Loading...
HomeMy WebLinkAboutAPA394POLICY ANALYSIS PAPER NO . 82 -13 Alaska Energy Planning Studies A review of three consultant studies submitted to Alaska state agencies in fiscal-year 1982 No v e mb e r 1 8 , 1 9 R 2 STATE OF ALASKA OFFICE OF THE GOVERNOR Division of Policy Development and Planning POUCH AD JUNEAU, ALASKA 99811 (907) 465-3577 POLICY ANALYSIS PAPER NO. 82-13 Alaska Energy Planning Studies A review of three consultant studies submitted to Alaska state a9encies in fiscal-year 1982 N·o v e mb e r 18 , 19 B 2 ALASKA ·ENERGY PLANNING STUDIES A review of three consultant studies sUbmitted to Alaska state agencies in fiscal-year 1982 for Division of Policy Development and Planning Office of the Governor State of Alaska by Arion R. Tussing and Gregg K. Erickson INSTITUTE OF SOCIAL AND ECONOMIC RESEARCH UNIVERSITY OF ALASKA 4 November 1982 {supersedes all previously-dated drafts) CONTENTS Introduction and Summary Introduction Summary BackgroWld to the Studies The Three Studies Acres' study of the Susitna hydro- electric project The Battelle "Alternatives" study "The Long Term Energy Plan" The Conceptual Framework for Comparing · Railbelt Generation Alternatives Future oil prices World oil prices and Alaska state revenues Consequences for Alaska . Load Forecasts The "moderate-growth" case The "low-growth" case Lessons for the future Long Term Energy Plan Alaska Fossil-Fuel Availability and Costs The relevance of "opportunity cost" to Railbelt natural-gas and coal' sup- ply Coal prices and oil prices Historical evidence The real world of coal-purchase con- tracts Gas-fired vs Susitna generating costs The Cook Inlet gas-price/worlc!_ oil- price nexus The world oil-price assumption once again Susitna Construction Costs Character of the project Alaska Energy Planning 10/26/82 Page i 1 1 2 3 .5 5 6 6 9 13 13 15 18 18 19 21 22 23 25 28 29 32 34 35 37 38 39 General economic conditions Institutional considerations: The· Road to WPPSS Non-Completion Risk Financing Issues Real discount' and interest rates NOTES APPENDIX A R Tussing, Reflections on the End of the OPEC Era Oil Pricesand Alaska's Economy The 1981 Turning Point Market Control by the Texas Railroad Commission (1935-1972) Panic Pridng in 1973-71+ and 1979 The Role of Spot Prices The Power of Saudi Arabia and OPEC The OPEC Mystique "Oil in the Ground is Better than Money in the Bank" The End of the OPEC Era The Flight from Oil What Have We Learned NOTES Alaska Energy Planning 10/26/82 Page ii 42 43 46 lf8 48 53 67 67 69 70 72 71+ 71+ 75 76 77 78 . 82 88 ALASKA ENERGY PLANNING STUDIES Introduction and Summary Introduction. During the first half of 1982, Alaska state agencies received three major energy-policy reports they commissioned in 1980 and 1981. The first of these is the second annual "Long Term Energy Plan" mandated by the Legislature.1 Two others specificaily address issues r aised by tlhe proposed Susitna hydropower project (hereinafter I 23 the "Ac r es" and ''·Battelle" reports). ' Unfortunately, these newly-delivered reports are al- ready largely obsolete. Their critical assumptions regarding price trends for various fossil fuels, the growth of population and economic activity in Alaska, and the resulting growth of energy demand in the state, are based upon a conventiona l wisdom about future energy. prices that subsequent experi- ence has made nearly untenable. 'The Division of Policy Developmen~ and Planning (DPDP) of the Governor 's Office engaged the University of Alaska Institute of Social and Economic Research .<ISER)4 to review the three study reports and to. identify and discuss those areas that are central to the reports' conclusions, particularly with regard to investment in new e l ectrical- generation facilities in Alaska's "Railbelt" (roughly the corridor from Fairbanks through Anchorage to the Kenai Peninsula). Readers should be aware that this paper is only a review and not intended as a successor to or substitute for any of the existing studies. The following pages are intended to cover a few crucial issues in sufficient depth to determine whether or not the reports make a solid case for their findings . In large, the answer is "no", but any new recom- mendations about an optimum energy-development strategy will have to await a new study or amendment of one of the existing studies . Alaska Energy Planning 11/03/82 Page 1 Summary. Briefly, the findings of this review are that: ! 