Loading...
The URL can be used to link to this page
Your browser does not support the video tag.
Home
My WebLink
About
Electric Power Generation For the Alaska Railbelt Region 1984
ELECTRIC POWER GENERATION FOR THE ALASKA RAILBELT REGION PART 1 Power Generation Alternative Selection PART 2 Financing Plan JANUARY 1984 Prepared For The Susitna Finance Committee by William Kent and Company es ary. i re KENTCO The following report presents the results of a six month study by William Kent and Company to examine the power genera- tion alternatives for the Alaska Railbelt and, most importantly, to construct a plan of finance. ELECTRIC POWER GENERATION FOR THE ALASKA RAILBELT REGION EXECUTIVE SUMMARY INTRODUCTION PART I - SELECTION OF POWER GENERATION ALTERNATIVE Introduction Demand for Railbelt Power Current Supply of Railbelt Power Power Generation Alternatives for the Future Introduction The Natural Gas Alternative The Coal Alternative The Hydroelectric Alternative Economic Comparison of the Alternatives KENTCO Page vr —=" 10 - 143 10- 11 1-75 37 - 48 49 - 143 49 - 59 60 - 92 93 - 110 a 123 - 143 PART II - PLAN FOR FINANCING POWER GENERATION INVESTMENT 144 APPENDIX BIBLIOGRAPHY - 221 KENTCO ELECTRIC POWER GENERATION FOR THE ALASKA RAILBELT REGION EXECUTIVE SUMMARY KENTCO EXECUTIVE SUMMARY This study has been designed to determine the most attractive and viable means of financing the lowest cost alter- native for supplying electric power to the Alaska Railbelt. Motivation for this work stems from the need to agree in the near term on a plan to meet future Railbelt power needs and to implement that plan. Cook Inlet natural gas, which fuels 70 percent of the Railbelt's power generation, continues to be depleted. Proven reserves are not expected to last through the end of this century. Current low cost contracts with the utilities will expire about 1995. Additional supplies from outside the Railbelt or from new discoveries are expected to be higher priced. If gas is to continue to be used, then gas generation, with this higher cost fuel, will have to prove to be more economical as a future power source than either coal or hydroelectric power. The time required to plan and construct an alternate means of generation for the Railbelt can be as long as 10 years if the hydroelectric option is chosen. Since that is about the time remaining until most of the power from the proposed Watana Dam on the Susitna River is needed, there is a real sense of urgency to reach a decision. Numerous professional consulting firms have done complex power supply studies covering the next 60 years. We have relied extensively on data from this work, particularly as it pertained to costs. In the engineering studies, a basic KENTCO question being answered was how best to supply the power needs of the Railbelt almost half way through the next century. In almost all cases the conclusion was that the Susitna Project was the lowest cost option. In the context of a 60 year power planning study it was appropriate for the consultants to look at the Susitna Project in its entirety. However, for the purpose of our work, it is essential to look at Susitna as two distinctly separate projects, Watana and Devil's Canyon. They are, in fact, separate, as one dam can be brought on-line without having to build the other. Since the Watana Dam has been slated to be built first and since its output just about evenly matches the projected Rail- belt demand for the mid-1990's, our first goal was to determine if that dam, by itself, was the most cost-effective means of supplying a specific amount of power equal to the output of the dam. We devised a basis of cost comparison for the coal, gas and hydroelectric options using a present value analysis. Since many costs are common to all three, we compared only the major variables which were capital costs, fuel costs and operations and maintenance costs. We concluded that the Watana Dam can produce 3500 gigawatts of power annually for its projected 50 year life at a price well below that of using either coal or gas. The second dam, Devil's Canyon, is estimated to cost only half as much as the first, due to much of the infrastructure costs having been absorbed. Its output is about the same as Watana and, therefore, the financial viability of Devil's EE KENTCO Canyon is almost assured. The only question concerns the future demand for its power. However, no decision to proceed with Devil's Canyon need be made for some years. In arriving at the above findings the following points make up integral parts of the conclusions: Demand for power in the Railbelt will continue to grow and will reach about 4000 gigawatts per year by the mid-1990's. This is sufficient to absorb the output of the largest power generating alternative, namely hydroelectricity. The future of petroleum prices will have a strong influence on the cost of generating power using thermal methods or on the opportunity cost of generating power with hydro. These same price trends will dictate the » levels of state revenues. Petroleum Industry Research Associates (PIRA), under contract to KENTCO, has estimated oil prices, in 1983 dollars, to decline slightly for the rest of this decade and then climb at 2 percent per year through the 1990's. The prices of coal and gas, after present low-priced contracts expire, will fall between production costs as a minimum and world prices for similar commodities as a maximum. In the financing section of this report is laid out a basis for structuring a financing program, a recommended plan of finance and several alternatives. III KENTCO Projected revenues from Susitna needed to be developed first. Available revenues will be tied to the power purchaser's ability and willingness to pay. In general, power purchasers are only willing to pay for the lowest cost alternative, and Susitna has been demonstrated to be the lowest cost alternative, long term. However, in the early years of the Watana project operation other alternatives are lower cost, and experience has shown that there are practical constraints to the willingness of purchasers to pay a higher near term cost. Therefore, rates to be charged to Susitna had to be both the lowest rate overall, and in addition, could not at any time diverge markedly from a short term cheaper alternative, in this case natural gas. The study used as its basis for revenues an initial rate for wholesale power of 5.9 cents, in 1983 inflation-adjusted dollars, beginning in 1996. That rate is estimated to be one cent higher (in 1983 dollars) than the projected 1996 cost of electric generation from gas at that time. In the long run the price for power from Susitna becomes much less than the cost of any of the alternatives and in 1983 equivalent dollars approaches only one cent per kilowatt hour. Given the revenues available, Watana could not be financed with 100 percent debt since the revenues forecast in the early years were less than the cost of standard debt service under today's financial conditions. Devil's Canyon would not present that problem and could be financed without equity money. The financing of Watana would require some equity investment assistance by the state, either as equity contributions to project construction costs or operating cost assistance. In IV KENTCO analyzing the possible ability of the State to participate in the Susitna Project, the report used its PIRA oil price forecast to project the amount of State revenues available for capital, programs, loans and new operating programs. This total averages $900 million per year for the next 10 years, suggesting an ability to provide some equity and operating assistance. We assessed the most likely forms of debt to be used in conjunction with State assistance. The two most advantageous were federal REA-guaranteed debt and tax-exempt, or municipal debt. Of the two, REA-guaranteed debt was preferred as it required less total State dollars paid and allowed a longer time period for State appropriations to be made. REA-guaranteed debt is available through a federal program for loans to REA generation and transmission cooperatives. Therefore, implicit in the use of this type of debt is the need to form a generation and transmission cooperative in the Railbelt. The report presents a preferred plan of financing which uses debt, State equity and State operating assistance. Of the debt, three-quarters is REA-guaranteed and one-quarter is municipal or tax-exempt. The plan requires $800 million (in 1983 dollars) of State equity investment at the time of con- struction and $778 million (in 1983 dollars) of State operating assistance spread over a number of years after construction is completed. We also show a possible variation in structuring the REA debt so that the total amount of State assistance is reduced further. We have presented an alternative plan of financing in which all debt is municipal debt. This requires a large amount Vv KENTCO of State equity at the time of construction and may require federal tax law changes. Following the above, a straightforward plan of finance for Devil's Canyon is presented which uses all debt and no State assistance. The plan of finance indicates the return on the State's investment. The return begins at a low level approximately 10 years after the project commences operating and becomes very substantial after the project debt is retired. This return resembles the idea propounded in the State of Alaska for a Capital Investment Fund. Some of the key decisions regarding Susitna and its financ- ing are policy decisions beyond the scope of an outside consultant. Development of a comprehensive State energy policy seems essential to the making of these decisions. This executive summary highlights some of the more important points from this report. However, we caution the reader to look to the body of the report for the underlying logic, details and recommendations. The subject is complex and any attempt to use this brief summary as a stand-alone report would be misleading. vI KENTCO ELECTRIC POWER GENERATION FOR THE ALASKA RAILBELT REGION INTRODUCTION KENTCO INTRODUCTION Goals of the Study This study has been designed to achieve two primary goals. First, to examine in an objective manner the conclusions previously drawn concerning the most cost-effective method for fulfilling the electric power generation needs of the Railbelt area of the State of Alaska for the next sixty years. Second, and most importantly, if that method is to build the Susitna Project, then to recommend the optimal strategy for financing the project. Scope of the Study The study concentrated on three major fuel alternatives as being capable of supplying large amounts of power with technology available today; namely, gas, coal and hydroelectric generation. Geographically, the study was concerned with the Railbelt area. (See Exhibit 1) The Railbelt contains 70 percent of the State's population and produces more than 70 percent of the State's power. The timeframe of our investigation spanned the period 1960 to 2045. The past 23 years were examined for historical per- spective and empirical data which could throw light on the present and future. The period until 2045 was used to analyze recommendations regarding power generation alternatives. It was selected to coincide with the minimum life of the longest generation option, namely hydroelectric. a KENTCO Exhibit 1 Railbelt Area of Alaska Showing Electrical Load Centers FAIRBANKS-TANANA VALLEY ELTA © GLENNALLEN PESO f. on . qa. ‘i rx CER REN) yt = 0 LOVE T KORG ASS Sens ZOOM Sepa PP LOI Bay LAE eS, ea KO “yy ) SASL K PSL w NBs Ai: hte eon < JERE AS Pes NeP song LTeres A UY. QR x S OF BSOS £9 x ts RRS eH VSO (OOS Lote REE a ROSEN ~w ti Ee 23 AS ‘ f2 9 ROHS az KS 2x Ry ° go a? i KENTCO The analysis in this report is concerned with the economic costs of the alternatives. It has not attempted to deal with issues concerning engineering, the environment or technical matters. Similarly, this report does not treat the energy issue on a Statewide basis. Therefore, State energy policies cannot be inferred from our recommendations. We strongly believe, however, that it is necessary and desireable for the State to define its long range energy policies. Motivation for this Study The motivation for this study comes from a need to make a decision soon on how electric power will be supplied to the Railbelt. The pressing need to plan stems from a number of factors. These include: e continuing growth in the population of the Railbelt e the probability that natural gas, the fuel currently used to generate electricity, will become unavailable or significantly higher priced before the end of the century e long planning and installation lead times which accom- pany alternative generation methods, especially the hydroelectric option e the fact that declining State revenues may require investment funds to be marshalled over a longer period of time than previously had been the case Numerous studies have been conducted concerning most as- pects of power generation in the Railbelt and vast amounts of data collected. Most of this work has been performed by excellent professional organizations. It was not the intent to have this study duplicate that work nor to challenge it. KENTCO Rather the study has drawn extensively on these reports for both technical and non-technical information. It was, however, the intent to utilize the existing data to answer several new questions. The study isolated the first decision, whether or not to build the Watana Project, and set out to answer that question. We distilled the necessary information from existing sources, developed an easy to handle formula that allowed comparison between the possible generation methods and reached a conclusion. It is important to understand that this is not a complete power planning study, but does draw from those already in existence. Rather it is a decision-oriented study. It was the perception of the Energy Committee of the Anchorage Chamber of Commerce that information needed to be consolidated and a clear answer developed to the questions con- cerning power generation in the Railbelt. Studies conducted to date have indicated that the Susitna Project is most suitable to the power production needs of the Railbelt. The Chamber was asking the next two questions; namely, is Susitna affordable and how can it be financed without burdening unduly the state. The Energy Committee of the Chamber incorporated a new corporation, Susitna Finance Committee, Inc., with a broad- based board of directors who had a common desire for answers to the questions posed in the preceding paragraph. In June 1983, the committee hired William Kent and Company to conduct a study aimed at answering these questions. Approach to the Study The steps followed in conducting this study have been as follows: erate KENTCO e review and confirmation of the facts surrounding the subject of supply and demand of power in the Railbelt today as well as for a period of time into the future equal to the life of the longest lived power supply option, hydroelectricity e identification of reasonable solutions (alternatives) e review, study and quantification of the principal costs associated with each alternative e@ projection of the costs of each alternative under a model of key cost components, demand and time period e identification of a best solution e review for financial feasibility e development of a finance plan Resources Used by Kentco Basic published sources provided Kentco with background and technical information necessary to establish a starting point. Some of the most important reference works allowing us to do this were: The Acres American engineering studies; the Harza-Ebasco update of September 1983; the Battelle Pacific Northwest Labora- tories ''Railbelt Alternatives Study"; The Alaska Power Authority submittal to the Federal Energy Regulatory Commission; and the Burns and McDonnell "Study of Alternate Power Supply Organiza- tional Structures". We have interviewed over one hundred people representing all sides of the power generation and energy issues. Within the State these interviews have included managers and directors of the Railbelt utilities, the senior staff of the Alaska Power Authority, state officials, federal representatives, researchers at the Institute of Social Research, representatives of the oil, gas and coal industries as well as members of the private sector, KENTCO the trade unions and many informed citizens. Outside the State of Alaska we have interviewed staff officials in Washington with the Senate Energy Committee, senior members of the Rural Electric Association and staffs of Senators Stevens and Murkowski. In New York, Paris and London we have had extensive discussions with American and international invest- ment bankers and utility financing experts. These discussions led us to discussions with Washington Public Power Supply System (WPPSS) officials and a look at what happened to the Bonneville Power Administration as well as investigations into the Canadian energy program and its goals. We have made use of outside experts when such opinions carried sufficient importance to the goals of the study. In particular, we have contracted with Petroleum Industry Research Associates, the internationally renowned oil and gas forecasting center, to give us their forecast for the world price of oil and gas. We have contracted with the Greenley Energy Corporation, an internationally recognized coal group, as well as Korean and Japanese experts, to provide us with a forecast for the price of coal. In addition to the above, we used numerous outside publica- tions which are listed under the Resource section. Nature of the Results We have endeavored at all times to guide the reader through each step of our analysis and conclusions in a factual, logical and clear manner. While the issue of energy and how to finance its generation in the Railbelt is complex, the results of our KENTCO analysis point toward some clear and understandable conclusions. First we have confirmed which of the available options for producing power in the Railbelt is the lowest cost alternative. Second, the model used to determine the lowest cost alter- native has been constructed in order to allow the reader to provide different inputs, such as power demand, fuel costs, etc., should he so desire, and, by working these through to the end, see what results would be achieved. Since this study concerns programs for the future and since some of the alter- natives have long lead times, we consider it essential that the model be useable at any future time when a decision is to be verified or changed. Third, and most important, the report develops a financial analysis of the selected option and a plan for financing the capital investment it requires. KENTCO GLOSSARY OF TERMS USED IN THIS REPORT Watt (w) = a measure of power Kilowatt (Kw) - 1,000 watts Kilowatt Hour (Kwh) - 1 kilowatt used for 1 hour Megawatt (Mw) = 1 million watts Megawatt Hour (Mwh) = 1 megawatt used for 1 hour Gigawatt = 1 billion watts Gigawatt Hour (Gwh) = 1 gigawatt used for 1 hour British Thermal Unit (Btu) = A measure of heat _MBtu = 1 thousand Btu's MMBtu = 1 million Btu's MCF = 1 thousand cubic feet BCF = 1 billion cubic feet TCF = 1 trillion cubic feet Peak Demand = the greatest amount of power needed at any one moment (usally expressed in megawatts) Annual Demand = total power usedin one year (usually expressed in gigawatt hours) Barrel of Oil = equal to 40 gallons MMB/D = million of barrels per day GNP = gross national product - total of all goods and services produced in one year by a nation Nominal Price = price that is expressed in current dollars. Current dollars reflect the effects of inflation over time. Nominal Dollars = see nominal price Real Price = price that is expressed in constant dollars. Constant dollars do not reflect the effects of inflation. Real Dollars = same as real price KENTCO F.0O.B. = Free on Board - generally price where goods leave producer/manufacturer CepLi aris Cargo insurance and freight, - generally price of goods delivered to customers port NCW = Non-communist world OECD = Organization for Economic and Cooperative Development LNG = liquefied natural gas TPY = tons per year MMT = millions of metric tons (2,200 lbs) KENTCO ELECTRIC POWER GENERATION FOR THE ALASKA RAILBELT REGION PART | SELECTION OF A POWER GENERATION ALTERNATIVE KENTCO INTRODUCTION A basic function of any state government is to help ensure that a sufficient supply of power exists at all times. Whether the government works through regulatory agencies to oversee privately-owned utilities or cooperatives, or whether the state plays a direct role in producing power, it still shares respon- sibility to see that power is available. Due to the long lead times connected with supplying power, the need for long range planning is essential. By concentrating on this task, the state will: e ensure sufficient power to meet demand e allow power needs to be met in the most economical manner e ensure that whatever programs are developed for the future are affordable In the case of Alaska, the State legislature and the admin- istration through the Alaska Power Authority and the Alaska Public Utility Commission have held the public portion of respon- sibility for meeting the growing demand for power. To date, demand in the Railbelt has been met without a great deal of forward planning due to the availability of sufficient natural gas in Cook Inlet to fuel all needs, and to the utilization of gas turbines that can be added in modest increments without long lead times. In recent years, longer planning horizons have become more essential and these can be expected to lengthen. The base of electricity users, both residential and commercial, continues - 10 - KENTCO to grow. The yearly increments continue to be larger and larger, necessitating greater additions of power for the same percentage growth rates. Moreover, the latest available infor- mation indicates that the proven gas in Cook Inlet will run out before the end of the century and new gas, if available, is likely to be much more highly priced. Clearly this brings into question the wisdom of adding gas-fired turbines to continue to supply the growing demand for power. In order to answer questions posed by this situation, we have taken three steps. These are to estimate the demand for power for the period of the analysis, to describe the current method and capacity for supplying electricity in the Railbelt, to compare alternative methods of supply, and to recommend that method which is most cost-effective. The sections of Part I which follow correspond to those three steps. @ Demand for Railbelt Power e Current Supply of Railbelt Power e Power Generation Alternatives for the Future -ll- KENTCO DEMAND FOR RAILBELT POWER - Historic and Current Demand (1960-1983) - The demand for power in the Railbelt since 1960 shows a steady, almost uninterrupted growth. The annual rate of growth in both total demand and peak demand has averaged 11 percent. Annual demand is the total amount of power used over the course of an entire year. Annual peak demand is the greatest amount of power used at any one moment in time during a year. For planning purposes, peak demand requirements dictate how much capacity has to be available at any time. Annual demand data provides the basis for calculating revenues, costs such as operating and fuel expenses, and the average amount of capacity required of the combined generation facilities. Between 1960 and 1982 peak demand, typically measured in terms of megawatts (Mw), rose from 60 Mw to 565 Mw. Annual demand typically measured in gigawatt hours (Gwh) went from 450 Gwh to 2937 Gwh. Because of the constantly growing base, the percentage increase in total demand has been slowing, but the trend is consistently upward. Exhibits 2 and 3 illustrate these trends. —|2=— KENTCO Exhibit 2 PEAK DEMAND 1960—1980 550 500 450 400 se Peak Demand 300 250 200 180 100 1960 1965 1970 1975 1980 Exhibit 3 ANNUAL DEMAND 1960—1980* 2.60 2.40 2.20 2.00 Annual Demand —__ 1.60 40 Gwh PER YEAR (Thousands) 1.20 0.80 0.60 0.40 4 1960 1965 1970 1975 1980 YEARS *Source: Alaskan Power Authority (APA) Federal Energy Regulatory Commission License Application July 1983 (hereafter referred to as APA-FERC 1983) ap KENTCC Exhibits 4 and 5 give year by year data for the most recent years. (See Exhibit 4 and 5). Exhibit 4 PEAK DEMAND 1976—1982* 370 560 550 540 530 520 510 500 490 480 470 460 450 Peak Demand 430 420 410 4 1976 1977 1978 1979 1980 1981 19382 Exhibit 5 ANNUAL DEMAND 1976—1982* t= YEAR sands) 1976 1977 1978 1979 1980 1981 1982 *source: APA-FERC 1983 Shae KENTCO The load centers of the Railbelt are Anchorage and Fairbanks. Anchorage accounts for over 80 percent of annual power demand in the Railbelt and Fairbanks less than 20 percent. These same ratios between Fairbanks and Anchorage apply to the individual components of annual versus peak demand. (See Exhibits 6 and 7). Exhibit 6 1982 PEAK DEMAND BY REGION” FAIRBANKS (19.0%) ANCHORAGE (81.0%) 1982 Total Peak Demand = 565 Mw = 100% Exhibit 7 1982 TOTAL DEMAND BY REGION* FAIRBANKS (16.7%) ANCHORAGE (83.3%) 1982 Total Demand - 2937 Gwh = 100Z% *source: APA-FERC 1983 Cae KENTC Anchorage has been fueling the growing need for power and is likely to continue to do so. (See Exhibits 8 and 9). Exhibit 8 PEAK DEMAND BY LOAD CENTER* 1976—1982 500 450 = 7 Anchorage 350 300 3 250 200 150 : Fairbanks 100 50 1976 1977 1978 1979 1980 1981 1982 YEARS o ANCHORAGE PEAK + FAIRBANKS PEAK Exhibit 9 ANNUAL DEMAND BY LOAD CENTER® 1976—1982 ' pYNNN dpUONDO=“bUdADOVOON=HUdU / Anchorage “Fhowwends) °. oo. °. °. Fairbanks °.. Ss oO. ———— T T T 1976 1977 1978 1979 1980 1981 1982 YEARS 5 ANCHORAGE AREA + FAIRBANKS AREA *Source: APA-FERC 1983 er Gus KENTCO The users of power in the Railbelt are divided about equally between residential and commercial with industrial and military making up the remainder. (See Exhibit 10). Exhibit 10 TOTAL DEMAND BY USER 1982 INDUSTRIAL/MIL (5.0%) HOUSES (47.5%) COMMERCE (47.5%) 1982 Total Demand - 2937 Gwh = 100% Source: APA-FERC 1983 === KENTCO The growth in power demand in the Railbelt has been a combination of an increasing population and an increase in the per capita usage of power. Between 1960 and 1980 the population of the region has doubled from 140,486 to 275,818. (See Exhibit 11). Exhibit 11 POPULATION AND ANNUAL DEMAND* 1960—1980 2.8 2.6 2.4 2.2 1.8 1.6 jousands) 1.4 1.2 Pore 00/000 DEMANDx1,.0 Gwh o.8 o.8 O.4 1960 1965 1970 1975 1980 YEARS a POPULATION > DEMAND As population has grown, so has per capita consumption of energy. (See Exhibit 12) Exhibit 12 PER CAPITA TOTAL DEMAND * 1960—1980 Per Capita Total Demand ——~ 1960 1965 1970 1975 1980 Source: APA-FERC 1983 KENTCO - Forecasting Future Demand - Forecasting Methods There are four basic ways to approach the forecasting of power demand in Alaska: e informed opinion based on experience, empirical evidence and estimations e trend lines e deferral of long term planning in favor of reacting to near term requirements e modeling techniques In the very short term, those working to meet the demand for power may,in fact,be able to provide the best assessment of near term demand through their knowledge of what is happening minute by minute and their experience that tells them what will happen a short time into the future. These horizons, however, are too short to be of use in long range planning. Trend lines, in the most recent past, have been relied upon heavily by utilities in the United States. As long as the same factors that drove demand in the past were present in the same order of magnitude in the future, it was possible to use trend lines. Unfortunately, many utility planners remained too long with this simple tool and, in recent years, have made serious, expensive errors. Deferral of planning is not meant to imply negligence. As we have discussed, it has not been necessary, nor perhaps even prudent, in the past to put into place long range power plans for Alaska. Growth has come regularly, the user base was rela- tively small, Cook Inlet gas was available at an attractive =_195— KENTCO price,and gas turbines can be put on-line in short order. On this basis, deferring any long range planning was a workable approach. Today, for the reasons stated earlier, the State no longer has the luxury of short term fixes and must engage in serious long term planning. Modeling is the only option of the four which has enough credibility to be useful as a means of arriving at a forecast of demand. Modeling has been a proven forecasting technique for many years. With the use of large computers its efficacy has increased greatly. Modeling has the ability to look at a vast number of variables and fixed elements; it has the flexibility to allow the user to test different assumptions by assigning different weightings to the inputs; and it permits evaluation of quantitative and qualitative factors in quantitative terms. The exercise of "what if" is easily handled by modeling so that the user can see the results of different variations that he might wish to enter. At any point in time, new data can be entered into the model, thereby allowing the usefulness of the model to continue into the future. As time goes on, the fine tuning of any model should increase its accuracy and usefulness. In the forefront of modeling in Alaska has been the Insti- tute for Social and Economic Research, a part of the University of Alaska. Their work is sophisticated and highly respected both within and outside of the State. ISER Credentials ISER was formed in 1962 and is a multi-disciplinary univer- sity-based research unit with a full time faculty and support - 20 - g KENTCO staff. They have undertaken economic and policy studies for almost every government agency in Alaska and are considered to be the resident authority on the Alaska economy. ISER is responsible for the development of many of the fiscal and analy- tical tools for the State government and in this role has created the power demand models used by the Alaska Power Authority. Forecasting demand by modeling has been going on in the State for some time, mostly in conjunction with the ISER work. Recently this work has provided the backbone of the demand fore- casts used by the Department of Revenue and the Alaska Power Authority. As various segments of the State face the realities of a rapidly changing economic climate and make their plans for the future, this work will become even more important. Importance of Oil Revenues and Pricing Revenue from oil, according to the models used and developed by ISER, is the single most important factor in the determination of demand for the Railbelt. As will be seen later in this report, it is also critical to the determination of the cost of generat- ing power by thermal methods. Given the importance of oil, a discussion of the future of oil prices is appropriate at this time. As the oil price increases, the amount of revenue the State of Alaska receives in the form of oil royalty revenue increases. The greater the State revenues, the more Alaska has to spend on State development in general. Such development fuels population growth, which in turn provides increased demand for power. This Osa KENTCO is simulated by the computer models previously mentioned. Alaska derives most of its State revenues from petroleum related royalties and taxes. The State receives a royalty which is generally 12% percent of the value of all oil pumped and this plus oil taxes has generated in recent years large revenues available for State spending. While in 1971, oil revenues accounted for 21 percent ($46.2 million) of the total State revenues, in 1981 they accounted for 89 percent ($3.3 billion). (See Exhibit 13) The State's wealth is tied to the price of oil on world markets. Since Alaska's producers pump oil at a constant rate of approximately 1.6 million barrels per day, the revenue to the State varies according to price. The price is tied directly to the world price which in the last year has declined about 15 percent. A drop of $1 a barrel causes a drop of. approximately $150 million in State petroleum-related revenues or 4 to 5 percent of the State's total revenues. The growth of oil revenues in the last five years has stimulated an unprecedented growth in State spending on both capital infrastructure projects and increased services. The larger revenue stream also has allowed the elimination of State resident income and sales taxes. The effect has been to increase the population of the Railbelt by attracting workers from the Lower 48 states and this, in turn, has stimulated the growth in demand for power. Since 1978 the State has made sizeable contributions to the Alaska Permanent Fund, a corporation designed to provide for the = 22. = KENTCO Exhibit 13 State of Alaska Petroleum Revenues Dollars Co) ACTUAL &Z2 ESTIMATE 188 78 Pp E R E N $8 i A G = =95 8 1975 1988 1985 FISCAL YEARS Fiscal Total General Fund 1 Total General Fund 1 State Oil Revenues Year Unrestricted Revenues Unrestricted Petroleum Revenues as Percent of Total 1971 220.4 46.2 Ze 1972 219.2 47.1 21 1973 208.1 49.3 24 1974 255k 71953) 31 1975 383505 87.6 26 1976 709.7 386.1 54 1977 874.1 472.5 54 1978 787.4 430.3 55 1979 2,278.5 819.0 69 1980 2,632.6 Zp22oeo 86 1981 377182 3,299.4 89 1982 4,108.4 3,574.7 87 1983* 3,630.5 3,054.5 84 1984* 3,071.2 2,666.3 87 1985* 3,165.9 nilsoal 88 1. After Permanent Fund deductions bd Estimate Source: Alaska Department of Revenue - 23 - _ KENTCO future of Alaska as the oil runs out. This money is saved and is not a direct factor in creating population growth or demand. The Institute of Social and Economic Research shows a 1-2 year lag between State expenditures and increases in population, with the resultant delay in higher demand for power. The 1983 power demand growth will be a record 12 percent, reflecting in part the high level of State spending of the previous two years. Since the increase in State revenues available for spending (total less Permanent Fund) have continued to rise through 1983, power demand likely will continue to rise through 1985 due to the lag effect. The Department of Revenue (DOR) projects a probable decline in State revenues (net of Permanent Fund) in 1984 and then a slow increase for a number of years thereafter. DOR projects oil production to peak at 1.8 million barrels a day (MMBD) in 1987 and then to decline steadily. Increases in the price of oil will offset the decline in part, but in real terms the total contribution to the State from oil revenues will decline over time. The exact amount of that decline will depend on the magnitude of price increases, the success of future exploration and the location of future production. World Oil Price Forecast William Kent and Company has contracted with Petroleum Industry Research Associates, Inc. €PIRA) to prepare a long term outlook for world energy prices. As stated previously, the forecast of oil prices is very central to energy planning deci- silly ees : KENTCO sions. Although significant work in oil forecasting has been done for the State, Kentco wanted the latest analysis possible from a most credible and independent source. PIRA is considered by the oil industry to be one of the leading oil forecasting firms in the world. John Lichtblau, founder and director of the firm, is frequently cited for his expert opinion by the Wall Street Journal, the New York Times and national news weeklies. PIRA has seventy retainer clients, in- cluding many of the major oil companies, refiners, gas transmission companies, chemical companies, financial institutions, and the governments of oil producing countries. Lichtblau is a member of the National Petroleum Council, Council of Foreign Relations and Chairman of the Crude Oil Advisory Council of the New York Mer- cantile Exchange. We believe PIRA's work is the best available for forecasting oil prices and we will use their forecast throughout our analysis. The complete PIRA report is presented in Appendix 1. General conclusions from the report are as follows: e OPEC will continue to administer oil prices worldwide e Economic growth rates worldwide will be as follows: SD Se * KENTCO Real Growth Rates of the World Economy (Percent per year) 1973-1981 1981-1990 1990-2000 2.8 3.0 3.0 energy demand in the non-communist world (NCW) will grow at a rate significantly less than total economic output nuclear energy in the NCW will grow rapidly (11 percent a year) through 1990 and then at half that rate until 2000 coal synfuels will not play a significant role in supply- ing world energy needs natural gas demand will decline 10 percent or more through the year 2000 oil's role in world markets will diminish from 48 percent of the total energy supply to 38 percent of total by year 2000 in the NCW price pressure on oil, in real terms, will be downward through 1990 and then upward PIRA Forecast World Oil Prices*, 1983-2000 (Dollars per barrel) Nominal Price* Real Price (1983 $'s)* 1983 $29.00 $29.00 1990 38.25 26.10 2000 75.25 31.80 (See Exhibit 14) OPEC marker crude price, Saudi Arabian Light, F.0.B. Ras Tanura - 26 - KENTCO Exhibit 14 PIRA OIL PRICE FORECAST 88 1985—2030(CONSTANT 1983 DOLLARS) 56 54 52 50 48 46 44 42 40 ss ss 34+ 32 30 285 26 1985 1990 1995 2000 2005 2010 2015 2020 2025 2030 YEARS $S/bbI Other Oil Price Forecasts The power demand forecasting that has been done by the Alaska Power Authority, using the ISER models, has employed a variety of oil price forecasts. The lowest forecast has come from the Alaska Department of Revenue which projected an average or DOR mean case, and a statistical median case called DOR-50. This latter case is the lower forecast of the two. Higher forecasts have been presented by Sherman Clark Associates, Data Resources, Inc. and the U.S. Department of Energy. As will be seen from Exhibit 15, the PIRA forecast is above the more conservative DOR estimates but well below all the others. Non-Oil Demand Elements There are a number of key ingredients that belong in a demand forecast other than the price of oil. None of these carry nearly Source: PIRA =| 27i\\2 Exhibit 15 ING/$ E86l- WO 4O 30d GTYOM KENTCO DOR 50% 2000 1990 1983 “ eS aes oo - lJ © = w e = gto ‘ace ca Lu Ww are te a. FERC 1983 PIRA APA- Source: KENTCO the same weight as oil. Population changes in Alaska have a strong effect on demand, but as noted above they are influenced by oil price and oil production. Employment opportunities encourage or discourage changes in population. Rising employment also has meant increased business activity from which has come a growing per capita consumption of power. State revenues pro- vide a stream of funds into the economy that determines the levels of economic activity which in turn influences the level of employment and thus population. Tourism puts seasonal demands on the supply of power as well as providing a growing part of the State's GNP. Cost of living becomes important for its effect on overall economic activity. Per capita use and conservation trends also affect total demand. There also are many less important elements to a demand model but none are significant enough to detail here. Several of the components of demand had been modeled and projected for specific purposes. These, in turn, are useful to provide input to a later model constructed for the express purpose of forecasting power demand in the Railbelt. The demand fore- casting used by the APA and its engineering consultants uses several models. The first model is the PETREV model (Petroleum Revenue Forecasting Model) designed by the Alaska Department of Revenue in order to forecast State oil revenues based on oil price fore- casts. This model has been used for budgeting purposes and is constantly being refined. It is not a complicated model and aOR KENTCO has had a very acceptable record for accuracy. The second in a chain of three models that determines power demand is the MAP model (Man in the Arctic). Developed in the 1970's by ISER under a grant from the National Science Foundation, this model too has had considerable update work performed and is presently believed to be an excellent working tool by ISER and the governmental groups who use it. It is far and away the most sophisticated forecasting technique available in Alaska. The MAP model has been used to project population growth and many forms of socio-economic activity. The MAP model simulates the state economy and projects eco- nomic and demographic factors based on quantitative relationships between elements of the Alaskan economy such as: e employment in basic industries e employment in non-basic industries e government employment .