.75 $0.79 $0.84 $0.89 * Includes estimated distribution margin. See Appendix C for the detailed components of these projections. TABLE 3.6 LARGE COMMERCIAL FUEL OIL PRICE PROJECTIONS FOR FAIRBANKS* (1987 $/GALLON) 1987 1990 1995 2000 2010 WORLD OIL PRICE SCENARIO CONSENSUS $0.68 $0.71 $0.82 $0.95 $1.18 CONSENSUS (LOW) $0.68 $0.67 $0.74 $0.81 $0.95 LOW PRICE SCHOOL $0.68 $0.57 $0.62 $0.67 $0.71 * Wholesale price at the Fairbanks airport. Large endusers in Fairbanks purchase oil directly from the refinery at roughly the wholesale price. Thus, there is no distribution margin. See Appendix C for the detailed components of these price projections. 20T00284 Page 3-26 4. NATURAL GAS DEMAND ANALYSIS In this chapter ICF-Lewin presents its analysis of the markets for natural gas and develops projections of the net-back value of gas by market segment. The chapter is divided into four sections. We begin with a brief overview of the key market segments. This is followed by an overview of demand trends. In the third section we present more detailed discussion of the individual market segments and estimate the value of gas in each segment. Finally, we combine these segments into an estimated demand curve for Cook Inlet gas. KEY MARKET SEGMENTS Natural gas from Cook Inlet is sold into five major markets involving a wide range of uses. Figure 4.1 shows the major market segments based on 1986 consumption. Table 4.1 presents the data used to generate the table. Out of a total use of 193 Bcf in 1986, about 180 Bcf were sold. About 5 Bcf were unaccounted for while the rest was used in field operations, including shrinkage and lease uses. Of the 175 Bcf sold and accounted for, the largest portion, 62 Bcf, was used to make liquefied natural gas (LNG) for the Japanese market at the Phillips-Marathon Plant on the Kenai peninsula. The next largest market was electric power generation, 44 Bcf. Most of this generation is by Chugach Electric Association and Anchorage Municipal Light and Power with about 4.4 Bef used by the military to generate power at Fort Richardson and Elmendorf Air Force Base (along with some space heating uses). About 36 Bcf was consumed in 1986 at Unocal’s ammonia/urea plant in the Nikiski area. The al Alaska Department of Natural Resources, Division of Oil & Gas, Historical and Projected Oil and Gas Consumption, January 1988, p. 18. The unaccounted for volumes are the "difference between ‘sold’ and the sum of listed ‘sold’ items." (p. 19). 20T00361 Page 4-1 FIGURE 4.1 GROSS MARKET SEGMENTS FOR COOK INLET PRODUCTION: 1986 Total Marketed Production = 175 BCF a Electric Utility Ammonia-Urea Residential/Commercial Producer fertilizer product from this plant is sold to Asia and the U.S. West Coast. About 23 Bcf was consumed in the residential and commercial market serviced by Enstar. Finally about 15 Bcf was used by producers for a variety of internal uses, e.g., field operations. This report does not analyze gas use by producers. 20T00361 Page 4-2 May not add due to independent rounding. Source: 20T00361 TABLE 4.1 Consumption of Cook Inlet Production: 1986 Field Operations Sold Vented and Flared Used on Leases Shrinkage Other Power Generation Public Military Gas Utilities 3 Residential Commercial LNG Ammonia/Urea Producers Unaccounted For Total Consumption 18. 3. 13.; i. 0. 174. 44. 39%, 4. 23. Ai. Li. 61. i 14. C3. 192. Alaska Department of Natural Resources, Historical and Projected Oil and Gas Consumption (January 1988); p. 18, Table 3.4, Railbelt Totals. ONMr Ff WONWOWONUNNDA ve} Division of OSI & Gas, Page 4-3 OVERVIEW OF DEMAND TRENDS Figure 4.2 presents historical consumption by the four major market sectors: LNG production for export, electric generation, residential/commercial use and ammonia/urea production for export. Gas sales to electric generation and the residential/commercial markets have generally tracked population growth in the Railbelt as well as economic development. Sales for LNG are tied to the contract terms between Phillips- Marathon and the Japanese. Ammonia/urea production has generally followed market conditions. Figure 4.3 presents projected gas consumption by market sector prepared by the Alaska Department of National Resources.2 Estimates of 2002-2010 consumption for Cook Inlet (domestic) users are based on these trends. Growth in consumption is projected to be modest in _ the residential/commercial sectors. Gas use in the electric utility sector should decline in the early 1990’s as the Bradley Lake Hydro project displaces some gas-fired generation. Thereafter, electric utility consumption will slowly rise to 1990 levels. The outlook for ammonia/urea is projected to be constant at levels matching peak historical production, which are above 1986 levels. ICF-Lewin review of the fertilizer market outlook confirms that the plant utilization should remain steady. Assuming that the current LNG export license is renewed as requested, demand for gas for LNG, as projected by Phillips-Marathon, should grow slightly.? 2 Ibid., p. 24. The LNG export estimates are based on projected volumes by Phillips-Marathon in their current filing before the Economic Regulatory Administration for an export license renewal. 3 See Phillips-Marathon, Application to Amend Authorization to Export Liquefied Natural Gas, ERA Docket No. 88-22-LNG, 4-7-88. 20T00361 Page 4-4 FIGURE 4.2 HISTORICAL RAILBELT GAS CONSUMPTION BY SECTOR Ammonia-Urea 200 7] ine L Electric Utility on nal Residential/Commercial 160 < < ~ = 100 o 3 a 60 1971 1973 1975 1977 1979 1981 1983 1985 Note: Exclusive of producer usage and unaccounted for volumes. 20T00361 Page 4-5 FIGURE 4.3 PROJECTED RAILBELT DEMAND FOR NATURAL GAS BY SECTOR ascunl= Ammonia-Urea | LNa Electric Utility 200 = || Residential/Commercial 160 -—- : Eyed BCF/YEAR Oe 1990 1996 2000 2006 2010 Note: Exclusive of producer usage and unaccounted for volumes. MARKET OUTLOOK BY SEGMENT -- DEMAND AND VALUE In this part of the chapter, each of the market segments is examined and a net-back wellhead value calculated based on the value of gas in that market. RESIDENTIAL AND COMMERCIAL DEMAND In 1986, the residential and commercial market was about 23.2 Bcf, with slightly over half of the market (11.9 Bcf) in the residential sector. The residential/commercial market uses gas principally for space heating, cooking, and water heating. This market is served by the Enstar Natural Gas 20T00361 Page 4-6 Company (a subsidiary of Seagull Energy Corporation since 1985) which supplies gas to the Matanuska-Susitna Valley, Anchorage, and Kenai Peninsula areas around Cook Inlet. As Figure 4.3 shows, growth in the residential/commercial sectors is projected by ADNR report to be modest -- rising from 23 Bcf in 1986 to about 28 Bef in 2010. To estimate the value of natural gas in this market, we examined the next best alternative -- distillate fuel oil. Table 4.2 presents our projections of fuel oil prices in the residential commercial market for South Central Alaska. These fuel oil prices were calculated as the refinery-gate price in Nikiski plus $0.15/MMBtu ($0.02/gallon) for transportation from the refinery to the distributor plus $1.80/MMBtu ($0.25/gallon) for the retail gross margin.* TABLE 4.2 FUEL OIL PRICE PROJECTIONS FOR SOUTH CENTRAL ALASKA Residential/Commercial Market (1987 Dollars/MMBtu)? 1990 1995 2000 2010 Consensus $6.43 $7.19 $8.08 $9.81 Consensus Low $6.06 $6:54 $7.07 $8.08 Low Price School $5.41 $5.71 $6.06 $6.42 To develop an estimate of the natural gas wellhead value from these prices, we used Enstar data on the variable non-gas costs of bringing gas from the wellhead to the burnertip -- that is, all gathering, transmission, distribution and "other" operating and maintenance expenses. ("Other" 4 See Chapter 3 for a discussion of the estimate of the Nikiski refinery gate price. These prices were converted from cents per gallon to dollars per MMBtu using conversion factors of 42 gallons per barrel and 5.825 MMBtu per barrel. 20T00361 Page 4-7 includes customer accounts, service and information, sales, and administrative and general expenses.) The fixed costs are not included because Enstar has substantial excess capacity and because these were considered sunk costs. The sources were Enstar’s 1984 cost-of-service study and the company’s 1986 annual filing to the Alaska Public Utility Commission. Table 4.3 summarizes and compares these costs and presents the cost per Mcf. TABLE 4.3 VARIABLE COSTS OF TRANSPORTING GAS FROM WELLHEAD TO BURNERTIP (Nominal Dollars/MMBtu) 1986 1985 1984 Production/Gathering $.0054 $.0063 $.0053 Transmission .0225 .0189 0234 Distribution . 1006 0929 . 1047 Other . 2843 -2365 2415 $/Mcf£ 41 35 .37 ICF/Lewin estimated 1987 variable costs at $0.40 per MMBtu. This was subtracted from the fuel oil costs in Table 4.2 to yield the following estimates of a net-back wellhead value of gas for this sector of the market. TABLE 4.4 ESTIMATES OF NATURAL GAS WELLHEAD VALUE FOR THE RESIDENTIAL/COMMERCIAL MARKET DEMAND (1987 Dollars/MMBtu) 1990 1995 2000 2005 2010 Consensus $6.02 $6.79 $7.68 $8.50 $9.41 Consensus (Low) $5.66 $6.14 $6.67 $7.16 $7.68 Low Price School $5.01 $5.32 $5.66 $5.84 $6.02 Current gas commodity rates on the Enstar System range from $3.31/Mc£® for residential customers to $2.63/Mcf for large commercial customers. 6 One Mcf equals about 1 MMBtu. 20T00361 : Page 4-8 ELECTRIC UTILITY DEMAND As noted, the electric utility demand projections are from the ADNR publication. The drop in the early 1990s as seen in Figure 4.3 results from the Bradley Lake hydropower facility. When Bradley Lake comes on line, it will back out some gas-fired generation. For this analysis, the value of natural gas in the electric utility sector is estimated to equal the cost of the next best alternative, distillate fuel oil. As new or replacement generating capacity is required in the future, coal-fired generation potentially could compete to serve future load. The extent to which this is possible is unclear. While at higher oil prices, oil/gas may be more expensive than coal, the economic choice must take into account capital investment as well. For existing gas- fired generation facilities the costs of retrofit to burn coal would be significant. However, the project did not conduct a detailed analysis of the potential for coal since this was outside the scope of the study. The projected costs of distillate to the electric utility sector are shown in Table 4.5. These prices differ substantially from those in Table 4.2 because the distribution margin of $1.80/MMBtu (distributor to end user) was not included, under the assumption that electric generators would purchase oil directly from the refineries. Similarly, for estimating a gas net-back value from Table 4.4, only the variable costs of transmission ($0.03/MMBtu) on Enstar were used, not the full transmission and distribution margin used for the residential/commercial sector in Table 4.3. The resulting net-back values for the three oil price scenarios are presented in Table 4.6. 20T00361 Page 4-9 TABLE :4.5 FUEL OIL PRICE PROJECTIONS FOR SOUTH CENTRAL ALASKA ELECTRIC UTILITY MARKET (1987 Dollars/MMBtu) 1990 1995 2000 2010 Consensus 4.62 5338) 6.28 8.01 Consensus Low 4.26 4.74 5.27 6.28 Low Price School 3/61 3392) 4.26 4.62 TABLE 4.6 ESTIMATES OF NATURAL GAS WELLHEAD VALUE FOR THE ELECTRIC UTILITY MARKET DEMAND - (1987 Dollars/MMBtu) 1990 1995 2000 2005 2010 Consensus 4.59 Sao 6.25 7.06 7.98 Consensus Low 4.23 4.71 5.24 Saue 6.25 Low Price School 3.58 3.89 4.23 4.41 4.59 Current Enstar commodity rates for the electric utilities it serves (Chugach and Anchorage Municipal Light & Power) are $2.00-$2.13 per Mcf. LNG EXPORT A major issue concerning the Cook Inlet gas marketplace is the extent to which supplies and prices are affected by the world LNG market. In this section, we explore the background of and outlook for gas exports from Cook Inlet as LNG. We also estimate net-back values for Cook Inlet gas from LNG market values in Japan. Japan will likely remain the clearing market for LNG trade in the Pacific Rim region during the period examined in this work. Background Since the unsuccessful attempts to establish an Alaska-California LNG trade in the 1970's, the only outlet for Alaskan gas as LNG continues to be 20T00361 Page 4-10 the Japanese market. / Phillips-Marathon commenced exports to Japan in 1969 and has continued to export at the annual rate of about 50 Bcf equivalent.® These sales in Japan are made to the Tokyo Electric Power Company and the Tokyo Gas Company under a 1967 agreement. The LNG is regasified and used for power generation and town gas production. ? Table 4.7 shows the export levels and prices since 1974. Under the export agreement, the Japanese pay a crude oil equivalent price for the LNG in Japan under a formula, which is presented later in this section. The price is calculated monthly, based on the previous month’s crude oil import prices. The Phillips-Marathon liquefaction plant, located at Nikiski Point on Kenai Peninsula, is one of the first baseload LNG plants ever constructed. Supplies for the plant come from Phillips-and Marathon-owned production both on and off shore in Cook Inlet. The Phillips volumes come from the North Cook Inlet field while the Marathon volumes come from the Kenai field. Phillips supplies about 43.6 Bcf/year, Marathon about 18.6 Bcf/year. LNG is shipped to Japan via two LNG tankers dedicated to this project. Here we refer to the Pacific Alaska/Pacific Indonesia proposal (1974) to transport Cook Inlet LNG to the California Market and the El Paso LNG proposal (1975) to transport North Slope LNG also to California and then by displacement to markets throughout the lower 48. Regasification terminal siting problems in California and economics turned against the Pac/Alaska deal; El Paso’s proposal was rejected in favor of the Alaska Natural Gas Transportation System. 51.1 MMBtu equals one metric ton of LNG; overall transport and liquefaction efficiency is about 83 percent. Town gas is a mixture of natural gas and other waste gases (coke gas for example) which have a lower Btu content. In this market gas is much like a feedstock. 20T00361 Page 4-11 TABLE 4.7 ALASKAN LNG EXPORTS TO JAPAN BY YEAR AND PRICE Price in Japan ($/MMBtu, Year Bef equiv. nominal) 1974 46 i> 1975 53 1.39 1976 49 1.69 1977 Si 201 1978 48 2.26 1979 50 2.34 1980 44 4.99 1981 5S 6.12 1982 49 5.97 1983 52 5.28 1984 52 5.20 1985 52 4.95 1986 49 3.71 Source: Japanese Ministry of Finance, Customs Statistics 1987. While the exports of Alaskan LNG to Japan have remained constant, the LNG market in Japan has grown enormously. In 1971, Alaska was Japan’s only supplier into a total market of 50 Bcf/year. By 1986, Japan was importing almost 1.5 trillion cubic feet (Tcf) equivalent. Figure 4.4 illustrates the growth of the Japanese market and identifies the market shares of the principal suppliers. It shows a diversified portfolio with Alaska accounting for 3 percent of the total in 1986 and no one country supplying more than 18 percent. Beginning in 1989, Australia will become Japan's sixth major supplier of LNG. The Japanese primarily use LNG for electric power generation (75 percent). Approximately 22 percent is used for town gas with the rest used as industrial fuels. 20T00361 Page 4-12 The strong growth in the Japanese market demonstrated by Figure 4.4 is not projected to continue.l9 In 1986, total Japanese natural gas demand was about 1.5 Tcf; by 1989/1990 the Japanese expect the total market to reach or exceed 1.7 Tcf and grow to about 2 Tcf by 2005. This implies an average growth rate of about 1.3 percent per year. By contrast, between 1974 and 1986, Japanese consumption of natural gas grew at an average rate of 20 percent per year. The additional growth in demand is expected to be in the town gas sector of the economy. Japan projects little additional demand in either the industrial or electric utility sectors. The only other current market for natural gas in the Pacific Rim is in Korea. Korea began to import LNG from Indonesia in October, 1986 for use in electric power production and town gas distribution. Imports are running at about 2 million tons/year (102 Bcf equivalent). The Koreans have announced plans to import up to about 5 million tons per year by 1996 (255 Bcf equivalent) . +1 ‘ Taiwan is constructing an LNG terminal on the southwest coast of the island. The Chinese Petroleum Corporation plans to commence purchasing 1.5 million tons of LNG (77 Bcf equivalent) annually from Indonesia beginning in 1991. The total potential market is estimated at 3.5 million tons (179 Bcf) annually. }2 Figure 4.5 summarizes the outlook for the LNG Market in the Pacific Rim countries through 2005. The Japanese market is clearly the dominant market. Growth should be modest over the 1990-2005 time frame. Absent developments 10 Japan Oil Company analyses provided to ICF, 1988. 11 Institute of Gas Technology, "Evaluation of the Feasibility of Exporting North Slope Alaska Gas as LNG," prepared for Yukon Pacific Corporation, June 1987, p. 15. ian Yukon Pacific Corp., “Application for Authorization to Export LNG from the U.S.," EPA Docket 87-68-NG (Dec. 1987) at p. 25. 