1. World Oil Prices and Alaska Energy Demand. The dramatic change in oil-price expectations since 1980 ·calls for reconsideration of the levels .of Alaska economic activi- ty and energy demand assumed in the AcreS and Battelle studies, and to a lesser extent in the Long Term Energy Plan. ~. Alaska. Coal and Natural-.Gas Supply. The assump- tiOns in the Battelle and Acres studies concerning the prices and av~ability of Alaska coal and natural gas for local electric power generation are not well supported. 3. Capital-Market Conditions. Recent high interest rates and capital-market conditions not dealt with by the contractors cast serious doubt on the Acres and Battelle COnclusions regarding the risks, costs, and financing ar- rangements of the Susitna hydroelectric project, and with respect to capital-intensive energy-supply projects general- ly~ 4. Implications for Susitna. Findings 1-3 imply signifi- cantly less favorable conclusions from those of Acres ·and Battelle regarding the r~ative economic attractiveness of the Susitna hydroelectric projects for serving electricity demand in the Railbelt region. 5. All of these findings point toward a conclusion that now is not the time for major initiatives in publicly financed power development in Alaska. Despite the erosion of some of their fundamental assumptions, the analytical framework and much of the data presented in the reports remain useful ---even essential ---to evaluating Alaska's choices with respect to the Susitna project in particular, and energy issues in general. Alaska Energy Planning 10/26/82 Page 2 BackgrOla'ld to the Studies Energy looms bigger ln Alaska's public-policy deliberations than in any other state. Elsewhere ---even in states with a history of economic activism, like Wisconsin or California ---no one would consider as even plausible a scheme to invest public funds equivalent to two o r three times expected annual state tax revenues (or about $15 thousand per capita) in all energy ventures as a class, let alone to qevote such funds to a single electric-power generation project like Susitna. Nor would the legislature of any other state countenance anything remotely similar to the energy-cost subsidy programs that Alaska now has on its books ---programs which in Fiscal-Year (FY-) 1983 can be expected to account for more than one-sixth of the state budget. 5 The sources of this unique perspective on the state's rol e in energy policy are not the focus of t his review, but they surely include the fact that oil production has ---almost painlessly ---put unprecedented fiscal resources at the comm and of state policy-makers. Regardless of its origins, the deep involvement of Alaska s tate government in energy decisions that would be left to the private sector in any other state has evoked a demand for information and analysis on an awesome range of .engineering, economic, and financial topics. Because the responsible state agencies (including the legislature and the governor's office) do not have the experience or staff to assem ble this information, evaluate it critically, or assimilate it effectively, they have had to depend on outside consultants to generate and process the relevant data, propose policies and programs, and monitor them. The legislators and executive-branch officials who promoted and authorized these three studies viewed them as complementary to one another ---overlapping in places so that decision-makers co uld view Alaska Energy Planning 10/26/82 Page 3 . ·' ,t•: ..• certain crucial issues from more than one perspective but, on the whole, dealing with different aspects or segments of an interrelated ·whole. These officials ex~ected that, together, the various reports would put the decisions they had to make in some kind of rational order, and ·resolve some of the uncertainties they faced in making these decisions. One hope, for example, was that rigorous engineering and economic analysis by nationally ren·owned experts would give them objective and politically acceptable answers, for each of the state's regions, about --- 1. The amount of electrical-generating capacity Ala- ska would need over the next two decades; 2. Which generation technologies and/or specific gene- rating projects would be most cost-effective; and 3. What was the optimum strategy for financing the chosen investments? Most of the information sought from these three studies is clearly relevant to the issues the state intended to adaress. And, although the quality of the three reports varies widely, as a group they present the bulk of the requested information ---in one place or another ---in a professionally