@ military employment e State revenue and spending e wages and salaries e gross State product e consumer price index e@ population e tourism The third model that completes the chain (see Exhibit 16) and delivers the forecasts of electric power demand in the Railbelt area is the Railbelt Electricity Demand model (RED). The RED model = 30F— Te Source: WORLD OIL PRICE FORECAST APA-FERC 1983 Exhibit 16 ALASKA POWER AUTHORITY DEMAND MODELS MAP MODEL |POPULATION] RED MODEL E reo a Soa ECONOMIC LOAD REVENUE FORECASTS : FORECAST PETROLEUM HOUSEHOLDS| FORECAST REVENUES ANNUAL OOLNAy KENTCO uses demographic and other data to project both total electric energy needs and peak load demand in the Anchorage/Cook Inlet and Fairbanks/Tanana Valley load centers for the period 1980-2010. This model was developed originally by ISER in May 1980. It was modified and expanded by Battelle Pacific Northwest Labora- tories in 1982 and has been further updated and improved in 1983. The latest update includes a validation of the model performance. . The RED model uses the output data from the MAP model, which in turn started with the PETREV conclusions and uses them in con- junction with: e housing demand e appliance use e fuel prices e commercial consumption of electricity The end product is a forecast of electric power demand in the Railbelt. The forecast also checks demand with representative costs of power indicated by that level of demand. After investi- gation, it is our conclusion that the models utilized have incorporated all the factors we consider critical in determining demand. Moreover, we have confidence in their ability to forecast demand accurately and have used them for our analysis. KENTCO - Railbelt Demand Projections - In the APA Susitna license submittal, the ISER MAP model of population growth and socio-economic activity was used as input to the RED model to forecast power demand in the Railbelt. Sixteen of the seventeen variables considered were held constant. One variable, oil revenues to the state, was varied in accordance with different outputs of the PETREV model. These outputs, in turn, were derived by using various oil price forecasts. A comparison of three oil forecasts is as follows: Oil Forecasts $/Barrel Dollars 1985 1990 1995 2000 2010 2020 2030 DOR-Mean 26.45 27.42 29.09 31.33 36.36 42.20 48.98 PIRA 27.10 26.10 28.81 31.80 38.76 47.24 27599 SCHA 26.30 27.90 32.34 37.50 50.39 64.48 74.84 Note: DOR forecasts Saudi medium prices DOR Mean price has been adjusted to reflect equivalent Saudi light price It will be recalled that DOR-Mean is a relatively conservative forecast and Sherman Clark's projection is higher. The study Kentco commissioned done by PIRA falls in between. We elected to base our analysis on the results of the PIRA forecast. By using the PIRA forecast of oil prices, we have derived the following forecast for population, annual power demand and peak demand to the year 2000. (See Exhibit 17) - 33 - KENTCO 1990 1994 2000 Population 354,831 337,5203 434,254 Annual Energy Demand (Gwh) 3332 3705 4356 Annual Growth Rate (Percent) = 2 ----------- 2.7------------- Peak Demand (Mw) 693 Titel 906 Exhibit 17 ANNUAL POWER DEMAND FORECAST USING PIRA OIL FORECAST Annual Demand —__ 1990 2000 2010 2020 2030 It is important to understand that the demand modeling done by APA, using ISER models, contains one major assumption. Namely, that State spending will keep growth at the levels set by the model. The resources potentially available to the State to do this are: e Permanent Fund income e taxation e new oil finds Source: APA-FERC 1983 NONICYU e oil royalties e oil price increases e royalties from non-oil natural resources It is our opinion that the above resources are sufficient to allow the State to maintain economic growth through about the ~~ year 2000. This means that those making a decision with regard to selecting the preferred Railbelt energy source can expect the State to influence demand growth and, therefore, the demand for power, at least until the time when the option with the longest lead time to construct can be built, namely a hydroelectric dam. Beyond the year 2000, all demand projections become less certain. This does not affect significantly the conclusions of this study, however. Exhibit 18 estimates Railbelt demand from 1960 to 2030.* We have dealt with the uncertainty of long range forecasts by consider- ing all power production alternatives in as small increments as possible. This is particularly important in the case of Susitna, where we have treated Watana and Devil's Canyon as stand-alone projects to be evaluated separately. Power demand for the Watana analysis needs to be known only until it reaches a level equal to the output of the dam. This level will be reached well before the year 2000. fe * All projections at consumption level, excluding military and self generations that cannot be supplied by Susitna or Railbelt utilities. Ju Exhibit 18 ANNUAL DEMAND HISTORY AND FORECAST 1960—2030 ~ ia eo 5 | Annual Demand—_ 0 fo a2 3 44 cf ao oO 1960 1970 1980 1990 2000 2010 2020 2030 YEARS Source: APA-FERC 1983 ONO TNAN ROENIULYU CURRENT SUPPLY OF RAILBELT POWER This part will describe power generation in the Railbelt by fuel source, the utility system for power distribution and the current price of power in the Railbelt. The Railbelt is heavily dependent on natural gas as a source of fuel. Exhibit 19 shows that over 78 percent of electric power generated in the Railbelt today is powered by natural gas. Exhibit 19 RABEL TSENG Aon les | mene iy he 1982 DIESEYORA, Gy 1%) COAL (15.5%) NATURAL GAS (78.4%) 1982 Total - 1253 Mw = 100% Source: APA-FERC 1983 KENICO The discovery of large inexpensive reserves of natural gas in Cook Inlet in the 1960's was the determining factor in the selection and development of gas-fired power generation. Ninety- two percent of the power generation in the Anchorage area is from natural gas. The rest is from diesel fuels (3 percent) and hydroelectric plants (5 percent). (See Exhibit 20) Exhibit 20 ANCHORAGE =GENERATION SY FUEL [TPE 1982 HYDRO (5.0%) DIESEL (3.0% NATURAL GAS (92.0%) 1983 Total - 898 Mw = 100% Source: APA-FERC 1983 Sein INeIVIWY The Fairbanks area generates its power primarily from coal (87 percent) with a secondary source being diesel (13 percent). (See Exhibit 21) Exhibit 21 FAIRBANKS GENERATION BY FUEL TYPE 1982 DIESEL (13.0%) COAL (87.0%) 1982 Total - 354 Mw = 1002 Source: APA-FERC 1983 = 39) = KENTCO - The Generation and Distribution System - The Railbelt includes two electrical supply centers; one in the Anchorage/Cook Inlet area and the other in Fairbanks. An intertie is under construction which will connect these two centers. It will permit the two supply centers to share power or, with expansion, construct central power generation facilities. The Anchorage/Cook Inlet Load Center Seven different utilities and agencies serve the Anchorage/ Cook Inlet area: (See Exhibit 22) e Chugach Electric Association, Inc. (REA) = e Municipality of Anchorage - Municipal Light & Power Department (Municipal Utility) e Homer Electric Association, Inc. (REA) e Matanuska Electric Association, Inc. (REA) e Alaska Power Administration (Federal Power Agency) e Elmendorf AFB (Military) e Fort Richardson (Military) Chugach Electric Association (CEA) is a Rural Electric Cooperative (REA) serving the Anchorage area, except those sections served by Anchorage Municipal Light & Power (AML&P). In addition to the Anchorage area, CEA serves many surburban areas of Anchorage plus Kenai Lake, Moose Pass, Whittier and Hope. CEA also sells wholesale power to AML&P and Homer Electric Association. CEA has a total of 478 Mw of peak capacity and 1. A Rural Electric Association Cooperative is owned entirely by its members who are also customers for the power. - 4O - KENICO supplies about 40 percent of the total Railbelt electrical demand, thereby making CEA the largest utility in the Railbelt system. Anchorage Municipal Light and Power (AML&P) is a city-owned utility which serves areas immediately within Anchorage not served by Chugach Electric Association. It has a total of 311 Mw of peak capacity and provides about 20 percent of the total Railbelt power needs. AML&P sells wholesale power to Elmendorf Air Force Base and Fort Richardson. Homer Electric Association (HEA) is an REA and serves the City of Homer as well as other @istomers ca the Kenai Peninsula. HEA purchases its power from CEA and maintains only emergency back- up generation capability. Along with the Seward Light Department, HEA distributes power to about 10 percent of the total Railbelt demand. Matanuska Electric Association (MEA) is an REA and serves the Matanuska Valley north of Anchorage. MEA purchases its power from CEA and the Alaska Power Administration supplies about 10 percent of the total Railbelt demand. The Alaska Power Administration is a federal power agency and provides bulk power to MEA, CEA and AML&P. This power is generated from the Eklutna Hydroelectric facility which the B30 Administration owns and operates. It produces BB Mw of peak capacity. The Fairbanks Load Center The Fairbanks Load Center is served by six utilities. (See Exhibit 23) - Al - KRENI CU e Fairbanks Municipal System (Municipal) e Golden Valley Electric Association, Inc. (REA) e University of Alaska, Fairbanks e Eielson AFB (Military) e Fort Greeley (Military) e Fort Wainwright (Military) Fairbanks Municipal Utilities System (FMUS) serves primarily the Fairbanks area although it does serve several residential areas immediately adjacent to the city. FMUS has an installed capacity of 68 Mw and provides about 5 percent of the total Railbelt energy supply. Golden Valley Electric Association (GVEA) serves por- tions of Fairbanks and the Fairbanks North Star Borough. Additionally, it serves the cities of North Pole, Fox and Ester and the military bases of Eielson Air Force Base and Fort Wainwright. GVEA also serves areas south of Fairbanks including Nenana, Healy, Clear, Anderson and Rex. GVEA has an installed capacity of 221 Mw and provides 12 percent of the total Railbelt energy supply. The University of Alaska, Fort Wainwright and Eielson AFB generate part of their own power (50 Mw). So Exhibit 22 INSTALLED CAPACITY OF ANCHORAGE/COOK INLET AREA HYDRO UTILITIES Hydro Alaska Power Administration 30.0 Anchorage Municipal Light and Power 0 Chugach Electric Association 15.0 Homer Electric Association 0 Matanuska Electric Association 0 Seward Electric Association 0 TOTAL 45.0 MILITARY INSTALLATIONS Elmendorf AFB 0 Fort Richardson 0 Subtotal 0 INDUSTRIAL INSTALLATIONS Subtotal 0 Total 45.0 1982 OIL NATURAL GAS Combustion Steam Diesel Turbine Turbine 0 0 0 0 215.6 0 0 463.7 0 2.6 0 0 0.9 0 0 232 0 0 9.0 760.1 0 2.1 0 31.5 7.2 0 18.0 9.3 0 49.5 9.6 16.0 0 27.9 776.1 49.5 Source: Battelle Pacific Northwest Laboratories, Facilities and Planned Additions for the Rai Alaska, Volume VI, September, 1982, Alaska Power Administra- tion 1983; updated by Harza-Ebasco Susitna Joint Venture, 1983. Revised by Tom Stahr, AML&P, September 1983 * Figure is for 1981, latest year that data was available. - 43 - KENTCO Total 30.0 215.6 478.7 2.6 0.9 5.5 7133.3 33.6 25.2 58.8 25.6%* 898.5 Existing Generatin [belt Region Of Exhibit 23 KENICO INSTALLED CAPACITY OF THE FAIRBANKS/TANANA VALLEY AREA - 1982 OIL Diesel UTILITIES Fairbanks Municipal Utility System 8.4 Golden Valley Electric Assn. 2359 University of Alaska 5.6 Subtotal So MILITARY INSTALLATIONS Eielson AFB 0 Fort Greeley Die) Fort Wainwright 0 Subtotal Da) INDUSTRIAL INSTALLATIONS Subtotal 2.8 Total 46.1 Source: Battelle Pacific Northwest Laboratories. HYDRO Hydro COAL Combustion Steam Turbine Turbine 30.1 30.0 172.8 YAO 0 L370) 202.9 68.0 0 15.0 0 0 0 22510 0 3720) 0 0 202.9 10550 Tota 68.5 221716 18.6 308.7 0. SP 22 42.5 2.8 354.0 Facilities and Planned Additions for the Railbelt Region o Existing Generating aoe Alaska, Volume VL, September 1982; Alaska Power Administra- tion 1983; updated by Harza-Ebasco Susitna Joint Venture, 1983. EGG RENIOCU Anchorage Capacity The total Anchorage generation capacity is now 898 Mw. Total 1982 peak demand in Anchorage was 469 Mw. (See Exhibit 24) Reserve capacity in 1982 was 91 percent. This level of reserve capacity is considered just sufficient to maintain proper generation. Reserve capacity is maintained so that a power generation system can continue to provide power even if serious equipment failures occur during the period of peak demand. Fairbanks Capacity The total Fairbanks generation capacity is 354 Mw. Total peak demand in Fairbanks was 110 Mw. (See Exhibit 24) Reserve capacity was over 200 percent more than current demand. This excess reserve occurred as a result of an overestimation of demand growth in the late 1970's. Exhibit 24 PEAK AND TOTAL GENERATION CAPACITY 1982 900 800 700 600 500 400 MEGAWATTS 300 200 CAPACITY PEAK CAPACITY ANCHORAGE AREA FAIRBANKS AREA Source: APA-FERC 1983 NRONIUU - Price of Power - The price of power in the Railbelt is dependent primarily on the type of fuel used for generation (natural gas, coal, diesel, hydro). For comparison purposes we have converted all fuel sources ‘o their energy equivalent, namely million British thermal units (MMBtu). The Anchorage area generates its power primarily from Cook Inlet natural gas. The current Cook Inlet natural gas selling price is an average of $0.86 MMBtu (See Natural Gas). This is much less than the Lower 48 price of $3.50 MMBtu and, as a result, the Anchorage area enjoys some of the least expensive power in the U.S. (See Exhibit 25) The Fairbanks area generates its power primarily from coal mined in Healy (near Fairbanks). This coal currently sells for $1.50 MMBtu and, as a result, the Fairbanks area power costs are almost twice that of power generated in the Anchorage area. This subject is discussed in full detail under the next section. There are small variations between the price of power charged by each utility within each area due primarily to the financial structure and distribution costs associated with each. Exhibit 26 lists the rates charged to customers by the utilities. 4a -ly- ELECTRIC RATE COMPARISON Exhibit 25 ALASKA GOLDEN VALLEY ELECTRIC MATANUSKA ELECTRIC HOMER ELECTRIC MUNICIPAL LIGHT & POWER CHUGACH ELECTRIC ASSOC. NATIONAL SEATTLE SAN FRANCISCO HOUSTON SAN DIEGO NEW YORK CITY Source: Chugach Electric Association 1983 50 75 100 DOLLARS PER 1000 KILOWATT HOURS OOLNS» Uj Es HL “I - fe) oO Qo oO Oo 9° Oo fo} oO oO o ° ° oO 0 O98 OF FR O WM SF YM N KF DGDUUUULULUmDDUDUDUULlUCULULULUUlUCODWUDUUDUDUCUCUCODWUUUUUCUOD put wo MPN KENTCO POWER GENERATION ALTERNATIVES FOR THE FUTURE - Introduction - Purpose This section will determine the most economic method of generating power for the Railbelt over the next several decades. It will describe, evaluate and compare methods of power generation which are available to the Railbelt. There are three reasonable alternatives which are: e combined cycle turbines fueled by natural gas e steam turbines fired by coal e hydroelectric turbines Other methods of generating power exist and were examined by us during the course of this study, but all were either imprac- tical or deficient in some way for use in meeting the Railbelt's needs. Approach There are five points about the analysis in this section of which the reader should be aware. First, we consider only the direct costs of each alternative. We do not evaluate other factors such as environmental, technical and political considerations. Second, the analysis is concerned with the costs and bene- fits of each alternative over the long term. Evaluating the alternatives over decades rather than years allows us to consider all of the viable alternatives and to calculate the true costs and benefits of each. - 49 - KENTCO Hydro and coal require long lead times for planning and construction and could not be considered properly in any short term analysis. The useful lives of all generating plants under consideration are long -- coal, 30 years; gas, 25 years; hydro, 50 years. The most accurate comparison will look to the entire life of the longest option. Planning for the long term produces a more complete and accurate analysis of costs and benefits because of differences in the behavior of costs over the long term. The operating costs of some alternatives, such as natural gas and coal plants, almost are certain to change over time due to changes in the price of the fuel. The costs of hydroelectric plants, however, are likely to remain quite stable. It is necessary, therefore, to compare the alternatives over a time period extensive enough to cover the longest useful life of any alternative. Thirdly, we are concerned with selecting the method of generating electric power at the lowest possible cost. We have attempted to do this by evaluating optimal gas, coal and hydro generation alternatives. We have based our analysis on cost and output figures which would be achieved by efficient generating equipment all operating efficiently. In the case of gas and coal, we assume the use of plants large enough to take advantage of economies of scale and new enough to take advantage of existing technology. In the case of the hydro alternative, there presently is only one viable project proposed which is capable of supplying sufficient amounts of power KENTCO to the Railbelt, the Susitna Hydroelectric Project. We have accepted the cost and electric output estimates of Acres American and Harza-Ebasco, the Susitna engineers, and have used them as representative of the most efficient way to generate large amounts of hydroelectric power in the Railbelt. Fourth, in the case of Susitna, we will treat the project as two separate projects, Watana and Devil's Canyon. We consider it essential that the first dam, Watana, be economically viable on its own as we believe it impossible, as well as unnecessary, at this time, to judge accurately when or if. the demand for power in the Railbelt will reach a point which requires that Devil's Canyon be built. As we will show, the economics of Devil's Canyon are so attractive, assuming Watana is built, that there is no question this dam would be the lowest cost means of providing the next 3500 Gwh of power once Watana is fully utilized. The real issue is whether or not Watana, by itself, can provide power at a lower cost than any of the thermal alternatives. The fifth point to make about our analysis is that it is an incremental analysis of the three alternatives for a given period of time and a given demand. It is not an overall power generation study which would include power needs beyond-the scope of our analysis. This report does not attempt to evaluate the engineering aspects of the hydro alternative or the actual sizing, design - 51 - KENTCO or timing requirements of gas and coal plants. Method Our method in this section will be to describe, evaluate and then compare the costs of generating power by means of coal, gas and hydro. In order to produce a valid comparison of costs, the power generated is assumed in the analysis to be held con- stant for all three alternatives. We do this by holding constant the period of the analysis and the level of power supplied annually by all the alternatives. The analysis begins with the first year in which physical capacity from all the alternatives could be put on-line, 1996. Gas plants require two to three years for construction. Coal plants require approximately four years and several additional years for planning and licensing. The Watana portion of the Susitna project would take a minimum of 10 years from 1984 to come on-line. The earliest Watana could start operating, therefore, is 1994. We believe the conservative estimate of 1996 to be more realistic. Our analysis of the costs of generation alternatives, therefore, begins in 1996. Another reason exists for choosing 1996 as a starting date for the analysis; 1996 is the first full year in which Railbelt utilities will not be able to buy inexpensive natural gas from Cook Inlet under existing contracts*. These contracts provide for prices at an average of $0.86 per MCF. Since there is no cheaper way of generating power available than with natural gas at Under the terms of existing contracts, Chugach is required to buy between 55 and 60 MMFper day. The total it can buy at contract prices in any event is 373 BCF. Chugach will have reached this limit by mid-1995 if it buys the minimum daily quantity and by early 1994 if it buys the maximum. - 52 - KENTCO such favorable prices, it is clear that any evaluation should be concerned with the post-contract period. We assume this to be 1996 onward. A third reason for choosing 1996 as a base year for this analysis is that it is in 1996 that demand grows to a level at which Watana could be fully utilized. Since the analysis uses optimal gas and coal generation systems, it is appropriate to consider the hydro alternative at its most efficient level as well. Affull utilization in 1996, the Watana portion of the hydro alternative would be at its most efficient level. The period over which the analysis takes place is the life of the longest-lived alternative. Since the hydro project's estimated useful life is 50 years, we compare the projected costs of each alternative over a 50 year period. The time apa of our analysis, therefore, is 1996-2045. The level of electric output used as the basis for compar- ison conforms to that of the largest single amount of power produced by any one of the alternatives. The Watana Dam would produce that largest amount or approximately 3500 Gwh per year. Natural gas and coal plants would produce power in smaller increments which can be combined to produce an amount equal to Watana, or 3500 Gwh per year. Therefore, we compare the costs of the three alternatives at the level of 3500 Gwh per year for each of the 50 years of the analysis. In the Power Demand section, the projection for power demand was such that 3500 Gwh of generation will be nearly sufficient - 53 - KENTCO beginning in the year 1996. Demand will grow in subsequent years and additional generation will be required. Our analysis treats the question of how to supply this incremental demand (demand over 3500 Gwh per year) as separate from the question of how to supply the base 3500 Gwh per year amount of power.(See Exhibit 2' After describing the gas, coal and hydro alternatives over the 50 year period we will arrive at the total cost of each for the individual years 1996-2045. By then comparing the alternatives we can identify the most attractive one from an economic standpoint. This comparison is based on the 3500 Gwh per year level of supply and demand. Costs of the Power Generation Alternative In the next chapter we set forth the costs of each alterna- tive power generation method. There are three key groups of costs which we need to quantify: e capital costs e fuel costs e operations and maintenance costs Together these three groups encompass the cost of information we will need to evaluate and compare the alternatives. Clearly there are other costs related to the supplying of power to the Railbelt. There are, for example, the administrative expenses of the utilities and the costs of maintaining distribution systems. These types of costs, however, will be incurred regard- less of the generation method selected. Their magnitude does not affect the decision and we do not attempt to evaluate them. The aS ie pee ea Gwh PER YEAR (Thousands) Exhibit 27 BASE LEVEL DEMAND ANALYZED 3500 Gwh Incremental Generation Thermal 3500 Gwh/yr. Genera- tion 1990 2000 2010 2020 YEARS Oo DEMAND +) OTMER HYDRO 2030 OOLN3Ax KENTCO only costs we do evaluate are those which are not common to all alternatives. The three categories listed above are the only significant costs which behave in this fashion. Capital costs are the fixed costs incurred when a power generation plant is installed. Capital costs include the cost of land, equipment, construction and interest incurred during con- struction. In the cases of the natural gas and coal alternatives, we have assumed capital costs to be those which would be incurred building plants of optimum efficiency. These figures were taken from the Harza-Ebasco Fall 1983 update of the Susitna Hydroelectric Project study. In the case of hydro, we have used the capital cost estimates for the Watana portion of the Susitna project, taken from the same Harza-Ebasco study. Fuel _Costs Fuel costs represent the cost of obtaining the energy re- quired to generate electricity. In the case of the two thermal alternatives, natural gas and coal, fuel costs are the annual expense of buying natural gas and coal. The magnitude of fuel costs depends on two factors; the cost of a unit of fuel and the amount of fuel used. Since our analysis assumes a constant amount of generation (3500 Gwh per year), and we can calculate the amount of fuel needed to produce this output, the only factor to cause year to year changes in fuel costs is the price of a unit of fuel. In the case of the hydro alternative the cost of the fuel, water, is zero. KENTCO A few words on the mechanisms which determine the prices of thermal fuels is necessary at this point. Prices for most of the world's natural gas and coal are determined by market forces which operate on a worldwide basis. The world market tends to price energy commodities according to their value as an energy source. Value as an energy source can be expressed in BTU's (British thermal units), a measure of the amount of heat produced by that source. A barrel of Saudi medium crude contains approximately 5.8 million BTU's (MMBtu's). One thousand cubic feet (MCF) of gas contains approximately one MMBtu. If the price of oil is $29 per barrel, the cost per MMBtu is about $5.00. As long as oil is $29 per barrel, the cost of natural gas is also likely to be $5.00/MMBtu, or $5.00 per MCF. A similar logic can be used to arrive at a price for coal. The prices of thermal fuels, therefore, tend to move in unison when they are determined by the world market. In attempting to project the price of energy commodities, such as natural gas and coal, it is possible to look at expected movements in the price of energy in general and to deduce from this the price of gas or coal. Oil prices, for several reasons, tend to be the dominant force in determining energy prices. Other energy commodities’ prices tend to be adjusted according to the movements of the price of oil. Therefore, one can use a forecast of oil prices as a good indicator of the future prices of gas or coal. This is, in fact, the method we have adopted for projecting the world market prices of natural gas and coal. ANAS 7 KENTCO Since trends in the price of energy have important effects on our analysis, we have attempted to obtain the most reliable world energy price forecasts possible. In addition to consider- ing published energy forecasts, we hired a highly respected firm of energy economists, PIRA*, to forecast oil prices for the specific time period with which we are concerned. PIRA's complete set of findings are included in Appendix 1. We discuss the impli- cations of the PIRA forecast later in this section. When discussing the effects of world prices on the Railbelt, we will use the term "net back price". A net back price is the price paid for gas or coal at the point of origin before adding conditioning, liquefying or shipping costs. Net back prices are calculated by subtracting the costs of conditioning, liquefying (in the case of gas) and shipping from the price paid for the commodity delivered to its Gest tewtton: Net back prices, there- fore, represent the amount received by the producer of a commodity such as gas or coal. World prices do not always govern the prices paid for gas and coal in Alaska. If no export market exists for a commodity, local prices will be set by local supply and demand factors. Among these factors are the costs of producing the commodity and the willingness of consumers to switch to an alternative fuel. Local market forces such as these will play a role in the analysis of fuel costs which follows. * Petroleum Industry Research Associates, Inc. - 58 - KENTCO Operation and Maintenance (0 & M) costs represent those operation and maintenance expenses associated with the generating plants of each of the alternatives. As in the case of capital costs, our assumptions on 0 & M costs for natural gas and coal correspond to those which would be required to service an efficient system of generation for each fuel alternative. These estimates were obtained from the Battelle Railbelt Electric Power Alternative Study done in 1982. The 0 & M cost estimates we use for hydro are those forecast by Harza-Ebasco. Use_of Constant _Dollars All costs in this analysis are described in constant 1983 dollars. This removes the effects of inflation from the costs shown in the analysis. It also should allow the reader to under- stand more easily the advantages and disadvantages of the alternatives we evaluate. We have assumed, for example, that the real cost of 0 & M services will not change over time. Therefore, we show the cost of these services for each alternative, stated in constant 1983 dollars, to be the same in each of the 50 years of our analysis. This does not mean that the actual price paid for 0 & M services will remain level in nominal terms, but rather that the value of the services in pre-inflation dollars will not change. KENTCO POWER GENERATION ALTERNATIVES FOR THE FUTURE - The Natural Gas Alternative - Capital Costs ’ Assumptions In estimating the capital cost of the natural gas-fired turbine alternative, we have made three important sets of assumptions. First, estimates of capital costs are based on the efficient size and type of generating plant appropriate to the conditions and level of demand in the Railbelt. As stated earlier, we make this assumption in order to approximate the ideal operation of a gas-fired generation system. We use the following data regarding the capital equipment itself. This is based on Battelle's work done in 1982 and included in the Federal Energy Regulatory Commission Application Revision of July 1983.* Each generating plant would Have the following characteristics: Type Combined cycle Peak Capacity 200 Mw Annual Capacity 1489 Gwh per year (at 85 percent utilization) Fuel Usage 12 million MCF year Plant Life 25 years Combined cycle plants, which employ steam boilers to cap- ture waste heat from the gas turbines are more efficient than simple gas turbines. According to the APA's Federal Energy Regulatory Commission application for Susitna, a 200 Mw combined cycle plant is representative of an efficient plant of the size * (See footnote on next page) - 60 - KENI OCU Harza-Ebasco in its December 1983 update of the material presented in the FERC application changes its assumptions somewhat regarding the design, size, cost and fuel usage of the combined cycle plants which could be employed in a natural gas-based generating expansion program. Hurza- Ebasco assumed that a somewhat larger plant could be used as part of a gas generating system. Such a plant would burn fuel at a slightly higher rate than the Battelle/FERC plant, which was used also by Harza-Ebasco in their Fall 1983 update, but would have a lower per kilowatt cost. It is not clear, however, that the Harza-Ebasco plant actually would cause the total configuration of a gas-based generating system to have a lower capital cost than the plant assumed by Battelle and the APA in the FERC License Application. The selection of the correct sizes of equipment is possible only when peak power demand, outage rates and other statistical factors are considered. For example, larger sized plants, although producing more power when operating, may require a utility to keep more reserve capacity on hand than if the utility had several plants, each providing smaller amounts of power. A complex model such as the Optimum Generation Planning Model which Harza-Ebasco uses, is generally required to provide such power expansion plans. We have assumed, ag the suggestion of Harza-Ebasco, that the 200 Mw ossaused sth plant specified in the FERC application is representative of a highly efficient gas-fired plant which could be used in the Railbelt and that the 200 Mw plant we use would not yield a significantly different actual capital cost than the one assumed in the Harza-Ebasco update. Moreover, it is important to recognize that, in practice, neither the 200 Mw plant, nor any other of its size, would be the only size plant to be used in the Railbelt. A real world configuration of equipment likely would include smaller, less efficient plants as either reserve or incremental capacity. The system, we assume, therefore, is more efficient, or less expensive,than would be possible in actuality. —S KENTCO and type which could be employed in the Railbelt. Given a peak capacity of 200 Mw and an assumption that it can be operated at an 85 percent utilization rate, the annual capacity of each plant is 1489 Gwh per year. If a plant produces 1489 Gwh of power, it will use approximately 12 million MCF of fuel, or 8.06 thousand cubic feet per Gwh. Battelle assumed a useful plant life of 25 years. This means that during the 50 year period of the analysis, each plant will be replaced once and will have reached the end of its useful life again by the end of the analysis. Our second assumption is that the cost of generating equipment does not change in real terms over time.. This is to say that the costs of acquiring and constructing each plant escalate at the same rate as inflation generally. Therefore, the cost of each plant, when expressed in constant 1983 dollars, is the same in 1996 and in 2021 when it is replaced. Third, we assume that it is possible to install a fraction of a plant at the pro-rata cost of a whole plant. We do this in order to arrive at an estimated capital equipment cost re- quired to generate 3500 Gwh of power per year. The number of plants needed to generate this amount of power can be calculated as follows: Annual demand + generation per plant = number of plants 3500 Gwh per year + 1489 Gwh per year = 2.35 plants The capital cost of generating 3500 Gwh per year, therefore, will be 2.35 times the cost of one plant. S62N = KENTCO This procedure does not take into account the supply of peaking capacity afforded by the Susitna Project which, in turn, would be required in the gas alternative to make the plants strictly comparable. This tends to understate the cost of the gas alternative. see The Battelle estimate for the cost of a single 200 Mw plant of the type described above converted to 1983 dollars is $213,215,000. Since 469 Mw is required, the initial capital cost of the gas generation alternative is $501,055,000 in 1983 dollars. After 25 years of operating, these first plants are assumed to be retired and a new set of plants replaces them. These also cost $501,055,000 in 1983 dollars. Therefore, total capital costs for the natural gas alter- native over 50 years are: Year Capital Cost 1996 $ 501,055,000 2021 501,055,000 Fuel Costs Assumptions A series of assumptions underlie our projection of fuel costs for the natural gas alternative. First, we accept the forecast of world oil prices done under subcontract to William Kent and Company by the Petroleum Industry Research Associates (PIRA). The PIRA forecast is more conservative than many of MG OMe KRENIUCU the projections used in the past by Battelle, Harza-Ebasco and the APA. Exhibit 28 compares PIRA with forecasts by the Alaska Department of Revenue (DOR) and Sherman Clark. The graph demonstrates that the DOR-Mean forecast is closer to the PIRA projection than any of the others. Only the DOR forecasts are lower than the one we use. The second assumption we make is that gas prices in the world market will continue to move in concert with world oil prices. This is a standard assumption which we explained in the introduction to this section. It allows us to derive a forecast for natural gas prices from a forecast of oil prices. We have done this using the PIRA oil price projections. The results are shown in Exhibit 29. A third assumption is that there will be no legislation which materially will affect the determination of gas prices in the Railbelt. This is to say that the market forces we describe, such as supply and demand, will govern the determina- tion of prices. The fourth assumption follows from the previous one. Cook Inlet gas producers, we assume, will stop exporting after current contracts with Japan expire, if the net back price to the producer fails to exceed their production costs. In other words, producers will not be willing to continue exporting if doing so would mean selling at a loss. Our fifth assumption states that a pipeline might be built in order to export North Slope gas if and when the world price of gas exceeds the cost of bringing North Slope gas to market. - 64 - S9 $ /MCF(1983 CONSTANT) Source: 1985 Exhibit 28 NATURAL GAS PRICE FORECASTS DOR,PIRA,SHCA 1990 1995 2000 2005 2010 2015 2020 2025 YEARS Oo DOR + PIRA © SHCA APA-FERC 1983 PIRA 2030 OOLNS» - yy - WORLD PRICE OF NATURAL GAS FORECAST DOLLARS PER MCF Source: 14 Exhibit 29 PIRA — CIF TOKYO Nu Wt ODN DO O O — O 1985 PIRA PIRA Forecast T T T 7 1995 2005 2015 YEARS 2025 —— 2035 2045 OOLNAY KENTCO including an allowance for risk. This could happen when the world price of gas reaches $5.85/MCF (in constant 1983 dollars). The sixth assumption is that there are unproven reserves of natural gas in Cook Inlet of 2 trillion cubic feet (TCF). This figure is found in an analysis done by the Alaska Depart- ment of Natural Resources (DNR) .2 Our last important assumption is that these unproven reserves will be developed independently of oil drilling. There- fore, revenues from the sale of this gas will have to cover all production costs. As in the case of exports, the owners of the unproven reserves will not produce this gas unless they can recover all their costs. The availability of natural gas is a key element in the analysis of the cost and viability of the natural gas alterna- tive. In this discussion we will establish a) that a reliable supply of natural gas will be available to the Railbelt through 2045 and b) that the source from which this gas is supplied has a strong impact on the price which electric utilities must pay for gas. a. Cook Inlet The only Alaska source of gas supply for electric utilities in the State today is Cook Inlet. At the end of 1983, Cook Inlet had approximately 3 TCF of 1. Trans Alaska Gas System = Economics of an Alternative for the North Slope Natural Gas - report by the Governor's Economic Committee on North Slope Gas, January 1983 2. Alaska Department of Natural Resources - Historical and Projected Oil and Gas Consim-*ion, January 1983 6a KENTCO proven reserves, according to Harza-Ebasco. We have accepted the Alaska Department of Natural Resources' estimate of an additional 2 TCF of undiscovered - reserves in Cook Inlet. Between proven and unproven reserves, therefore, Cook Inlet may hold a total of 5 TCF of gas at the present time. (See Exhibit 30) Exhibit 31 shows our estimates of consumption of natural gas from 1984 to 2016. The table indicates that proven reserves will be nearly exhausted by the year 1998. In 1996 proven and estimated but undis- covered reserves will total about 2.35 TCF. (See Exhibit 32) Consumption of gas from 1984 to 1996, the base year of this analysis, has been estimated by Harza-Ebasco to be 2.45 TCF. This estimate was based on assumptions of higher growth rates in demand for electricity and gas for heating than those we have forecast.