20T00361 Page 4-13 256 20 MILLION MT 16 10 Source: in California, FIGURE 4.4 JAPANESE LNG IMPORTS BY COUNTRY OF ORIGIN SARAWAK INDONESIA ABU DHABI BRUNE! BENGE ALASKA 5] oe oC} 350066 x] J m2 ee “ © K K 2 SZ oe o °, SS SO °, SE 5 Xs 4 RRR xo oS a8 oO O x? xs ZZ 2 4 S Q $5 2 $5 RP 3 252 S 2 25 $0585 <5 <5 xP <> Xs SZ SZ RS 2 Q 2 <s o <> 7°, b KS > bY . be s 1971 1975 1980 SX ZS ce V7 LLL 1986 Japanese Ministry of Finance, Customs Statistics, 1987. expected. Figure 4.5 also shows the Japan Oil Company LNG productive 13) capacity of plants having sales additional LNG demand much beyond these contracts levels is not estimates of the total with Japan (including Australia) to be about 41 million tons/year (2.1 Tcf equivalent) through 2005. By 1995, the spare LNG capacity in the market may be gone, suggesting that additional production from Cook Inlet (or the North Slope) may be feasible. 13 Figure 4.6 shows Japanese demand will exceed its current Japan Oil Company, analyses provided to ICF, 1988. 20T00361 Page 4-14 3000 2600 2000 1600 1000 600 Source: FIGURE 4.5 PROJECTED LNG DEMAND BY COUNTRY Bef Equivalent/Year KOREA EEE] TAIWAN ZZ % COOOT 1990 1996 2000 Japan Oil Company, 1988. 2006 CURRENT CAPACITY OF PLANTS WITH JAPANESE CONTRACTS contracted supply after 1990 slightly and by 4.5 million tons/year (230 Bcf equivalent) by 2005. would appear to have a market. Continued exports from the Phillips-Marathon facility The current Phillips-Marathon contract and export license will expire in 1989. Phillips-Marathon has filed an application with the Economic Regulatory Administration to extend the license for an additional 15 years 20T00361 Page 4-15 60 « < w = 40 3 z ° - a z Q 2 30 = 20 10 0 Source: at slightly higher export levels. FIGURE 4.6 COMPARISON OF JAPANESE LNG DEMAND AND CONTRACTED SUPPLIES MMMM JAPANESE DEMAND XXX «CONTRACTED SUPPLY CONSTANT (227A ACTUAL CONTRACTED SUPPLY %, xO O08 XK] SN % SS ee SoS ate RKO SONS L206 <> % < Ko x Ce 2 < 4 o, a x LOO oe SS Se x $5 SS RA? S $2 <4 x vas o, 5 oS SS Se eX x 2 % <> x o, 1%, KS KO “ <xS 4 % % % £55 KO x 0% ote! 8 0% vonenenee xX 4 50 S525 050 SRS SLL NM > “4 x < $0 BL < x x KX SL 1990 x S80 x8 Xx KO x KO oe et oe x? <3 ox <> $< oe x2 es 52 x Japan Oil Company, 1988 license extension from 1984 to 1989.) This study adopts the export levels presented in the current filing, namely 50.6 trillion Btu (TBtu) to 1989; 52.0 TBtu through 1993, and from then on, transportation efficiency is about 83 percent, this assumes gas purchases in about 57.9 TBtu per year. Cook Inlet of about 65 to 69 Bcf per year. 20T00361 (In 1982, <S ; is o, < es ses x? SS x <> SS %, XS RAL oe << x x? KS? oe x 1% x? x 1% % 6% <> o, x o, 5 % ‘s x < x °, “ ~~ $3 ‘ SxS 524 x? < 1% 8 <> % x? <x cx 55 KO x? ? x? <> > % wars < 5 204 x <> K? % i <> % x KS % % X o, ese <> BOC oN $05 x? LS <S 5 a es R? eo, o, x K % % em gx ? 1% $6 1% a < SS the plant received a 5 year Since the combined liquefaction and Page 4-16 Discussions with Phillips-Marathon indicate they have no plans to increase plant capacity. We are not aware of other plans to develop an LNG plant in South Central Alaska other than the Yukon Pacific proposal to liquefy North Slope natural gas. Estimate of LNG Net-back Value To estimate the value of natural gas in Cook Inlet as LNG sold in Japan we first established the value of the LNG in Japan and then subtracted the costs of transportation and liquefaction. This was done for two cases. The short run case considers continued use of the Phillips-Marathon facilities by subtracting only the variable costs of transportation and liquefaction. The longer run case examines the prospect of investment in additional LNG facilities by incorporating both capital investment and variable costs. The value estimated using the former assumes continued exports at current or near current levels using existing plant and equipment. Adding the capital costs gives a net-back value after investment in new capacity to sell more LNG. For both cases, the value of LNG is set by the world crude price. To a considerable extent, LNG sellers are price takers, thus the fully loaded cost (new capacity) estimate yields lower net backs to the wellhead. Over the time period covered in this report, LNG prices will remain linked to oil prices in Japan. The Phillips-Marathon contract and export permit extension application contemplate the continuation of this linkage. Furthermore, recent contracts with Australian producers incorporate this price setting mechanism. Japanese LNG importers appear interested in creating a spot market for LNG, but they have been unsuccessful so far. Although it is certainly not impossible that they will succeed, the capital intensive nature of LNG projects generates intense pressure for binding purchase contracts before project financing is completed. The possibility that excess LNG capacity may force prices below the oil price equivalent seems somewhat remote given the estimate of demand in the Pacific Rim and 20T00361 Page 4-17 existing capacity shown in Figure 4.5. Commoditization of LNG in the foreseeable future -- which could conceivably generate prices below crude equivalence -- would likely require significant contract abrogation. World oil price projections are the beginning point. We estimate net- back values under each of the world oil price scenarios (Consensus, Consensus Low and Low Price School) reported in ICF-Lewin’s earlier report to the Alaska Power Authority. 14 Since these represent values in the U.S. Gulf coast (from the Persian Gulf), to estimate delivered costs in Japan required calculating the difference between transportation costs from the Persian Gulf to the U.S. and to Japan. The difference, $0.68, represents the additional cost of transporting to the U.S. rather than to Japan. Subtracting this from the world oil price projections yields a landed price in Japan (Table 4.8). To convert this to an LNG price, we used the basic pricing formula from the pricing agreement between Phillips-Marathon and Tokyo Electric and Tokyo Gas (per amendment dated May 8, 1987) as follows: LNG price/MMBtu = $5.92 [Average Cost of Crude in $/bbl] + $0.17 34.48 To estimate a wellhead price for natural gas we drew upon earlier work performed by 1cFL5 in which transportation and liquefaction costs were estimated for a new LNG plant. These assumed a plant efficiency of 92 14 ICF-Lewin, Qutlook for World Oil Prices. 15 ICF Incorporated, Alaska Natural Gas Development: An Economic Assessment of Marine Systems, for the U.S. Department of Transportation, Maritime Administration, Sept. 1982. 20T00361 Page 4-18 TABLE 4.8 LANDED CRUDE PRICES IN JAPAN (1987 Dollars/Bb1) 1990 2000 2010 Consensus 19.32 29.32 39.32 Consensus Low 17.32 23.32 29.32 Low Price School 13:.32 17.32 19.32 percent for liquefaction and $1.18/MMBtu for fixed costs. 16 Transportation to Japan assumed 2 percent in losses (including fuel) and $0.58 per MMBtu for other costs. Table 4.9 presents the results of these calculations, and represents the value of natural gas as LNG with capital costs and variable costs subtracted to yield a net back value to the plant-gate. From this we also subtracted $0.03 for production/gathering and transmission costs. TABLE 4.9 ESTIMATED NATURAL GAS WELLHEAD VALUE FOR LONG RUN (NEW FACILITY) LNG DEMAND (1987 Dollars/MMBtu) 1990 1995 2000 2005 2010 Consensus 1.52 2 22 3.07 3479 4.62 Consensus Low 1.21 1.64 2.14 2.58 33307 Low Price School .59 .88 1.21 1.36 1552 In Table 4.10, the liquefaction costs were reduced to the variable cost components (efficiency factors) alone. Transport costs include the costs of additional ships to account for the expanded exports from the existing 16 These estimates assume a much larger facility than that now in operation -- over 1 Bcf/day output -- and reflect economies of scale. A smaller plant would likely have a higher per unit of output cost, thus, this estimate is an optimistic one. 20T00361 Page 4-19 TABLE 4.10 ESTIMATED NATURAL GAS WELLHEAD VALUE FOR SHORT RUN (EXISTING FACILITY) LNG DEMAND (1987 Dollars/MMBtu) Consensus 2.54 3.16 3.93 4.58 5.33) Consensus Low 2.26 2.65 3.09 3.49 3.93 Low Price School 1.70 1.96 2.26 2.39 254 plant. Also, the efficiency factor for liquefaction was adjusted to reflect the historic efficiency of the Nikiski plant. 17 It would appear that the LNG market is capable of sustaining the current exports from Cook Inlet for the foreseeable future. There may be opportunities to expand production in the future as well. AMMONIA/UREA EXPORT The Collier Carbon & Chemical Corp., which is a subsidiary of Unocal Chemicals Division, owns a large ammonia/urea plant in the Nikiski area. The plant has a combined peak productive capacity for both ammonia and urea of about 3400 tons per day. The plant was originally built to serve U.S. West Coast requirements, but also exports substantial quantities to other Pacific Rim destinations. 17 Phillips Marathon sells about 50 Bcf/year to Japan but uses about 60 Bcf from Cook Inlet Reserves (See Historical and Projected Oil and Gas Consumption 1987, Alaska Dept. of Natural Resources.) The overall implied efficiency (transportation and liquefaction) is about 83. percent. See also, National Economic Research Associates, Inc., An Economic Analysis of the Proposed Extension of the Phillips-Marathon LNG contract with Tokyo Gas and Tokyo Electric, May 1982, p. 15. 20T00361 Page 4-20 Use of natural gas at the Collier plant peaked in 1982 at 55.2 BcF, 18 (See Figure 4-2.) However, gas use declined to 35.7 BCF in 1986 and 40.4 BCf in 1987. The decline in gas use for ammonia/urea production in 1986 was due in part to a decline in world-wide demand and increased competition from low cost suppliers. In 1987, the Collier plant was shutdown for a time to correct corrosion problems and to improve the catalysts used in production. Having finished plant repairs, Collier is planning to operate the plant at full capacity in 1988 (near-historic peak levels) .19 Collier obtains the gas used at its ammonia/urea plant from its parent company, Unocal, from Kenai and Beaver Creek gas fields. The original contract with Unocal was signed in 1968 with a total commitment of 450 BCF. The term of this original contract was 20 years. A second contract with Unocal was signed in 1977 and increased reserve commitments to 585 BCF. Because the reserves remaining under these contracts are declining, Unocal recently entered into a new contract for additional reserves. Unocal reports that it viewed purchase prices much above $1.00 per mcf to be uneconomic. 2° Fertilizer Market Outlook U.S. production of ammonia/urea was depressed in 1986 and 1987. Domes- tically, consumption of nitrogen fertilizer?! declined as substantial farm acreage was taken out of production. In addition, the domestic producers’ share of the U.S. nitrogen fertilizer market has declined as foreign produ- cers, particularly the Soviet Union, have delivered ammonia/urea for less than U.S. domestic production costs. Similar developments have occurred 18 ADNR, "Historical and Projected Oil and Gas Consumption," January, 1988. p. 18. 19 Conversation with W. Nellis, Regional Gas Manager, Unocal, Anchorage, February, 1988. 20 Ibid. 21 Nitrogen fertilizer is the principal application of ammonia/urea. 20T00361 Page 4-21 world-wide in recent years, with the size of the total market shrinking and U.S. producers losing market share. 22 As a result, current ammonia/urea prices are substantially below full cost levels. However, both the World Bank and the United Nations project renewed growth in ammonia/urea demand and prices by the early 1990's.23 Capacity utilization at existing plants is projected to rise to peak levels as demand increases. Additional capacity will likely be brought on-line somewhere in the world. In order to attract new investment, ammonia/urea prices would have to rise from their current variable cost levels to near full replacement cost. Estimates of Fertilizer Net-back Value Based on the U.S. and World Bank analysis, ICF-Lewin has developed two separate projections of natural gas netback value related to Alaskan ammonia/urea production and export. The first set are variable cost netback prices for the existing ammonia/urea plant. The second set correspond to full cost netback values for a hypothetical new ammonia plant to be built in South Central Alaska. As noted above, the World Bank assumed that fertilizer prices would have to approach full cost levels in the early 1990’s in order to attract new investment. For this analysis, we have assumed that ammonia prices will rise from the current level of about 50 percent of full cost to 75 percent of full cost by 1990 and to 90 percent of full cost by 2000.24 22 U.S. Department of Commerce, U.S. Industrial Outlook, January, 1987, p. 13-2. 23 William F. Sheldwick, World Bank, World Nitrogen Survey, April, 1987, p. 74-75. Fertilizer Group, Food and Agriculture Organization, United Nations, as cited in U.S. Department of Commerce, op. cit. 24 ‘This subjective assessment is based on the observation that there appears to be a tendency in world fertilizer markets to overbuild capacity whenever prices begin to firm. This indicates that full cost recovery for marginal production facilities may not be achievable. 20T00361 Page 4-22 The starting point for the net-back calculations was the value of ammonia on the West Coast as determined by the full cost of production on the West Coast. These estimates were taken from the World Bank's World Nitrogen Survey .25 Next, South Central Alaska ammonia FOB prices were calculated as the West Coast value less transportation costs. 26 Variable cost netback prices were then calculated as the FOB price less non-gas variable costs. Non-gas variable costs were estimated to be $32.00 per ton based on average U.S. production costs in 1985 as reported by the World Bank.27 Full cost net-back prices were calculated as the FOB price less the full unit cost (excluding feedstocks) of a hypothetical new plant built in South Central Alaska. Based on prior work done by ICF, the full cost of ammonia from a new plant in South Central Alaska was estimated to be $129 dollars per ton. 28 Table 4.11 presents the resulting estimates of net-back wellhead values related to the existing ammonia/urea plant and for a hypothetical new plant in South Central Alaska. The variable cost netback values in the year 2000 range from a high of $4.78 per million Btu in the Consensus scenario to a low of $2.86 per million Btu in the Low Price School scenario. The full cost netback prices in the year 2000 range from a high of $1.75 per million Btu in the Consensus scenario to a less than zero value in the Low Price School scenario. 25 World Bank, op. cit., p. 73-81. The World Bank’s ammonia prices used in the netback calculation were adjusted to be consistent with the oil price outlooks used in this analysis. 26 ammonia transportation costs were taken from: ICF Inc., Alaska Natural Gas Development: An Economic Assessment of Marine Systems, p. 49. 27 World Bank, op. cit., p. 42. 28 ICF, Alaska Natural Gas Development, p. E-31, escalated to 1987 dollars using the GNP deflator. 20T00361 Page 4-23 TABLE 4.11 ESTIMATED NATURAL GAS WELLHEAD VALUE FOR EXISTING AMMONIA/UREA FACILITY DEMAND (1987 Dollars/MMBTU) 19900 s«1995 ss 20000—ti2005 2010 Consensus $2.31 $3\,32 $4.78 $5.52 $6.37 Consensus (Low) $2.06 $2.81 $3.82 $4.27 $4.78 Low-Price School $1.54 $2.10 $2.86 $302 $3.18 METHANOL FUEL MARKET ASSESSMENT Another potential market for Cook Inlet natural gas is as a feedstock to methanol fuel in vehicles. Since the 1970s, there has been sporadic interest in methanol-fueled vehicles both for environmental and fuel security reasons. Both Atlantic Richfield and SOHIO have identified methanol as an option for North Slope gas production. In this section of the report, we examine the outlook for and value of methanol in this market and calculate netback values. Background2? Methanol is an important chemical in the U.S. economy. Until recently, most methanol consumption was oriented towards methanol as a feedstock for producing other products, especially formaldehyde, acetic acid, polyesters and acrylic plastics. 29 This discussion is based on the following sources: ICF, Alaska Natural Gas evelopment: Econom ssessment of Marine Systems, Sept. 1982; JFA, "Methanol Prices During Transition," prepared for EPA, August 1987, and ChemSystems, "A Briefing Paper on Methanol Supply/Demand for the United States and the Impact of the Use of Methanol as a Transportation Fuel," prepared for the American Gas Association, Sept. 1987. 20T00361 Page 4-24 A potentially major change in demand for methanol relates to use as a vehicular fuel. So far, methanol consumption as a vehicular fuel has been paced by consumption of methyl tertiary butyl ether (MTBE). MTBE is a methanol derivative now used as an octane boosting additive in gasoline. MTBE use accounted for about 20 percent of the 1.3 billion gallons of methanol consumed in the United States in 1986. Methanol may also be used as a neat fuel (i.e., 100 percent methanol) or as a gasoline blend. M85, a blend that is primarily methanol (85 percent with 15 percent gasoline), has recently gained attention as an alternative fuel option. Today, the pressures for a new vehicular fuel such as methanol (or compressed natural gas) come from environmental concerns more than from price. Important Clean Air Act deadlines, approaching rapidly, require aggressive pursuit of additional pollution control measures. Methanol (as well as compressed natural gas and ethanol) can significantly reduce vehicular ozone formation because the emissions are far less reactive photochemically than are gasoline emissions.29 Estimated ozone reductions vary from 20 to 50 percent. 