* Over half of the proven reserves in Cook Inlet are already committed by contract. Chugach Electric, which buys most of the gas destined for use in power generation, has contracts which require purchasing between 50 MMCFand 60 MM@per day, but not more than 373 BCF over the life of the contract. Assuming Chugach continues to buy an amount within this range, * Our lower forecast of demand for electricity, it will be remembered,was derived from the PIRA oil price forecast. = 63 < Exhibit 30 Dee ne a MAN Mestre ee Res a cater eae es a DT | 12/83 KENTCO CG NM OO LT NNW HO TtTNHKH DMD OTN O Year 1982 1983° 1984 1985 1986 1987 1988 1989 1990 LOE 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Source: Phillips LNG 62 62 62 62 62 62 62 62 62 62 62 62 62 62 62 62 APA Updated Kentco,October 1983 Cook Inlet Natural Gas Usage Collier Urea aD 35 Do DD 55 25 55 55 55 55 5 55 a5 SP, 55 39 Exhibit 31 (BCF) Enstar 7, 18 .70 18. -96 19. 20. 21% ZN an Ze 24, 24, Zor 26. ie 28. 29) 30. Sie 32e 34. 35. 36. Sie 39% 40. 41. 43. 44. 46. 48. 49. De D3) Pen ie 32 62 or 02 76 52 3 12 97 84 75 68 65 65 69 77 88 03 22 45 13 05 41 83 29 81 38 00 68 42 22. 08 Ol = 70) Field Ops Military 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 Power Generation 38), 40. 43. 45. 47. 49. 46. 48. 50. ols D3): 54. on 2% 33) Se 34. Sos 3 36. Silos 40. 41. 42. 44. 46. 47. 48. 50. Ses 52. 54. 55 EY ic RPDNWOURNWORDRRRN COCO NOFUOUFOUUNYNAUN Of KENTCO Total 203. 206. 209. 227 214. lis 215) 213i, 220. 222. 220% 227. 229% 207. 208. 20% 95% 96. 98. 100. LOZ. 106. 108. ls 115, 117. 120. L235 L265 129% 1327 135% L387. 142. 87. ONONKDNUUNANOCOKFWODOWONONUONUWEOWOWNORNEHE Remaining 5,594 5,391 DI Loo 4,976 4,763 4,549 ans3L 4,116 3,898 3,677 3,454 317229 3,001 2,772 2,565 2,356 2,146 2,050 L953 T7855 1,754 1,651 1,544 1,436 1,325 1,209 1,092 971 847 215 591 459 325 184 42 (45) Exhibit 32 AT WA! SS) Te le WES CO Oi IN ET 1983 AND 1996 6 Total Gas Reserves = 5.1 TCF 5 \ NS Po} Total Gas Re ves = 2.35 TCH LO 1983 1996 TCF (Thousands) ZZ] PROVEN [\N) PROBABLE Source: APA-FERC 1983 OO INAM KENT CO its contract will expire in 1994 or 1995. Other entities, such as Enstar, also have contracts for gas to be used locally from Cook Inlet. Exports of natural gas from Cook Inlet are likely to cease by 1998. By 1998 exporters' commitments to export to Japan (Cook Inlet's main export customer) will have expired. Exporters will continue to export only if they can buy or produce gas at low enough prices to keep them competitive in the world market. This, however, is not likely to be the case. By 1998 proven reserves of cheap gas in Cook Inlet will be exhausted. Additional reserves would have to be proven and new wells drilled for exporters to obtain a supply of gas. Similar drilling in the Lower 48 pre- sently costs $3.50 per MCF. We have been advised by those in the industry that a comparable figure in Alaska would be $3.75 per MCF. The current cost of conditioning, liquefying and shipping Cook Inlet gas to Japan is $3.25 per MCF according to Battelle. The $3.75 per MCF production cost plus the $3.25 per MCF cost of conditioning, liquefying and shipping yield a total cost of $7.00 per MCF. On top of all the above, producers will want some allowance for profit and risk. The world price for gas in Japan at this time, however, is estimated to be only $5.00 per MCF. Cook Inlet gas, therefore, will not be competitive when proven reserves run out and export contracts expire in 1998. Exhibit -72- KENICO 33 shows that Cook Inlet exports will not be competi- tive until 2012. When exports cease the only remaining consumers of natural gas will be local buyers: Chugach Electric, Enstar and the military. Exhibits 34 and 35 show forecasts of their consumption after 1996, assuming they continue to use natural gas to fuel power genera- tion plants. Local users presumably will continue to use Cook Inlet natural gas if it is the only local source, or if it is cheaper than imports, or if it com- pares favorably in cost to other fuels and energy sources. At the usage rates forecast in Exhibit 34, Cook Inlet reserves will be completely exhausted by the year 2015. (See Exhibit 36) If, however, electricity were to be generated with a fuel other than natural gas, Cook Inlet reserves would remain available longer. For example, if coal or hydro power were used for generation beginning in 1996, the life of Cook Inlet reserves would be extended to 2025. (See Exhibit 37) This would allow the use of Cook Inlet gas for home heating use. The additional cost of gas to home heating customers if they were to source from the world market would repre- sent an additional element to the cost of gas-fired power generation. We consider this possibility later. b. The North Slope The North Slope holds proven reserves of natural gas of 30-33 TCF. At present, this gas is inaccessible to Railbelt consumers since no pipeline exists. Such =o $/MCF (CONSTANT 1983) Exhibit 33 COOK INLET VS WORLD GAS PRICE CIF TOKYO Cook Inlet Gas Price —}+——& Ss 8 - = S —& & & =a World Gas Price OOLNAY Exhibit 34 KENTCO FORECAST OF CONSUMPTION OF COOK INLET NATURAL GAS - 1997-2015 (BCF) Year 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Source: Phillips Collier LNG Urea Enstar 62 DD 29. 30. Sle 32. 34. Sr 36. S75 395 40. 41. 43. 44, 46. OS 49. Si a3 Ss 47 APA Updated Kentco, October 1983 = 752 65 69 76 87 02 21 45 72 04 41 82 29) 80 37 67 41 21 07 Field Ops Military 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 30 Power Generation 357 34, 3 555 36. o1 40. 41. 42. 44, 46. 47. 48. 50. pple D206 54. Dore 57. oc FnWOUFN WRF ADF FF NY CO Or uw Total 210. 95F 96. 98. 100. 102. 106. 108. LPS 115 TE 120. 23 126. EZ95 32% L350 L338) 142. NAN ADUUNUUNKDDODOOrF WU OO AN ON WUN Remaining 2,146 205A. 1,954 1,855 1,754 OD 2 1,545 1,436 L325 1,210 1,092 972 848 722 592 459 324 185 43 Exhibit 35 UMPTION OF NATURAL GA S S RAILBELT CON 1997—2015 OO OOTY. KX N OO OYA KX WN OO OY. WN NINN BO OY A;]”RW PO OAK WN Me ZK] NN OO IINWN BT. KIN 1050/3. NN Po OYA SS OO OYA KY OOK WN OO OYBRSNS SO OY/KIWYN PO OY.-MW,MNS YD POO OY RWW(D IIL ASS I) ODOonRONnNYTMNN HS weer err rer er ere 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 KENTCO XXX) KWH YEARS RS OOPS/MIL (ZZ) ENSTAR [XN] UREA : APA-FERC 1983 LNG - iL - Exhibit 36 COOK INLET GAS RESERVES 1985—2015 INCLUDES USE BY UTILITIES 5 ——- ona = 44 D 3 4 —Cook Inlet Gas Reserves U0 c ® Lo y On eo yh c Big & 24 a. 14 0 1985 1990 1995 2000 2005 2010 2015 2020 YEARS Source: APA-FERC 1983 OOLN]Ay - gL - Exhibit 37 COOK INLET GAS RESERVES 1985-2025 EXCLUDING USE BY UTILITIES IGE (Thousands) 1985 1990 1995 2000 2005 2010 2015 2020 YEARS Source: APA-FERC 1983 OOLNAy KENTCO a pipeline, if it were solely for the purpose of supply- ing gas to the Railbelt, would be prohibitively expensive. A pipeline to exploit North Slope gas can be justified only by the opportunity to sell large amounts of gas. This requires such gas to be competitive on the world market. When conditions justify the construction of a pipeline, the Railbelt can benefit from the fact that the pipeline gas will pass through the Railbelt on its way to the world market. Several schemes have been proposed to commercialize North Slope gas. Two have received serious considera- tion. They are the Alaska Natural Gas and Transportation System (ANGTS) and the Trans Alaska Gas System (TAGS). ANGTS would carry North Slope gas efrough Canada to the Lower 48. The proposed route would run close enough to Fairbanks for a branch line to be built to bring the gas to Fairbanks for power generation and heating. The entire ANGTS project would cost an estimated $25 billion.* Studies by Battelle and the U.S. Government Accounting Office? have indicated the price of ANGTS gas in the Lower 48 would be in the range of $8-10 MCF. This will remain significantly higher than the world price of gas through the first two decades of the next century. As a result, there is little prospect of ANGTS being econom- 1. Battelle, 1982 2. Battelle, 1982, U.S. Government Accounting Office II, 1983 = 79 = * KK KENTCC ically viable for the next 40 to 50 years. The project, therefore, does not figure in this analysis. TAGS would transport gas probably to tidewater on the Kenai Peninsula. This would allow the Railbelt to receive gas at the tidewater price. The Governor's Economic Committee concluded that the cost of North Slope gas at tidewater would be $4.27 per MCF in 1983 dollars. * This represents the following breakdown of costs per MCF: 1983 Dollars Gas at the Wellhead** 0 Conditioning and Pipeline (Construction & 0 & M) $2.38 Financing & Profit 1.89 Cost per MCF $4.27 These studies estimate that liquefication and shipping to Japan would cost $1.66 (in 1983 dollars). The tide- water price ($4.27) plus liquefcation and shipping ($1.66) equal the minimum price the gas would have to command on the world market in order to justify the construction of TAGS ($5.93). The PIRA forecast (see Exhibit 29) expects the world price will not rise above $5.93 until 2004. Earlier we made the assumption that North Slope gas could be made available on the world market only when justified by world prices. In practical terms this Trans Alaska Gas System = Economics of an Alternative for the North Slope Natural Gas - Report by the Governor's Economic Committee on North Slope gas, January 1983 Gas associated with oil is assumed to incur no incremental drilling or production costs. aie: Y Yau KENTCO means that investors may forecast the date at which TAGS becomes viable and invest so that TAGS comes onstream on that date. There is a possibility that TAGS or a similar project may make North Slope gas available to Railbelt consumers. Based on the foregoing analysis, we project that this could happen in the year 2004. c. Other Sources of Gas There is necessarily uncertainty inherent in project- ing the construction of TAGS 20 years into the future. However good the case for TAGS may seem today, TAGS might not be built for reasons which are not apparent now. Even if a TAGS line is not constructed, natural gas still would be available from other sources. There is an untested possibility that gas exists in the Gulf of Alaska. Should the Gulf not contain gas, the Railbelt will have the option of importing natural gas. Thus, if Cook Inlet gas were exhausted in 2015 and the North Slope, or other local sources of gas were not available, the Railbelt could import gas at Anchorage. We conclude, therefore, that a supply of natural gas always will be available to the Railbelt. Exhibit 38 shows Cook Inlet will provide a supply at least through 2015. North Slope gas will not be available until 2004, the first year it could come onstream. It then. would be available to the Railbelt significantly past the year 2045. Imported gas is available now and will be through- out the period analyzed here. - 81 - - 7 - Cook Inlet North Slope Imports Exhibit 38 POTENTIAL SOURCES OF NATURAL GAS FOR THE RAILBELT - 1984-2045 - 1984 1995 2005 2015 2025 2035 2045 OOLNSA» KENTCO Natural Gas Costs 1996-2045 This section will discuss the price which Railbelt electric utilities will have to pay for natural gas for each year during the period of our analysis, 1996-2045. It is useful to divide this timeframe into the periods before and after the TAGS line comes on-line. We assume a functional pipeline in the year 2004. (see page 81 ) By 1996, contracts now in existence between Chugach and Cook Inlet producers will have expired. (see page 72) Since these contracts will expire in 1994 or 1995, we would expect new negotiations to take place by that time. Since Cook Inlet re- serves proven as of today will run out by the beginning of 1998, we expect the price arrived at by new contract negotiations to be determined by conditions governing new reserves of gas which are assumed or have been proven to exist in Cook Inlet. Starting in 1998 the price of gas will be determined by negotiations between the Railbelt utilities and Cook Inlet pro- ducers. The minimum possible price will be the cost of production of the new reserves. This will be approximately $3.75 per MCF. If the utilities were not willing to pay $3.75/MCF or more, the owner of the gas would have no incentive to produce it. The maximum possible selling price of gas to the utilities will be that which the utilities would have to pay to buy gas on the world market. Exhibit 39 shows the world price of gas from 1996 through 2016 and compares the world price to the production cost of Cook Inlet gas. The shaded area between these two lines represents the range of prices on which the utilities and pro- =1|183i| = | - 93 - $ /MCF Exhibit 39 COOK INLET GAS PRICES MIN AND TO RAILBELT UTLITIES 1996—2016 MAX 8 m World 7.5 Price of Gas* (Max imum 7 Price) 6.5 6 Range of Possible Prices of 5.5 5 Cook Inlet Gas to Electric Utilities** 4.5 4 5 Hi Inlet Gas (Minimum oD r 7 Price) isos a086 ook Jeog onde 2016 YEARS oO MINIMUM PRICE + MAXIMUM PRICE Based on the PIRA oil price forecast ** Assuming no TAGS line Cost of Production of Cook =. OO.LNAY KENTCO ducers possibly could agree. The price of fuel to electric utilities from 1996 to 2004, when TAGS comes on-line, will be between $3.75 per MCF and the world price of natural gas. The latter will range from about $5.16 in 1996 to $5.93 in 2004 and over $7.50 by the year 2016 (in constant 1983 dollar terms). It is difficult to predict the prices on which the gas producers and utilities will agree in the mid-1990's, but a presumption must be made about these prices in order to arrive at estimates of future fuel costs for the utilities. We have assumed that utilities will pay the minimum price, $3.75/MCF, for each of the years 1998-2004. Because we assume the utilities will pay the minimum price of gas, any other price they might actually pay would mean a higher cost of the gas alternative than we estimate here. This would make the gas alternative correspondingly less attractive with respect to coal and hydro. About 2004 natural gas from the North Slope, intended for the world market, could pass through the Anchorage area. If this happens, the conditions which created the minimum/maximum price situation shown in Exhibit 39 will disappear. All Railbelt gas prices inevitably will tend to move into parity with this North Slope export net back price. If Cook inlet production costs are lower than the export net back price, the Cook Inlet producers will offer Railbelt utilities the same price as the export net back price. Either way, Railbelt utilities will pay the export net back price from 2004 onwards. - 85 - KENTCO Exhibits 40 and 41 summarize our projection of the price Railbelt electric utilities will pay for natural gas between 1996 and 2045. The reader should remember that the price during the period 1996-2003 is based on the assumption that the utilities will pay the minimum possible price for fuel. Prices after 2003 are based on the PIRA oil forecast and on our assumptions con- cerning the cost of conditioning, liquefying and shipping. The total annual cost of fuel under the natural gas alter- native is readily calculated from the data devexiand in Exhibit 41. The annual cost is the price of fuel times the amount of fuel needed to generate a base load of 3500 Gwh of electricity with the type of equipment assumed above. This is as follows for each year: Average price/MCF for the year X 28.2 MCF = cost of fuel for a year. Exhibits 41 and 42 show the projected annual cost of fuel for the gas generation alternative for 1996-2045. Operations and Maintenance Costs Assumptions In estimating the operations and maintenance (0 & M) costs of the gas alternative, we made two assumptions. First, we used Battelle's August 1982 estimates of the initial annual costs, coverting them to 1983 dollar terms by applying an infla- tion rate of 6.5 percent. Second, we assumed that O & M costs will not change in constant dollar terms during the 50 year period of the analysis. - 86 - - 1/8 - $ /MCF Exhibit 40 E/E eon a NE Lea ASSUMES TAGS LINE IN 2004 6 1996 2000 2004 2008 2012 2016 YEARS OOLNSy Year 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 Exhibit 41 FORECAST OF THE COST OF FUEL TO THE RAILBELT UNDER THE NATURAL GAS ALTERNATIVE, 1996-2045 Fuel Requirement (MCF millions) NNN NONNNNNNNNNNNNNNNNND (0 ©O 00 00 00 00 00 00 50 GO 09 & 0 DOH & WOOO MO NOUN NNNNNNNNNNNHKNHNHNHNNP Price of Gas Per MCF (constant 1983 dollars) $3.75 3.75 3.75 3.75 3.75 375. a79 3.75 4.27 4.39 POBDFLRPOBNUENPOBONAUW UOBrUBNIYRFANNWOUNDDODWE NDADADDADDUVUUUUYE PEE NNNN WOAUW Dorw 8.07 8.27 8.46 8.67 8.87 9.08 9.30 9.52 9.74 9.97 10.20 10.44 10.68 10.93 1L.12 11.44 ura Total Cost of Gas m: 1983) $106 106 106 106 106 106 106 106 120 124 127 131 134 138 142 145 149 153 157 161 165 170 174 178 183 187 192 197 202 207 212 217 222 228 233 239 244 250 256 262 268 275 281 288 294 301 308 315s 323 330 on KENTCO - 68 - MILLIONS CONSTANT 1983 DOLLARS Exhibit 42 FOWAD RIN Oge@ Telb ReAginis Gu NATURAL GAS 1996—2045 340 Te tenn) OO a] 320 300 280 260 240 220 200 180 160 140 120 100 Vn taat iy an ty nl pete agent get olan eae ad 1996 2004 2012 2020 2028 YEARS 2036 2044 OOLNAy KENTCO Because we assume no real escalation in 0 & M costs during the period of analysis, the 0 & M cost for each of the 50 years in the analysis is $9,926,000. Total Costs of the Natural Gas Alternative Exhibit 43 summarizes the cost data developed in the pre- ceding discussion of the natural gas generation alternative. It shows costs for each category in the year in which they are incurred. The cost estimates, it should be remembered, are based on a series of assumptions we made concerning capital, fuel and 0 & M costs. These assumptions were made in order to arrive at a basis on which to compare the gas generation alternative with those of coal and hydro. They are not intended to achieve the accuracy required, for example, in the development of an actual power expansion plan. , In general, the assumptions we made tend to understate the cost of the gas generation alternative. There are four assumptions we believe favor the gas alter- native. Should actual conditions be different from those we assume, the gas alternative most likely would be more expensive and ,therefore, less attractive. One such assumption, namely that world gas prices will move as predicted by PIRA's projections for the increase in energy prices over time, is substantially more conservative than most of those used in previous analyses of the Susitna Project. If the PIRA forecast is lower than actual future prices, the cost of fuel for gas turbine generation will be higher than we project YEAR 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 201€ 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 Exhibit 43 SUMMARY OF ANNUAL COSTS OF NATURAL GAS ALTERNATIVE (MILLIONS CONSTANT 1983 DOLLARS) CAPITAL COsTs 501 uw °o oooooocooceo°oc°oc~°ceocoeceooooco coc CCCoOOoOrFCOCOCCCOCOOCOCCOCOCCCOOCOCOCOCCOCCCOCO0Of FUEL costs 106 106 106 106 106 106 106 106 120 124 127 131 134 138 142 145 149 153 157 161 165 170 174 178 183 187 192 197 202 207 212 217 222 228 233 239 244 250 256 262 268 275 281 288 294 301 308 315, 323 330 O&M COSTS 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 KENTCO TOTAL costs 617 116 116 116 116 116 116 116 130 134 137 141 144 148 151 155 159 163 167 1a 175 179 184 188 193 698 202 207 242 a7 222 227 232 238 243 249 254 260 266 272 278 285 291 298 304 311 318 325 332 340 KENTCO after 2004. For example, for every dollar rise in the price of a barrel of oil, there would be a 17¢ rise in the cost of one MCF of natural gas. This would translate into an increase in the annual cost of fuel for the gas alternative of $4.8 million dollars. A second assumption which may favor the gas alternative is that gas-fired generation will be significantly more efficient in the future than it is in the present. We assumed that all gas generation will use scale-sized, technologically advanced combined cycle equipment and that such equipment would supply 100 percent of the 3500 Gwh per year of base demand considered in this analysis. We are well aware the actual mix would be less efficient. The cost of equipment needed to produce 3500 Gwh per year in reality could be higher than the optimum plant Battelle listed. Third, we assumed the minimum possible price of gas to electric utilities between 1996 and 2004. It seems to us quite likely that producers and utilities will come to a compromise price somewhere between $3.75/MCF and the export net back price. Lastly, we assume that the TAGS line will come on-line at the earliest economically viable time. If TAGS were not built, Cook Inlet reserves would be exhausted by the year 2015. The Railbelt would be forced to import natural gas, not only for power generation, but also for home heating. Given the. scarcity of local supplies of gas until 2015, heist price most likely also would tend toward the world import price. ay Cy) KENTCO POWER GENERATION ALTERNATIVES FOR THE FUTURE - The Coal Alternative - Introduction Coal has been selected as one of three viable alternatives for the production of power in the Railbelt over the next 50 years. Having developed a base case need to produce 3500 Gwh of power in order to meet the needs of the Railbelt commencing in the year 1996, we now will e establish the actual and forecasted supply and demand for coal from the present through 1996, when the 3500 Gwh are first fully needed e determine the generating capacity needed to produce 3500 Gwh and maintain this base supply for 50 years e develop the following costs relevant to supplying the base supply: - capital costs - fuel costs - operations and maintenance costs @ present the total costs for each of the next 50 years for this base production e discuss those significant factors which can affect the forecast Supply of Coal Alaskan coal reserves are estimated to be well in excess of one trillion tons, or enough to satisfy the current world demand for 5,000 years. However, all but 10 billion tons of these reserves lie under the North Slope where they are not economically recoverable in the foreseeable future. - 93 - KENTCO Most of the recoverable coal is located in two major fields: (See Exhibit 44) Nenana 7.0 billion tons Beluga 2.4 billion tons The quality of coal in both fields is rated at 7500 BTU/1b which is commonly considered a lower grade of coal from the heat value standpoint. The Nenana Field is located south of Fairbanks immediately adjacent to the Alaska Railroad on the edge of Denali National Park. The coal in this field is leased mainly to the Usibelli Coal Company. Usibelli produces about 800,000 tons of coal per year from this area to supply the Fairbanks utilities, military bases and the University of Alaska. At today's rate of consump- tion the Nenana Field would last in excess of 8,000 years. Total 1982 consumption from Nenana by user was: User Tons U.S. Government 417,000 Fairbanks Mun.Utility 172,000 University of Alaska 58,000 Golden Valley Utility 140,000 Storage 7,500 Other 18 , 000 802,000 In 1981 the Usibelli Coal Company signed a fifteen year contract with the Suneel Corporation in Korea and the South Korea Electric Company to deliver 810,000 tons of Nenana coal annually. This is Alaska's first export opportunity and will double the amount of coal produced from Nenana. The coal will be shipped by rail from the mine to the planned Seward coal terminal for shipment to Korea. 2) YAN E KENTCO Exhibit 44 Railbelt Coal Reserves ES3y FIELDS HAVING SUPERIOR POTENTIAL OTHER FIELDS C — L- NENANA FIELD Wary ‘¢ ; s ~ JARVIS CREEK ~~ ~SHBROADy U \y PASS ) GLENNALLEN a dle 3 C are) a } x : re WY o SS BELUGA ian NY Yeo . S TNR Ss — NS FIELD VALOEZ sf zZ g EX. a Of wise sov"o 6S Fix f g - 95 - KENTCO Together with the current Nenana production, the Korean contract will bring annual projected production to 1.6 million tons annually. We have been advised that the mine could produce at the rate of 4 million tons per year for over 50 years. The Beluga Field is located west of Anchorage across Cook Inlet. This field is not currently in production and no firm dates have been set to commence production. The Diamond Alaska Company, a part of Diamond Shamrock Corporation (DAC), holds leases on one billion tons of the Beluga coal. DAC expects to begin production of the mine about 1990, although this is de- pendent on the appearance of sustained export opportunities. In order to support the required infrastructure and environmental requirements, DAC envisions producing 10 million tons annually. Placer Amex also holds leases on some of the Beluga Field. Placer anticipates producing at the rate of 5 million tons annually and is hoping the export markets will develop to allow initial mining by 1990. Beluga has the advantage of being near tidewater so that intrastate shipping costs will be reduced. Beluga has approximately the same quality coal as Nenana. Local Demand for Coal The local demand for coal in Alaska comes entirely from the generation of power. In 1983 Usibelli expected to ship 830,000 tons for this purpose. No other production exists within this scale and today only 15 percent of the power needs of the Rail- belt are met by coal. Were the Railbelt to utilize coal as its primary source of fuel to generate power, it could source this coal from either the - 96 - KENTCO Nenana or Beluga fields. In 1983, two million tons would be needed to generate the 3000 Gwh of power being consumed. In 1996, our base year, 2.64 million tons would be needed to produce 3500 Gwh. Since Healy (Nenana Field) can produce at the rate of 4 million tons per year, the power needs of the Railbelt in the base year could be met from either mine. This is true even after taking into consideration the Suneel contract for 810,000 tons. Several important facts must be kept in mind. First the Healy mine is restricted practically to producing no more than 4 million tons per year, due to geological restraints of the seams. This means that there would be little leeway in accept- ing additional export demand and only marginal opportunity for cost reductions based on greater volume. Second, the Beluga Field is not likely to be opened for production levels under 3 million tons per year. (See Exhibit 45) Assuming a 20 percent return on investment, the cost per ton difference between pro- ducing 1 million tons per year at $3.76 per MMBtu and 3 million tons per year at $2.65 per MMBtu is a substantial 30 percent. The world coal market into which Alaska coal must be sold continues to grow, albeit at a slower rate recently, due to the world recession. Growth will be dependent principally on the future demand for electric power generation modified by the conversion worldwide away from such a heavy dependence on oil for both utility and industrial usage. The Japanese and the Europeans remain concerned about future oil costs and have established national policies which assure a continual and steady growth for coal and its supply. Since domestic coal production =197) Exhibit 45 BELUGA AREA HYPOTHETICAL MINE Summary of Selected Data (1983 Dollars) Case l Production Rate per year(tons) 1,000,000 Mine Life at full production (yrs) . 30 Average Stripping Ratio (BCY/ton) eos Personnel (Average) Operating 81 Maintenance 74 Salaried 33 Total 188 Tons per man-shift (average) 2S Initial Capital Investment Initial Capital Investment per annual ton Life of Mine capital required $101,041, 000 $101.04 $183 ,027,000 Average Annual Operating Costs (Per Ton) Drainage Control & Reclamation $0 .60 Stripping 9.19 Mining and Hauling Coal el Coal Handling & Transporting 305 Haul Road Construction & Maint. 1.24 General Mine Services ree. Supervision & Administration 296 Production Taxes and Fees 0.35 Total Cash Costs $19.72 Average Depreciation of Total Capital 6.10 Average Total Cost S25 R82 Average Coal Prices (Per Ton) At 7, Rate of Return $40.85 At 15% Rate of Return 47.99 At 20% Rate of Return 56.40 Average Coal Prices (Per MMBtu) (a) At 10% Rate of Return SE 72 57/72 At 15% Rate of Return BE20 At 20% Rate of Return 357.0 Note: (a) Assumes 7,500 Btu/1b. Source: Mining Cost Estimates and Coal Prices Beluga Area Hypothetical Mine Paul Weir & Co., 1983 = 98: - KENTCO Case 2 3,000,000 30 5.89 194 176 56 426 28.2 $186,321,000 $62.11 $353,450, 000 aoe Oe - 08 i: .65 oa) . 64 SD) $15 512 S97. $19.09 OrPOCOrRFr®AO $28.52 33092 7190 eS -65 NNF KENTCO in these same large consuming countries is limited by either economics and/or good quality reserves, growth in the importation of coal is expected. (See Below) World Demand The markets for imported coal for the rest of this century are estimated as follows: International Coal Trade C Thermal coal ) Importers 1985 1990 2000 (million of tons) Europe 98 148 290 Japan 20 Spe 100 Other Pacific Rim 23 49 98 Central & South America 6 8 LZ Other 1 L 5) U.S.A. 1 6 30 T49 262 545 Source: For 1985-1990 - Shell Coal International Ltd. For 2000 - Data Resources Incorporated While worldwide conversion to coal accelerated along with the rapid rise in oil prices during the 1970's, more recently it has slowed considerably. In 1981 coal forecasts for 1985 targeted consumption by the Pacific Rim countries at 60 million tons per year (TPY) and 100 million tons by 1990. Now, two years later, those same forecasts are projecting 40 million tons in 1985 and only 50 million by 1990. Along with the slowdown in demand has come more stringent buyer requirements concerning quality. The most noticeable is Japan's insistence on lower moisture content and higher heat values. In particular, this will prove troublesome for Alaska coal with its high moisture and low BTU levels. - 99 - KENTCO As demand has slowed, worldwide competition has grown and spot prices have slumped. Supplier countries are developing new mines, modernizing transportation systems and extending sales efforts. Indicative of this are: South Africa's export capability at Richards Bay which has doubled and will do so again to 44 million TPY by 1985. Plans are being made now to add even more by 1990. Poland's national policy of selling coal at any price so that they may continue to earn foreign exchange. Australia and Western Canada are rapidly adding to their production capability. The Lower 48 can increase production by as much as 20 percent in any given year. Starting in the early 1990's, certain Pacific Rim coun- tries themselves will be producers for export. Thirty-one coal projects were begun in China during the first eight months of 1983 with a combined annual output of 16 million metric tons. An additional 19 new mines will be started by the end of the year bringing the total new on-line capability this year to 30 million metric tons per year. Much of this new capacity is slated for export to other Pacific Rim countries and will compete very strongly with other and more traditional sources of supply. Along with increased coal availability can be expected increased price competition. Japan represents 15 percent of the world export market for coal, is its most sophisticated buyer and is an accurate bell- weather of prices. Exhibit 46 indicates what share of that market is held by which exporting nations. According to the Greenley Energy Corporation, U.S. coal generally is the highest cost coal delivered to Japan and enjoys a significant market only because of Japanese interest in maintaining a position in a number of diverse and stable markets. - 100 - KENTCO Exhibit 46 PROJECTED NATIONAL SHARES OF JAPANESE COAL MARKET FOR IMPORTS IN THE YEAR 1990* Nation Million Tons Percentage Australia QS 41.8 Canada 6.0 Ls United States 78 553) China Si. 16.0 USSR 216 5.6 South Africa Dyas 4.2 All Others 256 Ye: TOTAL 50.7 100.0 * Includes steam coal and metallurgical coal Source: Mitsubishi Research Institute, 1982 "Future Energy Demand in East Asia" Updated by Greenley Energy Corporation, 1983 - 101 - KENTCO Price of Coal The viability of Alaska coal both as a source of energy at home and as an export commodity is totally dependent on price and future demand for coal in the Pacific Rim countries. More than adequate supply exists as does a growing demand. Without a world market share to bring world price levels to Alaska coal, it now is being priced on production cost basis, i.e. cost plus a profit. This will not change unless the coal can be proven to be competitive on the world market. In such a case, the price would rise to meet the net back from world price and this would apply to all coal extracted in the state, includ- ing that used locally. The current long term contract price for coal delivered from the Usibelli mine at Healy (only operating mine in Alaska) to its customers is: Golden Valley Electric Assn. (Fairbanks) $1.16 MMBtu Fairbanks Municipal Utility (Fairbanks) $1.35 MMBtu Were contracts to be let today by Usibelli, Battelle estimates and local producers confirm that the price would be $1.50 MMBtu, based on a 50 percent operating capacity of the Healy mine or $1.40 MMBtu if capacity rises to 100 percent.* As previously stated, the Beluga mine might be opened either if Alaska became truly price competitive in the world markets or, if the state decided to use coal for most of its Railbelt Power needs. The minimum market price required for Beluga coal can be estimated by the price it would take to put Beluga into produc- 1, * Battelle Pacific Northwest Laboratories, "Railbelt Electric Power Supply Alternatives Study", Volume XIII, 1982 - 102 - KENTCO tion. This production cost has been estimated by Paul Weir and Company, a well known Chicago-based mining and geological con- sultant. In their work, a Beluga area hypothetical mine was modeled and estimates of cost were assumed including a rate of return for the investors. Based on a minimum rate of return of 10 percent, Weir estimated a delivered coal price of $1.90 MMBtu. Even at this very low rate of return, Beluga is not competitive with Healy ($1.40 MMBtu) for consideration as a source of fuel for Railbelt power generation. Since neither mine is competitive currently in the world market, even with the annual production volumes contemplated, we will use lower priced Healy coal as the logical and conservative choice for making our comparisons of coal, gas and hydro options to produce Railbelt power. In all likelihood it will be necessary to take into consider- ation some transportation costs associated with moving coal from the mine to acceptable generating sites. Below is a table of acceptable power plant locations selected by Battelle. This is not an exhaustive list nor does it consider moving coal from Beluga since it was assumed generating plants at this location would be sited at mine mouth. Therefore, this table calculates only the cost of moving coal from Healy.* Potential Transport Healy Coal Total F.0O.B. Plant Destination Charge Price Price Plant Site (Per MMBtu) (MMB tu) (MMBtu) Nenana $.32 $1.40 $1.72 Willow 51 1.40 1.91 Matanuska - 60 1.40 2.00 Anchorage .70 1.40 2.10 Seward .78 1.40 2.18 * Battelle, 1982 - 103 - KENTCO Even taking into consideration these transportation charges, the Healy option is the more attractive unless a very strong case could be made for hauling of coal to Anchorage. Such a case seems unlikely since the new intertie with expansion can serve the entire Railbelt from acceptable sites closer to mine mouth. In selecting the Healy mine (Nenana Field) for our compari- son, we are well aware that given the growth of demand projected in the Demand section, the Nenana Field would have insufficient capacity starting in 2013. However, were we to base our compar- ison on Beluga, the costs assocated with producing power from the coal alternative would have been higher from the start. Therefore, if any option other than coal, i.e. gas or hydro, proves to be most cost-effective, the comparison would only serve to further disadvantage coal if we selected Beluga costs. If coal does prove to be the most attractive option we will recal- culate from a Beluga base to see if the higher costs would affect the conclusion. Capital Costs Battelle, in its Railbelt Energy Alternatives Study, developed the economics of a 200 Mw, coal-fired steam electric plant as the most efficient means of producing power from coal. The tech- nology is well proven and the capital costs of bringing such plants on-line can be calculated with reasonable certainty. The specifications and costs of such a plant are depicted in Exhibit 47. Two hundred megawatts was chosen as the optimal size plant in terms of producing power at the most economical cost. - 104 - SOT Construction Labor Construction Equipment Equipment Permanent Total and Insurance Supplies _ Repair Labor Rent Materials Subcontracts Direct Cost 1. Improvements to site $ 350,000 $ 2,100 $ $ 901,000 $ 110,000 $ $ 1,363,100 2. Earthwork and Piling 2,100,000 13,000 5,400,000 16,000 7,529,000 3. Circulating Water System 2,561,000 174,200 2,391,000 1,235,000 11,500,000 17,861, 200 4. Concrete 5,982,000 540,000 1,091,000 2,387,000 10,000,000 S. Struct.Steel/Lift Equip./ Stack 1,757,000 7,155,000 8,912,000 6. Buildings 682,000 800,000 1,482,000 7. Turbine-Generator 1,800,000 19,500,000 21, 300,000 8. Steam Gen. & Accessories 15,662,000 138,000 12,000 21,800,000 37,612,000 9. Air Quality Control System 12,400,000 27,100,000 39,500,000 10. Other Mechanical Equipment 8,950,000 8,950,000 11. Coal and Ash Handling 1,937,000 18,000 150,000 5,785,000 7,890,000 12. Piping 14,435,000 9,000,000 23,435,000 13. Insulation & Lagging 441,000 46,000 11,000 1,049,000 1,547,000 14. Instrumentation 3,000,000 3,000,000 15. Electrical Equipment 12,720,000 1,150,000 800,000 18,000,000 32,670,000 16. Painting 1,142,000 58,000 25,000 575,000 1,800,000 17. Off-Site Facilities 4,827,000 3,600,000 3, 260,000 11,687,000 18. Waterfront Construction N/A 19. Substation-Switchyard 1,623,000 34,000 143,000 3,017,000 4,817,000 20. Indirect Const. Cost and Architect/Engr. Serv. (b) 54,943,000 42,560,000 2,882,000 2,617,000 9,000 103,011,000 Subtotal $135, 362,000 $44,733,300 $2,882,000 $17,141,000 $132,748,000 $11,500,000 $344, 366,300 Contractor's Overhead and Profit 21,000,000 9,000,000 30,000,000 Exhibit 47 TOTAL COST OF BRINGING 200 Mw, COAL-FIRED, STEAM GENERATING PLANT ON-LINE (a) Contingencies 47,000,000 TOTAL PROJECT COST $421, 366,300 - 1982 $448,754,000 - 1983 % N/A = Not Applicable (a) The project cost estimate was developed by S.J. Groves and Sons Company. No allowance has been made for land and land rights, client charges (owners administration), taxes, interest during construction or transmission costs beyond the substation and switchyard. (b) Includes $40,816,000 for construction camp, $31,300,000 for engineering services and $30,895,000 for other indirect costs including construction equipment and tools, construction-related buildings and services, nonmanual staff salaries and craft payroll related costs. pource: Battelle 1982 OOLNSM KENTCO A profile of a 200 Mw plant is as follows: Plant & Transmission Costs $547,161,000 (1983 dollars) Plant Utilization Factor 85% utilization Total Power Per Year 1489 Gwh Fuel Needs Per Year 1 million tons coal Plant Life 30 years To calculate the capital cost of producing the 3500 Gwh in 1996 that makes up our base case, we multiply 2.35 times the cost of 200 Mw plant ($547,161,000) to find the total capital cost of $1,285 million. Fuel Costs As we have shown previously, electric power generating plants calculate the delivered cost of energy for fossil fuel in dollars per million Btu's. It is necessary to convert this cost to dollars per ton of coal as this is how coal will be purchased and used in their plants. Beluga and Nenana coal averages 7500 Btu per pound or 15 million Btu's per ton (2,000 lb. x 7500 Btu/1lb). The price of coal was calculated by Battelle and Harza-Ebasco. Their analysis starts with the average de- livered price of $1.95 per MMBtu at the projected plant sites and escalates by 1.5 percent per year to reflect increasingly more expensive extraction costs. By this method a price of $2.37 MMBtu in 1996 has been calculated. Given this price of $2.37 per MMBtu, as it is projected to be in 1996, then multiply- ing 15 x $2.37 will give the price of a ton of coal. 15 x $2.37 = $35.55 per ton.* aa We use the word tons in the Alaska market to refer to short tons (2,000 lbs) and in the export market to refer to metric tons (2,200 lbs). - 106 - KENTCO The cost of coal to operate one 200 Mw plant for one year is: 1,000,000 tons X fuel cost (MMBtu) X 15 = annual cost. Since we calculate 2.35 plants (3500 Gwh will be needed in our base year of 1996) then the total annual cost of fuel for the Railbelt would be 1,000,000 tons X price per MMBtu X 15 X 2.35 = total annual fuel costs to produce 3500 Gwh of power. Fuel costs derived by the formula above, from 1996 to 2045 are shown in Exhibit 48. Operations and Maintenance Costs Operations and maintenance costs consist of staff costs and variable costs that include supplies, scheduled maintenance, materials, etc. These were calculated by Battelle for a 200 Mw plant. Staff costs $3,660,000 per annum Variable costs 971,000 per annum $4,631,000 per annum (constant 1983 dollars) For 2.35 plants (3500 Gwh)total operation and mainten- ance costs is estimated at $10,882,000. Variables The environmental issues surrounding the generation of power from burning coal are as yet unresolved. Therefore, we are not able to estimate any costs that might be associated therewith. - 107 - Year 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 Exhibit 48 ANNUAL FUEL COSTS OF THE COAL ALTERNATIVE Nenana $ 83. 84, 85. 23 88. 89. SL. 925 93. 95. 96. -26 99. 101. 102. 104. 105. 107. 109. 110. oS 114. 115. -48 ELOe 121. -85 124. 126. 128. 130. E32. 134. 136. 138. 140. 142. 144. 146. 149. 151. 153. 155.0 158. 160. 163. 165. 167. 170. Lar 112 117 122 42 67 94 54 86 21 58 97 38 81 73 75, 29 85 44 05 69 04 75 24 03 69 56 46 39 34 33 34 39 46 ah 1a 88 08 32 59 89 23 60 ol 46 94 46 02 108 (millions constant 1983 dollars) Beluga $113. TTS), 116. 118. 120. 122. 23). 12555 127. 129; 131. 133. 1335 137. 139. 03 143. 146. -20 150. 152; 154. Loe L597. 162. 164. 166. 169. L721 141 148 174 L793 231 36 06 79 54 32 12 95 81 70 62 56 53 54 57 63 85 ol 42 68 97 30 65 0S 48 95 45 99 ao 177. -85 182. 185: 188. 190. 193. 196. 199; 202. 205. 208. zit: 215% 218. 2a 224, 228. 20) 235. 19 55 29 07 89 5 66 61 60 64 72 85 03 26 53 85 23 is KENTCO KENTCO Total Costs of the Coal Alternative Total coal costs are portrayed in Exhibit 49 and consist of $1,285 million in capital costs, an escalating fuel cost measured in 1983 dollars (See Exhibit 48) and an operations and maintenance cost of $10,882,000 in 1983 dollars. - 109 - YEAR 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 SUMMARY OF ANNUAL COSTS OF COAL Exhibit 49 ALTERNATIVE (MILLIONS CONSTANT 1983 DOLLARS) CAPITAL costs 1285 128 ecoooooooocoooceococooCocCoOoOMoOoOCoCOOOCOCCOCOOCCOCOOCOCCCCOCCOCCeCeCCCOCOO FUEL costs 83 85 86 87 89 90 91 93 94 95 97 98 100 101 103 104 106 107 109 lll rr2 114 116 117 119 Lan 123 125 27 128 130 132 134 136 138 140 143 145 147 149 mS 154 156 158 161 163 165 168 170 a3 - 110 - O&M COSTS 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10. 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 10.9 KENTCO TOTAL costs 1379 96 97 98 99 101 102 103 105 106 108 109 Bee 112 114 Lis 117 118 120 122 123 E25 127 128 130 132 134 136 Si 139 1426 143 145 147 149 aap TS53 156 158 160 162 164 167 169 172 174 176 a9 181 184 KENTCO POWER GENERATION ALTERNATIVES FOR THE FUTURE - The Hydroelectric Alternative - Introduction Hydroelectric energy has been selected as one of the three viable alternatives for the production of power in the Railbelt over the next 50 years. Having developed a base case need for 3500 Gwh to meet the needs of the Railbelt commencing in the year 1996, we will e describe the operational, authorized to be built, and planned hydroelectric projects available to generate power in 1996 e determine the generating capacity needed to produce 3500 Gwh of power and maintain that base supply for 50 years beginning in 1996 e develop the following costs relative to supplying the base supply of power: - capital costs - operations and maintenance costs @ present the above costs for each of the next 50 years e discuss those factors which can affect the forecast Hydroelectric Installations in the Railbelt Hydroelectric energy has been one of Alaska's most debated and least used resources. Beginning in 1946, the Bureau of Reclamation (BOR) conducted studies that led to a Statewide reconnaissance of hydroelectric sites in Alaska. Based on their studies, the Eklutna Project was authorized for construction and built in 1950. Subsequently, the BOR began extensive studies of the Susitna River and recommended a feasibility study be under- taken in 1953. As a result of that study, BOR recommended that - lll - KENTCO the Susitna Project be authorized by the federal government for construction. In the late 1950's, a separate study was undertaken by the Army Corps of Engineers which recommended an alternate large dam strategy based on the Rampart Project to be built on the Yukon River. As a result, the Susitna Project was tabled in 1962 pend- ing the outcome of the Rampart analysis. The decision not to proceed with Rampart was mady by the Army Corps in 1971. The Army Corps had also undertaken some studies for smaller projects. Only the Bradley Lake Project near Homer was authorized for construction (under the 1962 Flood Control Act). However, it was never built because cheap natural gas in Cook Inlet became the dominant source for power supply in the Railbelt. Between 1962 and 1967 the BOR undertook an extensive study to inventory possible Statewide hydro sites. These results, published in 1967, included major evaluations of the Susitna, Rampart, Wood Canyon, Yukon- Taiger and Woodchopper projects. These studies reaffirmed the previous findings that the sites at Susitna and Bradley Lake were feasible. Little progress was made in the 1960's and 1970's in actually constructing dams because the availability of cheap natural gas as the energy source for the Railbelt precluded their need. The cancellation of Rampart and the subsequent Arab oil embargo with resultant oil price rises established renewed interest in Susitna as well as other hydro projects. Currently the Railbelt has two small operational hydroelectric dams. First is the Cooper Lake hydro station, owned and operated - 112 - KENTCO by Chugach Electric Association which produces 16 Mw of power. Second is the Eklutna Water Project which is owned and operated by the Alaska Power Administration, an agency of the federal government and not to be confused with the Alaska Power Authority (APA), a State agency. Eklutna produces 30 Mw of power which is sold to Chugach Electric and to the Matanuska Electric Association. Two additional projects have been recommended to the State government for construction and, according to the APA, will be funded, built and on-line in 1988. The Grant Lake project will produce 7 Mw of energy and is to be owned by the State. The Bradley Lake hydroelectric project will deliver 90 Mw of power to the Railbelt and also will be owned by the State. Total existing and authorized for construction Railbelt hydroelectric generation is shown below. Existing and Authorized Railbelt Hydroelectric Generation Capacity Total Annual Energy Capacity (Mw) Qutput (Cwh) Eklutna 30 118 Cooper Lake 16 34 Bradley Lake (1988) 90 315 Grant Lake (1988) 7 19 1988 Total 143 486 Source: Harza-Ebasco Update, September 1983 The existing dams (Cooper Lake and Eklutna) generate about 5 percent of the current annual Railbelt power needs. The addi- tion of Grant and Bradley Lakes will increase the total hydro generation percentage to 16 percent of the existing system. In the Demand section of this report we projected, by use of the ISER model, that the demand for power in 1996 will be - 113 - KENTCO 3925 Gwh. The hydroelectric capacity that will exist at that time from the above projects will produce 486 Gwh. Therefore, the Railbelt will need an additional 3439 Gwh of hydroelectric power by 1996. Again, since this timing coincides with the planned completion of the Watana Dam and since that dam will produce 3500 Gwh per year of power we are using 3500 GWH per year in 1996 as our base case. This allows proper comparison of the three viable alternatives; coal,zas and hydro, all of which can be cal- culated to produce the same power output. Proposed Railbelt Hydroelectric Projects There are currently only two new hydroelectric projects proposed in the Railbelt. The first, the Chakachamna Hydro Project, would produce 1300 Gwh per year. The second, the Susitna Hydroelectric Project, could supply 7000 Gwh per year. However, it is vitally important to remember that Susitna is two separate projects. The first, Watana, would produce 3500 Gwt, sufficient to supply the required Railbelt demand in 1996. The second, Devil's Canyon, also would produce 3500 Gwh per year. It should be built, however, when and if the demand growth is sufficient to warrant this additional amount of power. ahekechamns | Byero Eee Chakachamna Lake is on the west side of Cook Inlet. The project calls for the diversion of water from Chakachamna Lake via a tunnel to a power plant on the McArthur River. The project would have an installed capacity of 330 Mw, average annual energy generation of 1300 Gwh and is estimated to cost $1,438 million in 1983 dollars. - 114 - KENTCO Chakachamna Hydroelectric Project Installed Capacity - Mw 330 Total Capital Cost including transmission and interest during construction $1,438 million Operations & Maintenance Costs $2.0 million annually Annual Energy Production 1300 Gwh Source: Harza-Ebasco Update, September 1983 Should Watana be built, there is little need for Chakachamna. If Watana is not built, then Chakachamna would be reviewed as a potential partial solution to meeting Railbelt power demand. Susitna Hydroelectric Project The Susitna qyerosiechnes Project was first proposed in 1948 and has experienced a number of false starts since its initial consideration. Recommended for authorization by the Bureau of Reclamation in 1960, Susitna was later tabled pending the outcome of the Rampart Project. With the cancellation of Rampart in 1971, Susitna became the premier potential large scale hydroelectric project in the state. Acres American, an engineering firm, was commissioned in 1979 by the Alaska Power Authority to conduct a detailed feasi- bility study in order to determine the technical feasibility, economic viability and environmental impact of an optimal develop- ment in the Susitna basin. In March 1982, Acres published their report and concluded that Susitna was feasible on all three counts. The APA prepared a submittal to the Federal Energy Regulatory Commission in support of a license. This submittal was made in July 1983, hearings will commence in March 1984, and - 115 - - OIT - Exhibit 50 fh “ so Ors. \! \ a C ch N % \ Wa hae - eos vf s AY ihe Bt lod NF LM reget oT De YF oe Zr TRANSHESSION — | LIES |) x * Ye SRR o% § ago ay AER DAL en, eee eh OO POWER Se NON INSTALLATIONS eee | 7 \ a . Ss xe => ONSTRUCTION \D VILLAGE = : CONSTRUCTION: VILLAGE " A rAd Lay <==: 4 p , way 4 5 j > \ BA %, F ¢ rf S an ~79 2000 5 J ° 2000 4000 DEVIL CANYON SITE FACILITIES §=‘—~4;-——+ —- WATANA SITE FACILITIES ‘-—j2,—— | PLATE & [ PREFERRED DEVELOPMENT PLAN OOLNSY KENTCO a decision by FERC is expected in late 1986. If the license is granted, the State of Alaska will have a final opportunity to decide whether or not to proceed with the Susitna Dam. In this section we will summarize the information on Susitna and, based on extensive work done by Battelle, Harza-Ebasco, the APA and others, determine and present the capital costs, fuel costs and operations and maintenance costs related to the Watana Dam in our base year of 1996, when its full output is expected to be utilized, and for the 50 years beyond.* Description of the Project The Susitna Dam sites lie about 180 miles north of Anchorage in the Susitna River basin. Two dams would be established, one at Watana and one at Devil's Canyon. a. Watana The Watana Dam will be an earthfill structure with a maximum height of 885 feet. The dam will be among the highest in the world. Its 885 foot height will exceed that of the highest completed embankment dams in North America (Mica Creek in British Columbia -- 794 feet, and Oroville in California -- 771 feet). Watana - Pertinent Data Dam type Zoned earthfill Dam height 885 feet Reservoir length 48 miles Reservoir capacity 9,500,000 acre feet Installed capacity 1020 Mw # of Turbine Units 6 Average Annual Energy 3450 Gwh Source: Acres American, Inc. - Susitna Hydroelectric Project, 1982 * We have decided to use only 50 years life of the dam for this analysis. It is generally agreed the life could be considerably longer. . - 117 - KENTCO bul Devil "s'Canyon The Devil's Canyon dam will be a concrete structure with a maximum height of about 645 feet. The power output of Devil's Canyon is almost the same as Watana. Watana 3450 Gwh per year Devil's Canyon 3340 Gwh per year As we will show later, the cost of Devil's Canyon is less than half the cost of Watana because most of the infrastructure development (roads, control houses, transmission system) are all costs associated with Watana. Exhibit 50 details the location of both dams. Devil's Canyon can be built only when needed. Should the need never arise, the dam never need be built. Devil's Canyon - Pertinent Data Dam type Concrete arch Dam height 645 feet Reservoir length 26 miles Reservoir capacity 1,092,000 acre feet Installed capacity 600 Mw # of Turbine Units 4 Average Annual Energy 3340 Gwh Source: Acres American, Inc., 1952 Capital Costs The capital costs of Susitna occur at the time each dam is constructed. As presently planned, Watana would come onstream in 1996 and Devil's Canyon at some later date to be determined by need. The capital costs of each include: - 118 - KENTCO roads land rights power plant dams and spillways mechanical and electrical equipment transmission contingencies engineering and administration interest during construction design modifications approved by the APA board in Fall 1983 ; The cost estimates for both dams were originally done by Acres American and have since been refined by Harza-Ebasco. The capital cost upon completion of each is expected to be: Watana $3 43billion (1983 dollars) Devil's Canyon $1.62 billion (1983 dollars) Fuel Costs There are no fuel costs for hydroelectric projects. The source of fuel for the project is the Susitna River. Stream flow volume and reliability have been documented by the Alaska Power Authority. Source: Harza-Ebasco Update, September, 1983 Operations and Maintenance Costs Operations and maintenance costs are outlined in Exhibit 51. O & M costs for Watana will be $8,520,000 (stated in constant 1983 dollars) beginning in 1996 and for each year thereafter through 2045. Source: Harza-Ebasco Update, September, 1983 Total Costs The total capital, fuel and 0 & M costs for the Susitna Project are: - 119 - KENTCO Capital Costs (1983 dollars) Watana $3.43billion (beginning in 1996) Devil's Canyon $1.62 billion(beginning when re- quired after 1996) Fuel Costs No fuel costs Operations and Maintenance Costs Watana* $8.52 million (annually beginning in 199 Devil's Canyon* $2.51 million (annually beginning when required after 1996. Factors Affecting Capital Cost and 0 & M Forecasts There are few factors affecting the forecast of capital and O &Mcosts. As there are no fuel costs and 0 & M costs are relatively small, the major elements of uncertainty all relate to the dam construction and its financing. Capital Cost Overruns The issue of cost overruns has been studied by the various engineering firms under contract to the APA. Originally, Acres American calculated that there is a 7 percent chance that cost overruns would be experienced that would make their lowest cost thermal alternative more attractive. The Acres American estimates were verified as reasonable by Harza-Ebasco. Unlike nuclear plants, hydroelectric dam construction is an established technology. As an example, the Churchill Falls Dam in Canada will be completed in 1985 for slightly less than the estimate of $10.5 billion made in 1977. fe * The initial higher operations and maintenance costs have been assumed for the life of the project. - 120 - KENTCO We have been informed by the staff of the APA that Harza- Ebasco have been instructed all along to calculate their costs in a most conservative manner. Harza-Ebasco confirms that this has been done. - 121 - - 21 Exhibit 51 SUSITNA PROJECT OPERATION AND MAINTENANCE COST ESTIMATES ($1, 000/yr) Watana + Devil's Canyon Eventual Project Labor Expenses Subtotal ; Labor Expenses Subtotal! Labor Expenses Subtotal Power & Transmission 3300 990 4290 625 500 1125 2740 990 3460 Contracted Services? 900 900 480 480 1050 1050 Townsite Operations 625 180 805 400 55 455 285 180 465 Environmental Mitigation 1000 1000 Contingency (15%) 1045 310 895 Total January 1982 dollars 8040 +2370 6870 Escalation to 1983 dollars 6% 480 140 410 Total January 1983 dollars 8520 +2510 7280 1 For first five years of operation of each development; total of 10 years 2 Includes annual maintenance services, major maintenance overhaul, helicopter service and road maintenance. Source: arza-Ebasco Update, September 1983 OOLNAY KENTCO POWER GENERATION ALTERNATIVES FOR THE FUTURE - Economic Comparison of the Alternatives - This section will apply economic analysis to the three forms of power generation laid out in the preceding pages -- coal, gas and hydroelectric. The goal will be to determine the lowest cost means of producing power in the Alaska Railbelt. The section first will describe what economic analysis is, then review the previous cost data for coal generation, gas generation and hydroelectric generation. Following that we will show the analysis results, make comparisons and draw conclusions. Economic Analysis Economic analysis is a method of comparing the costs of several alternatives over a period of time. It synthesizes large amounts of cost and benefit information about each alternative into a single number for easy comparison with other alternatives. It allows expenditures to take place at very different times over the period being analyzed and yet be compared in today's terms. The form of economic analysis we will use employs two steps. The first is to set out, year by year, the costs of power gener- ation for the entire 50 year period of the analysis, 1996 through 2045. The result is a long series of cost information for each alternative. Those costs include the cost of building a gener- ation plant,the cost of operating and maintaining it and the cost of fuel. Since we will be using this analysis to compare the alternatives, any cost which is identical in all alternatives, ~-1235- KENTCO such as the cost of building and maintaining local distribution lines, is not part of the analysis. The second step in the economic analysis is a present value calculation. Costs shown from the first step are adjusted so that costs which occur earlier are given more weighting, or importance, and costs which occur in later years are given less weight. The method used to give less weight to future costs is to deduct a percentage from the cost, once for each year into the future that it occurs, such as 3 percent per year, 10 percent per year, or some other percentage. For example, using a 3 percent discount rate, a one dollar cost which occurs one year into the future is given a discounted value of approxi- mately 97 cents, and if two years into the future approximately 94 cents, and so on. The future results are thus "discounted", and the percentage used is called the discount rate. The discounted results for each alternative are then added up, and the resulting total is a single number which describes the cost of that alternative. This result is often called the "present value" of the costs. The use of a discount rate in an economic analysis shows a preference for costs which occur later rather than sooner. The most important reason for using a discount rate is the theory that money, until it is needed to pay costs, can be invested profitably. For example, if a $100 cost occurs immediately, there will be no opportunity to earn money with that $100, since it will be spent immediately. If the same $100 cost were to occur in five years, a smaller investment, say $85 or $90, could go into a savings account now, and together - 124 - KENTCO with five years' interest could pay the $100 expense. The discount rate attempts to capture this effect, and the $85 or $90 is a "present value" of the $100 future cost. Other reasons for using a discount rate can include the higher uncer- tainty that future costs will actually occur, or a general preference for delaying costs. It is very common for alternatives to be such that one alternative has higher costs in the early years but much lower costs in the later years, when compared to another alternative. When this occurs, the discount rate becomes highly important. For example, when analyzing an alternative with low early costs and very high costs later, high discount rates can make the larger, but later, costs look less worrisome and can show the alternative to be a good choice, while lower discount rates can keep the value of the future costs large, and make the alterna- tive as a whole look less attractive. This is the case in the economic analysis of gas, coal, and hydroelectric power. Hydroelectric power has large costs which occur early, and low costs later. Coal and gas have lower early year costs, and higher costs in later years. Our analysis will take into account the timing of these costs by showing the present value of the costs of each. The reader will then be able to make comparisons for the entire life of the project. Reasons for a 1996 Start Date Our economic analysis and our subsequent financial plan assume a 1996 start date for Watana. It will be remembered - 125 - KENTCO there are three reasons for choosing 1996. First, 1996 corresponds to the first year in which total Railbelt electricity demand is sufficient to fully utilize Watana and all other anticipated hydroelectric generation in the Rail- belt, including Grant Lake and Bradley Lake. Because there would be little cost savings in retiring existing hydroelectric generation (as opposed to the major fuel savings which accompany the retirement of thermal generation) all projected output from hydro should remain available after the advent of Watana. The RED model of forecasted power demand indicates all of this hydro power should be utilized by 1996. Second, 1996 corresponds to the expiration of contracts for low cost gas from Cook Inlet. Watana would have a more difficult job in competing as the low cost alternative during the years that such low cost gas is available. Third, we believe that 1996 is a realistic start date for the project, allowing some slippage in the present APA schedule for a 1993 start. The present schedule calls for FERC hearings to be concluded under an accelerated or "fast track" plan, in late 1986. The present APA plan also calls for all engineering work to be completed so that construction can start immediately upon award of the FERC license. This may be possible, but we note that there is no design money in the proposed APA budget for FY 1985, thus leaving only FY 1986 for all engineering work. A slippage in either the FERC hearings schedule, the engineer- ing, or the award of construction contracts could lead to completion in FY 1995, hence a 1996 start date. - 126 - KENTCO Review of the Alternatives Our analysis of the gas, coal and hydroelectric alternatives uses a model which makes the behavior of generation costs easy to understand and to compare. It groups the principal costs of electric generation, and describes how each group of costs changes over the analysis period. We use a 50 year period for our analysis because it corresponds to the assumed life of the longest option, hydroelectric. Also, for ease of comparison between the alternatives, we are presenting costs for generating 3500 Gwh per year in each alternative, which corresponds to the average annaul energy output of the Watana Dam and is very close to the average annual energy output of the Devil's Canyon Dam. In our analysis we have adjusted the cost information to exclude the effects of inflation, so costs are shown in 1983 dollars. Assumptions behind the model follow. The Gas Alternative Gas has three components of cost: capital cost, fuel cost and operations and maintenance cost. All costs are based on the efficiency and cost structure of the 200 Mw plants described before, as that size has achieved the economies of scale avail- able from large size. Capital costs of a gas-fired generation plant occur twice in the 50 year analysis period. In the first year capital costs of $501,270,000 occur (in 1983 dollars). A gas-fired plant lasts 25 years, so in the 26th year capital costs occur again. At the end of the 50 year analysis period there is no useful life remaining on the gas plants, so there is no need to give a value for remaining plant life. - 127 - KENTCO ne Fuel costs provide for the consumption of 28.2,MCF of gas per year -- sufficient to produce 3500 Gwh of electricity. We projected gas costs in the analysis laid out on page 88. In 1996 through 2004 gas costs follow the cost of production of new gas and escalate with inflation. In the year 2015 the Cook Inlet gas field runs dry and gas costs rise to either the world price of gas or, if the TAGS line is built, the tidewater price of North Slope gas. In either case, world prices of gas directly influence the price of Alaska Railbelt gas after 2015 at either the world price or at a net back from world price, and the price escalates two percent per year reflecting the forecasted real increase in world price of energy. Operations and maintenance costsinclude the expense of personnel, parts, repairs and so on. It amounts to $9,926,000 in 1983 dollars and remains at this level in constant 1983 dollars in subsequent years. There is one significant additional effect of the choice of power generation which, if it occurs, is too large to be excluded from the analysis. If the TAGS line is not built and the corresponding export market not developed, then gas in Cook Inlet will remain production-cost priced instead of world-priced. The difference in cost of gas between world pricing and produc- tion pricing benefits the majority of the Railbelt which uses gas for heating. The longer the gas reserves last, the greater the benefit to the Railbelt residents. The use of gas for electric generation has the effect of shortening the life of the Cook Inlet gas reserves by about 10 years, and therefore, shortens the period of time during which - 128 - KENTCO homeowners can use low cost gas for home heating. After the reserves run out, the price goes up to the import price. This cost difference, if it occurs, is significant. If included in the analysis it adds approximately $1.1 billion in net present value terms to the cost of using gas for electric generation. (See Exhibit 52) The total present value of using gas to generate Railbelt power is $4,709 million. (See Exhibit 53) The Coal Alternative Coal has three components of cost: capital cost, fuel cost and operations and maintenance cost. All costs are based on the efficiency and cost structure of large plants. At 200 Mw, coal plants have achieved most of the economy of scale available from large size. The cost per kilowatt hour of coal-generated elec- tricity is no less for any size plant above 200 Mw; therefore, the capacity (472 Mw) needed to produce 3500 Gwh produces power at the same cents per kilowatt hour as the 200 Mw plant shown in the preceding section on coal. Capital costs are the costs of constructing the coal plant. They occur twice in the course of the analysis. In the first year capital costs of $1,285 million (in 1983 dollars) occur. This represents the initial construction of the coal plant. This plant lasts 30 years, so in the 3lst year capital costs occur again to construct a replacement plant. At the end of the 50 year analysis period, we give a value for the remaining 10 year life of the second coal plant. Fuel cost is based on the consumption of 2,350,000 tons of coal per year. We use Healy coal delivered to Nenana. The aHL29 i= So Oeta Exhibit 52 PRESENT VALUE OF EXTENDED COOK INLET GAS RESERVES Assumptions: TAGS line not built Cook Inlet gas is priced at production cost Cook Inlet gas used for heating but not power generation Non-Cook Inlet gas is priced at world price Depletion date of Cook Inlet reserves if gas not used beyond 1996 for power generation: 2024 Cook Inlet gas not competitive in world markets after 1998 All prices shown are in 1983 dollars Gas Price If Gas Price If Heating Gas Total Production-Cost World Price Consumption Savings Year Based (per MCF) Based (per MCF) Difference (BCF ) ($000,000) 2016 3-3.75 Soe $ 3.77 57.0 $ 214.9 2017 S519 7.67 3. 92 59.0 231.3 2018 S70 7.83 4.08 61.1 249.3 2019 a7 5 7.98 4.23 63.2 267.3 2020 575) 8.14 4.39 65.4 287.1 2021 a7. S31 4.56 67.7 308.7 2022 5515 8.47 4.72 70.1 330.9 2023 3375) 8.64 4.89 Ta 354.5 2024 a1) 8.81 5 l 380.0 . 