31 The environmental pressures for alternative vehicular fuels is greatest in California. California is under considerable pressure to improve the air quality in those districts not meeting federal air quality standards.?2 The California Energy Commission has launched a program to promote methanol use in vehicles to meet air quality standards and to diversify the state’s dependence on gasoline and diesel fuels. One important market niche in California which has been targeted for alternative fuels is bus fuel. The Southern California Rapid Transit District recently purchased 30 methanol buses, and there is a 500 bus fleet goal statewide. The California Energy 30 Although ozone depletion is a worldwide problem, ozone concentration in urban areas is also considered an environmental problem. 31 Richard D. Wilson, "Alternative Fuels: Their Prospects for Fighting Smog," EPA Journal, October 1987, pp. 18, 19. 32 See California Energy Commission, Fuels Report, Dec. 1987. 20T00361 Page 4-25 Commission's methanol fleet demonstration program currently requires an estimated 300,000 gallons per year. By 1990, requirements may approach 900,000 gallons annually .33 The California methanol fuel market appears to be of greatest interest for Cook Inlet. First, California would be the closest major market. Moreover, California’s position at the forefront of the alternative fuel markets makes it likely to emerge as a major market before other areas. Methanol Market and California Price Outlook This analysis estimates the potential value of Cook Inlet gas used to produce methanol fuel by relating the value to methanol fuel prices in California. The analysis differs from the previous analyses in this chapter in that the methanol fuel price projections in California are based upon studies published in late 1987 by the CEC and based on market penetration studies. Thus, we have not developed alternative price projections for the Consensus, Consensus Low and Low Price School cases. Any adjustments which ICF-Lewin considered necessary are described where relevant. The general approach remains to estimate the value of natural gas as feedstock to the California methanol fuel market by netting out the costs and shrinkage of each step between the projected market price and the input to a methanol fuel plant in Cook Inlet. If methanol succeeds as a vehicular fuel, two important market phases are generally anticipated. In the first phase, methanol supply exceeds demand. In the second phase, demand growth will lead to a demand-supply balance for methanol, and supply would expand as necessary. The shift from the first to the second phase has important price (and value) implications. 33 Ibid., p. 33. 20T00361 Page 4-26 Through the mid-1990s, worldwide methanol capacity is likely to exceed demand. 34 When excess supply combines with the generally competitive conditions operable in the world methanol market, prices generally decline towards variable cost of production. These lower prices can persist as long as supply exceeds demand because producers would rather sell their product than shut down if they can cover the costs of production and perhaps gain some contribution to fixed costs. As demand grows to the supply level, the price must rise high enough to encourage suppliers to build new capacity. This new price level, therefore, would cover fixed as well as variable costs. This price, however, cannot exceed gasoline-equivalent pricing if methanol fuel must compete with conventional vehicular fuel. Of course, price rises could also stimulate over-construction, such as has chronically happened in the oil tanker market. Here, we adopt the assumption that prices cover fully loaded costs. This assumption is conservatively optimistic with regard to gas value in Cook Inlet. The primary source for the landed price in California is a December 1987 report by California Energy Commission staff, which they presented at the World Methanol Conference. >> Koyama, et al. present two price requirements that they judge necessary for methanol to penetrate the California vehicular fuel market: Year Methanol Price (1987 $) 1993 35 cents/gallon 1996-97 42 cents/gallon 34 available projections reviewed by ICF-Lewin may understate the duration of the period during which methanol suppliers may compete at variable costs. Demand figures appear optimistic, based on high oil prices (e.g., $27 per barrel in 1990) or simple growth rate assumptions, and capacity figures through 1990 appear to be based on scheduled construction and existing capacity. 35 Koyama, et al., op. cit. 20T00361 Page 4-27 These projections incorporate CEC’s assessment of both methanol production costs and petroleum price expectations. Basically, the figures presented above represent methanol prices that the CEC staff have judged as (1) attainable on the basis of methanol production and shipping costs (variable production cost in short term, full production cost in longer term); and (2) price competitive with unleaded gasoline. These prices are low because in the supply surplus market circumstances anticipated by the CEC, methanol suppliers would recover variable costs and not fixed costs. The price figures above are used to generate the following projections: Year Landed Methanol Price (California (cents/gal) ($/million Btu) 1990 35 5.83 1995 38.5 6.42 2000 42 7.00 These projections assume that until demand starts to balance with world supply, the variable cost of production plus shipping for the marginal supplies of methanol to California would remain constant in real dollars and would set the price for methanol. By 2000, a fully loaded cost basis for methanol fuel would be in effect. Between 1993 and 1997 the price would climb from a variable to full cost basis. Two major cost items must be netted-out of the landed price in California to determine the value in Cook Inlet of natural gas as a methanol fuel feedstock: e Shipping cost (Inlet to California) ° Methanol production cost CEC staff recently estimated shipping costs from Canada to Southern California at 5 cents per gallon ($0.84 per million Btu) . 36 This estimate 36 Koyama, et al. 20T00361 Page 4-28 was based on a 6000 DWT ship dedicated to bringing methanol from Canada to Southern California. The transport distance from Cook Inlet to Southern California is about double the Canada-Southern California distance. Assuming the shipping cost are primarily fuel and time costs, this approach indicates shipping costs would be about 10 cents per gallon ($1.67 per million Btu) .37 The cost components (excluding shrinkage) for methanol production are based upon a 1987 study for the EPA. 38 Capital cost $6.25 per million Btu O&M cost $1.57 per million Btu Shrinkage 30 percent The capital cost is adjusted by a 1.6 multiplier (versus standard construction in the U.S. Gulf Coast and Europe) . 39 The capital cost multiplier for South Central Alaska has been estimated as high as 2.1.40 These costs show that Cook Inlet gas cannot be used to make methanol fuel to compete in the California market. Even after methanol prices rise to fully-loaded costs, Cook Inlet natural gas would have no value as a feedstock for methanol fuel. 37 In 1982, ICF estimated the cost of shipping methanol from Valdez to California on Jones Act 120,000 DWT vessels (i.e., a much greater scale of operation) at $0.38 per million Btu (Jan. 1982 dollars). ICF, op. cit., p. F-1l. 38 The efficiency is not reported in op. cit. The 70 percent efficiency is based on ICF’s 1982 work. JFA, op. cit., p. 75ff 39 Western LNG Associates, January 1982; cited in ICF, Alaska Natural Gas Development, p. E-7. 40 Dow-Shell Group, Report to the State of Alaska; Feasibility of a Petrochemical Industry, Vol. 2, September 1981, p. 39. 20T00361 Page 4-29 Thus, over the time frame covered in this report, methanol would not be a factor in the gas supply/demand balance for Cook Inlet reserves. The markets for compressed natural gas (CNG) was not examined in this analysis, 41 but have been examined previously. Reviewing that report, however, does not suggest this market is any more viable than methanol for Alaskan supply. SUMMARY OF DEMAND ANALYSIS Table 4.12 presents the estimates of the natural gas net-back values in Cook Inlet based on the foregoing analysis. The volumes of gas shown for the residential/commercial, electric utility, and existing ammonia/urea sectors are those published by ADNR. The other quantities are based on ICF- Lewin estimates. (No methanol estimates are included since the netback values are negative.) Figure 4.7 illustrates the price-quantity relationships of the market sectors under the Consensus Low case for the year 2000. The sectors are arrayed in descending order, left to right, of their gas values as developed in this chapter. Figure 4.8 presents the same data in the form of demand curves for each of the oil price scenarios also in the year 2000. (A demand curve exists for each of the years projected.) These curves approximate demand curves for Cook Inlet gas. As is readily apparent, they differ from the usual theoretical demand curve in that the quantities demanded are the same at each of the price (or value) levels presented. For this analysis, only one set of demand projections was available, which we believe represents ADNR’s projections at price relationships not apparently different from current ones. The demand presented is therefore probably less elastic than may actually be the case. Nevertheless, in several key respects it is a good approximation at this point in time. 41 ICF, Alaska Natural Gas Development, Appendix B. 20T00361 Page 4-30 20T00361 TABLE 4.12 SUMMARY OF VALUES AND QUANTITIES FOR POTENTIAL GAS DEMAND BY SECTOR AND OIL PRICE FORECAST RESIDENT IAL/COMMERCIAL Quantity (Bcf/yr) Consensus Value ($/MMBtu) Consensus Low Value ($/MMBtu) Low Price School Value ($/MMBtu) ELECTRIC UTILITIES Quantity (Bcf/yr) Consensus Value ($/MMBtu) Consensus Low Value ($/MMBtu) Low Price School Value ($/MMBtu) LNG NEW PLANT Quantity (Bcf/yr) Consensus Value ($/MMBt!!) Consensus Low Value ($/MMBtu) Low Price School Value ($/MMBtu) LNG EXISTING PLANT Quantity (Bef/yr) Consensus Value ($/MMBtu) Consensus Low Value ($/MMBtu) Low Price School Value ($/MMBtu) NEW FERTILIZER PLANT Quantity (Bef/yr) Consensus Value ($/MMBtu) Consensus Low Value ($/MMBtu) Low Price School Value ($/MMBtu) EXISTING FERTILIZER PLANT Quantity (Bef/yr) Consensus Value ($/MMBtu) Consensus Low Value ($/MMBtu) Low Price School Value ($/MMBtu) METHANOL Quantity (Bcf/yr) Consensus Value ($/MMBtu) Consensus Low Value ($/MMBtu) Low Price School Value ($/MMBtu) 25.9 6.02 5.66 5.01 43.4 4.59 4.3 3.58 0.0 1.5: 1.21 0.59 63.5 2.54 2.26 1.70 0.0 0.00 0.00 0.00 50.0 2.28 2.03 1.51 0.0 0.00 0.00 0.00 26.5 6.799 6.14 5.32 40.4 5.35 4.71 3.89 50.0 2.22 1.64 0.88 69.0 3.16 2.65 1.96 0.0 0.00 0.00 0.00 50.0 3.29 2.78 2.07 0.0 0.00 0.00 0.00 26.8 7.68 6.67 5.66 43.3 6.25 5.26 4.3 175.0 3.07 2.16 1.21 69.0 3.93 3.09 2.26 50.0 1.72 0.75 0.00 50.0 4.75 3.79 2.83 0.0 0.00 0.00 0.00 45.3 7.06 5.72 4.41 325.0 3.79 2.58 1.36 69.0 4.58 3.49 2.39 50.0 2.39 1.14 0.00 50.0 5.49- 4.26 2.99 0.0 0.00 0.00 0.00 47.4 7.98 6.25 4.59 325.0 4.62 3.07 1.52 69.0 5.32 3.93 2.54 50.0 3.31 1.72 0.12 50.0 6.34 4.75 3.15 0.0 0.00 0.00 0.00 Page 4-31 FIGURE 4.7 CONSENSUS (LOW) DEMAND ESTIMATES FOR COOK INLET NATURAL GAS 2000 10 || Residential/Commerciai 6 VA Electric Utilities | existing Fertuizer Plant \ LNG Existing Plant LNG New Plant KX New Fertilizer Plant $/MMBtu o 100 200 300 400 600 Bot/year First, demand for the LNG export market is fixed within the price ranges identified. That is, LNG will be exported as long as the costs of production are covered; it cannot be increased if prices rise. Residential/commercial demand is also relatively stable. One would expect that at substantially higher prices, demand would decline (other things being equal), but higher prices would probably mean a healthier Alaska economy, which might offset such a decline. Electric utility demand would probably exhibit greater price elasticity, but this analysis is being done 42 elsewhere and was not available for this study. Finally the ammonia/urea production is probably sensitive to higher gas prices. In the next chapter, we integrate these demand curves with supply curves developed for Cook Inlet reserves. The resulting analysis will identify where the gas market clears for Cook Inlet production and whether this clearing market is linked to world energy markets. This market clearing analysis provides the basis for the natural gas price outlook. 20T00361 Page 4-32 FIGURE 4.8 OVERVIEW OF COOK INLET NATURAL GAS DEMAND OUTLOOKS CONSENSUS 10 8 = a 6 5 XS 4 A 2 - oe aes 0 100 200 300 400 600 Bef/year CONSENSUS LOW 10 8 2 6 a =z 4 ~ Af 2 0 0 100 200 300 400 600 Bef/year LOW PRICE SCHOOL 10 8 s 6 3 Ss 4 ~S a 2 9 be an oe L 0 100 200 300 400 600 Bcf/year 20T00361 Page 4-33 development would follow a least cost path; that is, the next well drilled will always be the next cheapest well. In the real world this is often not the case because developers do not really know what they will find next. Nevertheless, the assumption of least cost development does reflect the strategy that would be followed in trying to add reserves. The analysis takes into account the location of fields in onshore and offshore areas. A complete economic analysis was conducted, recognizing all significant cost items, the productivity of wells and realistic financial and tax assumptions. Exploration of drilling was evaluated on a full cost basis for each field. Developmental drilling was based on the economics of individual wells, excluding exploration risk and lease bonus. The marginal cost was defined as that product price which yields a zero discounted present value of the after-tax cash flows. Using a 10% (nominal) discount rate, equivalent to 6% without inflation effects, the marginally economic resource in the Cook Inlet area is shown in Table 5.2. TABLE 5.2 GAS PRODUCTION POSSIBILITIES IN COOK INLET BY PRICE INTERVAL Marginal Cost of Production, Wellhead Cost Developmental New Field Speculative ($_ 1987 /Mcf) (Bcf) (Bcf) (Bcf) $2.00 800 650 1500 $2.00 - $3.00 50 200 250 $3.00 - $5.00 55 300 400 Total < $5.00 905 1150 2150 Total Resource 1070 2100 3400 20F0148 . Page 5-5 GAS PRODUCTION BCF/YR FIGURE 5.2 SCHEMATIC DIAGRAM OF METHOD USED TO INTEGRATE GAS SUPPLY WITH DEMAND <= GAS DEMAND } FORECAST DELIVERY CAPACITY REQUIRED FROM NEW DISCOVERIES <s— ADDITION AND DEVELOPMENT OF PROBABLE RESOURCE ~<e— DEPLETION OF PROVED RESERVES YEARS 20F0148 Page 5-6 f These data are to be interpreted as "if discovered fields were to be found in decreasing order of size (largest first) then a total of 4205 Bcf of Nonassociated Gas would be economically viable at wellhead prices less than $5.00/Mcf." Figure 5.3 presents the outlook in terms of a supply curve. This well ordered, least cost to highest cost, pattern is unlikely to track with actual exploration and production experience. Nevertheless, ICF-Lewin believes that it presents a useful first approximation of the expected supply trends .2 TRANSMISSION DIFFERENTIAL ISSUES An important concern to the Alaska Power Authority is the cost implications of different locations for natural gas-fired generation. Before moving on to the gas supply and demand integration, it is important to examine this issue. From the existing gas fields, there is a well developed network of pipelines. Gas pipelines are of two general kinds: producer-owned gathering and transmission lines and Enstar transmission and distribution (the latter are not shown in any detail). From conversations with Enstar personnel, there appears to be ample capacity in the transmission system to accomodate current demand and growth. The principal volume dependent, variable cost of actual transmission of natural gas is the cost of fuel to operate compressors to move the gas. Other variable costs of transmission are added for rate-making purposes and include operating and maintenance expenses such as_ supervision and engineering, system dispatching and control, labor, communications and so forth. These are expenses that do not by and large, vary with the level of throughput. More sophisticated supply analyses, including play-analysis techniques are feasible but costly. They were beyond the scope of this preliminary supply analysis. 20F0148 Page 5-7 Marginal Cost ($1987/MMBTU) FIGURE 5.3 POTENTIAL SUPPLY OF COOK INLET GAS 8.0 Tas) 7.0 Smoothed Regression Equation : 6.5 Cost= 0.21267 * 104(3.2314e-4 Quantity) R42 = 0.984 6.0 5.5 5.0 4.5 4.0 Aggregate of: 5 Onshore and Offshore Regions 3.5 Probable and Possible Resource Base 3.0 Zo 2.0 1.5 1.0 0.5 0.0 Associated and Nonassociated Gas 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 Incremental Quantity of Resource (BCF) 20F0148 Page 5-8 Enstar has two compressor stations -- Gudenrath and Kalinofsky -- both of which are on the southern, Kenai to Anchorage line. (The Beluga line has no compressors and has substantial excess capacity.) These stations, according to Enstar personnel, are used very little, mainly on peak days to pressurize the lines to serve the Anchorage/Kenai loads. For most of the year, gas flows under the pressure from the production fields. In 1986, these stations consumed a total of 95,559 Mcf, or 0.2 percent of total gas purchases by Enstar. Therefore, it would appear that under current conditions there is little or no variable cost difference among the Beluga, Kenai and Anchorage locales. In the future, cost differentials could arise depending on whether the location of new generating capacity requires additional compression on the Kenai to Anchorage line. Fixed costs could be incurred only where new capacity (compression or pipe) would be required either to access new fields or to accomodate new load increments. For the purposes of our analysis, the estimations of new supply curves assume that the variable costs of gas transmission are the same throughout the region. (In our net-back calculations, ICF-Lewin estimated $0.03/MMBtu as the upper bound of all gathering and transmission costs.) Thus in the next section, the integration of the supply and demand analyses, location differentials are not considered. SUPPLY/DEMAND INTEGRATION In Chapter 4, demand curves were developed based on projections published by ADNR and values developed by ICF-Lewin. Curves were developed for each of the forecast years (1990, 1995, 2000, 2005, and 2010) under each of the oil price scenarios: Consensus, Consensus Low, and Low Price Schools. There are fifteen demand curves in total. Supply curves were estimated in a parallel fashion for each year under each scenario. 20F0148 Page 5-9 As discussed in the supply summary of this chapter, the schedule of required incremental additions and hence the supply price/quantity pairs depend on the level of demand. Our assumption has been that a reserve to production ratio of 10:1 must be maintained for new reserves.2 This study assumes that incremental reserves will only be added if demand exists to support them. Thus, to satisfy 35 Bcf of incremental demand in a year, 350 Bef must be added in the year immediately prior to the year of the demand and 35 Bcf (replacement) every year thereafter. ICF-Lewin has analyzed supply/demand balances under two alternative gas market scenarios: e demand with no new gas-based exports (isolated case) e demand with new gas-based exports (new link to world market case) The scenarios are used to capture the effect of additional demand and production on future costs of production in Cook Inlet. NATURAL GAS PRICE OUTLOOK The natural gas price outlook in the Railbelt region, specifically Cook Inlet, will depend to a great extent on whether or not Cook Inlet’s gas market is isolated from other energy markets or linked to them. This linkage, in turn, will depend on world energy price levels, as influenced by world oil prices. ICF-Lewin's analysis indicates that Cook Inlet is most likely to remain an isolated market unless world oil prices proceed along a fairly high price trajectory, such as the Consensus projection. 3 Note, however, that in the recent history of Cook Inlet, R/P ratios in some fields have approached 20:1 due to the lack of markets for gas. 20F0148 Page 5-10 This outlook emerges from consideration of Cook Inlet supply and demand factors. As an isolated market, Cook Inlet has gas supplies more than adequate to meet the local requirements. ICF-Lewin defines the local requirements as including existing export-oriented facilities. This price outlook includes two elements: wellhead price outlook and price differentials. The wellhead price outlooks result from the integration of the demand and supply information presented earlier. In general, the balance analysis that would generate demand/supply balances is a process that iterates until the results converge on market clearing prices at each point in the period examined. Such a market clearing price trajectory would represent the price path for which precisely enough supply would be forthcoming to serve the effective demand for that same price path. 4 Because of the nature of the South Central Alaska gas market, sophisticated balance analysis in which price exploration would continue until balance was achieved in every period was not possible. Thus, this integrative analysis bounds the likely prices with the estimated marginal cost of supply and marginal value of demand. In an idealized, competitive market, marginal cost would equal marginal value (and equal the theoretical market clearing price). The Cook Inlet natural gas market differs from the idealized market in important ways. Perhaps most importantly, the number of potential buyers and sellers of natural gas is relatively small. Equally important is that demand is constrained below market clearing levels due to the limited size of the local market and its isolation from the broader national or world market. This "thin" market means that the actual gas price at a particular point in time, as reflected in a new or renegotiated contract, can depend on Such balance analyses necessarily incorporate assumptions about fore-knowledge (i.e., what a producer or purchaser would know about future prices) and other analytic issues. 20F0148 2 Page 5-11 expectations about future market conditions (or other aspects of "market psychology"), the relative negotiation skills of the buyer and seller, or other non-competitive factors. Current contracts incorporate the results of all these factors and reportedly are in the $1.50 per million Btu range (plus or minus 10 percent). TABLE 5.3 PRICE RANGES FOR NEW COOK INLET NATURAL GAS (Isolated Case) (1987 $/mmBtu) Oil Price Marginal Cost/ Marginal Scenario_ Year Reserve Addition Value _ Consensus 1990 -- af 2.28 1995 0.47 3.16 2000 0.86 3.93 2005 1.63 4.58 2010 Bele, 5.32 Consensus Low 1990 -- af 2.03 1995 0.47 2.65 2000 0.86 3.09 2005 1.63 3.49 2010 Sake 3.93 Low Price 1990 -- af 1.51 1995 0.47 1.96 2000 0.86 2.26 2005 1.63 2.39 2010 3.12 b/ 2.54 a/ No reserve additions required until 1991. b/ The clearing price would not exceed marginal value. Table 5.3 presents the bounds of wellhead prices for new natural gas in Cook Inlet through 2010, by oil price scenario. In each instance, the value of the lowest value local use identified exceeds the estimated marginal cost 20F0148 : Page 5-12 of production. In such an isolated market with a surplus supply, price could slip towards the marginal cost of production. For a variety of factors prices appear to be remaining well above those minimum levels. Likely explanations include the facts that actual exploration and production costs will not follow exactly the cost supply curve developed for this preliminary analysis, that the number of market transactions and participants is rather small compared to competitive conditions, and that these participants’ expectations not only include some implicit expectations about price increases but also probably include expectations about new links to world markets. The surplus supply situation led ICF-Lewin to explore the potential for new export-oriented projects that would further link the Cook Inlet gas market to Pacific Rim, and ultimately, world energy market conditions. The demand analysis presented earlier identified three major candidates for potential new export projects: LNG, fertilizer, and fuel methanol. Of these candidates, LNG presented the highest potential value to Cook Inlet gas producers. ICF-Lewin first explored the potential of Cook Inlet to serve the full projected and uncontracted-for requirements for major Pacific Rim LNG markets. The production costs of serving this large market (potentially 2250 Bef annually by 2000), however, far exceeded the potential netback value at Cook Inlet. Next, ICF-Lewin explored the possibility of adding one new LNG export facility requiring 50 Bcf annually. This new plant would be roughly the size of the existing LNG facility. Only the highest oil price scenario (Consensus) would generate netback values in Cook Inlet that could possibly support LNG exports for the roughly twenty years required for a project to be economical. In this case, for the net back to Cook Inlet in the final two years to balance with the costs of new reserves would require lower margins on LNG. production or transportation. Such discounts would be likely to maintain the export 20F0148 Page 5-13 TABLE 5.4 PRICE RANGES FOR NEW COOK INLET NATURAL GAS (New Links to World Market) a/ (1987 $/mmBtu) Oil Price Marginal Marginal Scenario Year Cost of New Reserves Value Consensus 1990 -- b/ 2.28 ¢/ 1995 0.73 2.22 2000 1.65 3.07 2005 3.54 3.79 2010 §.33¢/ 4.62 Consensus Low 1990 -- b/ 2.03 c/ 1995 0.73 1.64 2000 -- e/ 2.14 2005 -- e/ 2.56 2010 -- e/ 3.07 Low Price 1990 -- b/ 1.51.¢/ 1995 0.73 0.88 2000 -- e/ 1.21 2005 -- ef 1.36 2010 -- ef 1.32 a/ Linked by a new 50 Bcf per year LNG export facility. b/ No reserve additions required before 1991. c/ The 1990 marginal value is for a lower total quantity without new exports (which are assumed to begin in 1995). It is the netback value for the existing LNG facility. d/ Any clearing price would not exceed marginal value. e/ Adequate reserves cannot be developed to support additional exports for this price scenario. 20F0148 Page 5-14 activity. The analysis found that the new LNG export value would exceed new fertilizer export value. The balance analysis indicated that a plant of this size would be economically attractive in the Consensus oil price scenario. For the other oil price scenarios, however, such an additional LNG demand would drive Cook Inlet production costs above net back value for LNG before the plant could achieve twenty years of operation. Table 5.4 presents the results of this examination of new links to world energy markets for Cook Inlet. Thus, the gas price outlook emerges from the Consensus Low and Low Price sections of Table 5.3 and the Consensus section of Table 5.4. Tables 5.5A and 5.5B provide additional data for the comparative cases with and without a potential new export project. PRICE DIFFERENTIALS To project natural gas prices at specific locations or for specific facilities, information about likely price differentials must also be taken into account. The incremental economic costs of moving natural gas to facilities already on the Cook Inlet natural gas transmission and distribution network are minimal. ICF-Lewin interviews with parties in the Cook Inlet region indicated that in general, no major bottlenecks existed in the network and that compression to move the gas to market was required only during brief periods of peak deliveries. Other factors could lead to actual differentials in prices offered at specific potential sites for electric generation facilities. The major such factors would be regulated rates and existing gas contract prices. Regulated rates include significant contributions towards fixed costs and do not represent incremental costs in an economic’ sense. Gas supply contractsreflect past gas market conditions and there is significant renegotiation activity underway at this time. Assessing the specific 20F0148 Page 5-15 outlook for either rate changes or contract renegotiation was beyond the scope of this project. Table 5.6 provides the information specific to the three sites in the Cook Inlet of greatest interest to the Alaska Power Authority. 20F0148 Page 5-16 TABLE 5.5A SUPPLY/DEMAND DATA WITHOUT NEW EXPORT PROJECTS CONSENSUS CONSENSUS LOW LOW PRICE SCHOOL POTENTIAL POTENTIAL POTENTIAL 1987 DEMAND SUPPLY 1987 DEMAND SUPPLY 1987 DEMAND SUPPLY YEAR DOLLARS (MCF) (MCF ) DOLLARS (MCF) (MCF) DOLLARS (MCF) (MCF) 1990 2.28 183 479 2.03 183 464 1.51 183 424 2.54 133 49% 2.26 133 478 1.70 133 440 4.59 69 573 4.23 69 562 3.58 69 540 6.02 26 610 5.66 26 601 5.01 26 585 1995 3.16 186 428 2.65 186 404 1.96 186 363 3.29 117 433 2.78 117 410 2.07 117 371 5.35 67 498 4.71 67 481 3.89 67 455 6.79 27 530 6.14 27 517 5.32 27 497 2000 3.93 189 ‘377 . 3.09 189 344 2.26 189 302 4.75 120 402 3.79 120 372 2.83 120 332 6.25 70 439 5.24 70 415 4.23 70 386 7.68 27 467 6.67 27 448 5.66 CG 426 2005 4.58 192 314 3.49 192 277 2.39 192 231 5.49 123 338 4.24 123 303 2.99 123 261 7.06 73 372 5.72 73 344 4.41 3 313 8.50 27 397 7.16 er 374 5.84 27 351 2010 5.32 194 265 3.93 194 225 2.54 194 171 6.34 125 289 4.75 125 250 3.15 125 200 7.98 75 320 6.25 a 287 4.59 Te) 251 20F0148 Page 5-17 TABLE 5.5B SUPPLY/DEMAND DATA WITH POTENTIAL NEW EXPORT PROJECTS CONSENSUS CONSENSUS LOW LOW PRICE SCHOOL POTENTIAL POTENTIAL POTENTIAL 1987 DEMAND SUPPLY 1987 DEMAND SUPPLY 1987 DEMAND SUPPLY YEAR DOLLARS (MCF) (MCF ) DOLLARS (MCF) (MCF ) DOLLARS (MCF ) (MCF ) 1990 2.28 183 469 2.03 183 454 1.51 183 414 2.54 133 484 2.26 133 468 1.70 133 430 4.59 69 563 4.23 69 552 3.58 69 530 6.02 26 600 5.66 26 591 5.01 26 575 1995 2.22 236 365 1.64 236 324 0.88 236 248 3.16 186 412 2.65 186 389 1.96 186 355 3.29 117 418 2.78 17 395 2.07 17 363° 5.35 67 483 4.71 67 466 3.89 67 447 6.79 27 515 6.14 27 502 5.32 27 489 2000 1.72 414 227 0.75 414 124 0.55 414 86 3.07 364 304 2.14 364 265 1.21 364 192 3.93 189 338 3.09 189 314 2.26 189 276 4.75 120 363 3.79 120 342 2.83 120 306 6.25 70 400 5.24 70 385 4.23 70 360 7.68 27 428 6.67 27 418 5.66 27 399 2005 2.39 567 171 1.14 567 80 0.81 567 59 3.79 517 233 2.58 517 190 1.36 517 128 4.58 192 258 3.49 192 230 2.39 192 204 5.49 123 283 4.24 123 257 2.99 123 234 7.06 73 317 5.72 v6) 297 4.41 73 286 8.50 27 341 7.16 27 327 5.84 27 324 2010 3.31 569 130 1.72 569 67 1.11 569 33 4.62 519 174 3.07 519 145 1.52 519 75 5.32 194 193 3.93 194 178 2.54 194 144 6.34 125 217 4.75 125 204 3.15 125 173 7.98 75 248 6.25 5 240 4.59 75 224 9.41 28 270 7.68 28 268 6.02 28 260 20F0148 Page 5-18 Location Beluga Anchorage Municipal Bernice 20F0148 TABLE 5.6 LOCATION-SPECIFIC PRICE DATA Marginal Cost of Supply Wellhead + 3¢/MMBtu Wellhead + 3¢/MMBtu Wellhead + 3¢/MMBtu Current Price Current Pricing Basi: $0.30/Mcf Wellhead Contract, $2.00/Mcf $2.13/Mcf in renegotiation Tariff Tariff Page 5-19 6. CONCLUSION This analysis of fuel oil markets in the Alaska Railbelt region has found two markets especially important to the economics of electric generation and transmission. These markets are the fuel oil market in Fairbanks and the natural gas market in Cook Inlet. Each of these markets has important features that make it behave differently than classic, theoretical competitive markets. The Fairbanks fuel oil market is a truly unique market. A major market factor is the special arrangement to receive North Slope crude from the TAPS and to re-inject refinery bottoms into TAPS. Anchorage refinery economics play a role in the heating oil market in Fairbanks because Tesoro appears likely to remain the marginal supplier there. For oil-fired electric generation, however, the GVEA/Mapco/TAPS arrangement appears likely to maintain a low refinery margin plus North Slope crude acquisition costs. For the Cook Inlet natural gas market further direct linkage to world energy markets appears unlikely. Such linkage would develop directly only if actual or expected high world oil prices led to investment in an additional LNG facility. The Cook Inlet region appears capable of supporting as much as one additional LNG export facility of a size roughly comparable with the current facility. Development of North Slope gas through the proposed Trans Alaska Gas System might indirectly link the Cook Inlet market to world energy developments, especially if an Anchorage-- Fairbanks line were constructed. Finally, this analysis has found that there is considerable uncertainty and potential for volatility in the fuel markets relevant to the Alaska Power Authority. Careful monitoring of new developments should be an important element of the application of the price paths crafted in this report. 