06 75. Present value of gas not used for power generation, assuming a discount rate of 3% percent = $1,097 million. OOLNS» Exhibit 53 COSTS OF GAS GENERATION* - Annual Costs and Total Present Value Cost - Year 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 (millions constant 1983 dollars) Capital Cost 501.27 501. SCOCCOCOCOCOCOCOCOOOOCOOOOOOCOOCOCO: COOOCOOCOCOCOCOCOOCOOOCOOCOOCOCOOCOOCOOCOCOOCOOCODSD 27 Fuel, Operations and Maintenance Cost 112.14 112.14 112.14 112.14 112.14 112.14 112.14 112.14 126.11 129.92 133.30 137.25 140.92 145.15 149.10 152.77 156.72 160.67 164.62 168.57 172.52 176.75 181.26 185.21 190.57 194.80 199.60 203/55 209.19 213.42 219.07 224.43 229.50 23 241.07 246.43 252.36 258.28 264.21 269.85 276.34 282.83 289.31 295.80 302.86 309.63 316.68 323313 S29 5L9) 336.29 Present Value of Gas Generation Assuming a discount rate of 34 percent = $4,709 million. * 3500 Gwh per year of power alcpl KENTCO KENTCO cost of coal follows the analysis of the price of coal presented on pages 102-104, in which the cost of coal rises at 1.5 percent above the rate of inflation. First year (1996) cost of coal is $83,420,000 in 1983 equivalent dollars, then rises at 1.5 percent per year. Operations and maintenance costs are the costs for per- sonnel, parts and repairs, and so on. These costs appear constant in our inflation-adjusted analysis, but with the effects of inflation the costs would rise in nominal terms. Annual costs for operation and maintenance are $10,882,000 in 1983 dollars. (See Exhibit 54) The Hydro Alternative Hydroelectric costs are comprised of capital costs and operations and maintenance costs. There are no fuel costs associated with hydroelectric power -- the downward flow of water is free. Hydroelectric power would be provided by the two dams in the Susitna Project: Watana and Devil's Canyon. Watana is assumed to come on-line in 1996, while Devil's Canyon comes on-line at a future date when demand grows sufficiently to justify it, probably some date well beyond the year 2000. Watana capital costs include two elements: the construc- tion cost and the expense of retiring existing thermal utility generation. Construction costs amount to $3,432 million in 1983 dollars. Retirement of existing generation refers to the way in which Watana will replace much of the existing utility genera- tion the day that Watana starts generating power. Watana will - 132 - Exhibit 54 COST OF COAL GENERATION* - Annual Costs and Total Present Value Cost - Year 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 (millions constant 1983 dollars) Capital Cost 1285 ~ N RCCSCCDDOCCOCOCOCCOCOCOOCOMODDODDOOOOOOCOOCOOOOCCOOCOOCOOCCOCOOOCCOOOCOO ‘ S Fuel, Operations and Maintenance Cost 94 95 96 98 99 100 102 103 105 106 107 109 110 112 113 114 116 117 119 121 122 124 126 127 129 131 133 134 136 138 140 142 144 146 148 150 150 152 154 159 161 163 165 168 170 172 175 178 180 182 Present value of cost of coal generation assuming a discount rate of 34 percenc = $4,431 million * 3500 Gwh per year of power - 133 - KENTCO KENTCO provide 3500 Gwh of power in its first year of operation which, the year before, was provided by other means. The Railbelt utilities have borrowed to pay for their generation plants, and will be relying on revenues from the sale of power from those plants to retire the debt. Full use of Watana may require these utilities to reduce their usage of thermal plants. In this case, the utilities will need a replacement stream of cash flow to retire their thermal plant debt. Therefore, we assume in our analysis that Watana's capital cost will include the cash necessary to retire debt on surplus generation equipment. Whether such cash will be needed and whether it is properly a cost of the project will depend on needs at the time to provide for peaking capacity, standby generation and growth in demand. Utilities will have time between now and 1996 to plan for the retirement of their thermal plants, and to make the retire- ment coincide with the start-up of Watana. Therefore, the debt outstanding at the time of start-up may not be as large as it is today. Present long term debt, other than debt which is clearly to finance distribution lines (which will not be made obsolete) is approximately $247 million. The amount which will be outstanding at the start-up date of Watana is unknown, but for purposes of this analysis we assume it to be $200 million in the year 1996. It is entirely possible that our approach is too conservative as most of the thermal capacity could be required for peaking needs. The operations and maintenance costs for Watana cover personnel, repairs and various other small costs. The costs are $8.5 million in all years of operation (in 1983 dollars). Se134e= KENTCO Exhibit 55 COST OF WATANA HYDROELECTRIC GENERATION® - Annual Costs and Total Present Value Cost - (millions constant 1983 dollars) Operations and Year Capital Cost Maintenance Cost 1996 3432 8.5 1997 0 8.5 1998 0 8.5 1999 0 8.5 2000 0 8.5 2001 0 8.5 2002 0 8.5 2003 0 8.5 2004 0 8.5 2005 0 8.5 2006 0 8.5 2007 0 8.5 2008 0 8.5 2009 0 8.5 2010 0 8.5 2011 0 8.5 2012 0 8.5 2013 0 8.5 2014 0 8.5 2015 0 8.5 2016 0 8.5 2017 0 8.5 2018 0 8.5 2019 0 8.5 2020 0 8.5 2021 0 8.5 2022 0 8.5 2023 0 8.5 2024 0 8.5 2025 0 8.5 2026 0 8.5 2027 0 8.5 2028 0 8.5 2029 0 8.5 2030 0 8.5 2031 0 8.5 2032 0 8.5 2033 0 8.5 2034 0 8.5 2035 0 8.5 2036 0 8.5 2037 0 8.5 2038 0 8.5 2039 0 8.5 2040 0 8.5 2041 0 8.5 2042 0 8.5 2043 0 8.5 2044 0 8.5 2045 0 8.5 Present value of the hydro alternative, assuming a 34 percent discount rate = $3,631 million. * 3500 Gwh per year of power - 135 - KENTCO The total present value costs of the Watana Dam are calculated to be $3,631 million. (See Exhibit 55) Devil's Canyon costs include capital costs and operations and maintenance costs. Capital costs of Devil's Canyon include the same two com- ponents: construction cost and cost of retiring existing utility debt. Devil's Canyon construction costs are calculated by Harza- Ebasco to be $1,552 million in 1983 dollars. In retiring existing utility debt the situation may be very similar to that of the 1996 start-up of Watana. As demand grows beyond the capacity of Watana, utilities would need to build additional generation, probably thermal, until demand grew to a level sufficient to justify Devil's Canyon. Utilities would need to finance this additional generation with debt, and at the date of start-up for Devil's Canyon a means would have to be provided to pay off that debt. Utility debt retirement is assumed to be the same, inflation-adjusted, as for Watana. Exhibit 56_ Estimated Current Utility Long Term Debt Outstanding, Excluding Distribution Debt Debt Outstanding Chugach FFB (as of) June 30, 1983 127,903,000 Mun.of Anchorage " June 30, 1983 52,290,000 MEA CFC " Sept.26, 1983 7,810,000 GVEA FFB and CFC " Dec. 31, 1982 36,210,000 HEA CFC a Dec. 31, 1982 6,410,000 FMUS ad Nov. 30, 1983 9,000,000 246, 623,000 - 136 - KENTCO Devil's Canyon operation and maintenance costs are $2.5 million per year in 1983 dollars. These costs are lower than Watana's because the Devil's Canyon operation shares some Watana facilities. Results of the Analysis and Comparison Exhibits 52 through 55 show the costs of the gas, coal and hydroelectric power generation alternatives, year by year, and the present value calculations. It will be remembered that pre- sent value weighs earlier costs greater than later costs. Before presenting the findings, we need to define a term, "real discount rate". Real discount rate refers to the amount by which the nominal discount rate in the economic analysis exceeds inflation. It is a useful concept because it does not change if the assumptions about the future rate of inflation change, and the need to calculate inflation for economic analysis is eliminated. The concept of real discount rate is also used in several other Susitna studies. For consistency in our eco- nomical analysis, all costs for the alternatives are shown in real, or inflation-free dollars and our analysis uses a discount rate which is 3% percent above the rate of inflation, i.e. a real discount rate of 3% percent. This is the same rate used generally in the financial community as well as in other recent Susitna studies, including the Harza-Ebasco September 1983 Susitna update. Conclusions The economic analysis shows the 50 year costs of Watana to have a present value of $3,631 million. The present value cost - 137 - KENTCO of coal generation is $4,431 million. The costs of gas gener- ation are $4,709 million, if the TAGS line comes onstream in the year 2004. If the TAGS line comes onstream earlier, the cost of gas generation is greater because the price of gas rises to a net back world price sooner. If the TAGS line comes on- stream later or is not built, the cost of gas generation is again higher because of the lowered gas availability for heating. Therefore, the assumption that the TAGS line comes on in 2004 is a conservative assumption. In the case that the TAGS line is not built at all, the cost of gas generation is $5,806 million. (See Exhibit 57) When compared to the lower cost case for gas (TAGS line is built), Watana compares favorably at any real discount rate up to 4.9 percent. When compared to coal generation, Watana has a lower cost at any real discount rate up to 5.1 percent. When compared to the no-TAGS gas case, Watana has a lower cost at any real discount rate up to 6.1 percent. All of these rates are significantly above the norms of the present or past. Devil's Canyon is very low cost in comparison to any of the alternatives. For 50 years of operation, Devil's Canyon can produce 3600 Gwh of power for a total present value cost of $1,773 million. Costs of generation by thermal alternatives varies depending on the actual starting date of the Devil's Canyon Dam, but since their costs rise over time they will be similar to or higher than their costs as shown in the Watana analysis, iii Tee tis clear from this comparison that Devil's Canyon displays very favorable economics. eS Oh KENTCO TOTAL COSTS OF ELECTRIC GENERATION - Present Value of Fifty Years Operation - Present Value (millions of dollars) Watana 3631 Coal 4431 Gas, with TAGS built and export market available 4709 Gas, no TAGS line 5806 Assumptions: Real Discount Rate - 3% percent Start Date: - 1996 Period of Analysis:- 50 years KENTCO The Susitna Project as a whole, then, includes Watana with a favorable cost structure, and Devil's Canyon with a highly favorable cost structure. The actual present value of the two together depends on the timing of Devil's Canyon and the cost of bridging between the two. However, the analysis clearly shows that the project has a lower cost than any thermal alternative. Discussion Economic analysis can include the costs and benefits of effects other than the cost of power. Although such analysis is beyond the scope of this report, we wish to highlight several additional considerations to the reader. These are multiplier effects, environmental considerations and several considerations of risk. Multiplier Effects Construction of the Susitna Project will provide a large number of jobs and a large amount of investment during the multi-year construction of each stage, while construction of coal plants or natural gas generation plants will employ fewer workers and involve less investment. The employment and investment would result in secondary benefits to the state economy. Environmental Effects Generation alternatives will have environmental effects. For example, hydroelectric power affects the environment by flooding land behind the dam and altering stream flows below. - 140 - KENTCO Coal burning can have effects on air quality, and coal mining can have effects on the terrain. An elaborate analysis could try to quantify these environmental effects through assumptions about people's preference or the cost of additional mitigation, but this is beyond the scope of our study. Risk There are two important elements of risk for the reader to keep in mind when reviewing Susitna's economics. For one, the project, even Watana by itself, would become the dominant supplier for the Railbelt's electric power needs. At the time that Watana comes on-line, and again when Devil's Canyon comes on-line, the Susitna Project may be providing as much as 90 percent of the Railbelt's electric energy. Thus, at the two critical times when each project comes onstream, Susitna power is close to being the sole source of Railbelt power. At these critical times a breakdown in the project, or a change in the economics of the alternatives which served to make the alternatives cheaper, could leave the Railbelt with little flexibility to switch to a different alternative. On the other hand, the Susitna Project can protect the Railbelt from those changes in the economics of the alternatives which would make the alternatives more expensive. For example, an unexpected rise in the world price of gas or coal might open up Alaskan exports earlier and raise the price of gas or coal at home, thereby raising the cost of the thermal alternative. This is an uncontrollable variable. Thus while in one sense Susitna entails some risk because of its size, in another = 7A « KENTCO sense Susitna reduces risk because it is not affected by changes in fuel prices. The manner in which Susitna guards against the risk of increases in the cost of alternatives takes on additional mean- ing when the economics of Devil's Canyon are considered. The construction of Watana sets the stage for the subsequent con- struction of Devil's Canyon, a project which has a very favorable cost structure. Watana provides stable-cost energy and in addition makes possible the construction of another stable-cost project, with better economics, to meet future demand. In this sense Watana serves to reduce risk. Conclusions = ie have seen that the three principal alternatives for generating power in the Railbelt are gas, coal and hydroelectric power. Also we have seen that the hydroelectric alternative is a two-dam system on the Susitna River, Watana and Devil's Canyon, of which Watana would be built first. Economic analysis shows that the first stage, Watana, has the lowest present value cost when compared to the thermal alternatives of gas and coal over a 50 year period. Further, it shows that if the Trans-Alaska Gas line is not constructed, and North Slope gas is not available at a tidewater price, Watana's advantage, excluding any consideration of Devil's Canyon, would be increased. The economic analysis of the second stage, Devil's Canyon, shows a very low cost when compared to any of the alternatives. KENTCO Overall, based on our forecasts, we find the Susitna Project including either both dams or Watana alone to make economic sense and to be the most attractive means of generat- ing power in the Alaska Railbelt. - 143 - KENTCO ELECTRIC POWER GENERATION FOR THE ALASKA RAILBELT REGION PART Il PLAN FOR FINANCING POWER GENERATION INVESTMENT KENTCO INTRODUCTION The second half of our report presents the financing considerations of the Susitna Project, having concluded that this alternative is more attractive than any other. In this section we will take the information developed previously, especially the economic analysis, and present it in financial terms. We will show the project's need for capital, and the project's ability to collect revenues over the years. Following that, we will introduce the various types of capital which may be available for loan or investment to Susitna. We will analyze them, especially as to their requirement for rate of return and risk. We will show that different sources of capital vary widely in their requirements for rate of return and risk, and that some are more appropriate to Susitna than others. From our analysis of the various sources of capital, we will present a plan of finance. The plan of finance will high- light the key assumptions behind the plan, and the major events which must occur before the financing takes place. Following the presentation of a plan of finance we discuss the State's role in the project and the need for any State financial participation to correspond to a long term energy plan. thh - KENTCO DEFINITIONS In this report we will use several terms in a specific way. We define them as follows: Capital Capital refers generally to money which is used to pay for something which is permanent such as a hydroelectric project. In the case of Susitna, we refer to capital as all up-front money used to cover start-up of the project. Including construc- tion costs and any start-up expenses not paid for through the sale of power. Capital includes grants, loans and equity investments. Rate of Return Rate of return is a measure which compares an investment to the benefits received from that investment. As used in this report the benefits are generally a cash income to a lender or investor, although in certain cases benefits may also include a reduction in costs or other benefits. The rate of return compares the size and timing of the benefits to the amount and timing of the investment, and gives a measure expressed as a percentage. The larger or earlier the benefits relative to the investment, the higher the rate of return. Risk Risk is an expression of the uncertainty of the benefits of an investment: how much the benefits will be, when they will occur, and whether they will occur. Risk is difficult to quantify, and in our report we will deal with risk without attempting to assign specific values. - 145 - KENTCO Nominal Dollars Nominal dollars refers to amounts of money which increase with inflation. This method of presenting dollars lists them as they are spent, so that the number of dollars spent in the future for a certain item is greater than the number of dollars spent at present for the same item, assuming that inflation exists. Real Dollars Real dollars refersto amounts of money which are adjusted for the effects of inflation. Under this method of presenting dollars, the cost and price of an item generally stay the same over time. Only when the price of an item rises faster than inflation, or more slowly than inflation, does the real dollar price of that item change. DETERMINATION OF CASH FLOW We start the development of a financing plan by determin- ing the cash flow to be expected with the Susitna Project. Cash flow refers to the amount of dollars put into a project, and received from a project, on a year-by-year basis. Cash flow is directly related to the economics of the project. In the following pages we will briefly review Susitna's economics, then convert those economics to a cash flow. Economics of the Susitna Project Our analysis in the last section concluded with an economic model which compared the construction and operating costs of Susitna to the construction and operating costs of the best alternative means of generating Railbelt power. Our review of alternative thermal generation showed a cost of fuel, - 144 - KENTCO operations and maintenance, and a related cost of generation which started at some value and then rose with inflation or in excess of inflation. In contrast, Susitna had a cost which basically did not escalate in constant dollar terms due to the absence of. fuel costs. Our net present value analysis showed that the long term cost of Susitna was less than the thermal alternatives. It showed Watana to be the economically preferable alternative up to a real discount rate of 4.9 percent and Devil's Canyon to be economically preferable (assuming Watana is built) up to a real discount rate of 9.5 percent. Susitna power is economic because of forecasts of a need for power and certain costs associated with the alternatives. Conversion of Economics to Cash Flow In the long run, unless there are government grants, Susitna will be paid for by consumers through their utility bills. Therefore, a key element in Susitna's cash flow is to determine what amount Susitna can collect in revenues from consumers. Two key constraints on what Susitna can collect are the consumer's willingness to pay and his ability to pay. Recalling the foregoing economic analysis, we know that Susitna exists in a world of generating alternatives. We believe that utilities and retail users generally will be willing to pay an amount for their electric power equal to or less than the lowest cost of any alternative, but not much more than the cost of alternatives. Our interviews with utility managers confirmed this: utilities are interested in Susitna power only if it is preferable in cost to the alternatives. Our interviews did - 147 - KENTCO reveal, however, some willingness on the part of consumers to pay rates higher than the alternative, for a short period of time, if doing so would clearly lead to long term benefits. We have assessed the maximum amount which utilities may be willing to pay as approximately 10 percent above the cost, at retail, of an alternative source, but even then only for a temporary period and if the long term benefits are clear. Calculating that roughly half the costs of retail power are distribution costs and half are generation and transmission costs, the 10 percent margin at retail corresponds to approximately 20 percent of the cost of generation and transmission. Revenues from Susitna are also constrained by the consumer's ability to pay. If power rates, even though lower than any alternative power costs, are so high as to be unaffordable, then the power cannot be sold. At some level of rates, or with certain types of industrial consumers, ability to pay could be an important constraint on potential Susitna revenues. However, we do not believe this to be a problem in the range of rates contemplated. Projected Susitna power rates, if priced equal to or less than the thermal alternative, will provide real power costs which are lower than those paid currently in many parts of the United States where per capita income is lower than Alaska's. " The constraints of "willingness to pay" and "ability to pay'' provide a strong link between the economics of a power project and its cash flow. The example of the Washington Public Power Supply System (WPPSS) default earlier this year - 148 - KENTCO is a good demonstration of that link. In the WPPSS situation, demand shortfalls and cost overruns led to a situation where a number of nuclear plants, as part of a major power supply pro- ject, were no longer economic. The consumers were unwilling to pay for these uneconomic projects in their power rates, and rebelled, even though they contractually were obligated to pay. As the cash flow requirements of the WPPSS financing diverged from good economics, the cash flow failed, and the financing subsequently failed. In developing a finance plan for Susitna, therefore, we have established the following criteria: 1. The project must make economic sense to the utilities and their customers. Consumers will be unwilling to pay power rates which support an uneconomic project. 2. The power rates must offer long term savings when compared to the thermal alternatives. 3. In general, power rates in any year should be equal to or less than the thermal alternative. There is some leeway for rates exceeding the thermal alternative in the short run, if long term benefits are clear. 4. Cash flow from Susitna and power rates are directly related. Exhibit 58 shows the estimated costs of the gas thermal alternative, in nominal dollars, commencing in the year 1996. We use gas because it is, in the short run, the lowest cost alternative. Exhibit 59 shows the maximum cash flow available from Watana assuming that power rates are allowed to exceed - 149 - KENTCO the thermal alternative by 10 percent at retail (20 percent of generation and transmission cost) in the short run. The exhibits suggest a first year maximum wholesale price for Watana power of 5.9 cents, in 1983 dollars. We will work with this number in our report; however, it is important to note that the cost of the alternative could be different depending upon each utilitie's assumptions about the future. What is important is whether, at the time to commit to a financing plan, utilities will be willing to commit to a 5.9 cents (in 1983 dollars) wholesale price. - 150 - Exhibit 58 NOMINAL COST OF GAS GENERATION Assumptions: - Costs are based on cost structure for 200 Mw or larger plants - Costs shown are for generation of 3500 Gwh per year - Capital costs assume level debt service financing in 1996 - Interest rate on debt is 11.2 percent - Capital cost of gas generation plants is $501 million in 1983 dollars - Level debt service is $137 million for 25 years - Costs are shown only for the first 15 years ($000,000) Fuel and Fuel and Level Debt Total Cost FY Operations, Real Operations, Nominal Service, Nominal Nominal 1996 112 254 LST 391 1997 112 271 137 408 1998 ily 2RR 137 425 1999 i 307 137 444 2000 112 327 137 464 2001 112 348 137 485 2002 112 371 137 508 2003 112 395 17, S32 2004 126 473 137 610 2005 130 519 137 656 2006 133 567 137 704 2007 137 622 137 759 2008 141 680 37: 817 2009 145 746 137 803 2010 149 816 13). 953 OOLNS™ - @ST - Exhibit 59 MAXIMUM REVENUES FROM WATANA (Real Dollar Equivalent, 1983 Dollars) Assumptions: - Revenues tied to cost of gas generation alternative - Maximum Watana revenue assumed to be 20 percent over gas alternative - Inflation assumed at 6.5 percent per year - Only first 15 years shown GAS WATANA ($000,000) ($000, 000) Cents ($000,000) ($000,000) Cents Cost of Gas 1983 Per 120% of Gas 1983 Per FY Alternative Nominal Equivalent Kwh Alternative Equivalent Kwh 1996 391 172 4.9 469 207 5.9 1997 408 169 4.8 490 203 5.8 1998 425 165 4.7 510 198 5.7 1999 444 162 4.6 533 195 5.6 2000 464 159 4.5 557 191 5.5 2001 485 156 4.5 582 187 5.4 2002 508 154 4.4 610 184 5.3 2003 532 151 4.3 638 181 by? 2004 610 163 4.6 732 195 5.6 2005 656 164 4.7 787 197 5.6 2006 704 165 4.7 845 198 5.7 2007 759 167 4.8 911 201 5.7 2008 817 169 4.8 980 203 5.8 2009 883 172 4.9 1060 206 5.9 2010 953 174 5.0 1144 209 6.0 OOLNA» KENTCO COMPARISON OF CASH FLOW WITH COST OF DEBT The revenue from Susitna developed in previous pages starts at a low level and subsequently increases each year. In this section we will compare Susitna's cash flow with representative debt requirements. To start, we need to review the concept of level debt service. Basically, level debt service refers to a system of repayments similar to that on a conventional home mortgage, in which a borrower repays a certain amount of dollars, ona regular basis, for a number of years until the debt is paid off. Level debt service is by far the most common type of long term (anything over approximately 5-10 years) debt. It is a tradi- tional form of debt service which developed during the United States' long history of lower inflation and lower interest rates. Level debt service does not rise or fall with inflation. Al- though in today's environment of higher inflation and interest rates, level debt service makes less sense, it remains easy to account for, easy to describe and understand, and it remains the most acceptable form of debt. Exhibit 60 shows a typical level debt service. In this case it shows the payments necessary to repay $6.5 billion of debt, at an interest rate of 11.8 percent, over 34 years. We will show why we used the $6.5 billion amount, the interest rate of 11.8 percent and the 34 year time period in the follow- ing sections. . Exhibit 6CAcompares Watana's cash flow as developed on pages 151-152 and the cash flow which would ke reanired to Ses 2 KENTCO service level debt shown in Exhibit 60, Exhibit 60A shows that Watana's cash flow falls short of meeting level debt ser- vice in the years 1996 through 2005, although in later years Watana can meet this debt service. We can conclude that Watana may present some financing problems because of the early year shortfall of cash flow to meet normal debt service, in those early years. Devil's Canyon does not present the same problem. The capital investment in Devil's Canyon, in inflation-adjusted dollars, is much less than that of Watana. Further, at the time Devil's Canyon is to come on-line in the early twenty- first century, the cost of the alternatives is higher. Exhibit 61 shows the maximum cash flow available from Devil's Canyon assuming an on-line date of 2010, and the cost of level debt service again under conditions representative of today's market. Devil's Canyon does not appear to present a financing problem. However, the very attractive economics and financing abilities of Devil's Canyon occur only after Watana is built, and there- fore this favorable situation does not in itself solve Watana's problem. We conclude, therefore, that the principal financing issues of the Susitna Project center around the Watana Dam, and much of the remainder of the financial discussion will focus on Watana. - 154 - Millions $'s (Nominal) Assumptions: is Exhibit 60 REPRESENTATIVE LEVEL DEBT SERVICE Principal Amount: $6 Interest Rate: 11.82 Term of Debt Issue: Level Debt Service: $785,000,000 Annually 000- 800- 600- 400- 200- 1996 2000 Watana ,200,000,000 34 Years + 2010 YEARS - 155 - 2020 KENTCO + 2030 Debt Service Level Millions $'s (nominal) KENTCO Exhibit 60A COMPARISON OF LEVEL DEBT SERVICE TO WATANA'S ABILITY TO PAY Assumptions: - Level debt service of $785 million - Watana's ability to pay equal to thermal alternative plus 20 percent - Watana's ability to pay is net of $8.5 million in operations and maintenance expense (in 1983 dollars) Ability to Pay 1,000- Level 800- Debt Service 600- 400- 200- 1996 2000 2010 2020 2030 YEARS = Op Millions $'s (nominal) Exhibit 61 COMPARISON OF LEVEL DEBT SERVICE TO DEVIL'S CANYON'S ABILITY TO PAY Assumptions: 25 Level debt service KENTCO Devil's Canyon ability to pay equal to the thermal alternative plus 20 percent Devil's Canyon ability to pay is net of $2.7 million in opera- tions and maintenance expense (in 1983 dollars) 000- 1,600- 1, 200- 800- 400- Ability to Pay Level Debt Service 2010 2015 2020 2025 YEARS - 157 - KENTCO FINANCIAL STATUS OF WATANA In reviewing the status of Watana from the previous pages, the following points stand out: i Watana, as the large reservoir dam, must be built first in Susitna's two dam system. Projections for load growth indicate a 10 year or greater lag between the time that Watana is built and the time that Devil's Canyon is needed. Watana carries most of the capital cost of the two dam system, which is reflected in the different economics of Watana and Devil's Canyon. Watana's economics, and corresponding cash flow, indicate a cash flow deficit for the eary years of operation when compared to the likely cash flow requirements of debt financing . Devil's Canyon's more favorable cash flow begins too late to relieve Watana of its cash flow problem. A financial analyst or an investor will look at the project differently depending on which part of the investment world he represents. All investors will see a relatively slender Watana cash flow in the first five or ten years, compared to the size of the investment, depending on interest rates and inflation. They then will see potentially greater Watana cash flows, (plus potentially large Devil's Canyon cash flows), developing some time after the year 2000. With that in mind, we will review the three most likely sources of investment capital, briefly review several possible - 158 - KENTCO but less likely sources, and determine which if any are appropriate sources of capital for Watana. STATE EQUITY State equity is money which would be invested directly into the Susitna Project through appropriations from the State's General Fund. Such a cash investment represents an equity, or ownership, interest in the project. Given proper legal authority or framework, it also would represent a right to cash flow from the project after payment of all other expenses. The State of Alaska can make equity investments in the Susitna Project through legislative appropriations, and may re- ceive a return on its investment through any formula or under any system provided by law. In our report we will assume that any State investment is made from the General Fund, and any return is to the General Fund, although the use of other spec- ially created investment funds could also be utilized if all legal and constitutional questions are resolved. Present Alaska law, AS 44.83.390 through 44.83.398 provides for equity investments in power projects and for return on investment in some instances, but these statutes have undergone revisions in the past three years. Based on our interviews with various State officials, we believe that these statutes could be revised, if justified, to allow new plans for equity investment and return on investment. Therefore, we will con- sider State investment beyond the constraints of the present statutes, although we recognize that legislative action would be necessary for any such changes. - 159 - KENTCO The most unique financial feature of State equity investment is the flexibility possible in designing the return on invest- ment. Both the rate of return and the timing of return are adaptable. Since a governmental body may make investment decisions based on a number of criteria other than financial rate of return, such as economic development or improvement of living standards, the requirements for financial return may vary greatly from the return required by private investors. For similar reasons, the government may allow deferral of its return for several years or more. Therefore, the return on State equity investment may be quite different from the regular and highly certain return require- ments of level debt service. With such potential flexibility, a State equity investment might be made withacash return, if any, patterned to match the cash flow capabilities of the Susitna Project. The availability of State equity for investment, suffers from three major elements of uncertainty. First, the amount available for appropriation to the Susitna Project is not known, as it depends on future State revenues. Second, the amount actually to be appropriated is unknown as it depends on future legislative action. Third, constitutional requirements preclude one legislature from making binding commitments upon future legislatures, so that a partial appropriation by one legislature does not assure continued appropriations by subsequent legis- latures. Therefore, the total amount of equity participation remains uncertain even after an initial decision to invest is made. - 160 - KENTCO Regarding the uncertainty over State revenues and funding available for appropriations, our revenue estimates suggest that there will be insufficient funds to make a 100 percent equity investment to Susitna. The State of Alaska's revenues in the next two decades will not grow much over today's levels. This is based on average-revenue* estimates by the State of Alaska Department of Revenue and by the tate revenues resulting from the Lichtblau oil price forecasts. These revenues will have to meet ever-increasing demands for State monies. Increasingly there is more demand than available dollars. In the decade from 1984 to 1993 which would encompass appropriations for Watana's construction, State revenues under our forecast will average $3,021 million per year in 1983 dollars. Appropriations from the General Fund for the operating budget, debt service and dividends were approximately $2,100 million in Fiscal Year 1984. The difference of approximately $900 million in 1983 dollars will then be available each year over the next 10 years for all other State spending, including growth in the operating budget and all other capital projects in the State which may compete with Susitna for appropriations. Although the total State money available may not accommodate all of Watana's cost, it may accommodate a portion of the cost. The actual amount will vary according to the priorities of the State government. We cannot forecast government priorities, but our interviews with State officials and legislators indicate an uneasiness with one proposed Susitna financing plan which calls * Based on the State of Alaska Department of Revenue "mean" forecast for future revenues, dated September 1983. - 161 - KENTCO for appropriations totalling $1.6 to $1.8 billion (in 1983 dollars) over the next six to ten years. However, an amount less than that, over a longer period, may be acceptable. The amount of dollars available in the period beyond 1994, which encompasses Devil's Canyon construction, is much more uncertain as it depends greatly on future oil discoveries, State taxing policies, State spending and Permanent Fund income. The Lichtblau forecast shows very firm oil prices for the period, but Alaska production is uncertain. Therefore, we believe that State equity investment dollars may not be readily available for Devil's Canyon. In conclusion, we regard State equity as an attractive, although limited, possibility for the Susitna Project. It offers potentially high flexibility in its rate of return requirements, and it may accept such features as a delayed return. Practical constraints limit the total amount of State equity available. Investment levels of $1.6 ‘to $1.8 billion during the next decade for Watana construction are met by some apprehension from State officials. Also, State equity available for investment in Devil's Canyon in a later period is highly uncertain. MUNICIPAL DEBT The municipal market is a large, active, nationwide network of financial institutions and individuals who buy and sell State and local debt. A unique characteristic of the market is that interest paid on debt is tax-free. Interest rates on debt in the municipal market are generally lower than interest rates on debt in other markets. Since the interest received by an investor is tax-free, the rate of return = 162 — KENTCO of municipal debt can be greater than the after-tax rate of return on other taxable debt. Investors are therefore willing to accept lower interest rates on municipal debt than on other debt. To borrowers, raising money in the municipal market is attractive because the lower interest rate means lower interest costs. The tax advantage is so strong that at times the interest cost of municipal debt is lower than the rate of inflation, al- though that has not been true of recent long term municipal borrowings. Exhibit 62 is a chart which shows the "real" interest rate, at the time of issuance, of long term municipal debt over the past 50 years. Real interest is the bond interest rate minus the rate of inflation. Under present market conditions the interest rate at which Susitna could borrow is likely to be between 10% percent and ll percent. A recent Merrill Lynch Index of tax-exempt electric revenue bonds, which reflects present market rates of a sample of municipal bonds similar to what Susitna bonds would be indi- cates a market average of 10.54 percent. The rate on Susitna bonds in this market will be higher or lower than such an average, depending on the investors' perception of Susitna's credit worthiness (ability to pay debt service in a timely fashion) and technical factors such as the size and timing of Susitna's entries into the market. There are problems associated with municipal financing of Susitna. Chief among these are Susitna's present ineligi- bility for tax-exempt financing, the effect of debt service coverage requirements, and the conservative nature of the municipal market. ea eon Exhibit 62 KENTCO REAL INTEREST RATES Percent 20 | Municipals | a3 oes 4930 4340 49350 4960 41970 1980 - 164 - KENTCO Susitna's Ineligibility for Tax-Exempt Debt Susitna is presently not eligible for tax-exempt financing. To explain this, we will start with a brief review of several important rules surrounding tax-exempt financing. The legal basis for tax-exempt financing lies in the rela- tionship between the federal government and the states. The federal government, through the Internal Revenue Service, taxes income wherever it can do so. However, there is a constitutional protection which prevents the federal government from taxing state governments. Under this constitutional umbrella individual states, municipalities and their agencies and authorities are protected from federal income tax and therefore these entities can issue tax-exempt debt. (States and municipalities which have income taxes of their own routinely exempt any federally tax-exempt debt from their own income taxes, so the tax-exemption is generally complete.) However, there are problems at the fringes of the tax-exempt world, where the debt issuer is a tax-exempt entity, but the user of the debt monies is a trade or business which is not a tax- exempt entity. This happens when a state or local government or agency borrows money, and then either re-lends the proceeds to a trade or business, or builds a facility which is used by a trade or business. The trade or business then makes payments to the government agency, which in turn uses the payments to pay principal and interest on the bonds. 