20T00360 Page 6-1 SECTION 3 APPENDICES APPENDIX A. APPENDIX B. APPENDIX C. APPENDIX D. 20T00360 TABLE OF CONTENTS Cook Inlet Supply Analysis: Proved Reserves Cook Inlet Supply Analysis: New Supplies Fuel Oil Price Projections for Fairbanks Comments on the Draft Report and Responses APPENDIX A Technical Appendix A Cook Inlet Gas Supply Analysis Proved Reserves I. Overview The supply analysis subtask is charged with the mission of forecasting (a) production from proved reserves; (b) production from inferred reserves; and (c) production from future discoveries. This report provides data on the productive potential of known reserves in active fields, and thus serves as a baseline for natural gas productive capacity. Briefly, the potential recovery of natural gas from active Cook Inlet fields is: 1990: 152 Bef 1995: 102 Bef 2000: 72 Bcf 2005: 52 Bef 2010: 40 Bcf These estimates include both Nonassociated and Associated-Dissolved gas and are derived from a production decline analysis of individual fields. II. Methodology Data on Oil, Associated Gas, Natural Gas Liquids and Nonassociated Gas production were taken from the annual reports of the Alaska Conservation Commission during the period 1974-1984 (the series was not published after 1984.) Annual and cumulative production data were assembled. A total of eight fields which produce Nonassocaiated gas and six fields which produce oil and associated gas were analyzed. Because the entire production history of the field is not available, and data are limited to annual and cumulative production, the field production decline coefficient was estimated on the basis of trends in the ratio of annual to cumulative production under the assumption of a constant reserves to production (R/P) ratio, as illustrated below: Production model: (1.0) PCY) = R(Y) * K where: P(Y): Production in year "Y" R(Y): Beginning of Year Reserves in year "Y" K: fraction of reserves produced (constant over field life) Equation (1.0) may be transformed to express reserves in each year as a function of time since initiation of production: (2.0) R(Y) = R(Y-1) - P(Y-1) or (2.1) R(Y) = R(Y-1) * (1-K) Substituting G=(1-K), we can establish that: (2.2) R(Y) = R(1)*G(¥-)) and, incorporating equation (1.0), the production in any year is given by: (3.0) PCY) = [R(1)*G(¥-1]*(1-G) _... annual production By similar algebraic manipulation, including the sum of the geometric series, it is possible to demonstrate that the cumulative production during the first Y years of production is: (4.0) P(1) + P(2) + .... + P(Y) = R(1)*(1.0 - GY) Combining equations (3.0) and (4.0) we arrive at the following expression for the ratio of annual to cumulative production: (5.0) AP/CP = [G¢¥-1)(1-G)]/(1-GY) By using the data values for the Annual and Cumulative production for each field over the period 1974-1984, it is possible to estimate a production decline coefficient (K= 1-G) which reflects the underlying capacity of the field. The ICF-Lewin proprietary nonlinear regression program was used to make the specific calculations. No problems were encountered and R-squared values were all above 0.90, indicating that the constant R/P model adequately describes production from Cook Inlet Fields. Figure 1 illustrates the results obtained from the fitting process. Given a production decline coefficient, the future production capacity was extrapolated, based on reserve estimates published in the 1984 Alaska Conservation Commission Report, by using equation (1.0). A=3 Annual Production Divided by Cumulative Production (x100) Figure 1 Illustration of Calculation of Production Decline Coefficient Granite Point Field: Decline Rate of 5% of Remaining Reserves/Year 1974 1976 1978 1980 1982 Prepared by: ICF-Lewin Energy Service, 1968 1984 A-4 III. Results The following table provides the estimates of the production decline coefficient for the oil and gas fields which were studied, as well as the 1984 reserves base which was used in extrapolating productive potential: 1984 Est. Decline Reserves Coefficient Field (Bcf) (Fract)* Nonassociated Gas Beaver Creek (Gas) 221 0-15 Beluga River 676 0.04 Kenai 751 0.10 Mc Arthur River (Gas) - 600 0.06 N. Cook Inlet 812 0.06 Sterling 23 0.02 Trading Bay (Gas) 29 0.15 West Fork 6 O95 Associated/Dissolved Gas Beaver Creek (Oil) 1 0.15 Granite Point 18 0.05 Mc Arthur River (Oil) 29 0.15 Middle Ground Shoal 9 0.14 Swanson River 259 0.13 Trading Bay (Oil) 2 0:15 * Decline coefficient is "K" in Equation (1.0); specific values reported here were determined by fitting Equation (5.0) to the ratios of Annual to Cumulative production over the life of each field, as reported by the Alaska Conservation Commission Figure 2 shows the projected trend in total gas production and Associated-dissolved gas production. Gas markets in the Cook Inlet Area have been depressed during the period 1984-1987, and production has not kept pace with historical potential. Accordingly, the production forecasts were adjusted to reflect proved reserves on December 31,1987 as reported in: Historical and Projected Oil and Gas Consumption, a January,1988 report to the Alaska legislature As shown in the table below, near-term production capacity estimated on the basis of 1987 data is somewhat higher than projections based on 1984 data, largely due to market-related conditions: Year 1984 Base 1988 Base 1990 120 Bcf 152 Bcf 1995 82 Bef 102 Bcf 2000 60 Bef 72 Bef 2005 45 Bcf 52 Bcf A-6 Figure 2 Forecast of Production Decline From Existing Cook Inlet Fields Based on Appraisal of Individual Fields Capacity Analysis without Consideration of Markets 200 150 = S s : 2 100 3 ¢ a 3 5 50 1984 1988 1992 1996 2000 2004 Prepared by: ICF-Lewin Energy Service, 1988 2008 APPENDIX B Technical Appendix B Cook Inlet Gas Supply Analysis New Supplies I. Overview and Summary The supply analysis provides forecasts of (a) production from proved reserves; (b) the incremental cost of new discoveries; and (c) production from future discoveries. Technical Appendix A provides forecasts of production from proved reserves. This report provides cost-supply curves for wells and fields which remain to be discovered. In turn, these curves are an input to the process of modeling future reserve additions and production, as discussed in the main body of the report. The undiscovered oil and gas resource was analyzed in three "certainty" categories: ¢ Developmental: resource which could be added through extension of existing fields (Most Certain - 1070 Bcf of gas); «New Field Exploration: resource which could be added through additional exploration in formations known to be productive (2100 Bcf of Gas, 1000 Mil Bbl of Oil); and e Speculative: resource which could be added through exploration in formations which have not yet been proven to be productive (Least Certain - 3400 Bcf of Gas). The location of fields in onshore and offshore areas was also considered. A complete economic analysis was conducted, recognizing all significant cost items, the productivity of wells and realistic financial and tax assumptions. Exploration drilling was evaluated on a full cost basis for the field. Developmental drilling was based on the economics of individual wells, excluding exploration risk and lease bonus. The marginal cost was defined as that product price which yields a zero discounted present value of the after-tax cash flows. Using a 10% (nominal) discount rate, equivalent to 6% without inflation effects, the marginally economic resource in the Cook Inlet area is estimated to be: Marginal Cost Developmental New Fleld Speculative ($1987/Mcf) (Bef) (Bef) (Bef) SSSSSSSSSSSSSeSSSSSSSSSSSSsssesssssssssssssssssssesesesesceszz $2.00 800 650 1500 $2.00-$3.00 50 200 250 $3.00-$5.00 §5 300 400 Total <$5.00 905 1150 2150 Total Resource 1070 2100 3400 These data are to be interpreted as "if undiscovered fields were to be found in decreasing order of size (largest first) then a total of 4205 Bcf of Nonassociated Gas would be economically viable at wellhead prices less than $5.00/ Mcf." Although this resource is the theoretical maximum available at the $5/Mcf price, it is considerable and could support consumption at current levels of about 140 Bef/Yr for 30 years. Although the drilling and market study is required to establish the precise details, the analysis shows that the Cook Inlet Region has abundant low-cost gas resource. Il. Methodology The study relied on published data and established theory to estimate the marginal cost of successive increments of the undiscovered Cook Inlet Oil and Gas resource base. The analysis was conducted in four phases: *Phase 1: Characterize the known fields by size and well productivity; «Phase 2: Establish the statistical distribution of the total resource (known plus estimated undiscovered) by field size; «Phase 3: Calculate the number of fields remaining to be discovered, by size, product,location and certainty; and «Phase 4: Estimate the marginal cost of developing new fields, by size. Each of the four phases is described in detail below. A. Characterize Known Fields by Size and Well Productivity Data on Oil, Associated Gas, Natural Gas Liquids and Nonassociated Gas Production, well counts andestimates of remaining proved reserves were taken from the annual reports of the B-2 Alaska Conservation Commission (ACC) during the period 1974- 1984. (the series was not published after 1984.) Field size was defined as the sum of cumulative production and remaining proved reserves. The details of this calculation are shown in Table 1; as shown, the data provided estimates of size for six oil fields and eighteen gas fields. For each field, the average well size was calculated by dividing field size by the number of wells. Field deposition theory holds that wellsize is highly correlated with field size through the net pay variable. Specifically, Arps and Mortada demonstrated a logarithmic relationship between field and well-size in several Lower-48 basins. As shown in Figures 1 and 2, the same relationship is found in both oil and gas fields in the Cook Inlet Area. Log-linear regression was used to quantify this relationship in order to estimate the productivity of wells in fields remaining to be discovered. The equations are: (1.1) WSoi = 0.4478 x [FSoi19-4314] R2=0.97 (1.2) WSgas = 0.6735 x [FSgas?-7885] R2=0.94 where: WSoi: Typical Oil Well Size WSgas:Typical gas well size FSoii. Oil Field Size (Mil BOE) FSgas: Gas Field Size (Bcf) Relationships of similar form have been developed by ICF- Lewin for Lower-48 producing regions as part of the Domestic Oil Replacement Cost system. These equations, which are statistically reliable, based on ACC data, and consistent with established theory are critical to the economic analysis phase. Specifically, large fields will also have highly productive wells, which will, in turn, be economically viable at low prices. B. Establish the Statistical Distribution of the Total Resource by Field Size Established statistical theory indicates that the log geometric function describes the frequency distribution of total fields (i.e. known and undiscovered) by size. Numerous authorities have used the log-geometric distribution as the fundamental model of resource deposition, including: B-3 Table 1 Classification of Cook Inlet Fields by Size Oil Fields (A-D (A-D (Oil) (Oil) Gas) Gas) Cume Rem Cume Rem USGS Prod Resv's Prod Resv's Class Field Name (MMBbI) (MMBbI!) (Bef) (Bef) (Code) SSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSsSSSSSsSSSSSssesssssssesessessrszzzez Beaver Creek 3.0 1.0 1.2 136 t2 Granite Point 97.8 25.0 85.3 18.0 ues McArthur River 498.9 58.0 182.1 29.0 19 Middle Gr. Shoal 144.0 14.0 i ¢ 9 17 Swanson River 201.0 18.0 96.7 259.0 18 Trading Bay 86.1 2.0 S7.8 2 16 Gas Fields Cume Rem USGS Prod Reserves Class Fleld Name (Bef) (Bef) (Code) SSSSSSSSSSSSSSSSSSSSSSsSssSSsSsSsssSsSssssessesszsseseseeseesesqsqzqza Beaver Creek 18.9 221.0 14 Beluga River 181.3 676.0 16 Kenai 1540.8 751.0 16 N. Cook Inlet 687.8 812 ig, Mc Arthur River 107.9 600.0 16 Sterling 2.1 23.0 11 Trading Bay 2.0 29.0 11 West Fork 1.5) 6.0 9 Birch Hill 0.1 11.0 10 Falls Creek 0% 113.0 10 Ivan River QO. 26.0 11 Lewis River On7 21.0 11 Nicolai Creek lieve 3.0 8 North Fork 0.1 12.0 10 West Foreland QO. 210). 11 Albert Kaloa 0.1 -- 3 Cannery Loop 0.01 -- 1 Moquawkie 1.0 -- 3 Figure 1 Relationship Between Gas Field Size and Gas Well Size Cook Inlet Nonassociated Gas Fields 1000 Weil Size= 0.6735 * Field Size*0.7885 R = 0.94 100 = oO Gas Well Size (Bcf) .01 .001 .001 -01 AL 1 10 100 1000 10000 Prepered by: ICF-Lewin Energy Service, 1968 Gas Field Size (Bcf) os ' uw Figure 2 Relationship Between Oil Field Size and Oil Well Size Cook Iniet Active Oil Fields 100 Weil Sizes 0.4778 * Field Size*0.4314 Ri 20.97 Fletd and Well Sizes inctude Oil and Associsted Gas 1 10 100 Oil Fleid Size (Million BOE) Prepared by: 1CF-Lewin Energy Service, 1988 az ° a c 2 z McArthur River e 10 ; =| 3 = GD — Middle Ground Shoe! 2 © recog Bey 1000 a * The Gas Research Institute's Hydrocarbon Model ee Information Agency's Intermediate term Forecasting System (IFS) *The National Petroleum Council's representation of fields in the Arctic Oil and Gas Study The log-geometric distribution is also part of the ICF-Lewin Domestic Replacement Cost System (REPCO). Briefly, this statistical form is described by the following equation: (2.0) Nj = mNj;-; where: Fields have been grouped by size into discrete,mutually exclusive classes; Nj : Number (frequency) of fields in class "j" m : aconstant, generally in the range 1.5< m < 1.8 The recursion in the definition may be eliminated by the following: (2.1) Nj = Ny x [mi] The number of fields in the largest class (N;) has a special significance, and is discussed below. The log-geometric function is part of a class of statistical forms which describe highly skewed populations, of which oil and gas fields are only a small subset. The reader is referred to the following source for more details about the underlying theory which supports application of this distribution: Simon, H.A., Models of Man, Ch. 9 "On a Class of Skew Distribution Functions", 1955, Pergamon Press (pp 145- 164) Because the log-geometric function is widely used in oil and gas analysis, a standard methodology for assigning fields to size classes has been developed by the United States Geological Survey. This system is illustrated in Figure 3, and is described by the following equation: (3.0) Sj = $).1/2.0 B-7 where: Sj : average size of fields in class "j" In the USGS class system, each class is made up of fields that are half the size of fields in the next larger class. As shown in Figure 3, a broad range of field sizes are considered. The range is essential for statistical purposes; however, only a small portion of the range describes fields which are economically viable. The recursive definition may be eliminated by: (3.1) Sj= Sy/[2i-1] Application of the log-geometric distribution with the USGS class system is relatively straightforward, requiring an estimate of the size of the largest field and an estimate of the total resource base, as illustrated by the following derivation: The resource in any of the individual USGS field size classes is given by the product of the typical field size for the class multiplied by the number of fields: (4.0) Rj =Njix S| where: Rj : total resource in class "j", which also may be expressed by (4.1) below by substituting equations (2.1) and (3.1). (4.1) Rj = [Ny x (mi-1)] x [S4/(2i-1)] or (4.2) Ry = NyS, (m/2)i-1 The Total Resource (TR) is the sum of the resource in each of the size classes: (4.3) TR = Ry + Ro + Ro +... Rig by substituting Equation (4.2), we have: (4.4) TR = N,S; [1.0 + (m/2)! + ... (mm/2)18] However, equation (4.4) is a geometric series, and has a finite sum. Thus, the total resource may be expressed as: (5.0) TR = NyS; [1 - (m/2)19]/[ 1 - (m/2) ] B-8 Figure 3 The USGS Field Size Classification System For Statistical Purposes, a Broad Range of Field Sizes is Considered 10000 1000 Most of the Known Fieids sre Class 16 and sbove 100 10 Fleld Size (Milllon BOE) 12 3 4 5 6 7 8 9 1011 1213 1415 1617 1819 Prepared by: ICF-Lewin Energy Service, 1968 USGS Size Class Number B-9 Equation (5.0) was used in conjunction with the following sources of data to solve for the parameter m, thus establishing the log-geometric distribution for Cook Inlet Fields: * Sj : taken from the standard field classification system given in USGS Clrcular 828 ¢ N1: Taken from ACC data on known fields; since the Cook Inlet has been subject to exploration activity since the 1940's, it was assumed that the largest field had been found. ¢ All resource, known and undiscovered was converted to Barrel of Oil Equivalent (BOE), thereby assuming that the entire Cook Inlet area was a single exploration "play"; the conversion factor was: 5.8 Mcf = 1.0 BOE ¢ TR: The total resource was estimated as the sum of the known resource (i.e. cumulative production plus remaining reserves in existing fields) and the Undiscovered resource, as estimated by the Potential Gas Committee and the USGS. (Discussed in detail below). Given N; and m, it is possible to construct the frequency distribution for the total resource, by applying equation (2.0) for each of the 19 size classes, thus yielding the total number of fields in each class. Similarly, N;,S;, and m_ permit calculation using equation (4.2) of the total amount of resource in each of the size classes. C. Calculate the Number of Fields Remaining to be Discovered by The Undiscovered Resource is the difference between the Total Resource (calculated using the methodology of Section B, above) and the Known Resource (Section A). Two equations describe this process: (6.0) Niu = Nj) - Nix where: Nik : Number of Known fields in class "|" -Nj,u .: Number of Undiscovered fields in class "j" (7.0) Ri,u = Rj - Rik - Rj,p B-10 where: Rj,o : Development Potential of Known Fields in class "j" Rj,« : Resource in Known fields in class "j" Rj,u : Resource in Undiscovered fields in class "j" The development potential of known fields ( Rj,p ) was estimated by distributing the Probable Resource into the size classes in proportion to the amount of known resource: (7.1) Ri,p = Rorob [ Rik / Rknown] where: : PGC Estimate of probable resource : Total Known Resource = Rix + Rox + ...Ri9k The average field size for undiscovered fields ( FS) ) is given by the ratio of the resource to the number of fields: (8.0) FS,u = Riu / Niu Fields were apportioned to the major analysis categories of Product (Oil, Nonassociated [NA]Gas)location (Onshore, Offshore), and certainty (Possible, Speculative) on the basis of the contribution of these sources to the total undiscovered resource estimate, as shown below: Table Entry is Percent of Undiscovered Resource in Each Category Certainty Onshore Offshore Probable § 539 324 Possible Oil 28.14 18.76 Possible NA Gas 11.30 5.68 Speculative NA Gas 19.42 8.07 Thus, in the analysis, 28.14% of the undiscovered fields were treated as onshore oil fields and so forth. Following allocation of fields to Product, Location, and Certainty categories, Equations (1.1) and (1.2) were used to estimate the productivity of the typical well for each size class. The ratio of field size (Figure 3) to well size yields an estimate of the number of wells required to develop the field. B-11 D. Calculation of the Marginal Cost of Developing Fields within the Undiscovered Resource Base Typical fields from each size class analyzed to determine the minimum price required for economic viability. This price was defined to be the marginal cost for all fields in the size class. Across all size classes, the pairing of cost and resource in class forms the cost-supply relationship for the Product/Location/Certainty group; these results were then integrated to form the aggregate cost-supply curve for the entire Cook Inlet area. The ICF-Lewin Replacement Cost Model was used to perform the actual calculations. The operation of this model and specific worked examples of calculations are provided in: U.S. Department of Energy, Supply Analysis Methodology of the Domestic Oil and Gas Replacement Cost Model,..Publication DOE/FE/30014-1 (Vol 2), July 1985; Prepared by ICF-Lewin for the U.S. Department of Energy, Assistant Secretary for Fossil Energy under contract No. DEA CO 1-82FE-30014. This section provides the basic structure of the analysis, which involved four steps, as described below. 1. Scheduling of Field Exploration, Development, and Production At the end of Phase 3, the work had made possible the apportionment of the undiscovered resource into typical field sizes. The productivity of a given well could be estimated and therefore, the number of productive wells (i.e. as field size divided by well size) Our first briefing paper reported on methods for estimating the yearly production potential of wells. Thus, the first step of the economic analysis was to represent the major events of field exploration, development and production in a time schedule. B-12 “Exploration: The process of exploration begins in "Year 1",with the Lease Bonus. During Year 2,Geological and Geophysical Reconnaissance is conducted. In year 3, exploration wells are drilled, composed of one successful wildcat well and a number of exploratory dry holes, calculated on the basis of a 25% exploratory success rate; In offshore projects, all exploratory wells are assumed to be drilled from mobile platforms and therefore are not producable. For onshore projects, the successful wildcat well may be put on production. *Development: For Onshore projects, development drilling begins in Year 4. Offshore projects require platform construction prior to development drilling, which occupies Years 5 and 6, and thus, offshore development begins in Year 7. The time required for field development is a function of the total number of wells, and ranges from 1 to 3 years. A developmental success rate of 75% is assumed. Thus, the total development drilling involves both successful and dry wells. The development schedule is set to maximize production from the field. Oil wells are assumed to have a primary and secondary production phase; injection wells are drilled to service oil production wells at the rate of 0.5 injectors per producer. It is assumed that secondary production commences four years after primary begins; thus, injection wells are drilled three years after an oil well is put on production. *Production: for Onshore fields, the successful wildcat may be put on production in Year 4; development wells are put on production in the year that they are drilled. For Offshore fields, the platform must be complete before development can begin, and, due to space limitations and safety considerations, production cannot commence until all development wells are in place and production facilities are in place. Wells drilled in each year constitute a "vintage"; the analysis program sequences production for the field, aggregating across "vintages". A production stream of 35 years is calculated. The production model is: B-13 (9.0) P(Y) = R(Y) * K where: P(Y): Production in year "Y" R(Y): Beginning of Year Reserves in year "Y" K: fraction of reserves produced (constant over field life) values of "K" have been presented in our initial briefing The Probable resource base reflects developmental opportunities in known fields and was handled differently. The individual well was the unit of analysis. No exploration activity was attributed to the project, and drilling and production commence immediately in Year 1. The exploration and development activity for the field (or well) is recorded in a data array which contains the time-sequence of physical quantities (wells required, oil and gas production, etc.). The next stage of the analysis transforms these physical quantities into expenditures, as discussed below. B-15 2. Calculation of Costs of Exploration and Development The Cost Calculation routines estimate the expenditures required in all aspects of the field's life. The following costs (expressed in $1987) were used in the analysis: Cost Item Calculation Method SSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSssSessssessesessesesqesesesessqseqseqzzqza Geological and Geophysical Costs Exploration Phase 12.8% of Exploratory Drilling Cost Development Phase 0.3% of Development Drilling Cost Onshore Succ.Exploration Well [6.2410°8(DD2-83}41,593,000 Offshore Exploration Well [-0297DD1-95}4 4,100,000 Onshore Development Well [6.2410°5(DD2-83)}41,593,000 Offshore Development Well 291.6*DD * 1,290*°WD + 1,850,000 Onshore Dry Hole 75% Successful Well Cost Offshore Development Dry Hole 87% Successful Well Cost Onshore Oil Well (Primary) $1.044 x (Peak Prim Bbi/Day Prod) Onshore Oil Injection Well $0.14 x (Peak Sec. Bbi/Day Prod) Onshore Oll Well (A/D Gas) $0.99 x (Peak Mct/Day Prod) Offshore Oil Wel! (Primary) $3.654 x (Peak Prim Bbi/Day) Onshore Gas Well $0.74 x (Peak Mcf/Day Production) Offshore Gas Well $2.59 x (Peak Mcf/Day Production) Onshore Oil Wells (Prod and Inj) $84,200 per well per year Offshore Oil Wells $294,700 per well per year Onshore Gas Wells $84,200 per well per year Offshore Gas Wells $294,700 per well per year Onshore Oil Wells (Primary) $0.06/Bbl Onshore Oll Wells (Secondary) $0.13/Bbl Offshore Oil Wells $0.21/Bbi Onshore Gas Wells $0.06/Mcf Offshore Gas Wells $0.21/Bbl Offshore. Production Platform [292,000*°WD1.11]+5,060,000 The cost calculations are included in the data arrays which form the simulation of the field's development. Noticeably absent from these costs is the lease bonus payment, which is modeled as a percentage of the gross revenue from the project. The bonus payment is therefore added by the financial model, discussed below. 3. Financial Calculations The project data, compiled in steps 1 and 2 above, is converted to a project pro-forma budget. For any particular oil or gas price, the production stream may be transformed to revenues. The financial model calculates royalty, Alaska and Federal taxes, and computes after-tax cash flow for each year of project operation. The yearly values of cash flow are discounted and summed to yield the present worth of the prospect. An iterative algorithm varies the prices in order to solve for that price which results in a zero present value for after tax cash flows. This is the price at which the after-tax cash available from the project will recover the expenses incurred (recognizing the time value of money), and is thus the minimum price required for economic development. The financial routines reflect Federal and State tax law as of June 1987. A 12.5% royalty is assumed for onshore projects and a 16.67% royalty is used for offshore projects. Lease bonus is set at 3% of gross value for onshore projects; the percentage varies in proportion to field size for offshore projects. The specific details of the financial analysis model are provided in the system documentation. 4. Development of Cost-Supply Curves The resource theory yields an estimate of the amount of recoverable resource (i.e. supply) lies in each field size class; the financial analysis yields an estimate of the marginal cost for each field size class (i.e. cost). Thus, the cost-supply curve is determined by pairing supply and cost on a field-size class basis, or: B-16 (10.0) S = F(C) , approximated by {S;,Cj} where: { Sj,Cj } the set of pairings of Resource and Marginal Cost Sj : Cumulative resource through class "” = R; + Ro + ...+ Rj as defined in equation (7.0) Cj; : Marginal cost for Class "|" as determined by the Financial Analysis model described above F(C): smooth curve approximation developed by least squares fitting based on {Sj,Cj} The cost-supply curve (F(C)) is the "“least-cost" curve, and assumes that the resource can be found in the order of decreasing field size (i.e. largest fields first). As such, it is both appropriate and inappropriate: ¢ It is an accurate representation of the incremental cost of finding new resource because it considers the disposition of resource in fields of varying size and the field-related costs of exploration and development; ¢ It is an innacurate representation of the actual cost of oil and gas, to the extent that the discovery process does not find fields in their exact precise ordering by size. Because the cost-supply curve (F(C)) accurately represents incremental costs, it may be used in conjunction with an exploration model to compute a more realistic “average cost supply curve” which reflects differences from the “least cost" curve assumption of field discovery in order. Thus, F(C) represents the maximum supply available at prices equal to marginal cost "C". B-17 lll. Results The 1986 Report of Potential Gas Committee (PGC) was the source for estimates of undiscovered nonassociated gas, and the United States Geological Survey (USGS) Circular 860 was used for undiscovered oil resource. Figure 4 shows the disposition of this resource by product, location and certainty. Use of these estimates rests on several assumptions: « A recent report to the Alaska legislature (Historical and Projected Oil and Gas Consumption, January 1988) indicates that Cook Inlet Oil fields are very mature (80-90% depletion). Figure 5 indicates that oil field size (Cumulative Production plus Remaining Reserves) has been relatively constant over the period 1975- 1984. Hence, we conclude that there is_ little development potential remaining in existing oil fields and that the probable oil resource is zero. * The USGS does not make an oil estimate that is comparable to the PGC "Speculative" category. Hence, we have not analyzed any speculative oil resource. « All resource ( Oil, Associated Gas, and Nonassociated Gas) was converted to Barrel of Oil Equivalent (BOE), and the entire Cook Inlet Area was analyzed as if oil and gas fields were part of a single hydrocarbon play. The undiscovered resource, by product and location is shown below: Total Onshore Offshore (Mil BOE) (Mil BOE) (Mil BOE) Seeseseesesessesssesesseessssssesessseesssseseseseseeseseserez= Probable NA Gas 184 115 69 Possible Oil&AD Gas 1000 600 400 Possible NA Gas 362 241 121 Speculative NA Gas 586 414 172 B-18 Undiscovered Resource (Million BOE) 1600 1400 1200 1000 800 600 400 200 Figure 4 Cook Inlet Undiscovered Resource Gas: Potential Gas Committee (1986) Oil: United States Geological Survey Moe Fak pte moe inal Probable Possible Moet Certain Prepered by: ICF-Lewin Energy Service, 1968 Speculative Least Certain Field Bite (Cumulative Production + Remaining Reserves) BCE Fistd Bie (Cumuative Produstion + Remsining Reserves) ta! Bet Legertrmis Seue Figure 6 B-20 Trends In Oil Fleld Size Cook Inlet Active Fields (1979-1984) Fieid Size Has Remained Constant, indicating That the Process of Development is Compiete 1000 100 —e tewrw cee i teos mss j ee ei we } en Meee ee a eee —O- rere ey 10 1 1979 1980 1981 1982 1983 1984 Trends In Nonassociated Gas Fleld Size Cook Inlet Active Fields (1979-1984) 10000 1000 —e- merce mew Fi ~_— 100 errs tw set eee ceca —o am —=> “my 10 ——- sare 1 1979 1980 1981 1982 1983 1984 Meenas wy 1K Lemin Ener gy Service, 't08 Figure 6 Frequency Distribution of Cook Inlet Fields By Size Class Showing Fields Which Are Known and Those Which Remain to Be Discovered Although the theory posits a large number of fieids Onty @ small number sre economically viable 10000 [1 UNDISCOVERED m@ KNOWN Log-Geometric Distribution, m21.685 Number of Fleids (Frequency) 10 12-9 4 5 6. 7 c8h OS 10 TH 2 SS" 14 15°16 17, Prepared by: ICF-Lewin Energy Service, 1968 USGS Size Class is) ess 18 19 Based on the ACC data reported in Table 1, there is a total of 2962.6 Mil BOE of resource in the 22 known fields in the Cook Inlet Area. The distribution of this known resource by field size class is: USGS Total Oil Gas Class (Mil BOE) (Mil BOE) (Mil BOE) SSSSSSSSSSSSSSSSSSSSSssssssessseesssssesesesesessesqrzrezc=z 19 835.3 835.3 0 18 873.2 478.0 395.2 17 729.1 470.8 258.6 16 445.9 176.0 269.9 15 and Below 79.1 8.0 71.1 Total Known 2962.6 1967.8 994.8 Using the methodology described above, a log-geometric distribution was fitted to the ACC data on known fields and the PGC/USGS estimates of undiscovered resource. Figure 6 shows this distribution, with coefficient m [of definitional equation (2.0)] equal to 1.685. This value compares favorably with other values for other geologic basis in the U.S. and Canada as achieved by Arps, Arps and Roberts, the USGS and ICF-Lewin, all of which lie in the range of 1.65 - 1.75. Figure 7 illustrates the apportionment of undiscovered resource by field size. The calculated values for each of the components of the resource base are shown in Table 2,below: The analysis indicates that the Cook Inlet Region is immature and that a considerable amount of the undiscovered resource lies in large fields. The financial analysis calculated the minimum required price for a well (Probable Resource) or Field (Possible and Speculative Resource). Prices below $0.75/Mcf for gas and $5.00/Bbl for Oil were not recorded, and the values were used as "floors". Similarly, gas resource was not considered "viable" if the required price exceeded $12.00/Mcf; the oil price "ceiling" was $50.00/Bbl. Given this adjustment, the minimum required prices are: B-22 Figure 7 Cook Inlet Hydrocarbon Resource Resource (Known and Undiscovered) by Field Size 1000 - Showing Product (Oil and Gas) and Risk Category (Probable,Possible, Speculative) 900 800 700 600 500 400 (Average Field Size x Number of Fields) 300 Total Resource In Class (Million BOE) 200 100 W293 4 §& (67° 8) 9 HO-11 2 13 W455 16 74 1819 Prepared by: ICF-Lewin Energy Service, 1968 USGS Size Class Table 2: Undiscovered Resource by Field Class Nonassociated Gas Undiscovered Resource Typical PROBABLE POSSIBLE SPECULATIVE USGS Well Onshore Offshore Onshore Offshore Onshore Offshore Class _ (Bef) (Bef) (Bef) (Bef) (Bef) (Bef) (Bef) BSSSSSSSSSSSSSSSSSSSsSSsSSssSsSesssssessssessssssssssssseseseezczz 19 672 0 0 0 0 0 0 18 389 118 71 0 0 0 0 17 225 86 52 0 0 0 0 16 130 84 50 0 0 425 177 15 76 68 41 247 124 311 129 14 44 53 32 181 91 302 126 13 25 48 29 176 88 246 102 12 15 42 25 143 72 191 79 14 9 35 21 111 56 174 72 10 5 30 18 101 51 150 62 9 3 25 18 87 44 127 53 8 2 21 13 74 37 108 45 7 1 18 11 63 32 91 38 6 0.5 15 9 53 27 76 32 5 0.3 12 7 44 22 65 27 4 0.2 10 6 38 19 55 22 3 0.1 0 0 32 16 44 18 2 0.05 0 0 25 13 37 15 1 0.03 0 oO 21 11 17 8 Oil and Associated Gas Undiscovered Resource USGS Typical Well Onshore Offshore Class (Mil BOE) BOE) BOE) SSSSSSSSSSSSSSSSSSSSSSSsSSSsSsSssssesssssssessessseszseeesezseez=a 19 9.8 Qo 0 18 7.3 0 0 17 5.4 0 0 16 4.0 0 0 15 3.0 106 71 14 2.2 78 52 13 1.6 76 50 12 1.2 61 40 11 0.9 48 32 10 0.7 43 29 9 0.5 37 25 8 0.4 32 21 7 0.3 27 18 6 0.2 23 15 5 0.15 19 13 4 0.11 16 11 3 0.08 14 9 2 0.06 11 7 1 0.04 9 6 B-24 Table 3: Minimum Required Price, By Field Class Field Gas Prob 18 549 0.75 0 Ld 274 0.75 0 16 137 0.75 0 15 69 0.75 0 14 34 0.75 0 13 17, .,.0.76 1 12 9 1.32 2 11 4 2.30 3 10 2 3.96 6 9 1 6.86 10. 8 0.5 11.81 N Note: N.E. = Note Economic at analysis ceiling price The results reported in Table 2 constitute the "S|" and the data shown in Table 3 are the "C;" of Equation (10.0). Figures 8 through 15 show the cost-supply curves achieved by plotting the set of pairs {Sj,Cj}. Figure 16 shows the aggregate of the individual curves to form a total cost-supply curve for the entire Cook Inlet Basin. Least squares regression was used to plot a smooth curve through the points to achieve the following equations for the cost- supply relationship: Probable, Onshore Gas: Probable, Offshore Gas: Possible , Onshore Gas: Possible, Offshore Gas: 75 78 75 75 -75 +18 -06 -57 14 64 E. Speculative, Onshore Gas: Speculative, Offshore Gas: Possible, Onshore Oil: Possible, Onshore Oil: Note that: Gas Poss&Spec $/Mcf $/Mct 0.75 0.75 0.75 0.75 0.75 0.75 0.77 1.15 1.33 2.04 2.27 3.46 3.96 6.02 6.87 10.46 11.81 N.E. N.E. N.E. N.E. N.E. S = 460.7 C0.135 S = 275.1 €0.119 S = 297.5 C0.616 S = 131.1 €0.622 S = 821.8 C0.397 S = 315.8 C0.391 S = 4.156 C1.017 S = 0.767 C1.224 S : Supply (Bcf for Gas, Mil Bbl for Oil) C : Marginal Cost ($1987/Mcf for Gas, $1987/Bbl for Oil) Equations are limited to the ranges of: Gas: 0.75< C < 12.00 Oil: 5.00< C < 50.00 OIL USGS Size Onshore Offshore Onshore Offshore Onshore Offshore Class (MMB) $/Mcf $/Mcf $/Bbl $/Bbl SSSssssssessssssssssssssssasssssssssssesssssesssssesessseee== 11.43 17.74 15.41 23.87 20.80 32.20 27.92 43.25 R2 = 0.98 R2 = 0.99 R2 = 0.95 R2 = 0.99 R2 = 0.96 R2 = 0.99 R2 = 0.97 R2 = 0.98 zzzzzzz minim mim imm Cumulative Resource Economic at Price (BCF) w =72.0) Figure 8 Cost-Supply Curve, Cook Inlet Region Onshore Probable Resource Base Evaluated at the Cost of Developing Individual Wells 6% (Real), 10% (Nominal) DCF Rate of Return 600 550 Total Resource: 670 Bcf 630 Bcf Economic at Prices below $12/Mct 500 450 0.75 1.50 2.25 3.00 3.75 4.50 Prepared by: ICF-Lewin Energy Service, 1988 Marginal Cost (91967/Mct @ Wellheed) Figure 9 Cost-Supply Curve, Cook Inlet Region Offshore Probable Resource . Evaluated at the Cost of Developing Individual Wells 350 6% (Real), 10% (Nominal) DCF Rate of Return 7 300 Total Resource: 400 Bcf 367 Bct Economic at Prices Below $12.00/Mct 250 Cumulative Resource Economic at Price (Bct) 200 0°75 1.50 2.25 3.00 3.75 4.50 5:25 6.00 Prepered by: ICF-Lewin Energy Service, 1968 Marginal Cost ($1967/iict @ Weltheed). 