1. Pollock v. Farmer's Loan & Trust Co., 157 U.S. 429,586 (1895) - 165 - KENTCO In these cases, where a non-exempt entity uses a facility and in effect pays for the debt on the facility, the financings are classified as industrial development bonds, and are generally taxable,* unless they fall within some specific exclusion under federal law. Rural Electric Associations, such as Chugach, GVEA, MEA and HEA are non-exempt trades or businesses in the eyes of the IRS. Under present law and utility structure the REA's would use and pay for a sufficient amount of Susitna power to categorize any Susitna revenue bonds as industrial development bonds. Also, the Susitna system Ue under any special exemptions, and therefore the bonds would not be tax-exempt even though they are issued by a state agency. We see two approaches to changing Susitna to a tax-exempt project. The first is to change the Susitna power purchase and delivery system to meet the federal exemptions, and the second is to change federal law to exempt Susitna as it is. In the former approach, the Railbelt could reorganize its utilities into municipal entities (such as AML&P and FMUS), so that the power would not be used by an IRS-defined trade or business and the bonds would not be industrial development bonds. Another possible approach, being discussed by the APA under- writers but still to be tested, is to bypass the REA's entirely and charge debt service directly to the residents in the util- ities' areas. Individuals are not "trades or businesses" and the bonds could therefore be tax-exempt. 2. Internal Revenue Code & 103 (b) (1) - 166 - KENTCO The other approach, that of federal tax law changes, can take many forms. The simplest change would be to expand the industrial development bond exemptions of "local furnishing"? of electric energy to include the Susitna situation. Susitna presently does not qualify for the local furnishing exemption because it will serve more than two counties. Other alternatives include changes to exempt hydroelectric projects generally, regardless of area or customers served; to redefine REA's as exempt entities; or any specific federal language giving a special exemption to Susitna. Any such measures will be difficult, as the trend in Congress is to narrow rather than broaden the exemptions for industrial develop- ment bonds. As of this writing, there is a bill in the U.S. House of Representatives which would restrict the volume of industrial development bonding to a small amount per State, based on population. Such a bill would be especially restrictive in a low-population state such as Alaska. Debt Service Coverage A second problem with municipal finance is that revenue bonds in the municipal market typically require a project being financed to charge rates higher than the amount necessary to pay debt ser- vice and other expenses. This additional requirement is called debt service coverage. The requirements typically require rates sufficient to provide revenues after operating expenses which are 110 percent to 150 percent of debt service. Debt service 3. Internal Revenue Code & 103 (b) (4) (E) 4. Most people we talked to assume that the IRS will treat boroughs as equivalents to counties. - 167 - KENTCO coverage requirements can therefore aggravate the shortfall between the revenue requirements of the debt and the cash flow capabilities of the project. Looked at another way, debt ser- vice coverage requirements cause rates to be higher than they would be if the rates did not have to provide such additional coverage. Conservatism A third problem is that the municipal market is unwilling to accept very much risk. The municipal long term bond market is made up of generally conservative investors. There is an extreme reluctance in this market to purchase bonds with anything less than a top credit rating, a similar reluctance to accept bonds which offer anything other than level debt service, or which offer any other unique features. Part of the reason is that most of the investors in the municipal bond market are individuals who are relying on a fixed, steady income from their investments. Another factor is the practical requirements of the highly refined, vast network of investment bankers and bond traders necessary to reach all of these individuals. Ordinary level debt service bonds trade efficiently through this network because they are easy to describe and to understand. Bonds which take too long to explain may have a lower acceptability, and reach only a part of the potential investors, at higher rates. In today's environment the municipal bond market is less willing even to take risk than in other times. The financial environment is changing so rapidly that investors are nervous. Inflation since 1980 has varied from over 10 percent to less - 168 - KENTCO than 5 percent and average interest rates on tax-exempt bonds have ranged from below 8 percent to over 13 percent. Other expectations such as a steady upward trend in oil prices and electricity demand suddenly reversed themselves, throwing the electric industry in disarray. On top of all this, the Washington Public Power Supply System defaulted on a record amount of mun- icipal electric revenue bonds, which heightened investor consciousness and skepticism regarding projections of energy demand and the enforceability of contracts for the sale of power. Based on interviews with investment bankers and other pro- fessionals in the field regarding the present investment climate, we believe that investors for tax-exempt bonds, who are conserva- tive and risk-averse at any time, have at present an especially high level of uncertainty about the credit worthiness of electric revenue bonds. This does not mean that power projects are unfinanceable, but rather that the demonstration of long term credit worthiness will need to be especially stringent. This market will require a demonstration of ability to pay debt service, as well as a willingness to pay, under present conditions and under a conservative forecast of future conditions. It will not accept, or will be very skeptical toward, a debt service which relies heavily for payment on a forecast of future in- creases in the cost of alternative generation, on inflation generally, on growth in demand, or on future legislative appro- priations. Finally, it will not rely solely on power sales contracts to force payments to meet debt service unless it is convinced of the underlying willingness to pay, ability to pay and demand. - 169 - KENTCO The underwriters of the Alaska Power Authority have recommended several credit-enhancing features for Susitna revenue bonds, including power sales contracts, debt service coverage (at a level of at least 110 percent), and a State moral obligation feature. The moral obligation feature is a mechanism by which a shortfall in cash needed to meet debt service causes a report to go to the Alaska State Legislature. The legislature then is aware of the shortfall and has an opportunity, but no obligation, to appropriate money to make up the shortfall and to prevent a default. Finally, the underwriting team has recommended level debt service, with a State equity contribution which will lower the amount of debt financing of Watana to a point where Watana's cash flow can meet level debt service, including debt service coverage, in every year including the first full year of opera- tion (1996). Our own analysis (see the Finance Plan section) supports the above conditions if the tax-exempt route is taken. In today's environment the municipal revenue bond market will only accept level debt service, and only with credit-enhancing features which permit an "A" or better credit rating, such as debt service coverage, power sales contracts and the State moral obligation pledge. In addition, we believe that the power sales contracts used should be court-tested in advance for validity. This court test can give additional assurance to bondholders that the debt will be repaid, and can be worth from 0.50 to 0.60 of one percent in interest rate. - 170 - KENTCO To summarize, the municipal market is attractive in its lower rate of return requirements, but unattractive in its willingness to accept risk. It must be kept in mind that use of municipal debt presents a tax-exemption problem which must be solved before financing is possible. FEDERAL ELECTRIFICATION DEBT In this section we will review the federal government's involvement as a potential lender to Susitna, through the Rural Electrication Administration, or REA. The REA is an agency of the federal government created within the U.S. Department of Agriculture. The agency has been active in the financing of power projects since the 1930's when it was created under the Emergency Relief Appropriation Act of 1935. Susitna could qualify for a loan through one of the pro- grams of the REA, in which the REA gives a federal guarantee to long term bonds used to finance bulk power supply facilities. Since the bonds have the same security as other long term bonds issued by the federal government, the bonds carry similar interest rates. These bonds are not tax-exempt, but because of the very high security they offer with a U.S. government guarantee, investors require a lower rate of return than for other taxable bonds. The interest rate on REA guaranteed bonds is not as low as the interest rate on A-rated tax-exempt bonds, but since REA guaranteed bonds do not have an additional debt service coverage requirement the effect on utility rates of loans under the REA program can be very similar to the effect - 171 - KENTCO of tax-exempt bonds requiring 10 percent or higher debt service coverage. Under present market conditions, financing from REA- guaranteed long term bonds issued by the Federal Financing Bank carries an interest rate of 11.78 percent. There are two restrictions on REA debt which may present problems for Susitna's financing. First, the REA program is designed for REA borrowers. For example, the type of loan pro- gram under consideration here exists to provide loans to a type of REA utility called a generation and transmission cooperative, which then sells the power to REA retail utilities known as distribution cooperatives. No Railbelt-wide generation and transmission cooperative presently exists to serve as the conduit for Susitna financing and to wholesale Susitna's power output to the retail utilities. Such a cooperative would have to be estab- lished. Also, approximately 25 percent of the retail purchases of electricity in the Railbelt are through utilities other than REA cooperatives: Anchorage Municipal Light and Power, Fair- banks Municipal Utility System and Seward Electric System. REA will guarantee loans for bulk power supply facilities only when those facilities will eventually serve the needs of REA distribution cooperatives. What this means is that REA will guarantee the loans to finance only the percentage of Susitna which is used by REA distribution cooperatives, presently about 75 percent, or if it finances more than that percentage it will Le only if justified to meet the future load growth of the REA cooperatives. Sales to non-REA utilities would be permitted only until REA cooperatives grew to absorb the output being financed. - 172 - KENTCO If, for example, Susitna were 100 percent REA financed, the non-REA utilities would be able to purchase power only until Chugach, GVEA, MEA, etc. grew to absorb all of the Susitna output. The problem of sales to the Railbelt presently made by non- REA utilities can be addressed in two ways. The first would be to finance only the appropriate percentage of Susitna with REA funds and to use other financing for the remainder. It is possible to divide a project's cost, output and operating expenses on a percentage basis, and to treat the project as if it were two separately financed projects. This has been done elsewhere in the country both for REA financings and for tax-exempt financings. A second alternative is to finance Watana as an REA project and to finance Devil's Canyon on another basis. In this case, the non-REA utilities would purchase power from Watana on a temporary basis until all of Watana's output is utilized by the REAs. At that time the non-REAs would look to Devil's Canyon for power. We believe that this is a less desirable alternative because of the uncertainty about load growth and the timing of Devil's Canyon. Under this alternative the non-REA utilities could lose the availability of Watana's output before Devil's Canyon comes on-line, if the demand of REA utilities grows. We investigated a third alternative for REA financing which would reorganize the Railbelt distribution utilities so that all areas were served by REA utilities. This would require the expansion of GVEA, CEA and possibly HEA to serve the areas presently within the boundaries of FMUS, AML&P and Seward, and it would require APUC approval. However, in our meetings with REA officials we determined that such an expansion, even if it = 173. - KENTCO were to occur, would not render the new areas served eligible for the REA guarantee program. We therefore rejected this alternative. REA funding depends on Congress approving an amount of funding for REA large enough to cover REA's requirements. Susitna could be only one of those requirements. The approval is not a federal appropriation, but is is a dollar guarantee amount which must be agreed to by Congress. There is precedent for large REA financings. REA is participating in the Oglethorpe coal and nuclear project in Georgia with a $3.9 billion commit- Ment over several years. In 1983, the total REA guarantee ceiling from Congress was $4.6 billion and this year it is $3.25 billion. The reduced amount is not an indication of reluctance by Congress to appro- priate funds as much as it is an absence of projects REA wishes to pursue. Offsetting the above potential problems are several advan- tages of using REA. REA officials are familiar with Alaska, having cooperated in the financing and construction of power systems all over the State. Foremost among REA's advantages is the agency's possible willingness to accept the risk of the Susitna Project. Based on our discussions with William Davis, Western Area Director for REA, they expect to be a "participating agency" in the Federal Energy Regulatory Commission's (FERC) license review of Susitna, and they will be likely to use FERC's eventual findings regarding the need for power, environmental studies and so forth. Therefore, if the Susitna FERC license is granted, REA is likely to support a request for authorization - 174 - KENTCO to guarantee loans for the project. It should be kept in mind that REA presents its budget request through the Department of Agriculture, which may alter it. The Department of Agriculture request is then subject to further revision in Congress, and thus dependent on fiscal policies generally as well as subject to the lobbying efforts of the National Rural Electric Association, Inc. (NRECA). A second major advantage of REA financing is that the borrowing for the project occurs in a different market from that in which the State of Alaska does its other borrowing, and it occurs without a State pledge or guarantee behind it. This keeps the debt off the balance sheet of the State, and thus may help maintain the bond ratings and interest costs of State debt in the municipal market, such as other Alaska Power Authority bonds or bonds of other State agencies. A third advantage is that REA does not require ownership of the power projects it helps finance. Therefore, the State of Alaska can continue its role as owner and sponsor of the Susitna Project for licensing, engineering and so on. A final and major advantage of REA debt is that it will accept State subsidies as a supplement to the rates being charged to cover expenses. Other debt markets may be unwilling to lend unless rates alone, exclusive of State assistance, are sufficient to cover debt service. This is particularly true since under present Alaska law one legislature cannot bind another. REA is accustomed to the subsidies currently employed in rural Alaska, where the State gives substantial monies through the PPCA Program. Therefore, since it is desirable - 175 - KENTCO to bring down early-year power costs in the case of Susitna, the State can appropriate operating subsidies annually as an alternative to investing equity dollars into the construction cost of the project. In contrast, the tax-exempt market will not accept such subsidies if they depend on future legislative appropriations. We have calculated that a combination of equity investment and operating subsidies reduces the total cost of the project in the early years to a point where power rates, even then, are at the levels of the thermal alternative. Additionally, this combination increases the number of years available to the State to raise the total tate money required and, on a present value basis, reduces the total cost to the State. (See Finance Plan section) In summary, we find that REA debt ranks high in terms of rate of return requirements, although slightly below tax-exempt financing, and it ranks very high in terms of willingness to accept risk. The combination makes it a most attractive, cost- effective means of financing Susitna. OTHER INVESTMENT MARKETS This section reviews other sources of debt and equity capital. We concluded that none of these sources are appropriate for Susitna, generally because of higher rate of return require- ments. The markets reviewed below represent those large enough to accommodate the needs of the Susitna Project. There are innumerable other types of financing and investors, although most of these are too small to be practical for the large part of Susitna's needs. At the time of financing che project, a combination of cost reduction options should be studied and - 176 - KENTCO perhaps employed. Their appropriateness cannot be judged this early and, at best, they will have only a small effect on the overall financing costs of the project. Debt Other sources of debt include corporate bonds, Eurodollar bonds and other foreign loans. The corporate bond market is a large market of investors and borrowers of taxable long term debt. It is the market used, for example, by major industrial corporations and service indus- tries for debt financing. Of interest to Susitna is a segment of this market known as the public utility bond market, a market which finances the projects of private, for profit utilities serving the public. These bonds have neither the benefits of tax exemption nor of any U.S. government guarantee, and the rate of return demanded by davestors reflects this. An "A" rated long term utility bond under present market conditions probably will carry an interest rate between 13 and 13% percent. The corporate bond market does accept some bonds with other than level debt service, especially floating-rate bonds. Floating- rate bonds carry interest charges that move up and down with the market under a formula described on the bond. The interest rate is generally tied to the movement of some short term interest rate. Although under present market conditions floating-rate bonds have interest below those of fixed interest rate long term bonds, they can be risky because interest costs fluctuate greatly from year to year. In the past four years short term interest costs have varied by as much as 12 percent, from a low Sara KENTCO of under 8 percent to a high of over 20 percent. We believe that the use of variable-rate bonds for Susitna would cause major changes in cost of power from year to year, and would make rate- setting very difficult. However, there may be a role for a small portion of variable-rate financing, at certain times, as part of Susitna's overall financing plan. We will address this again later. Eurodollar bonds are loans issued outside of the U.S., but denominated in U.S. dollars. Investors in Eurodollar bonds have similar rate of return requirements to investors in corporate bonds. Legal requirements regarding the nature of the bonds and the source of capital for the investors vary from country to country, so small differences in rate of return do exist, but in general Eurodollar bonds offer little or no interest cost advantage over domestic corporate bonds. Another major problem with this market is that it lends primarily for terms of 7 to 10 years or less. We dismissed further review of this market because of the mismatch between the seven to ten year willing- ness of the lender and the 30 to 50 year needs of Susitna as borrower. Other foreign debt includes debt which is denominated in currencies other than the U.S. dollar. Revenues from the Susitna Project would be exchanged into a foreign currency to pay debt service. We rejected this type of debt from con- sideration because revenue requirements, and rates, would be at the mercy of foreign exchange rate fluctuations. - 178 - KENTCO Private Equity Private equity is a type of investment in which an investor acquires an ownership interest in a project. The equity investor generally has rights to the cash flow from a project after other expenses and obligations, and often has some voting power or other control over the project. The requirements of equity investors covers a wide range, including investors who are will- ing to take extremely high risks. The chief advantage of private equity is its potential flexibility regarding the timing of its return. Assuming that Susitna can pay, eventually, the rate of return demanded by a private equity investor, the equity return may be tailored to match Susitna's cash flow. However, private equity investment presents problems. First, the rate of return demanded by private equity investors is gen- erally quite high -- higher than the rate of return demanded by investors in any of the major debt markets -- and it would represent the most expensive source of capital for Susitna. Further, private equity investors would probably demand some degree of control over the operation, the setting of rates, and the use of cash flow from Susitna, with a corresponding loss of control by the State of Alaska, the utilities, or the consumers. For these reasons, we dismiss private equity as a promising source of capital. - 179 - KENTCO REVIEW OF THE MARKETS In reviewing the capital markets, we find that three show promise for Susitna: State equity, tax-exempt financing and REA debt. The remaining markets large enough to accommodate the bulk of Susitna's needs are less desirable, primarily because of higher rate of return requirements. In the table below we ranked the sources of capital for Susitna. Since the lower rates of return required by investors allow greater benefits to be passed to consumers, we ranked the lowest rate of return as the most desirable, and the higher rates of return as less desirable. We also showed the effect of each type of financing on first-year power rates if 100 percent of Watana were financed by that method with no state subsidy. Table Ranking of Alternative Sources of Susitna Financing For Rate of Return Financial Effect On First Rate of Return Year Power Rate Required If 100% Financed (nominal dollars) l. State Equity 0% return or higher O¢/Kwh and up 2. Tax-Exempt Debt 10.54% interest rate 13.7¢/Kwht 3. REA Guaranteed Debt 11.78% interest rate 13.8¢/Kwh 4. Corporate Debt & Private Equity Over 13% Over 15¢/Kwh 1. YWncludes minimum 10 percent debt service coverage. - 180 - KENTCO We see no advantages from corporate debt or private equity financing to offset the disadvantages of high interest cost, so we will dismiss these from further consideration. Of the three remaining alternatives, there are large dif- ferences in willingness to accept risk. State equity investments may be made in any type of project, even speculative or experi- mental projects with no immediate prospect of financial or other return. Federal REA guarantees will accept less risk than the State can and would not be made on speculative projects. They could be used for loans financing the best power project for an area, even under the uncertainty about future economic conditions or alternatives that applies in the case of Susitna. Tax-exempt financing will accept the least risk of the three options being discussed and will be available only under conditions of demon- strably high certainty about future repayment. Table Ranking of Alternative Sources of Susitna Financing For Willingness to Accept Risk 1. State Equity 2. REA Guaranteed Loans 3. Tax-Exempt Debt Overall, State equity appears to be the most desirable in terms of rate of return and risk acceptance. After State equity, the choice between REA financing and tax-exempt financ- ing shows each with its respective advantages. REA financing has a clear advantage over tax-exempt financing in the willing- ness to accept risk, its corresponding willingness to accept State subsidies in lieu of up-front State capital to meet first - 181 - KENTCO year power rate targets, and its advantage of preserving tate borrowing capability. Tax-exempt financing has an jocecaut rate advantage, but only a marginal advantage in its effect on early year power rates. On balance, we believe that REA financing is the preferred alternative for debt financing. Based on the above, we recommend a plan of finance which uses State equity within practical and necessary limits, and debt financing for the remainder. We believe that REA debt is pre- ferable to municipal debt, and therefore, we recommend maximizing the use of REA debt, and using municipal debt only for that portion of the project which is ineligible for the REA guarantee program. - 182 - KENTCO FINANCE PLAN This section presents first a financing plan for Watana, then a separate financing plan for Devil's Canyon. It presents the assumptions behind the financing, then presents the plan, then reviews the major events which must occur before the financ- ing takes place. We also discuss alternative financing plans, and how the assumptions and the necessary events would change. Following this section we have included a separate discussion of the issues surrounding State equity investment. Assumptions The key assumptions underlying the financing of Watana are as follows. All of these assumptions are derived from previous parts of the report: £ 1. Cost. Watana will cost $3.432 billion in 1983 dollars. 2. Use of debt. The Watana financing will use REA guaranteed debt, at an interest rate of 11.78 percent and tax-exempt debt at an interest rate of 10.54 percent together with a debt service coverage requirement of 10 percent. The use of debt will be divided 75 percent REA, 25 percent municipal. 3. Power Utilization. Watana will be 100 percent utilized when it comes on-line for the first full year of opera- tion, 1996. Implicit in this projection is a growth from present demand levels of about 31 percent, or a compound growth rate of about 2.2 percent per year. 4. Rate of Inflation. Inflation is 6% percent per year. First year target power costs. First year target wholesale power cost will be 5.© cents per kilowatt hour, in 1983 dollars. wa - 183 - KENTCO 6. Increase in Watana power rates. Power rate increases do not exceed the rate of increase of alternative generation costs. 7. Rate equality. Rates charged to municipal and REA utilities are the same. 8. State investment. State capital is assumed to be available for either equity investment at the time of construction or for power cost subsidies. Equity will be used first in the municipal share (25 percent) of the project, to the extent necessary to allow level debt service. Remaining equity will be applied to the REA utilities' share. Power cost subsidies will only be applied against REA debt service to the extent necessary to equalize rates. 9. Operation and maintenance cost. Operation and main- tenance cost for Watana is .5 million per year, in 1983 dollars. 10. Revenues in excess of debt service. Revenues in excess of debt service are pledged through bond convenants to the State. The financing plan considers these revenues in calculating the net cost of any operating subsidies. Financing Plan Exhibit 65 shows a financing plan for Watana. It requires an $800 million State equity investment in the construction cost, in 1983 dollars, and a State subsidy of Watana power rates in 1996 equal to $145 million (in 1983 dollars), declining to a subsidy of $7 million in 2006. There would be no subsidy thereafter. Total State participation in Watana in both equity investment and subsidy is $1,576 million in 1983 dollars. In this plan power rates need increase after the first year at approximately the rate of inflation for only 10 years. No further rate in- crease is necessary to pay the debt service on Watana. If the State requires a higher return on its equity investment, then ' the power rate would have to increase further, with the addi- tional revenues providing a return on State equity. In either - 184 - ms G)O tees Exhibit 63 COST OF WATANA, INCLUDING INTEREST DURING CONSTRUCTION Assumptions: - Construction cost of Watana is $3,432 million, in January 1983 dollars - Inflation is 6.5 percent per year - Equity investment in project is $800 million in January 1983 dollars - Construction cost paid first with equity, then with debt financing - Interest during construction is charged annually against all debt financing from prior years, plus on-half year's interest against current year debt financing needs - Interest rate for construction financing is assumed at 9.7 percent - Eight year construction period with real construction cost in each year constant ($000,000) Nominal Less: Nominal Plus: Previous Total FY Construction Cost State Equity Financed Amount Interest Financing 1988 588 588 0 0 0 1989 626 541 0 4 89 199 On. 667 0 89 41 797 L991 710 0 797 Ht) 1619 1992 756 0 1619 193 2568 L993 805 0 2568 288 3661 1994 858 0 3661 396 4915 1995 914 0 4915 Byall 6350 Total Debt Financing for Project: 6350 Add: Pay Off Debt for Existing Generation: - 200 Total to be Financed: 6550 OOLNA» Sma EFL i Exhibit 64 MAXIMUM REVENUES AVAILABLE FOR DEBT SERVICE Assumptions: - Revenues based on 120 percent of gas alternative - Operations and maintenance expense is $8.5 million annually in 1983 dollars - Inflation is 6.5 percent per year - 25 percent of project power used by municipal utilities - 75 percent of project power used by REA utilities - Only first 15 years of revenue shown ($000, 000) Nominal Less: Operations & Total Available FY Total Revenues Maintenance Expense for Debt Service Municipal REA 1996 469 19 450 aE) 338 1997 490 20 470 118 352 1998 510 22 488 122 366 1999 533 23 510 128 382 2000 557 25 - 532: 133 399 2001 582 26 . 556 139 417 2002 610 28 582 146 436 2003 638 30 608 152 456 2004 1e2: 32 700 175 525 2005 787 34 155) 188 565 2006 845 36 809 202 607 2007 911 39 872 218 654 2008 980 41 939 235) 704 2009 1060 44 1016 254 762 2010 1144 47 1097 274 823 OOLNSAY L81 Exhibit 65 PLAN OF FINANCE Assumptions: - Municipal debt service: $102 million annually - REA-guaranteed debt service: $677 million annually - Maximum project revenues based on gas alternative plus 20 percent - Project revenues available for debt service are net of operations and maintenance expense - Excess of municipal revenues over municipal debt service (return on equity) is revenue to the State ($060,000) Municipal State Assistance Equivalent Debt Municipal REA-Guaranteed REA State Net of Coverage Cost of State FY Service Revenues Debt Service Revenues Assistance Revenues Assistance Cin 1983 $"s) 1996 102 112 677 338 339 329 145 1997 102 118 677 352 325 309 128 1998 102 122 677 366 SLL 291 113 1999 102 128 677 382 295 269 98 2000 102 LSS 677 399 278 247 85 2001 102 139 677 417 260 223 72 2002 102 146 677 436 241 197 60 2003 102 152 677 456 221 171 48 2004 102 E75: 677 525 152 79 21 2005 102 188 677 565 112 26 7 2006 102 202 677 607 70 30* 2007 102 218 677 654 23 93* 2008 102 226 677 677 0 124% 2009 102 226 677 677 0 124% 2010 102 226 677 677 0 124* ' ' ' ' ' ' ' ' ' ' ' ' ' 2026 102 226 677 677 0 124% 2030 0 226 0 677 0 903* * Return on Equity Present Value of State Equity Contribution: Present Value of State Assistance through 2005: Present Value of Total State Assistance $ 800 million 778 million $ 1,576 million OOLNAY KENTCO event the State will receive a considerable annual return in later years. (See Exhibit 65) The recommended finance plan can be modified to use more equity investment and less subsidy, or less equity investment and more subsidy. We recommend the financing plan as shown in Exhibit 65 as it tends to minimize the total amount of State capital required and spreads out the State capital requirements over a large number of years. Other combinations of investment and subsidy would require largerappropriations in fewer years, placing a strain on the State budget. Alternative Watana Finance Plans A promising alternative to the plan of finance shown above is a plan which uses "climbing-coupon" debt. Under this plan some of the REA-guaranteed debt used to finance the project starts with a low interest rate, such as two percent, and in each succeeding year the interest cost increases until it reaches a market rate, such as 12 percent. The debt sells at a discount to compensate the lender for the lower, early interest return and to therefore give him overall a rate of return equal to the market rate. The benefit of such a plan is that with a lower early interest cost, the debt service more closely matches the ability to pay of the utilities. Based on our interviews with Mr. Laurence Bladen, Financing Policy Specialist for REA, such debt appears eligible for the REA guarantee if other aspects of the debt are un- changed from the level debt described previously. We believe - 188 - KENTCO that a guarantee such as the REA guarantee would be absolutely necessary for the successful mixkering of climbing-coupon debt. The value of a finance plan which uses climbing-coupon debt is that it reduces the amount of State operating assistance necessary. A finance plan similar to our recommended plan, but which uses climbing-coupon debt as a portion of the total REA- guaranteed debt, requires $667 million (in 1983 dollars) for State operating assistance and $1,467 million total State assistance, as opposed to $778 million in operating assistance and $1,578 million total under our recommended plan. However, we cannot state with certainty that REA-guaranteed climbing-coupon debt can be sold. Our interviews with various finance specialists on Wall Street and elsewhere encourage us that it could be sold, but they, too, have not tested this debt on their customers and, therefore, cannot answer with certainty. Since the availability of climbing-coupon debt is uncertain, we are not recommending its use as our preferred finance plan. However, we strongly recommend that the idea be pursued, as it may prove to be a valuable cost reduction feature of Watana's financing. Exhibit 66 shows a plan of finance which uses climbing- coupon debt, along with level debt and equity. Other alternatives for financing Watana include the use of tax-exempt debt, the use of a different combination of REA and tax-exempt debt, and the use of special debt features at the time of financing. Tax-exempt debt has the advantage of a lower interest cost. Although the effect on power rates of lower interest cost may be largely or totally offset by the debt service coverage - 189 - - O61 Exhibit 66 PLAN OF FINANCE, WATANA Assumptions: - Municipal Debt Service: $102 million - REA-guaranteed level debt service: $520 million - REA-guaranteed climbing-coupon debt service: as shown below - Project revenues based on gas alternative plus 20 percent - Revenues available for debt service are net of operations and maintenance expense ($000,000) State Assistance 1983 Cquivalent Mun. Debt Municipal REA Level REA Climbing Total REA REA State Net of Coverage Cost of FY Service Revenues Debt Service Debt Service Debt Service Revenues Assistance Revenues State Assistance . 