6575 [S) N Cumulative Resource Economie at Price (Million Boi) Prepared by: ICF-Lewin Energy Service, 1968 Figure 10 Cost-Supply Curve, Cook Inlet Region Onshore Possible Oil Resource Base Evaluated at Full Cost of Field Exploration and Development 6% (Real), 10%(Nominal) DCF Rate of Return , 200 150 100 Total Resource: 600 Million Bbi 213 Million Bbi Economic at Prices Below $50/Bbi 50 20 25 30 35 40 Morginel Cost ($ 1987/80! @ Wettheed) Annual Production (Bcf/year) FORECAST PRODUCTION DECLINE FROM EXISTING COOK INLET RESERVES FIGURE 5.1 260 260 [= 180 100 ;— Tn 1987 1990 1996 2000 2006 New resources. 20F0148 2010 investments in production capacity would develop undiscovered The study classified the undiscovered resource base into three categories based on increasing levels of uncertainty associated with the resource estimates and costs: Developmental: This class includes those resources which could be added through the extension of existing fields. The probability of success here is fairly high. On the other hand, the potential size of finds is low. New field exploration: This includes resources that additional exploration in formations known to _ be productive could add. The probability of success is considered moderate, with the size of the potential resource finds somewhat higher. Speculative: This class, which is the least certain category in this analysis, includes resources which could be added through exploration in formations not yet proven productive. While the probability of success is Page 5-3 low, the potential for finding very large reserves is best in this category.! To develop the supply curves for the undiscovered resource, ICF-Lewin began with data on known fields in conjunction with well-established statistical theory to estimate the distribution of the remaining resources. This analysis had four steps: Ls Characterize the known field by size and well productivity. 23 Establish the statistical distribution of the total resource (known plus estimated undiscovered) by field size. 30 Calculate the number of fields remaining to be discovered, by size, product, location and certainty. 4. Estimate the marginal cost of developing new fields, by size. The resulting supply information becomes a major input, along with the demand curves, for the integrative analysis. Figure 5.2 illustrates this integration process. As the figure shows, the difference between production from proved reserves (shown as 1 in the figure) will be supplied from increasingly less certain and more costly supplies (2 in the figure). Changes in the demand forecast (3 in the figure) can affect the schedule of required additions as well as the costs of additions. The more costly supplies represent both the increased difficulty in finding and development as well as uncertainty of success. This approach is necessarily a simplified one, and differs from real world development in one key aspect. It is based on the assumption that Potential Gas Committee, 1986 Report. These terms are based on Potential Gas Committee "mostly likely" figures. The PGC "mostly likely" figures can be interpreted as being roughly a 50 percent chance that these quantities can be produced. 20F0148 Page 5-4 5. GAS SUPPLY AND DEMAND INTEGRATION In this chapter, we present an integration of the natural gas supply and demand analyses. The demand estimates under the three oil price scenarios were discussed in Chapter 4. This chapter first summarizes the supply analysis -- presented in more technical detail in the appendix. This is followed by a discussion of natural gas price differentials around Cook Inlet. Next we overlay the demand and supply curves to locate the clearing market segments. The resulting market-based prices are then presented. SUPPLY ANALYSIS SUMMARY The supply analysis appraises the marginal economics of future discoveries from the Cook Inlet oil and gas resource base. Just as for the demand side, any significant natural gas supplies within the Railbelt region are around Cook Inlet and not anticipated for other Railbelt areas. The end products of the supply analysis are sets of price-quantity pairs representing production possibilities in the specified years. The first step in the analysis estimated the production profile from known reserves in active fields without consideration of economics or markets, that is, assuming no incremental development of new wells. This served as the baseline for natural gas productive capacity. Table 5.1 provides estimates of the production decline coefficients (the percent of remaining reserves produced per year) for the oil and gas fields studied; Figure 5.1 illustrates the decline in production from these reserves, without new development. As this figure shows, new investments in production are needed to meet future demands for gas. 20F0148 Page 5-1 ESTIMATES OF RESERVES AND PRODUCTION DECLINE COEFFICIENTS Field Nonassociated Gas Beaver Creek (Gas) Beluga River Kenai McArthur River (Gas) N. Cook Inlet Sterling Trading Bay (Gas) West Fork Associated/Dissolved Gas Beaver Creek (Oil) Granite Point McArthur River (Oil) Middle Ground Shoal Swanson River Trading Bay (Oil) TABLE 5.1 FOR COOK INLET FIELDS 1984 (Bef) 18 29 259 1984 a Bef. ONNOOr Est. Decline Reserves Production Coefficient —(Fract) _ oooooo°o$]9n ooooo$n ao -04 -10 06 02 ah) bel ~15 eo -05 aS -14 TES ko NOTE: The estimated decline coefficient is the estimate of the fraction of reserve that will be produced per year over the long-run. As in any given year actual production may depart from the long-run decline coefficient as reserve may be shut in or the table shows, more produced to satisfy short-run needs. 20F0148 Page 5-2 Figure 11 Cost-Supply Curve, Cook Inlet Region Offshore Possible Oil Resource Base Evaluated at the Full Cost of Field Exploration and Development 150 6% (Real), 10% (Nominal) DCF Rate of Return _ 125 100 75 50 Total Resource: 400 Million Bbi 25 Cumulative Resource Economic at Price (Million Bbi) 30 35 40 45 50 55 60 Prepared by: ICF-Lewin Energy Service, 1968 Marginal Cost ($1967 @ Weltheed) 8-30 Figure 12 Cost-Supply Curve, Cook Inlet Region Onshore Possible Nonassociated Gas Resource Base Evaluated at the Full Cost of Field Exploration and Development 900 6% (Real), 10% (Nominal) DCF Rate of Return 800 700 600 500 400 Total Resource: 1400 Bcf 300 200 Cumulative Resource Economic at Price (Bct) 100 0.75 1.50 2.25 3.00 3.75 4.50 5.25 6.00 6.75 7.50 Prepared by: ICF-Lewin Energy Service, 1968 Marginal Cost ($1987 @ Weltheed)’ B-31 Figure 13 Cost-Supply Curve, Cook Inlet Region Offshore Possible Nonassociated Gas Resource Base Evaluated at the Full Cost of Field Exploration and Development 6% (Real), 10% (Nominal) DCF Rate of Return 400 350 300 250 200 Total Resource: 700 Bcf 150 100 Cumulative Resource Economic at Price (Bcf) 50 0.75 1.50 2:29 3.00 3.75 4.50 5.25 6.00 6-75 Propered by: ICF-Lewin Energy Service, 1968 Marginal Cost ($1967}act @ Weltheed) Figure 14 Cost-Supply Curve, Cook Inlet Region Onshore Speculative Nonassociated Gas Resource Base Evaluated at the Full Cost of Field Exploration and Development 1700 6%(Real), 10%(Nominal) DCF Rate of Return 1600 S = 1500 i = 1400 § 1300 1200 Total Resource: 2400 Bcf 3 1100 NOTE: The Specuistive Resource lies in formations = 1000 that have not yet been shown to be productive 3 and theretore is the leest certain resource. However ¥ 8 sizable low-cost component. 800 700 0.75 1.50 2.25 3.00 3.75 4.50 5325) 6.00 6.75 7.50 Prepared by: ICF-Lewin Energy Service, 1968 Marginal Cost ($1987/Mct @ Weltheed) B-33 Figure 15 Cost-Supply Curve, Cook Inlet Region Offshore Speculative Nonassociated Gas Resource Base Evaluated at the Full Cost of Field Exploration and Development 650 6% (Real), 10% (Nominal) DCF Rate of Return 600 550 500 450 400 350 300 250 200 150 0.75 1.50 2.25 3.00 3.79 4.50 §.25 6.00 6.75 Prepared by: ICF-Lewin Energy Service, 1968 Marginal Cost ($1987/Mct @ Weltheed) Total Resource: 1000 Bcf Cumulative Resource Economic at Price (Bct) Marginal Cost ($1987/MMBTU) Figure 16 Cook Inlet, Alaska Aggregated and Smoothed Cost-Supply Curve 8.0 19 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 Smoothed Regression Equation : Costa 0.21267 * 104(3.2314e-4 Quantity) RA2 = 0.964 Aggregate of: Onshore and Offshore Regions Probable and Possible Resource Base Associated and Nonassociated Gas 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 incremental Quantity of Resource (BCF) Prepared by: ICF-Lewin Energy Service, 1968 APPENDIX C APPENDIX C FUEL OIL PRICE PROJECTIONS FOR FAIRBANKS This Appendix documents the detailed components of the fuel oil price projections discussed in Chapter 6 of this report. Table C-1 through Table C-3 present the components for the electric utility fuel oil price projections. Table C-4 through Table C-6 present the components for the residential/commercial fuel oil price projections. All prices presented in these tables are in 1987 dollars. Prices are presented in dollars per barrel units unless noted as dollars per gallon. The remainder of this Appendix summarizes the sources and assumptions used for these price projections. World Crude Oil Prices and Crude Quality Adjustments The world crude oil price projections for this analysis were taken from prior work done by ICF/Lewin in February 1988 for the Alaska Power Authority as contained in the report entitled: "Outlook for World Oil Prices: Analysis of Alternative Schools of Thought 1987-2010." The quality differential between the world crude oil price, which corresponds to the value of a medium quality crude such as Saudi Light in the U. S. Gulf, and Alaska North Slope crude oil is also taken from this report. The quality differential between ANS crude oil and Cook Inlet crude oil was calculated based on the actual prices of these two crudes in September, 1987. The Cook Inlet crude oil price was taken from the Alaska Department of Revenue Report: "Petroleum Production Revenue Forecast," (p. 14) which was published in October, 1987. The ANS price for September, 1987 used to calculate this quality adjustment was the average wellhead value of royalty oil sales of ANS crude oil as derived from data supplied by the Alaska Department of Natural Resources plus the TAPS tariff from the North Slope to Valdez. 20T00365 Page C-1 Transportation Costs Transportation costs used for this analysis were taken from various sources. Current (1987) tariffs for the TransAlaska Pipeline System (TAPS) were provided by the Alaska Department of Natural Resources. Projected TAPS tariffs for TAPS were taken from estimates provided by the Alaska Department of Revenue. Tesoro and the Golden Valley Electric Association provided current tariff estimates for petroleum product movements through their own pipelines. Current transportation costs for movements between the lower 48 and Valdez were calculated as the difference between the average value of ANS crude oil in the Gulf of Mexico as measured by cracking yields reported in Platts Oilgram and the average value of state royalty oil in Valdez. Tanker tariffs for shipments from Valdez to the Tesoro refinery in Nikiski were taken from the report by Louis Delong and Lloyd Pernella entitled: "Alaska Petroleum Product Pricing" (p. 99),which was published in February 1983. This tanker rate was updated to 1987 dollars using the GNP deflator. All tanker costs were assumed to remain constant at their 1987 levels. Current Fuel Oil Prices Current wholesale heating oil prices in Nikiski, Anchorage and Fairbanks were provided by Enstar. Current retail heating oil prices in Fairbanks were taken from the Fairbanks North Star Borough’s publication entitled: "Fairbanks Community Research Quarterly " (p. 81), which was published in Fall 1987. The Golden Valley Electric Association provided estimates of their current fuel oil costs. Gross Margins Current gross margins including processing costs were calculated as the difference between the wholesale price of fuel oil at the refinery (rack price) and the estimated refiner acquisition costs. Gross distribution 20T00365 Page C-2 margins for heating oil sales were similarly calculated as the retail price less the refiner wholesale price. 20T00365 Page C-3 20T00365 MAPCO COSTS SsSssssesss= WORLD CRUDE OIL PRICE (MID-QUALITY CRUDE U.S. GULF) ~ QUALITY ADJUSTMENT ~ TANKER TARIFF (VALDEZ TO U.S. GULF) - TAPS TARIFF (PUMP ST. @1 TO VALDEZ) = NORTH SLOPE WELLHEAD PRICE + TAPS TARIFF (PUMP ST. #1 TO MAPCO) + GVEA TARIFF = REFINER'S ACQUISITION COST + RETURNED OIL PENALTY + REFINER MARGIN (No. 4) ($/GALLON) REFINER RACK PRICE (No. 4) No. 4 ($/GALLON) TABLE C-1 ESTIMATED ELECTRIC UTILITY FUEL OIL PRICES CONSENSUS WORLD OIL PRICE SCENARIO (1987 DOLLARS) 1987 1990 2000 2010 $18.36 $20.00 $30.00 $40.00 $0.24 $0.28 $0.42 $0.56 $2.93 $2.93 $2.93 $2.93 $3.93 $2.43 $1.% $2.26 $11.26 $14.36 «= $25.29 $34.25 $2.40 $1.99 Sill 1.85 $0.10 $0.10 $0.10 $0.10 $13.76 = $16.45 = $26.50 $36.20 $0.87 $0.87 $0.87 $0.87 $3.87 $3.87 $3.87 $3.87 $0.092 $0,092 $0.092 80.092 $0.44 0.505 80.744 = $0. 975 Page C-4 TABLE C-2 ESTIMATED ELECTRIC UTILITY FUEL OIL PRICES _ CONSENSUS LOW WORLD OIL PRICE SCENARIO (1987 DOLLARS) MAPCO COSTS 1987 1990 2000 2010 WORLD CRUDE OIL PRICE $18.26 = $18.00 = $24.00 = $30.00 (MID-GUALITY CRUDE U.S. GULF) ~ QUALITY ADJUSTMENT $0.24 $0.25 $0.34 $0.42 - TANKER TARIFF $2.93 $2.93 $2.93 $2.93 (VALDEZ TO U.S. GULF) - - TAPS TARIFF $3.93 $2.43 $1.36 2.26 (PUMP ST. #1 TO VALDEZ) = NORTH SLOPE WELLHEAD PRICE $11.26 $12.39 $19.37 $24.39 + TAPS TARIFF 2.40 $1.99 $1. $1.85 (PUMP ST. #1 TO MAPCO) + GVEA TARIFF $0.10 $0.10 $0.10 $0.10 = REFINER’S ACQUISITION COST $13.76 = $14.48 = $20.59 $26.34 + RETURNED GIL PENALTY $0.87 $0,87 $0.87 $0.87 + REFINER GROSS MARGIN (No. 4) $3.87 $3.87 53.87 $3.87 ($/GALLON) $0,092 0.092, $0,092 $0092 REFINER RACK PRICE (No. 4) No. 4 ($/GALLON) $0,441 $0,458 = $0,003 $0740 20T00365 Page C-5 TABLE C-3 ESTIMATED ELECTRIC UTILITY FUEL OIL PRICES LOW PRICE SCHOOL WORLD OIL PRICE SCENARIO (1987 DOLLARS) MAPCO COSTS 1987 1990 2000 2010 WORLD CRUDE OIL PRICE $18.36 $14.00 $18.00 $20.00 (MID-QUALITY CRUDE U.S. GULF) - QUALITY ADJUSTMENT $0.24 $0.20 $0.25 $0.28 - TANKER TARIFF . $2.93 $2.93 $2.93 $2.93 (VALDEZ TO U.S. GULF) - TAPS TARIFF $3.93 $2.43 $1.% $2.26 (PUMP ST. #1 TO VALDEZ) = NORTH SLOPE WELLHEAD PRICE $11.26 $8.44 «$13.46 $14.53 + TAPS TARIFF $2.40 $1.9 $t.il 1.5 (PUMP ST. #1 TO MAPCO) + GVEA TARIFF $0.10 $0.10 $0.10 $0.10 = REFINER’S ACQUISITION COST $13.76 = $10.53 $14.67 $16.48 + RETURNED OIL PENALTY $0.87 $0.87 $0.87 $0.87 + REFINER GROSS MARGIN (No. 4) $3.87 $3.87 $3.87 $3.87 ($/GALLON) $0.092 $0,092 $0.0972 $0,092 REFINER RACK PRICE (No. 4) No. 4 ($/GALLON) $0,441 $0.364 $0,462 $0,505 20T00365 Page C-6 TABLE C-4& ESTIMATED HEATING OIL PRICES FOR FAIRBANKS CONSENSUS WORLD OIL PRICE SCENARIO (1987 $) TESORO COSTS 1987 1990 2000 2010 ‘SssssssSss5 REFINER COSTS & MARGINS COOK INLET CRUDE OIL $16.27 $17.95 = $28.46 = $38.97 ANS CRUDE OIL $15.73 $17.55 $27.19 $37.05 WEIGHTED AVERAGE ACQUISITION COST $15.86 $17.41 $27.24 = $37.07 + REFINER GROSS MARGIN $8.49 $8.49 $8.49 $8.49 = REFINER RACK PRICE ($/GALLON) $0,580 $0,617 $0,851 = $1085 TRANSPORTATION COSTS ($/GALLON) + PIPELINE TARIFF $0.020 $0,020 $0,020 $0.020 (NIKISKI - ANCHORAGE) + ARR TARIFF $0,077 $0,077. = $0,077 — $0.07 (ANCHORAGE - FAIRBANKS) TOTAL TRANSPORT COST $0.097 = $0,097 80.097 0.097 = WHOLESALE PRICE $0,677 $0,714 = 80.948 $1. 182 + DISTRIBUTION MARGIN 0.173 0.173 0.173 0.173 = DELIVERED PRICE $0,650 90.887 $1,121 $1. 20T00365 Page C-7 TABLE C-4.a TESORO REFINER ACQUISITION COST - ANS CRUDE OIL CONSENSUS FORECAST (1787 $/ BBL) 1987 1990 2000 2010 Seesssssessssssssssssssssss52=5S=5 WORLD CRUDE FRICE :CONSENSUS) $18.26 = $20.00 520,00 B40, 00 (MID-QUALITY CRUDE U.S. GULF) = QUALITY ADJUSTMENT 50,24 $0.28 50,42 50.56 ~ TANKER TARIFF $2.93 52.93 $2.93 Rn ‘WALDE2 TO U.S. GULF) > TAPS TARIFF $3.95 52.43 $1.26 82.26 (PUMP ST. #1 TO VALDEZ) ~ ADJUSTMENT 50.21 $0.21 50.21 50,21 = NORTHSLOPE WELLHEAD PRICE $11.05 814.15 825.08 534.04 + PIPELINE TARIFF g $3.95 $2.43 51.26 $2.26 (PUMP ST. #1 TO VALDEZ) + TANKER COST 50.75 50.75 $0.75 $0.75 (WALDEZ TO NIKISKID = REFINER’S ACQUISITION COST $15.73 $17.53 527.19 $37.05 20700365 Page C-8 TABLE C-4.b TESORO REFINER ACQUISITION COST - COOK [MET CRUDE OIL CONSENSUS FORECAST (1787 $/ 3BL) 1987 sissssesssss2s22s52s252225225Ss5S5S255== SNS PRICE AT VALDEZ 14.98 QUALITY ADJUSTMENT $1.02 TANKER COST ‘weST COCK INLET TO NIKiSKI) 50,27 SEFINER ACQUISITION COST $16.27 20T00365 1990 10.58 $1.10 50,27 517.95 2000 to. 44 $1.75 $0, 27 528.46 a2) . 165 s2.41 50,27 £38.97 Page C-9 TABLE C-4.c WEIGHTS FOR TESORO CRUDES 1997 1990 2000 no AOOR FORECAST OF CI PROD “EBL; 0) $8.233 25.479 7.671 Leder TESORO SHARE 37 0,57 0,27 0.57 COOK INLET AEIGHT 9.24 12 9,04 9.01 ANS WEIGHT 0,7 0,88 0.96 099 WORLD CRUDE PRICE-REFINER COST S115 51,24 51.41 1.58 20T00365 Page C-10 20T00365 TESORO COSTS REFINER COSTS & MARGINS COOK INLET CRUDE OIL ANS CRUDE OIL WEIGHTED AVERAGE ACQUISITION COST + REFINER GROSS MARGIN = REFINER RACK PRICE ($/GALLON) TRANSPORTATION COSTS ($/GALLON) + PIPELINE TARIFF (NIKISKI - ANCHORAGE) + ARR TARIFF (ANCHORAGE - FAIRBANKS) TOTAL TRANSPORT COST = WHOLESALE PRICE + DISTRIBUTION MARGIN = DELIVERED PRICE TABLE C-5 ESTIMATED HEATING OIL PRICES FOR FAIRBANKS CONSENSUS (LOW) WORLD OIL PRICE SCENARIO (1987 $) 1987 1990 2000 $16.27 $15.84 = $22.15 $15.73 $15.36 $21.27 $15.86 $15.42 $21.31 $8.49 $8.49 $8.49 $0,580 $0,569 0.710 $0.020 $0,020 $0.020 $0.077 $0.077 $0.07 $0.097 $0.097 $0,097 $0.677 $0,666 0.807 0.173 . 0.173 0.173 $0,850 90.839 $0,980 2010 $28. 46 $27.19 $27.20 $8.49 $0.850 $0.020 $0.077 $0.097 $0,947 0.173 $1.120 Page C-11 TABLE C-5.a TESORO REFINER ACQUISITION COST - ANS CRUDE OIL CONSENSUS (LOW) FORECAST (1987 $/ 8BL) 1987 1999 2000 2010 222222 => >>=- === SS WORLD CRUDE PRICE (CONSENSUS LOW) $18.76 518.00 574.00 $30.00 ‘MID-QUALITY CRUDE U.S. GULF) ~ QUALITY ADJUSTMENT 50.24 50.25 50.34 $0.42 - TANKER TARIFF 52.93 $2.93 52.93 52.93 VALDEZ TO U.S. GULF) - TAPS TARIFF $3.93 52.43 $1.26 52.26 (FUMP ST. @1 TO VALDEZ) ~ ADJUSTMENT 50.21 $0.21 50.21 50.21 = NORTHSLOPE WELLHEAD PRICE $11.05 $12.18 $19.16 = $24.18 + PIPELINE TARIFF $3.93 52.43 $1.56 $2.26 ‘PUMP ST. #1 TO VALDEZ) + TANKER COST $0.75 50.75 $0.75 $0.75 (VALDEZ TO NIKISKI) = REFINER’S ACQUISITION COST $15.73 $15.56 821027) $27.19 20700365 Page C-12 SS22eeesssss225S2522s=S2222522=>5=5==ssS ANS PRICE AT VALDEZ CUALITY ADJUSTMENT TANKER COST (WEST COOK INLET TO NIKISKI) SEF INER ACQUISITION COST 20T00365 TABLE C-5.b TESORO REFINER ACQUISITION COST - COOK INLET CRUDE OIL CONSENSUS (LOW) FORECAST (1987 $/ BBL) 1987 14.98 $1.02 $0.27 $16.27 1990 14,608 50.97 50.27 515.94 2000 20.524 51.26 50,27 $22.15 2010 20.44 $1.75 $0.27 5B. Page C-13 20T00365 TABLE C-5.c 1997 ADOR FORECAST OF CI PROD (MBBL/D) 45.233 TESORO SHARE 0.37 COOK INLET WEIGHT 0.24 ANS WEIGHT 0.76 WORLD CRUDE PRICE-REFINER COST $1.15 WEIGHTS FOR TESORO CRUDES 1990 22.479 0.27 9.12 0.68 $1.25 2000 7.971 0.37 0.04 0.96 51.04 2010 1,727 0.37 2.0L 09 51.45 Page C-14 20T00365 TESORO COSTS sauaezesszss REFINER COSTS & MARGINS coox INLET CRUDE OIL ANS CRUDE OIL WEIGHTED AVERAGE ACQUISITION COST + REFINER GROSS MARGIN = REFINER RACK PRICE ($/GALLON) TRANSPORTATION COSTS ($/GALLON) + PIPELINE TARIFF (NIKISKI - ANCHORAGE) + ARR TARIFF (ANCHORAGE - FAIRBANKS) TOTAL TRANSPORT COST = WHOLESALE PRICE + DISTRIBUTION MARGIN = DELIVERED PRICE TABLE C-6 ESTIMATED HEATING OIL PRICES FOR FAIRBANKS LOW PRICE SCHOOL FORECAST (1987 $) 1987 1990 2000 $16.27 $11.64 $15.84 $15.73 $11.41 $15.36 $15.86 = $11.44 = $15.38 $8.49 $8.49 $8.49 $0,580 $0,475 = $0. 568 $0.020 $0,020 $0.020 $0.077 = $0.077 =: $0077 $0.097 = $0.097 = $0,097 $0.677 = $0.572 $0,665 0.173 0.173 0.173 90.850 890.745 90.858 2010 $17.95 $17.33 $17.44 8.49 $0.615 $0.020 $0.077 $0.097 $0,712 0.173 $0. 885 Page C-15 TABLE C-6.a TESORO REFINER ACQUISITION COST - ANS CRUDE OIL LOW PRICE SCHOOL FORECAST (1987 $/ BBL) 1987 1990 2000 2010 SSSsssessssessss=s2ssssS555=>53 WORLD CRUDE PRICE (CONSENSUS LOW) $18.56 = $14.00 518.00 $20.00 (MID-QUALITY CRUDE U.S. GULF) ~ QUALITY ADJUSTMENT $0.24 $0.20 $0.25 $0.28 ~ TANKER TARIFF 52.93 52.93 52.93 2.93 (VALDEZ TO U.S. GULF) - TAPS TARIFF 53.93 $2.43 $1.4 $2.2 (PUMP ST. $1 TO VALDEZ) ~ nDJUSTMENT $0.21 $0.21 50.21 $0.21 = NORTHSLOPE WELLHEAD PRICE 511.05 $8.23 $13.25 $14.32 + PIPELINE TARIFF $3.93 $2.43 $1. $2.26 (PUMP ST. #1 TO VALDEZ) + TANKER COST $0.75 $0.75 50.73 $0.75 (VALDEZ TO NIKISKI) = REFINER'S ACQUISITION COST $15.73 SIL. 4 815.31. 20T00365 Page C-16 TABLE C-6.b TESORO REFINER ACQUISITION COST - COOK INLET CRUDE OIL LOW PRICE SCHOOL FORECAST (1987 $/ BBL) 1987 Ssemsssassessssesssssaessssszsssszss2==s ANS PRICE AT VALDEZ 14.98 QUALITY ADJUSTMENT $1.02 TANKER COST (WEST COOK INLET TO NIKISKI) $0.27 REFINER ACQUISITION COST $16.27 20T00365 1990 10.604 50.71 $0.27 511.64 2000 14,008 1.97 $0.27 $15.94 2010 16.58 $1.10 $0.27 $17.95 Page C-17 TABLE C-6.c WEIGHTS FOR TESORO CRUDES 1987 1990 2000 2010 ADOR FORECAST OF CI PROD (MBBL/D) 45.233 23.479 7.671 1.727 TESORO SHARE 0,37 9.37 0.37 0.37 COOK INLET WEIGHT 0.24 0.12 9.04 O.0L ANS WEIGHT 0.76 0.88 0.96 0.99 WORLD CRUDE PRICE-REFINER COST $1.15 1.21 $1.27 $1.31 20T00365 Page C-18 APPENDIX D ¢ x Alaska Power Authority State of Alaska Date: June 17, 1988 To: Distribution fy From: Richard Emerman be v Senior Economist Alaska Power Authority Subject: Review of Draft Report -- Long-Term Price Outlook for Cook Inlet Natural Gas and Fairbanks Fuel Oi] The Alaska Power Authority is presently studying the feasibility of certain electrical transmission projects in the Railbelt, including improvement of the connection between Anchorage and Fairbanks and between Anchorage and the Kenai Peninsula. Our policy is to distribute draft reports prepared in conjunction with such studies to affected utilities, local governments, state and federal agencies, public land managers, and other interested parties for review and comment. Enclosed is a draft report, prepared for this study by ICF Incorporated, presenting long-term crude oil price scenarios plus price forecast information for Cook Inlet natural gas and Fairbanks fuel oil keyed to each scenario. The report does not include a judgment about which crude oil scenario should be considered most likely. For purposes of Power Authority studies, that judgment remains the responsibility of our Board of Directors. You are invited to review the report and send us your comments at the following address: Alaska Power Authority Attn: Richard Emerman P.O. Box 190869 Anchorage, Alaska 99519-0869 All written comments received by the Power Authority prior to July 20, 1988, will be included within an appendix to the final report. Enclosure as stated. RE:aa PO. Box AM Juneau, Alaska 99814 (907) 465-3575 "$0897 SEBEL 704 East Tudor Road Ancnorage. Alaska 99519-0869 (907) 561-7877 01 A34LH STATE OF ALASKA /~---~ OFFICE OF THE GOVERNOR P.0. 80x AD JUNEAU, ALASKA 99811-0199 OFFICE OF MANAGEMENT AND BUDGET — DIVISION OF POLICY July 20, 1988 Alaska Power Authority Attn: Richard Emerman, Senior Economist P.O. Box 190869 Anchorage, Alaska 99519-0869 Dear Dick: Thank you for the opportunity to comment on ICF's draft report on crude oil, natural gas and fuel oil price scenarios. The value of providing comments on the report at this point is not entirely clear, given your statement to the APA Board of Directors in May that it was too late in the study process to change the crude oil price assumptions. Perhaps the natural gas and fuel oil price scenarios are not similarly “locked in” yet, although they appear to be based mainly on the crude oil price assumptions. In any event, the following comments are provided for the record. With respect to the crude oil price scenarios, | view the "Low Price” scenario to be excessively optimistic — or insufficiently cautious — as a "downside" case. Two other members of the Division of Policy who reviewed these price scenarios share this view. Because these price scenarios are to be used as a basis for major investment decisions by the State of Alaska, the low-range oil price assumptions should reflect a relatively low probability of occurrence. We believe that the probability of oil prices being lower than those presented as the "Low Price” scenario is too high for these prices to be prudently used as a low-case scenario. In addition, the crude oil scenario report is inconsistent in its description of the "Low Price" school of thought. On page ES-1, the report states that forecasters of this school "believe that real oil prices will rise very little, if at all, over the 1990-2010 period.” Yet the "Low Price” forecast projects a 43 percent real increase in oil prices over the period. While this is considerably less than the 100 percent increase projected in the "Consensus" forecast, a 43 percent increase does not match the "Low Price” description of real prices increasing very little, if at all. The analyses of natural gas and fuel oil price scenarios are generally well-done, given the scope of the study. However, because the price forecasts for these fuels are based mainly on the world oil price scenarios, we question the validity and prudency of the "Low Price” natural gas and fuel oil forecasts, for the same reasons cited above. Richard Emerman July 14, 1988 Page 2 With reference to the second report titled "Fuel Price Outlook for the Alaska Railbelt Region: Oi] & Natural Gas", it is one of the more comprehensive treatments of the subject. As a general comment, I believe that the Executive Summary could be shortened and made more concise leaving the discussion & analysis to the appropriate chapters. On a minor note, on page 2-8, first paragraph, reference is made to domestic or Cook Inlet consumption. Is this a valid definition in view of gas generated electricity sales to Fairbanks? Also with respect to the alternative fuel for the electric utility there is insufficient discussion of why the choice is not coal or hydro. On pages 4-5 and 4-6, figures 4.2 and 4.2, the legend should be the same for both graphs in order to preclude confusion. On page 4-8, table 4.4 add a footnote to indicate that one MMBTU is approximately equal to one MCF. On page C-5, table C-2, I believe that the table is for the consensus (low) forecast rather than consensus only as is table C-1. Please let me know if there are any questions. KD: bsp 1) 2) 3) 4) 5) 6) 7) 8) REPLY FROM ICF Comment noted regarding "Consensus (Low)" scenario. As now stated in the report, this scenario is simply intended to represent an intermediate outcome that could emerge if reality falls somewhere between the "Low Price" and "Consensus" expectations. There is nothing more than that to the derivation of the "Consensus (Low)" scenario. The comment regarding the assumptions underlying the Low Price School forecasts is noted. ICF has attempted to represent, as best we understand, the rationale for the Low Price School forecasts. Our report makes similar criticisms of the rationale. Comment noted on Executive Summary, but upon review we have declined to shorten it. On page 2-8, "Cook Inlet" consumption has been changed to "Railbelt" consumption. The overall forecast of gas use by electric utilities published by ADNR has not, however, been adjusted. There are a number of factors that could affect that forecast, but the impact of those changes on the overall gas market analysis would be quite small. On page 4-9, the discussion of alternative fuel for electric utilities has been clarified. The question of coal use in electric generation was not fully explored because it was outside the scope of this study. On pages 4-5 and 4-6, the legends have now been made the same. On page 4-8, footnote added as suggested. Correction made on page C-5. 20W00028 REPLY FROM ICF ON REPRESENTATION OF THE SCHOOLS OF THOUGHT The purpose of the world oil price portion of ICF’s contract was to identify alternative schools of thought on future oil prices. One of these alternatives was the "Low Price" school of thought. The key problem in presenting this school of thought was to define the forecasters who comprise it. One obvious member is Arlon Tussing Associates, Inc., who is on record stating that the future oil price will be in the $10-20 per barrel range ($10.29 to $20.58 per barrel in 1987 dollars per barrel). The average ARTA price is $15.44 per barrel. ICF decided that a school of thought should be based on more than one proponent, however, so we included in the Low Price School the three lowest price forecasts collected, i.e., ARTA, the Alaska Department of Revenue, and the World Bank, as shown in Table 1. Except for ARTA, the other Low Price proponents forecast rising prices over time. For this reason, the Low Price School forecast developed for the study rises over time. TABLE 1 Components of the Low Price School of Thought (1987 dollar per barrel) 1988 2000 2010 Arlon Tussing 15.44 15.44 15.44 Alaska Dept of Revenue* 15.18 17.81 -- World Bank 14.42 20.95 -- Low Price School $14.00 $18.00 $20.00 * 9/87 Forecast An issue has been raised as to whether the Low Price of though used in the study is low enough. Obviously, the future oil price level is a controversial subject, and the future level is unknown. Conceivably the future price could be below the Low Price forecast. Nevertheless, ICF believes that the low price forecast developed is representative of the Low Price school of thought. An issue has also been raised as to whether the Low price forecast used in the study is consistent with a belief that real oil prices will rise very little, if at all over the 1987-2010 period. The 1987 oil price was about $17 per barrel. The 2010 Low Price forecast is $20 per barrel. ICF believes that a $3 per barrel price increase over 23 years is an extremely small increase in the context of recent history. 20T00454 Page 1 PO. Box V State Capitol Juneau, Alaska 99811 Phone: (907) 465-3114 Senate Advisory Council MEMORANDUM TO: Richard Emerman Senior Economist Alaska Power Authority FROM: Kurt S. Dzinich Senior Advisor Senate Advisory Cduncil DATE: July 14, 1988 SUBJECT: Review of Draft Report -- Long-Term Price Outlook for Cook Inlet Natural Gas and Fairbanks Fuel Oi] After reviewing the final draft of “Outlooks for World Oi] Prices: Analysis of Alternative Schools of Thought" by the ICF - Lewin Energy Group - dated June 1988, following comments are offered. Overall, this report represents one of the better attempts at coming to grips with the problems of forecasting future oi] prices. The report is brief and to the point. Rather than quibble with semantics I would like to discuss a couple of areas of concern. With respect to the three crude oil price forecasts (table ES-2, etc), while there is good development of the rationale leading to the "consensus" and "low price school" forecasts, the case for the "consensus (low)" scenario is almost non-existent. I would therefore recommend that the later case be expanded in order to allow readers to ascertain how the result was obtained. The second area concerns two of the assumptions supporting the "low price school" forecast. On the energy efficiency gains it appears relatively obvious that there is an inverse relationship between the amount of efficiency or conservation gains and the cost of energy. It therefore seems improbable that low or lower energy prices would result in increasing conservation or efficiency gains. The second assumption dealing with substitute fuel viability in the transportation sector seems to be much too optimistic. While there are alternatives based on demonstrated technology, that is a long way from widespread commercial use even at $30 oil, let alone $18 or less! Railbelt Study Comments Page 2 The repor correctly notes that the Cook Inlet natural gas market is not a typicai competitive market, and that the best indicator of future gas prices is the current contract price, which the report estimates to be in the $1.50/mmBtu range (1987 dollars). | understand that Chugach Electric is on the verge of signing contracts for major new gas supplies at a reported price of $1.35/mmBtu, delivered in 1992. Assuming five percent annual inflation, this price is approximately 35 percent lower than the $1.50 figure cited in the report. | would hope that the actual Chugach contract price can be incorporated in the fuel price analysis when it is available. Sincerely, — . ck Kreinheder Senior Analyst cc: Mary Halloran Reply from APA on Overall Approach The approach to long-term oil price forecasts used here was to develop a quantitative representation of the major schools of thought based on a survey of individuals, firms, and government agencies recognized for their work on the subject, and to discuss and contrast the reasoning and evidence put forward in support of each. This discussion was intended to assist the Power Authority Board of Directors in assigning probability estimates to each of the identified "scenarios" or schools of thought. Rather than rely exclusively on the collective judgments expressed among oil price forecasters around the world or on the particular views found among Power Authority Board members, this approach was designed as a compromise that would enable the Board to impose its imprint within a well documented framework. Once the crude oi] scenarios were identified early in the study process, a substantial amount of work was invested in developing long-run forecasts of natural gas and fuel oi] prices as well as Railbelt employment and population forecasts consistent with each crude oil scenario. While the probability estimates were open to debate in May, changing the crude oi] scenarios themselves would have required additional time and money to redo that associated work. This was not anticipated in the study schedule or budget. Neither the "Low Price" nor the "Consensus" scenario was intended by the Power Authority to be a boundary case. Sensitivity testing will be performed for any project or program evaluated within the current Railbelt studies that appears to be economically feasible. Such testing will include evaluation at fuel prices and load forecasts that are judged to be closer to boundary levels. The system analysis for the Railbelt studies will incorporate natural gas price forecasts consistent with negotiated contracts to the extent possible, to be supplemented by forecasts based on the ICF work to the extent necessary. If the Chugach negotiations are not complete when the system analysis is performed, it is intended that the new contract recently negotiated by Enstar and Marathon will be used as a proxy. Please note that neither of these contract negotiations were complete at the time the ICF work was commissioned, nor did the Power Authority have any assurance that they would be complete in time to incorporate in the current Railbelt studies. 3608/870(1)