1996 102 112 520 40 560 338 222 212 94 1997 102 118 520 60 580 352 228 212 88 1998 102 122 520 80 600 366 234 214 83 1999 102 128 520 100 620 382 238 212 77 2000 102 133 520 120 640 399 241 210 72 2001 102 139 520 140 660 417 243 206 66 2002 102 146 520 160 680 436 244 200 60 2003 102 152 520 180 700 456 244 194 55 2004 102 175 520 200 720 525 195 122 33 2005 102 188 520 220 740 565 175 89 22 2006 102 202 520 240 760 607 153 53 12 2007 102 218 520 268 788 654 134 18 4 2008 102 235 520 268 788 704 84 49* 2009 102 254 520 268 788 762 26 126* 2010 102 263 520 268 788 788 0 161* . ' . ' ' ’ ' . . ’ ' . ‘ ' ' . . ’ 2029 102 263 520 268 788 788 0 161* 2030 0 263 0 0 oO 788 0 1051* * Return on Equity 1983 Equivalent of State Equity Contribution: $ 800 million 1983 Equivalent of State Assistance Through 2007: 667 million 1983 Equivalent of Total State Assistance $ 1,467 million OOLN3S» KENTCO requirement, the amount which is collected as coverage can remain available for other purposes, such as purchasing new equipment or retiring bonds early. In addition, future market conditions may change, making tax-exempt interest costs more attractive relative to REA debt. However, tax-exempt financing also has several disadvantages, which we have set forth earlier along with our reasons for recommending the extensive use of REA-guaranteed debt. Should conditions change at the time of financing, the State may want to consider higher percentages of tax-exempt financing as an alternative. Exhibit 69 shows a financing plan for Susitna which uses exclusively tax-exempt debt and State equity. The assumptions underlying this plan are the same as for the REA financing plan, with the necessary exceptions that tax-exempt interest rates and debt service coverage requirements are different, and State money is used in larger amounts for capital investment in the project and not for power cost subsidies. Also, this plan assumes that Susitna has solved the tax-exemption problem, and that if necessary the utilities have reorganized as exempt entities. There is also an alternative of using different combina- tions of tax-exempt and REA debt. As we stated previously, more than 75 percent of Watana could be financed with REA debt, but only in order to reserve additional Watana capacity for growth in demand of the REA utilities. The savings in State equity requirements are minor; the only advantage of such a financing plan is that it could be available if Susitna does not solve SO Lie KENTCO Exhibit 67 DEBT CAPACITY OF PROJECT IF WATANA IS FINANCED WITH TAX-EXEMPT BONDS Assumptions: - Project revenues based on gas alternative plus 20 percent - Debt interest rate is 10.54 percent - Debt service coverage required is 10 percent - Level debt service - Operations and maintenance expense is $8.5 million in 1983 dollars - First year of project operation is FY 1996 ($000,000) 1996 Project Revenues Available: 469 Less: Operations and Maintenance 19 1996 Revenues for Debt Service and Coverage 450 Maximum Tax-Exempt Bond Issue: S753) Level Debt Service, 34 year term: 409 Plus: 10 percent Debt Service Coverage 41 Revenue Requirement from Project: 450 - 192 - ako) Assumptions: Exhibit 68 - Interest during construction is 8.9 percent - Equity investment needed is $1,884 million WATANA PROJECT DEBT USING $1,884 MILLION EQUITY TAX-EXEMPT FINANCING ALTERNATIVE - Equity is applied against project cost first, then debt financing used for the remainder - Interest applied against previous year's debt, plus one-half of current year financing needs ($000,000) Construction Less: State Plus: Amount YS Cost, Nominal Equity Previously Financed Interest .- Total Financed 1988 588 588 0 0 0 1989 626 626 0 0 0 1990 667 667 0 0 0 1991 710 710 0 0 0 1992 756 295 0 2a 482 1993 805 0 482 80 1367 1994 858 0 1367 161 2386 1995 914 0 2386 253 3993 Total Debt Financing for Project: 3553 Add: Pay Off of Debt for Existing Generation: 200 Total to be Financed 3753 OOLNSY Assumptions: WATANA ALTERNATIVE FINANCE PLAN Exhibit 69 TAX-EXEMPT DEBT KENTCO - Tax-exempt debt is used, level debt service, 10 percent coverage requirement - Excess of municipal revenues over municipal debt service is revenue to the State - Equity investment necessary is $1,884 million 1996 1997 1998 2029 2030 Mun. Debt Service 409 409 409 409 0 ($000, 000) Municipal Revenues Net of O& 450 450 450 450 450 - 194 - State M Assistance oO] =] -oo ce Excess Revenues Available for Return on Equity 41 41 41 41 450 KENTCO its tax-exemption problem.* Lastly, less than 75 percent of Watana debt could be REA guaranteed, and the remainder municipal debt, but these alternatives would require greater State equity, plus a possible partial reorganization of the Railbelt utilities into tax-exempt entities. In comparing the recommended finance plan and the tax-exempt alternative finance plan, it can be seen that greater amounts of necessary equity require less escalation of power rates after the start date of the project. At equity investment levels approaching $2 billion (in 1983 dollars) such as shown in the tax-exempt alternative, rates can remain constant through the life of the project. Thus, while such greater amounts of equity can serve to increase the total amount of State assistance given to the project, there is a benefit of lower long term rates if the higher levels of equity are invested. In addition to the basic financing alternatives presented above, there are a number of small improvements to be assessed and possibly used at the time of financing which can lower the cost of power or reduce the amount of State equity needed. These include the possible use of capitalized interest, deferred principal payments or floating rate financing. Capitalized interest is a financing technique in which the amount being lent to a project also includes an additional loan amount to pay for interest payments in the first one or two years of project operation. The additional loan amount, called capitalized interest, is repaid by the borrower later, but it We believe that 25 percent of the project serving municipal entities should be eligible for tax-exempt financing under present IRS regulations. However, the actual determination of tax exemption must be made by recognized bond counsel. - 195 - KENTCO does have the effect of lowering debt service in the first one or two years, which are the years in which it is most difficult to match alternative power costs. Deferred principal payments are a loan feature in which the borrower pays interest only for the first few years of the pro- ject's operation, thereby again lowering debt service in the early years. Floating rate financing is a type of loan in which the interest cost moves up or down, or floats, with market conditions. As we discussed earlier, this type of financing can cause wide swings in interest cost and can play havoc with utilities' rate- making. However, since floating rates generally are lower than long term fixed interest rates (but not always), the use of a small amount of floating rate financing in conjunction with a large amount of fixed rate financing can reduce interest costs while keeping overall swings in interest acceptably small. - 196 - KENTCO EVENTS NECESSARY BEFORE THE FINANCING TAKES PLACE A host of events must occur between now and the time that Watana goes to market. We group these into four categories: State Investment, Education, Utilities and Future Economics. State Investment State investment refers to the State of Alaska's partici- pation in Watana as a provider of funds for construction and operating costs. 1. Availability. One of the earliest events necessary for Watana's financing is to determine if adequate money will be available for the proposed amount of State invest- ment. This must occur even before adopting the proposed plan of finance. Our review of State revenue and spending forecasts shows approximately $900 million per year, in 1983 dollars, to be available for appropriation to capital projects, loan programs, other new programs, or increases in the operating budget between now and the year 1993. Our review also shows adequate money may be available for operating cost subsidies in the years after 1996 under certain State fiscal assumptions, although we caution that any forecasts a decade or more into the future are highly uncertain. 2. Criteria for state investment. Closely related to the likely availability of State funds is the question of whether a project meets State criteria for funding. The - 197 - KENTCO State should establish criteria, then judge the Susitna Project against those criteria. We will have more to say on this in our closing section. 3. Statutory framework. The State legislature may need to establish some type of savings account to accumulate appropriations to Watana over a number of years, and it may need to establish a system for appropriating early- year operating cost subsidies to the project. Further, it may wish to establish a system for receiving a return on its equity investment in future years. All of these need a legal framework, yet to be developed. 4. Appropriations. Under the proposed REA financing plan 178 the State will have to appropriate from $B million to AL $67 million per year in nominal dollars over the next Five eget years toward the construction cost of Watana, and roughly the equivalent amount in operating cost assistance in the years following start up of the project. It should be noted that if interest income is not retained in the construction fund the appropriations are not protected against inflation, more than $800 million in real value of appropriations may be necessary to achieve $800 million real dollar equity at the time of the investment. If interest income is retained and, as is likely, the interest income exceeds inflation, then less than $800 million in real value of appropriations may be necessary to achieve the desired investment. In the two cases approximately $882 million in real value of appropriations is necessary without interest - 198 - KENTCO income retention, and $694 million in real value with interest income retention, (Exhibits 70-71) Education The second group of events to occur before the financing takes place is educational in nature. We believe that good widespread understanding of the benefits, risks and possible consequences of the Susitna financing are necessary to reduce the risk of future financial problems. 1. Education of ratepayers. Utilities and ratepayers must be made aware of the finance plan for Watana, what happens to rates, and why. Utilities must understand and agree that the undertaking of Watana is to provide the most economic alternative for their source of power, and that rates to be paid are lower, long term,than the expected costs from alternative power sources. 2. Risk. The State administration and legislature must understand and be willing to accept the risk that State appropriations needs could be larger than those presented in the finance plan. If cost overruns occur, then meeting the first-year power cost target could require a larger State equity investment or a larger amount of annual operating cost assistance. Utilities Concurrent with all of the above, several actions must be taken by utilities before this plan is put into effect. 1. Creation of an REA generation and transmission cooperative. As we stated earlier in the report the - 199 - KENTCO Exhibit 70 APPROPRIATIONS TO WATANA CONSTRUCTION FUND INTEREST INCOME RETURNS TO GENERAL FUND Assumptions: - $800 million (in 1983 dollars) State equity invested in project - $1,130 million (in nominal dollars) State equity invested in project ($000, 000) State 1983 Cumulative Construction BY Appropriations Equivalent Total Drawdown 1985 226 199 226 1986 226 187 452 1987 226 176 678 1988 226 165 904 588 1989 225 55 542 542 Results: $882 million (in 1983 dollars) State equity appropri- ated to project. - 200 - KENTCO Exhibit 71 APPROPRIATIONS TO WATANA CONSTRUCTION FUND IF INTEREST RETENTION ALLOWED Assumptions: 1985 1986 1987 1988 1989 $800 million (in 1983 dollars) State equity invested in the project $1,130 million (in nominal dollars) State equity invested in the project Interest is earned at 9.7 percent ($000,000) Cumulative Less State 1983 Nominal Construction Appropriations Equivalent Interest Total Drawdown 178 tS7 Ly 195 178 147 SIT, 410 178 138 58 646 178 ! 130 81 905 588 178 122 47 542 542 Results: $694 million (in 1983 dollars) State equity appropriated to project - 201 - KENTCO Railbelt will need a G & T cooperative to serve as a conduit for REA guaranteed loan financing and to be a wholesaler of power from Susitna to the distribution utilities. It will be the G & T cooperative which files the applica- tion for the REA guarantee. 2. Negotiation of power sales contracts. All utilities purchasing Susitna power will do so through power sales contracts. These contracts will need to be in place at the time this plan is activated. The contract will specify the rates to be charged or the rate-setting formula, and the amount of power to be purchased. In general, the contracts will provide an understanding between the utilities and the wholesalers of how power will be purchased. Court tests of the contracts will achieve a better interest rate on the municipal portion of the debt. Future Economics Finally, the events which must occur prior to financing include the continuation of economic parameters along the lines of this January 1984 forecast. Oil prices should continue to produce adequate State revenues to provide for State invest- ment, demand should continue to grow such that the forecast for 1996 demand indicates full utilization of Watana, and the fore- cast for alternative forms of generation should still indicate a cost of alternatives similar to that used in this report. Finally, the outlook must continue to show a combination of interest rates and inflation in which the investment can - 202 - KENTCO produce affordable power. If the spread between the rate of interest and the rate of inflation widens, the project can become more difficult to finance. All of these parameters should appear at the time the financing begins at the end of this decade. We believe that the framework for analysis set forth in this report is applicable for analysis at any time between now and the time for financing. If a forecast of any of the above parameters changes, it should be a very straightforward matter to recalculate the economic and financial analysis, and to verify or change our recommended finance plan. Under the alternative of tax-exempt financing, the events necessary are essentially the same with the following excep- tions: 1. State funding. State funding will be higher in the tax-exempt financing alternative, $1,884 million as opposed to $1,576 million in 1983 dollars. More important, the larger amount of funding necessary occurs over a much shorter period of time, averaging $370 million per year for eight years (Exhibit 72) 2. Resolution of tax-exempt issue. For tax-exempt eligibility, no more than approximately 25 percent of the output of Susitna may be purchased by non-exempt entities such as REAs. The Railbelt would have to reorganize its utility structure so that exempt utilities such as munic- ipal utilities serve most or all consumers, or failing that find some other solution such as a direct charge to the consumers or a special federal exemption. Depending - 203 - KENTCO on the solution chosen, it could be adviseable to secure a specific ruling from the IRS which certifies that the project is tax-exempt before implementing the finance plan. - 204 - Exhibit 72 KENTCO APPROPRIATIONS TO WATANA, TAX-EXEMPT ALTERNATIVE Assumptions: INTEREST INCOME RETURNS TO GENERAL FUND - Equity investment in Watana is $1,884 million in 1983 dollars - Equity investment in Watana is $2,886 million in nominal dollars oN Appropriation 1985 1986 1987 1988 1989 1990 1991 1992 Results: 370 370 370 370 370 370 371 295 ($000, 000) 1983 Equivalent 326 306 288 270 253 238 225 167 $2073 million (in 1983 dollars) priated to project. = 205) — Cumulative Construction Total Drawdown 370 740 1110 1480 588 1262 626 1006 667 710 710 295 295 State equity appro- KENTCO Exhibit 73 APPROPRIATIONS TO WATANA, TAX-EXEMPT ALTERNATIVE INTEREST INCOME RETAINED IN CONSTRUCTION FUND Assumptions: - Equity contribution to project: $1,884 million in 1983 dollars - Equity contribution to project: $2,886 million in nominal dollars - Interest earned on unspent appropriations retained in construction fund - Interest rate is 9.7 percent ($000,000) 1983 Cumulative Constructio _FY Appropriation Equivalent Interest Total Drawdown 1985 288 254 28 316 1986 288 238 59 663 1987 288 224 93 1044 1988 288 210 131: 1463 588 1989 288 197 114 1277 626 1990 288 185 92 1031 667 1991 294 178 52 710 710 1992 295 169 0 295 295 Results: $1,655 million (in 1983 dollars) State equity appropriated to project. - 206 - KENTCO DEVIL'S CANYON PLAN OF FINANCE Economic analysis shows Devil's Canyon to be a very strong project. It produces 3600 gigawatts of power annually, and has a construction cost of $1.6 billion in 1983 dollars. Compared to Watana, Devil's Canyon can deliver the same amount of power at half the investment. In part this favorable situation re- sults from Devil's Canyon relying on the large reservoir and transmission infrastructure already developed for Watana. At 100 percent utilization and 100 percent level debt service financing, Devil's Canyon can produce power at 5.7¢ per kilowatt hour, in 1983 dollars, which is within the ability to pay of the power purchasers in all years of its operation. Over time the cost of Devil's Canyon power declines in real terms to less than one cent per kilowatt hour, while the cost of thermal generation climbs to nearly 10 cents. When we convert the economics of Devil's Canyon to cash flow capabilities in the same fashion as previously done for Watana, it is clear that Devil's Canyon can generate sufficient revenues to pay level debt service, even if financed with 100 percent debt and no State equity. Therefore, we conclude that a plan of finance will be to use all debt financing, with no need for State equity or other State financial assistance. Of interest in Devil's Canyon is its timing, since Devil's Canyon is economically feasible at less than full utilization. However, under partial utilization, the cost of Devil's Canyon power will be higher per kilowatt hour sold since the costs of the dam are the same in any case. If less power is used, each - 207 = KENTCO kilowatt hour sold must carry more of the costs of the dam. Partial utilization would raise the cost of powerfrom Devil's Canyon beyond the guidelines we have established for ability and willingness to pay, i.e. 10 percent at retail above the cost of the thermal alternative, corresponding to 20 percent above the cost of alternative generation when comparing genera- tion alone. Qur plan of finance assumes that Devil's Canyon is brought on-line at about the time that Devil's Canyon power would be 100 percent utilized. The State may wish to accelerate the development of Devil's Canyon for economic reasons, and appro- priate operating subsidies or equity to bring down early-year cost in the same manner as we have prescribed for Watana. Such a move could give the State additional flexibility to match Devil's Canyon's timing to the retirement of obsolete thermal generating plants, to the depletion of alternative fuel sources, or to take advantage of favorable financial conditions. Assumptions The following assumptions underly the plan of finance for Devil's Canyon: (Exhibit 74) Cost of Devil's Canyon. Devil's Canyon will cost $1,622 million in 1983 dollars. This is a combination of construction cost and interest during construction. In addition the project will refinance existing utility generation debt for obsolete equipment, assumed to be $88 million in 1983 dollars. - 208 - KENTCO Use of debt. Devil's Canyon is assumed to use REA guaranteed debt at an interest rate of 11.78 percent and municipal debt at an interest rate of 10.54 percent. As stated above, the choice of debt financing may change depending on future market con- ditions. Total debt financing needed including interest during construction and retirement of obsolete generation debt is $9,366 million. Power Utilization. Devil's Canyon will be 100 percent utilized when it comes on-line for the first full year of oper- ation. Implicit in this assumption is a growth in Railbelt demand to approximately 7500 gigawatt hours annually in the year 2025. Rate of inflation. Inflation is 6% percent per year,but since the finance plan does not rely on an increase in power rates, inflation is not a critical assumption. State equity. None required. Operation and maintenance costs. Operation and maintenance costs for Devil's Canyon are $2.7 million in 1983 dollars, per year. - 209 - ~ ULG 7 deste KK Exhibit 74 DEVIL'S CANYON PLAN OF FINANCE Assumes Project Comes On-Line in 2010 Exhibit Shows First Year and Representative Years Thereafter Assumptions: - Devil's Canyon is 100 percent utilized - Financing is 25 percent municipal, 75 percent REA - Rates charged are sufficient for debt service plus coverage - Level debt service, 34 year term - Total municipal financing: $2,342 million @ 10.54 percent - Total REA-guaranteeed financing: $7,024 million @ 11.78 percent - Municipal debt has 10 percent coverage requirement - Operation and maintenance expense is $2.7 million in 1983 dollars ($000, 000) 1983 Equivalent Mun. Municipal REA REA Total 1983 Rate Per FY Revenues * Debt Service Revenues* Debt Service Revenues** Equivalent Kilowatt Hour 2010 281 255 847 847 1117 204 5.7 cents 2015 281 255 847 847 1122 150 4.1 2020 281 255 847 847 1130 110 Bak 2025 281 255 847 847 1140 81 Zuo 2030 281 255 847 847 1154 60 Loe 2035 281 255 847 847 1173 44 2 2040 281 255 847 847 1200 33 0.9 2045 281 0 281 0 1236 25 On7 2050 281 0 847 0 1286 19 0.5 2055 281 0 847 0 1353 15 0.4 2059 281 0 847 0 1425 12 On > m Zz * Net of operations and maintenance expense a Inclusive of operations and maintenance expense Oo must KENTCO EVENTS NECESSARY BEFORE DEVIL'S CANYON FINANCING Before Devil's Canyon can be financed, a number of events occur: 1. Growth in demand. Total Railbelt demand must grow to an annual level of approximately 7500 Gwh before Devil's Canyon is financially viable without State finan- cial assistance. The actual level of demand necessary may be higher or lower, depending on interest rates and other financial conditions at the time. 2. Reorganization of Railbelt utilities. Reorganization of Railbelt utilities would not be necessary if the financ- ing for Devil's Canyon uses the same type of debt as that used for Watana. If the type of debt financing for Devil's Canyon is different, then reorganization may be necessary to meet eligibility, as explained in the Watana finance plan. 3. Negotiation of power sales contracts. As in Watana, all utilities intending to purchase power from Devil's Canyon will sign power sales contracts in advance provid- ing for price, quantity and other terms of sale. 4. Continuation of economic parameters. Economic conditions must continue to indicate the feasibility of Devil's Canyon. Because of the present very favorable indications for the dam, the economic conditions are likely to be satisfied given adequate Railbelt demand. However, at some low cost levels of alternative thermal generation, or at some very high levels of interest rates, or combinations of the two, - 211 - KENTCO the project may not be financially feasible without State assistance. STATE EQUITY We must consider the nature of State investment in the Susitna Project. In particular, we need to review the role it plays, an understanding of its effects, the expected return on the State's investment, and the motivation for State investment in Susitna. In concluding, we discuss the need for a Statewide energy policy as a framework for State energy investment. In all of the Susitna finance plans presented above, State equity plays a large role. The REA plan of finance requires a State investment of $800 million in 1983 dollars as up-front equity, plus $776 million in 1983 dollars for operating assistance. The tax-exempt finance alternative requires an up-front State investment of $1,884 million in 1983 dollars. These amounts of investment represent a sizeable percentage of the State's projected capital available for all investments. In the 1983 legislative session, for example,the total State budget for all capital projects was approximately $700 to $800 million. At present there are no contractual provisions for the State to receive a cash return on its Susitna investment, and there is no appropriate statutory framework for a cash return.. Further, it is not clear whether, if the State requires a return, it intends to collect that return through charges to the pro- ject or its users, or whether the State is content to leave all benefits of the project and of-the State's investment with the - 212 - KENTCO users, or some combination of user benefits and user charges. The State may be called upon to provide additional cash beyond that called for in the plan of finance, if cost overruns occur or economic conditions change such that more equity is needed to keep the project financially sound. In essence, the State is assuming cost overrun risk or alternative-cost underrun risk. According to our calculations, using our cost numbers, the project could suffer a 20 percent cost overrun before it approached the cost of the most attractive thermal alternative. For all the above reasons, it is clear that the State is taking some risk in its investment. We assume that the State would do so through a combination of its responsibility to assist in planning for the provision of power in the Railbelt, and the prospect of cash returns on its investment. A final consideration is that’the Susitna Project serves the majority, but not all, of the State's population. Residents outside of the Railbelt will not receive power from Susitna. This presents a question for the state in its allocation of funds. Specific recommendations regarding the state's attitude toward risk, rate of return and allocation of funds is beyond the scope of this report. In all other respects this report has been able to create the framework for a decision regarding investment in future power generation for the Railbelt, includ- ing evaluation of the alternatives and criteria for drawing conclusions. Our economic discount rate, and our financial analysis were within a comparable framework for all alternative sources of capital. Our finance plan choices looked at rate eNO Re KENTCO of return and risk. In the case of State investment, however, we are unable to create such a framework. The alternatives for State investment are political, and the rates of return require- ments are political. We, therefore, cannot draw conclusions about the advisability of the State's potential investment other than what has been said before. We believe that a framework should exist for evaluation of Susitna, and that framework should be a statewide energy policy. We believe that a decision on State investment can be made better after the tate has established its priorities and criteria. Features of such a policy would include the follow- ing: Elements of a Statewide Energy Policy 1. Relative Priority of Energy to Other Capital Needs. Out of the State capital available for investment, the state should clarify the priority of energy projects in relation to other capital needs such as roads, ports, buildings, loan programs and so on. 2. Definition of State's Energy Needs. The State should have as much information as is practical regarding energy needs. This includes demand forecasts, resource informa- tion such as the quantity of gas, oil and coal reserves available for in-State use, together with their prices. This is important to Susitna because the state should be able to monitor the economic and financial progress of the project. - 214 - KENTCO 3. State's Planning Horizon. The State should decide over what period it is selecting the best alternative; 10 years, 30 years or 50 years. We selected Susitna as the best long term alternative but our analysis might have changed with a shorter planning horizon. 4. Criteria for State Investment. The energy policy should identify the most desired benefits for State investment. These benefits may be future revenues to the State, or they may be- benefits to energy users such as lower rates, stability of energy costs, or assurance of energy availability. In any event, these benefits are the return for the State's investment and the statewide energy policy should set return standards, or requirements, for evaluating invest- ments. 5. Distribution of Benefits. If equitable distribution of of benefits is an important consideration to the State, then the statewide energy policy should address it. Should per capita benefit be standard statewide? Should rate of return on state energy investments be uniform statewide? If the state collects a cash return, should this return be distributed statewide? 6. Power Pricing. If state investment allows the State to have a role in power pricing, then the energy policy should address pricing policy. Different pricing policies can have different effects on energy conservation. Also, pricing policies have an effect on the cash flow from an investment, which in turn has an effect on financial - 215 -° KENTCO feasibility and return on investment. We believe that a statewide energy policy can provide answers to many of the questions about Susitna, and can provide the answer to the advisability of State equity investment in this project. - 216 - KENTCO CONCLUSIONS The primary objective of our study was to develop a viable finance plan for the Susitna Project once comparative analysis had shown that particular project to be the most economical means of providing power to the Alaska Railbelt Region. In accomplishing this, we developed several additional findings and techniques which will be useful to the State of Alaska and other interested groups in their energy and financial planning. This report developed a straightforward model for evalua- tion of Susitna's two projects. This model will be useful in continuing to monitor the project and the appropriateness of the plan for Railbelt power generation. In the Susitna decision itself, the model identified the major alternatives for power generation, and then described the key cost elements of each alternative as well as how the costs would behave in the future. The model then allowed us to make a forecast, analyze what the costs of the alternatives would be under our forecast and reach a decision that the Susitna Project, given today's best estimate of future conditions, is the most cost-effective means of supplying power to the Alaska Railbelt in the future over the long term. Given the economic viability of the project and estimates of the cost of power in the absence of Watana and Devil's Canyon, we took a hard look at financing. The study has identified the alternative sources of capital and modeled the key elements of each source, chiefly rate of return and risk. - 217 - KENTCO We then selected the most appropriate sources of capital and prepared a finance plan, based on present financial conditions. One of the most important features of the finance plan is its ability to accommodate the willingness to pay and ability to pay of the power purchaser with a minimum requirement for state equity investment. Exhibit 75 compares costs of power under the finance plan to costs of power using the thermal alternative of gas. In both the economic analysis leading to a decision of how to best meet Railbelt power needs and the financial analy- sis leading to the best means of financing the project, the models above can accommodate changes in economic or financial parameters as new information becomes available. The decisions then can be verified or changed. The study identified and quantified the value of retaining Cook Inlet gas reserves for home heating. This turned out to be a significant factor in the economics of the generation alternatives, adding an opportunity cost (foregone benefit) of over one billion dollars to the cost of using gas to generate electric power. The use of the remaining Cook Inlet reserves should be carefully considered in any evaluation of Railbelt generation alternatives. The study identified a way to finance a project in which the requirements for payment closely follow inflation. The inappropriateness of level debt service in a time of high inflation has been a hindrance to large scale projects nationally. Alaska's wealth allows the state to pick up some of the debt service early, recapture it later, - 218 - = 61 = Assumptions: Gas generation costs follow development on previous exhibits Exhibit 75 COST OF GAS GENERATION COMPARED TO RATES CHARGED Gas generation plants replaced and refinanced in year 2021 Watana costs follow development shown on previous exhibits FOR WATANA POWER 1996 AND REPRESENTATIVE YEARS THEREAFTER Watana operations and maintenance costs are $8.5 million annually in 1983 dollars ($000, 000) Nominal Cost 1983 Equivalent 1983 Equivalent 1983 of Gas Equivalent Cost Per Watana 1983 Cost Per FY Generation Cost Kilowatt Hour Revenues Equivalent Kilowatt Hour 1996 391 172 4.9 cents 469 207 5.9 cents 2000 464 159 4.5 557 191 5.5 2005 656 164 4.7 787 197 5.6 2010 953 174 5.0 950 173 4.9 2015 1402 187 5.3 967 129 Bail. 2020 1959 191 5.5 990 96 2.7 2025 3668 260 7.4 1023 73 2.1 2030 5313 275 7.9 1067 55 1.6 2035 7795 295 8.4 1128 43 1.2 2040 11,628 321 9.2 1211 33 0.9 2045 17,249 348 9.9 1325 27 0.8 OOLNS» KENTCO and ease the burden on the users of the project. This type of financing should be useful not only to Susitna but to other hydroelectric projects and large scale projects in the State. The plan of finance developed for Susitna reduces the amount of up-front state equity required. Given limits on the amount of capital dollars available for state investment, this reduction in equity requirements can increase the number of projects receiving State assistance. The recommended sources of capital remove a potentially large quantity of debt from the State's balance sheet and from the State's total borrowings in the municipal market. Although difficult to quantify, the reduction of billions of dollars of revenue bond debt under the State of Alaska name may improve the interest rate or market acceptance of other state agency revenue bond debt. The plan of finance creates a long term source of revenue to the State. Beginning approximately 10 years after the start of Watana, the State can receive a return on its equity invest- ment. Once the outstanding debt on the project is retired, 35 years after the start date of Watana, the return is greatly increased even with no increase in rates (which at that time will be approaching only one-tenth of the cost of the thermal alternative). Thus, the State's equity investment resembles the ideas propounded for a Capital Investment Fund, or a second, "mini" Permanent Fund. With the U.S. Government guarantee behind the REA debt, the plan of finance creates a very high quality instrument - 220 - KENTCO which can be eligible for in-state Alaska Permanent Fund invest- ment. The Permanent Fund, because of its requirements for prudent diversification, may not be able to purchase a very large portion of the total Susitna debt. However, this finance plan can repre- sent a start toward in-State Permanent Fund investing. Our interviews with Mr. David Rose, Executive Director for the Alaska Permanent Fund, support our belief that some REA-guaranteed debt could be sold to the Permanent Fund, including, possibly, climb- ing-coupon debt. Finally, the study gives clear support to the need for a statewide energy policy. Both policy statements and information regarding the state's energy resources and energy needs are insufficient to make timely, clear decisions about major energy projects. The need for a statewide energy policy was particu- larly evident in questions regarding state equity investment, its magnitude and its requirements for return. Susitna is a very large project. The State definitely does not want to slide into Susitna. Rather, all financial and operational aspects of the project and its construction must be carefully planned from the start. This study is one more step in the large amount of preparation needed for Susitna. tt KENTCO ELECTRIC POWER GENERATION FOR THE ALASKA RAILBELT REGION APPENDIX 1 PETROLEUM INDUSTRY RESEARCH ASSOCIATES, INC. 122 EAST 42ND STREET + SUITE 515 NEW YORK, N. Y. 10168 (212) 867-0174 A_LONG TERM OUTLOOK FOR WORLD OTL PRICES Prepared For: William Kent and Company November 3, 1983 eles ToL Iv. Vv. PIRA TABLE OF CONTENTS PAGE INTRODUCTION ak ASSUMPTIONS IN THE ANALYSIS 6 NON-COMMUNIST WORLD ENERGY SUPPLY/DEMAND BALANCES 10 OIL SUPPLY/DEMAND AND PRICE EXPECTATIONS 13 A COMMENT ON WORLD NATURAL GAS PRICES 20 PIRA I. INTRODUCTION In this memorandum PIRA outlines its view on future world oil price developments for the remainder of this century and speculates about oil price trends beyond this period. The perception of the world oil supply/demand outlook has shifted rather dramatically in recent years from one of scarce supply and rising real prices to one of protracted excess capacity and falling demand and real prices. Although PIRA's analysis and projections have for some time supported crucial aspects of this revisionist view of international energy balances and their obvious impact on interfuel pricing, we have also continued to emphasize the fact that oil's role in the emerging energy future, though diminished, will nevertheless be important through at least the end of the century. The price of oil like any other commodity will be determined by the complex interaction between supply and demand,neither of which can be independently determined without an initial price expectation. Therefore, the first step in estimating future price developments is to set out a reasonable price assumption which can then be utilized to independently estimate a demand and supply outlook. The resulting balance between supply and demand determines the appropriateness of the initial Brice assumption; that is if supply is greater than demand at the assumed price then that price is too high or alternatively too low if demand is greater than supply. In the case of oil this methodology is complicated by the existence of an organization whose primary function is to set PIRA world oil prices. Controlling over three-quarters of the non- communist world's (NCW) proved oil reserves and a major portion of the world's oil supplies, the Organization of Petroleum Exporting Countries (OPEC) has been the dominant factor in world oil price determination for over a decade. By controlling available supply via their production policies, OPEC has successfully managed to maintain oil price levels substantially above those which would be present in a free market. The concentration of huge oil reserves in but a few of the member countries in the Middle East has to date provided the necessary leverage for OPEC to set world oil prices despite the existance of substantial surplus producing capacity in the Organization. Given the huge current account surpluses these countries were able to accumulate in the 1970's because of their limited internal capital absorption capability, and their overriding preference for higher then free market prices, substantial reductions in production were accommodated to support OPEC price levels. Thus while OPEC crude production has declined from a peak of almost 31 million barrels per day (MMB/D) in 1979 to about 17 MMB/D this year, the Organization has remained an effective price determining body. The 14 MMB/D reduction in OPEC crude production over the last four years reflected the combination of: (1) reduced oil demand (8 MMB/D) due to conservation and substitution of other energy sources for oil stimulated by the quantum Lace neHee ano) prices in both 1973/1974 and 1979/1980 periods as well as depressed economic activity; (2) increased non-OPEC supplies (4.5 MMB/D) and (3) inventory swings when comparing 1979 (+ 1 MMB/D) PIRA and 1983 (-0.5 MMB/D). Given these circumstances it is not Surprising that the degree to which OPEC controls world oil prices has weakened considerably. Its surplus producing capacity is quite large, and in addition the Middle East OPEC members have required that reductions in production be to some extent shared among all members, and those outside the Middle East are in the weakest financial situation to do so. While to date the latter have agreed to produce significantly below their capacity, the incentive to exceed currently allocated output levels is constantly present. Consequently, oil price forecasting is inherently frought with uncertainty. Any price level above that in a free market is arbitrary and while OPEC has set $29/Bbl currently as their marker crude price, it could -just as easily be $15, $17, $23 or $34 per barrel as long as whatever price level chosen is supported by OPEC's production policies. Given OPEC's substantial excess producing capacity, any near term price adjustments are likely to be downward rather then upward. A lower price would encourage economic expansion and oil use and thus would relieve to some extent the capacity overhang. But with oil demand being inelastic in the short term, lower oil prices would obviously reduce oil revenues and worsen the current financial plight of many OPEC countries. Thus the price is not likely to be reduced unless the Organization's ability to control its members' output collapses, a possibility but not an inevitable one. The price reduction in the marker crude last March of this year from $34 to 29/Bbl. was engineered by the Middle East OPEC producers, especially the largest producer, PIRA Saudi Arabia, and signalled a recognition that oil prices had moved too high, too quickly. In addition, it is significant that the price reduction also provided a degree of shock therapy for those in OPEC who were undercutting the price structure. From OPEC's perspective it was hopefully a lesson which will not be easily forgotten. The shock therapy was also painfully felt in the United Kingdom and Mexico, both of which gave informal support to the OPEC price structure established last March. These non-OPEC producers, as well as all others, have benefitted greatly from the price umbrella established by OPEC. By slightly underpricing their crudes relative to those of OPEC, they have ensured maximum market penetration for virtually all of their available oil supplies. Mexico is the only major non-OPEC producer which has actually shut-in upwards of ten percent of its production capacity in support of the OPEC price structure. While, with the exception of Mexico, non-OPEC oil producers have to date not taken upon themselves to reduce production, many of them have a vested interest in not seeing prices further substantially reduced. It is ironic that the bulk of the world's excess producing capacity is located in the world's most prolific and reserve ricn producing countries which are all situated in the volatile and politically unstable Middle East. Moreover, as noted, non- OPEC oil production has, in general, been at capacity and therefore these reserves have been, and continue to be, depleted at a relatively faster rate than those in the Middle East. An eventual return to increased reliance on this region is PIRA inevitable and consequently overtime the oil pricing power of the Middle East OPEC members will be enhanced. It is for this reason that PIRA believes that even if oil prices decline further in nominal terms over the next year or so, these producers will be in a position to make up for lost ground later on. PIRA II. ASSUMPTIONS IN THE ANALYSIS The most important assumption is, of course, that for crude oil prices. Based on the considerations outlined earlier, PIRA has defined a reasonable price outlook as the basis for developing NCW energy balances: The OPEC marker crude price--Saudi Arabian Light (34° API, 1.7% S)--is assumed to remain at $29/Bbl (FOB Ras Tanura) through 1985 and subsequently rise at the rate of inflation in the principal importing countries through 2000. It is the appropriateness of this price assumption which PIRA tests in its energy supply/demand analysis. This price path assumes implicitly that OPEC, or some other form of inter-governmental price setting mechanism or arrangement, will continue to keep world oil prices substantially above the level which would prevail andes conditions of unconstrained competition. We recognize that this assumption is currently more open to challenge than at any time in the last 10 years and that the challenge has a valid base. Nevertheless, we consider a scenario of administered oil prices more plausible than one in which oil prices will over an extended period of time be determined solely by market forces. However, the policy, form and mechanism of the price administration could be quite different from that of the last 10 years. A major factor underlying our view is that in almost all oil exporting countries prices are determined by political bodies whose basic orientation is towards price protection rather than market competition. Given their demonstrated success of administering oil prices, these bodies can be expected to retain or regain, as the case may be, some form of price control. This 6 PIRA does not mean that the setters of future world oil prices can ignore market conditions but, rather, that they should be able to modify underlying market pressures to their advantage, but much less so than. during the period 1973-82. In the last half our forecast period, rising marginal production costs should reduce the gap between our assumed administered price and the price which would otherwise prevail since incremental production will increasingly be supplied from higher cost, frontier areas. We have also made assumptions about economic growth rates in major regions of the world which are consistent with the oil price trajectory. These are shown in Table l. Table 1 ANNUAL REAL GNP GROWTH RATES (% per year) 1973-1981 1281-1990 1290-2000 U.S. 2.3 2.7 ASE! Other OECD 2.4 2.8 3.1 LDC's 4.9 4.3 4.2 CPE's 3.1 2.8 223 World 2.8 3.0 3.0 These rates are probably below the political aspirations of any of these regions but they are high enough to permit significant per capita growth rates in all of them. In the LDC's which will have the fastest population growth rates, the GNP rises about twice as fast as the population under our assumption. Next, we determine the relationship between economic growth and energy demand to find the total amount of energy required. We then estimated the amount of non-oil energy sources available to fill this requirement, with the balance being supplied by the PIRA oil which, under our price assumption, would continue to be the swing fuel in world energy. Our assumption regarding the energy/GNP relationship reflects the recent fairly significant decline in world oil prices which we believe will over time accelerate economic growth and energy demand from what both would have been if the 1981-1982 price level had been maintained. The lower price will also retard the development of some new energy supplies. However, it is our view that, notwithstanding the lower oil prices, energy demand will continue to grow at a significantly slower rate than GNP in all regions except the LDC's. One reason is that the move towards more energy efficiency which was triggered by the two oil price shocks of the 1970's has now taken on a-life of its own. New automobiles will continue to become more fuel efficient, new homes and other buildings will be designed to conserve energy and the same will be true of new industrial equipment. New electronic control technology will be an important tool in this development. Another reason is the ongoing structural change in industrial production. In the industrial countries the contribution to GNP of the so called "smoke stack" industries which are generally energy intensive is declining while that of high technology industries with relatively low energy requirements is increasing. We expect this trend to continue. Table 2 shows our energy efficiency projections. PIRA Table 2 ENERGY EFFICIENCY, 1973-2000 Metric Tons of Oil Equivalent per Thousand Rate of Change 1975 $GNP (% per year) 1973- 1981- 1990- 1973 1981 1990 2000 198) 1990 2000 U.S. 1.16 0.96 0.82 0.72 -2.3 -1.7 -1.4 Other OECD 0.73 0.63 0.59 0.52 -2.0 -0.7 -1.3 LDC's 0.72 0.75 0.74 0.69 0.4 0.2 =0.6 Average 0.87 0.75 0.69 0.62 -1.8 -0.9 -1.2 Table 3 shows the total volume of energy required, based on the energy/GNP relationships developed in the previous table. Table 3 OECD AND LDC ENERGY CONSUMPTION, 1973-2000 Million Metric Tons Oil Growth Rates Equivalent (MTOE) ($ per year) 973 — 1981- 1990- 1973 1981 1990 2000 1981 1990 2000 U.S. 1823 1808 OTT 2152 -0.1 1.0 0.8 Other OECD 1817 1870 2255 2695 0.4 21 1.8 LDC's 626 251 1365 1944 S24 Asl 3.6 Total 4266 4629 55,977 6791 1.0 2.1 2.0 PIRA III. NON-COMMUNIST WORLD ENERGY SUPPLY/DEMAND BALANCES To derive a forecast of oil demand based on the assumed oil price trajectory PIRA has made estimates of the likely availability and consumption of non-oil energy resources. These estimates are summarized in Table 4. Table 4 1981-2000 MTOE GROWTH RATES (% PER YR.) 1981 1990 2000 1981- 1990- 1990 2000 Coal 973 1260 1678 2.9 2.9 Nat.Gas 894 1097 1225 2.3 Lk Nuc. Pow. 172 440 75 121 5.0 Hydro & Other 349 456 571 3.0 2.3 . Total, Non-Oil 2388 3253 4189 3.5 2.6 Total Energy 4629 5597 6791 Zel 2.0 Required Oil Supply 2241 2344 2602 0.5 1.0 Memo Item: Million B/D Required Oil Supply* 47.0 49.4 54.7 *Converted to barrels per day taking into account likely volumetric gain from refinery processes. Of the various non-oil energy sources we expect nuclear power to rise most rapidly: at least 11% annually to 1990 and then at about half that rate to 2000 as the backlog of plants currently under construction iscompleted andthe substantial 10 PIRA reduction in the number of new plant orders in recent years reduces plant completions in later years. Next to nuclear power the fastest growing major energy source will be coal. Since both these fuels, as well as water power which will also grow faster than total energy, are primarily used to generate electric power, their growth rates are an indication of the increasing role of electricity in the world energy supply pattern between now and the end of the century. The conversion of coal into synthetic fuels is not likely to, in general, be economically attractive at current or projected fuel prices. Hence, coal synfuels, as well as most other synthetics, will not play a significant role in supplying world energy needs before the end of the century. The growth in coal demand in the industrial countries other than the U.S. will have to come almost entirely from increased imports. Thus, the growth rate in international coal trade is- likely to be more than twice that of non-communist world coal demand. Natural Gas in the U.S. accounts for nearly 60% of total NCW consumption. Between 1981 and 2000 U.S. natural gas demand is forecast to decline more than ten percent largely as a result of declining production. A portion of the output decline will be offset by rising gas imports from adjacent countries. In the other OECD countries gas consumption is expected to increase by around 60% between 1981 and 2000, with increased volume coming largely from imports from OPEC members, other LDC's and the Soviet Union. In LDC countries, where gas demand is likely to increase almost threefold, indigenous supplies will account for ll PIRA most of the expansion. The various non-oil energy sources, which supplied 52% of NCW energy requirements in 1981, are expected to increasingly contribute to meeting the world's energy needs. By 1990 some 58% of the NCW energy consumption will come from these sources; by 2000, 62%. Oil's role will obviously diminish overtime, but, nevertheless, will continue to be, by far, the most important source of energy to the non-communist world for at least the rest of the century. The estimated energy balance is, of course, based on the assumed direction of future oil prices. We turn now to determining the appropriateness of this assumption. 12 PIRA IV. OIL SUPPLY/DEMAND AND PRICE EXPECTATIONS Oil demand increases very slightly from 1981 to 1990, at an average annual rate of 0.5 percent per year. But this masks the sharp reductions in demand which have occurred in 1982-83, where oil demand declined some 4 MMB/D or 6.4% as shown in Table 5. From 1983 to 1990 oil demand growth will average about 1.7 percent per year, but will then drop back to 1.0 percent per annum growth in the 1990's. Table 5 PROJECTED NCW OIL CONSUMPTION, 1981~2000 Growth Rates Million Barrels Per Day (& Per Year) 1981 1983 1990 2000 1983- 1990- (prelim.) 1990 2000 47.0 44.0 49.4 54.7 1.7 1.0 In both the 1980's and the 1990's, the growth in demand will occur primarily outside the industrialized countries. In the U.S., oil consumption will rise in the next couple of years and then remain about flat during the remainder of the century. Relatively flat U.S. oil demand after 1985 will be due to a combination of further oil conservation, particularly in the transportation sector, further reduction in oil sales to the residential/commercial heating markets, higher demand for petrochemical feedstock and diesel fuel and a slowing of the displacement of fuel oil in the electric power and industrial markets. In the other industrial countries all growth will take place in the petrochemical and transportation sectors. In all industrial countries the historically unique drop in oil 13 PIRA consumption which started in the U.S. in 1979 and in the other OECD countries in 1980 is expected to end in 1983. Now let us consider whether our calculated oil volumes are likely to be available at our assumed price. It does not require much analysis to determine that the 49.4 MMB/D requirement for 1990 should be readily available. In 1979 oil supplies in the NCW were nearly 3 MMB/D higher than our projected level,and in 1982 when supplies amounted to 42.7 MMB/D there was enough available spare capacity to produce about 49 MMB/D despite the temporary production constraints imposed by the Iran-Iraq war. This together with the 1.4 MMB/D net communist bloc oil exports would have been adequate to meet the 1990 level last year, although with little spare capacity left. What a 49.4 MMB/D requirement will do to the industry's spare capacity by 1990 will largely determine the price of oil at that time. In PIRA's view, oil production will increase sufficiently to supply that requirement and still retain a significant volume of Spare capacity, although considerably less than at present. We estimate that crude oil and NGL production outside OPEC will rise by about 2.6 MMB/D between 1982 and 1990 or less than half the growth of the previous eight years. This will be partially offset by an anticipated decline in net communist bloc oil exports. The supply balances for both these years and also for 2000 are shown in Table 6. 14 PIRA Table 6 NON-COMMUNIST WORLD OIL SUPPLY AND DEMAND, 1982-2000 (Million Barrels Per Day) 1982 1990 2000 Oil Demand 45.3 49.4 54.7 Inventory Changes -1.2 00.0 00.0 New Supply Requirements : 44.1 49.4 54.7 Oil Supplies U.S.* 10.8 10.0 Canada 1.2 1.3 Mexico 3.0 3.7 North Sea 2.7 3.3 Other* 526 1.6 Subtotal 23.3 25.9 25.9 Net Communist Oil Exports 1.4 0.4 - Total Non-OPEC Supplies 24.7 26.3 25.9 Required OPEC Oil Production 19.4 23-1 28.8 Of Which: Crude 18.5 21.5 26.8 *Includes processing gains and minor quantities of synthetics. Estimates of OPEC's commercially available crude producing capacity as well as its "preferred" production level vary among experts. But there is general agreement that both are substantially in excess of the 21.5 MMB/D required for 1990. Consequently, absent a major physical interruption, the price pressure is likely to be downward until then. At the same time, the projected 1990 level represents a substantial improvement from this year's expected extremely depressed level of 17.2 MMB/D. Thus our initial assumption that the real world oil price in 1990 will be bel } 1 : £1983 iipeie; even if one were to assume a somewhat higher OPEC production level because of inventory build-up and restraints in production 15 PIRA increases by some non-OPEC exporters. For the year 2000 our 1% oil demand growth rate for the 1990's results in a demand level of 54.7 MMB/D. Tf Od) production outside OPEC does not increase during that period OPEC would have to supply 28.8 MMB/D in the year 2000, including nearly 27 MMB/D of crude oil. The possibility of such a development is very likely, particularly since present price expectations are likely to slow the development of some offshore and arctic area prospects, because of their long lead times and very large capital expenditures. Yet, most of the world's remaining reserves outside the Middle East are probably located in those areas. It is important to recall in this connection that we have assumed that production outside OPEC will, in eee remain at capacity, as it has since the early 1970's, so that as noted earlier, non-OPEC reserves are likely to be drained at a much faster rate than those of OPEC, particularly the group's Middle East (Arabian Gulf) producers which currently account for over 60% of NCW oil reserves. An eventual return to the Middle East Must therefore be expected. The speed of this development will be inversely related to the price of oil. But under any realistic assumption the production of Middle East oil will increase substantially during the remainder of this century from last year's level of 12 MMB/D and so will its share of world production. Based on current proved and probable OPEC oil reserves, the 28.8 MMB/D requirement for OPEC oil in 2000 is not expected to present any resource or technical constraint on OPEC's collective 16 PIRA produceability. However, since most of OPEC's 5.7 MMB/D required output increase from 1990 will have to come from Middle East suppliers, the 28.8 MMB/D requirement for OPEC oil could exceed some of these suppliers' politically or economically desirable production levels. In that case real world oil prices would rise again, regardless of whether there is still an effective price administering organization by then. The probability of such a development depends in part on the world oil buyers' degree of confidence in the political and commercial availability of the Middle East's excess producing capacity when needed. A high degree of confidence could act as a disincentive to oil exploration and development in other areas, all of which are higher-cost. A low degree of confidence would have the opposite effect. A related aspect is the future stake of foreign companies in Middle East oil. To the extent to which these companies are, or remain, excluded from participating in the economic rent generated by the low cost of Middle East oil through equity ownership or preferential access, they will seek their own economic rent by continuing looking for oil elsewhere at prevailing administered prices. This would tend to increase the volume of non-Middle East oil, and probably total non-OPEC oil, available in the 1990's. On balance, PIRA believes it can be reasonably expected that 1d i : ily ri : 1 : } 1990's. The primary cause will be increased market power in the hands of the Middle East OPEC producers, who in the past have not hesitated to use this power when it has existed. Whether real oil prices rise ay, PIRA one, two or five percent per annum in the 1990's is open to a great deal of uncertainty. Unfortunately, we have no real conviction about the specific real price escalation but believe it is more likely to be toward the lower end of the range than higher. Consequently, we have chosen a real oil price escalation of 2 percent per year in the 1990's. Based on the foregoing PIRA's world oil price forecast is shown in Table 7. Table 7 WORLD OTL PRICES*, 1983-2000 (Dollars Per Barrel) Nominal Price* fen Eres (1983 $)+* 1983 29.00 29.00 1990 38.25 26.10 2000 75.25 31.80 Note: Inflation in major oil importing countries is assumed to be 5.1% in 1984 and 1985, 5.7% per annum in the 1985 to 1990 period and 5.0% per year in the 1990's. *OPEC marker crude price, Saudi Arabian Light (34° API, 1.7% S), FOB Ras Tanura. Beyond the year 2000, projections of oil prices take on an even more speculative and uncertain nature since unforeseen technological changes can dramatically influence oil supply/demand balances. Barring a dramatic technological advancement in how oil is utilized or recovered from new sources of supply, continued oil price escalation in real terms is probably a reasonable assumption. Such a conclusion was recently reached by the U.S. Department of Energy (DOE) in its "Energy Projections to the Year 2010" published this past September. In its analysis, the DOE forecast 3 to 4.5% real oil price increases 18 PIRA per year between 2000 and 2010, off a year 2000 price which ranged from $36 to 80/Bbl in 1982 dollars. One of the important reasons behind the real oil price increase to 2010 was the DOE's estimate of the cost of unconventional oil sources such as shale oil and coal liquids which it estimated would be in the $50 to 80/Bbl range (1982 dollars). 19 PIRA Vv. A COMMENT ON WORLD NATURAL GAS PRICES Natural Gas will continue to compete at the margin with oil products in major end-use markets. Thus higher or lower oil prices will correspondingly affect the value and,in turn, price of natural gas. In most industrialized countries natural gas currently competes at the margin in boiler fuel markets with residual fuel oil since many of these facilities have dual-fired oil/gas capability. Many of the major oil producers outside the United States are expanding their capacity to absorb internally associated and non-associated gas production. In many instances their gas absorption capability is limited and some presently are planning or considering gas export projects in the form of LNG exports (See Tables 8 and 9). These projects are capital intensive and they have, in general, required considerable outside financing because of internal financial constraints in these countries. To date many planned projects have not come to fruition because of either an inability to get external financing or disappointment with expected netback returns on the gas. Only recently are potential gas exporters coming to the realization that to stimulate demand for LNG exports, the gas must be priced competitively with alternative fuels at the end-user burner tip. The relatively low netbacks implied by competition with residuai fuel oil at the burner-tip to allow absorption in boiler fuel markets, has tended to stall further major development of gas export projects. 20 T2 avi LNGtrade Contract Volume (BCF/YR) Comments Under Suspended/ Operat'l Constr. Expired To U.S. from Algeria Distrigas (Boston, MA) 45.50 - - Trunkline (Lake Chas., LA) 177.00 - - El Paso (Cove Point, MD, Elba Isl., GA) - - 353.00 Suspended over pricing dispute in '80. Total to U.S. 222.50 - 353.00 To France from Algeria Gaz de France (Le Havre) 20.20 = a (Fos sur Mer) 136.50 - 3 (Montoir) 202.30 = - To Spain from Algeria ENA Gas (Barcelona) 177.00 = iD To Belgium from Algeria Distrigaz (Montoir, Volumes rec'd in Montoir until Zeebrugge Zeebrugge) 202.30 - - completed (1986). To Spain from Libya ENA Gas (Barcelona) 50.60 - - To Italy from Libya SNAM (La Spezia) = - 126.50 Suspended in '82 over pricing dispute. To UK from Algeria British Gas Corp. (Canvey Isl.) - - 55.60 Contract has expired. Total to Europe 788.90 = 182.10 To Japan U.S.: Tokyo Gas, Tokyo Elec. (Negishi) 50.60 = = Brunei: Tokyo Electric, Tokyo Gas, Osaka Gas ; (Chiba, Yokohama, Sakai) 283.20 = - Abu Dhabi: Gas is associated production from Zakum and Tokyo Elec. (Chiba) 111.20 a - Umm Shaif fields. Indonesia: Chubu Elec, Kansai Elec, Osaka Gas, Kyushu Elec Nippon Steel (Himejii, Tobata, Chita, Osaka) 380.00 - - Badak & Arun volumes: 152 & 228 respectively. Malaysia: Tokyo Gas, Tokyo Elec. (Chiba) 303.50 - = Deliveries began in '83; full volumes not due until '86. Petronas 65%, Shell, Mitsubishi ea. 17.5%. WORLD LNG TRADE: FIRM PROJECTS (continued) Contract Volume (BCF/YR) Comments Under Suspended/ Operat'l Constr. Expired To Japan (continued) Indonesia (Badak): Chubu Elec., Kansai Elec Toho Gas, Osaka Gas (Chita, Himejii, Osaka - 161.80 co Scheduled to begin mid-'83. Indonesia (Arun): Tohoku Electric, Tokyo Electric (Higashi Niigata, Chib - 161.80 - Scheduled to begin in '84. Canada (Dome) : Chubu Elec., Osaka Gas, Kyushu Elec., Chugoku Electric, Toho Gas - 146.60 - Scheduled for '86 start-up; deliveries uncertain. Australia (N.W. Shelf): Not yet under construction. Tokyo Elec., Chubu Elec. Kansai Elec., Chugoku Elec., Kyushu Elec., Tokyo Gas, Osaka Gas, Toho Gas 7 - 303.50 - Scheduled for '87 start-up, plateau volume in '90. Not yet under construction. Total to Japan 1128.50 773.70 - All 2139.90 773.70 535.10 rm nr €2 WORLD LNG TRADE: POTENTIAL LNG PROJECTS Exporter/Market Indonesia/So. Korea Qatar/Far East, Europe Abu Dhabi/Japan Thailand/Japan Indonesia Nigeria/Europe? Camer oon/Europe? Chile Argentina Canada (Arctic Pilot)/ U.S., Europe? Trinidad/U.S., Europe Norway/U.S., Europe USSR (Sakhalin) /Japan Total Potential *Maximum volumes currently planned. Volume Start-Up (BCF/YR) 100 1986 300 1988-90 250 75 1988-90 150* 300 1990's 500* 300 1990 350* 150 1988-90 75 1990 200 125 1990 225 1989 300 late 1990's 150 1988-90 2550 2875* ee OV GL Comments Sales agreement firm. Supply from Arun plant to be used primarily for electric power generation by Korean Electric Power Co. Source, the offshore North Field, could be as large as 300 TCF (DeGolyer & MacNaughton). C. Itoh studying the development of onshore reserves. Market could also be Korea. Markets not established; new reserves from Natuna Isl. Ruhrgas, Gaz de France, SNAM discussed as potential buyers; volumes have been scaled down and partnership reorganized. Reserves volumes have been revised downward. Consortium Segazcam will develop. Potential markets: Spain, U.S., Japan. Technip and SNAM Progetti are studying. Problematic (political, economic). U.S. is a possible market. NEB has delayed consideration pending consummation of a sales contract. Another project (High Arctic Islands) is under consideration by Canadian suppliers and German buyers; development dependent on Arctic Pilot. Tenneco has explored U.S. markets, using existing terminals (e.g. Cove Point, MD); Gaz de France also a potential customer. Option for ‘development of northern Norwegian reserves. Some sources assess the project as problematic, with no firm end-use buyers, and possible technical delays. Equipment supply has been affected by the U.S. embargo; a French supplier may now be used. Japanese consortium, Sakhalin Oil Development Co., has agreed to import volumes. A political decision will be required to go ahead. Notes/Sources For Tables 8 and 9 Notes: Other projects which have been discussed, such as: Pac Indonesia to the U.S. West Coast, Nigerian volumes to the U.S., Alaskan volumes to Japan, are not included in the list here or in most other recent lists. Pac Indonesia partners have announced an indefinite delay, Nigeria's total export volume has been scaled back, and the Alaskan project is in nascent stages, not yet firm. Sources: Hydrocarbon Processing, Pipeline & Gas Journal, Japan ve Petroleum and Energy Weekly, World Gas Report, other trade publications and the OECD's Natural Gas: Prospects to 2000. KENTCO ELECTRIC POWER GENERATION FOR THE ALASKA RAILBELT REGION BIBLIOGRAPHY BIBLIOGRAPHY Acres-American, Inc., Susitna Hydroelectric Project Development Selection Report. Prepared for the Alaska Power Authority, ToaT Alaska Department of Revenue, Petroleum Revenue Division, Petroleum Production Revenue Forecast, : Alaska Office of Mineral Development, Alaska Division of Geological & Geophysical Studies, Alaska's Mineral Industry, 1982 Alaska Offshore Publications, Monthly News Publication Alaska Oil & Gas Association, Oil and Gas Leasing Hand- book, 1983 Alaska Oil & Gas Conservation Commission, 1981 Statistical Report ee Alaska Power Authority Monthly Issues, The Susitna Hydro Studies, April 1982, January 1982 Alaska Power Authority, 1982 Year End Report Alaska Power Authority, Susitna Hydroelectric Project, Volume 1, Exhibits A,B,C,D, 1553 . Alaska Power Authority, Susitna Hydroelectric Project, Volume 2, 1983 Alaska Power Authority, The Susitna Hydro Studies, 1981 Alaska Power Authority, Susitna Hydroelectric Project Summary Report Draft, Alaska Power Authority, Susitna Hydroelectric Project, Task IL: Financing Options, 1982 Alaska Power Authority, Energy Program for Alaska House Bill 9, Cost of Power Analysis, 1983 Alaska Power Authority, Board Packet, Copy #1, 1983 Alaska Power Authority, Board Packet, Copy #2, 1983 Alaska Public Utilities Commission Act, Chapter 5, Public Utilities and Carriers Alaska State Legislature, Markets for Alaskan Coal, 1981 Alaska State Legislature, Committee Report on Susitna Project, State Bills 68,69,70,71, 1983 Arthur Young & Co., A Concept for Power Production Assistance to Electric Utilities, 1980 Averitt, P., Coal in the United States Mineral Resources. U.S. Survey Professional Paper 820, U.S. Government Printing Office, Washington, D.C., 1973 Bates, John C., Doty, R.W., The Impact of 1982 Legislation on Tax-Exempt Financing Battelle Pacific Northwest Laboratories, Volume I: Rail- belt Electric Power Alternatives Study: Evaluation of Railbelt Electric Energy Plans, 1582 Battelle Pacific Northwest Laboratories, Volume II: Selection of Electric Ener Technologies for Consideration in Railbelt figsreic Energy Plans, 1982 Battelle Pacific Northwest Laboratories, Volume III: Battelle Pacific Northwest Laboratories, Volume IV: Candidate Electric Ener Technologies for Future Application in the Ratlpelt Region of Alaska, 1982 Battelle Pacific Northwest Laboratories, Volume V: Preliminary Railbelt Electric Energy Plans, 1982 Battelle Pacific Northwest Laboratories, Volume VI: Existing Generation Facilities and Planned Additions for the Railbelt Region of Alaska, [982 Battelle Pacific Northwest Laboratories, Volume VII: Fossil Fuel Availability and Price Forecasts for the Railbelt Region of rasta. 1982 Battelle Pacific Northwest Laboratories, Volume VIII: Railbelt Electricity Demand (RED) Model Specifications, Battelle Pacific Northwest Laboratories, Volume Ix: Alaska Economic Projections for Estimating Electricit Requirements for the Railbelt, 1982 Battelle Pacific Northwest Laboratories, Volume X: Community Meeting Public Input for the Railbelt Electric Power eteecaie twee L982 Battelle Pacific Northwest Laboratories, Volume XI: Comment Draft Working Paper No. 1.1, Cook Inlet Natural Gas, L981 Battelle Pacific Northwest Laboratories, Volume XII: Coal-Fired Steam Electric Power Plant Alternatives for the Railbelt Region of Alaska, 1982 Battelle Pacific Northwest Laboratories, Volume XIII: Natural Gas-Fired Combined Cycle Power Plant Alter- Natives for the Railbelt Region of Alaska, 1982 Battelle Pacific Northwest Laboratories, Volume XIV: Chakachamna Hydroelectric Alternative for the Railbelt Region of Alaska, 1982 Battelle Pacific Northwest Laboratories, Volume XV: Browne Hydroelectric Alternative for the Region of Alaska, 1982 ? Battelle Pacific Northwest Laboratories, Volume XVI: Wind Energy Alternative for the Railbelt Region of Alaska, 1982 Battelle Pacific Northwest Laboratories, Volume XVII: Coal-Gasification Combined Cycle Power Plant Alter- Natives for the Railbelt Region of Alaska, L982 Battelle Pacific Northwest Laboratories, Air Quality e Impacts of a Coal-Fired Power Plant in the z Cook Inlet of Alaska, 1979 Beluga Coal Company and Diamond Alaska Coal Company, Overview of Beluga Area Coal Developments, 1982 Burns & McDonnell, Study of Alternative Power Suppl Organizing Structure, 1983 Commonwealth Associates, Inc., Anchorage-Fairbanks Transmission Intertie-Transmission System Data. Prepared for the Alasxa Power Authority, 1980 Community Research Center, The Energy Report, Volume IV No. Final iissue Congressional Quarterly, Weekly Reports, The Politics of Water (1551), Volume 41, No. 30, I983 Demonstrated Reserve Base of Coal in the United States of January 1, 1980, U.S. Department of Energy, ashington, D.C. Department of Education, Division of State Libraries and Museum, Alaska Blue Book, 1983 Diamond Shamrock Corporation, The Cook Inlet Story Ebasco Services Incorporated, Coal-Fired Steam Electric Power Plant Alternatives For the Railbelt Region of Alaska, Volume X11, 1982 Ebasco Services Incorporated, Comparative Costs of Alter- nate Coal Source Generating ants, Goldsmith, P.I., Susitna Hydroelectric Project, Volume 2B 1983 Government Finance Research Center, Review Draft, A Review of Debt Capacity and Debt Management, 1983 House Research Agency Report No. 80-4, Coal Policy Paper, Leasing and Taxation Institute of Social and Economic Research, University of Alaska. Electric Power Consumption for the Railbelt: A Projection o egquirements, Technical Appendices. Prepared jointly for State of Alaska House Power Alternatives Study Committee and Alaska Power Authority, 1980 McGinty, Thomas E., Project Organization and Finance, 1981 Mitsubishi Research Institute, Tokyo, Japan. © Future Energy Demand and Supply in East Asia, 1982 North American Electric Reliability Council, Electric Power Supply and Demand 1983-1992, 1983 Placer Amex, Beluga Status Report, 1979 Public Utilities Commission, Thirteenth Annual Report to the Legislature for Year Ending December 31, 1982 Report to Department of Commerce and Economic Development, A Delphi Forecast of Alaska's Development/The Year 2000 E Beyond, 1983 Resource Development Council for Alaska, Inc., Resource Review (monthly publication) Resource Development Council for Alaska, Inc., Polic Statement No. 6: Coal Development ,1983 (draft) Sanders, Robert B., Coal Resources of Alaska, 1980 De ee eens Secrest, T. and Swift, W., Railbelt Electric Power Alter- natives seu Fossil Fue vailability an rice Forecasts, Sorg Printing Co., The Use of Equity Features in Tax- Exempt Financing Stanford Research Institute, Natural Gas Demand and Supply to the Year 2000 in the Cook Inlet Basin- South Central Alaska, 1975 State of Alaska, Department of Natural Resources, Division of Minerals and Energy Management, Historical and Projected Oil and Gas Consumption, 1983 State of Alaska,. Department of Commerce & Economics, A Discussion of Considerations Pertaining to Rural Energy Policy Options, 1979 State of Alaska, Department of Natural Resources, Division of Geological and Geophysical Surveys, Alaska Coal Resources, 1979 State of Alaska, Department of Commerce & Economic Develop- ment, Division of Energy and Power Development, Appendix I-1983 Long Term Energy Plan State of Alaska, Department of Commercial and Economic Development, Division of Energy and Power Develop- ment, Appendix II, 1983 Long Term Ener Plan’ (working draft) State of Alaska, Department of Commercial & Economic Development, Alaska Power Authority, Cost of Power Analysis-Ener Program for Alaska, House Bill 9, 1983 (draft) State of Alaska Public Utilities Commission, Alaska Village Electric Coop. Cost of Service Study - Rate Design, 1977 U.S. Office of Technology Assessment, Coal Exports and Port Development - A Technical Memorandum, 1981 U.S. Department of Agriculture, Rural Electric Borrowers, REA Bulletin Number 1-1, 1982 Statistical Report Weir, Paul & Company, Mining Cost Estimates Beluga Area Hypothetical Mine, 138% Woodward-Clyde Consultants, Final Report on Seismic Studies for Susitna Hydroelectric Project, 1982. Prepared for Acres American, inc. Woodward-Clyde Consultants, Volume IX: Alaska Economic Projections for Estimating Electricity Requirements for the Railbelt, 1980 William Kent and Company, Inc. One East Putnam Avenue Greenwich, Connecticut 06830 Telephone: (203) 661-3420 Telex: 643685 (KENTCO GRCH)