competent fashion. Nevertheless, these studies; together with the march of events since they were commiSsioned, have conspired to leave the responsible state officials fac.ing even more uncer- tain and contradictory signals than when the various studies were commissioned. Alaska Energy Planning 11/03/82 Page 4 The Three Studies Acres' study of the Sus.itna hydroelectric project. The Feasibility Study of the Susitna Hydroelectric Project prepared by Acres American Inc. was conceived as a detailed examination of the technical and economic feasibility of the the proposed project. In addition, it was to provide searching analyses of the project's environmental and social im- pacts. The studies leading up to the report were carried out over a two-year period at a cost to the state of nearly $40 million. The report itself is organized in three hierarchical tiers, a Summary Report of 56 pages, a .main report titled Draft Susitna Hydroelectric Project Feasibi- lity Report, consisting of three weighty volumes and four equally weighty appendices, and a multitude of "task reports" which, unlike the others, have not been widely circulated. Our review has focused on the Summary Report, volume 1 of the Feasibility Report, and the the Task 11 Reference Report: Economic, Marketing and Financial Evaluation. The centerpiece of the Acres study is a "multivariate risk analysis", which uses the probabilities the investigators have attached to different assumptions about the key variables (fuel prices, construc- tion costs, interest rates, etc.) to produce an array of economic judgments (about whether the Susitna projects are the least-cost approach to serving Railbelt electrical demand, for example) ranked by their respective probabilities. Of the three works reviewed here, the Acres study deserves the greatest praise. Not only is it physically the largest, but it is also --- particularly in the Summary Report ---the most carefully and readably written. In most areas of interest a reader has the option of delving deeply or superficially, and in either case will usually find a dear and appropriately detailed explanation of the assumptions used, the evi- dence supporting those assumptions, and the methodology by which they were ~ncorporated into the analysis. Alaska Energy Planning 10/26/82 Page 5 ·~: .: The· fact that the Acres report is analytically the most interesting of the three studies ---and will clearly be the most influential ""--has caused us to devote more attention to it than to the others ---and to emphasize_ its faillngs. Readers should not be misled by this concentra- ti'on~· The methodology by which the Acres team evaluated the project's economic feasibility is elegant, and largely sound. While the report's errors come at sufficiently critical points to invalldate Acres' "bottom line", namely the economic ranking of the various ·electrical generation alternatives for the Railbelt, most of these errors are correctible, and Acres' general approach will survive them. The Battelle "Alternatives" study. Both authors . of the present ·review were professionally involved in the process that led to the chol'ce of Battelle to conduct a Railbelt generation-alternatives study. This involvement gives them a special insight into what was expected of the study, but it inescapably colors their assessment of the work that resulted. Readers should be aware of this fact, and draw t _heir own conclusions taking it into account. The Battelle study has generated several documents, but we have reviewed only two of them here: .Railbelt Electric Power Alternatives Study: Evaluation of Railbelt Electric Energy Plans (February 1982), and Railbelt Electric Power Alternatives Study: Fossil Fuel Availabili- ty and Price Forecasts. Although the former volume is labeled "com- ment draft", we understand that it is in substance the final report. Because Acres and the other contractors were directed to use scenarios and vital assumptions from Battelle, we have dealt with the Acres and Battelle analyses of individual issues, like load forecasting and coal prices, in one place. "The Long Term Energy Plan". In 1978, Alaska adopted legislation requiring the state Department of Commerce and Economic Develop- Alaska Energy Planning 10/26/82 Page 6 ment, in conjunction with the Alaska Power Authority, to prepare an annual"long-term energy plan". The law (AS 44.83.224) mandates that the plan shall contain: (1). an "end-use" study of Alaska energy consump- tion, (2) a plan for meeting "projected energy needs", (3) a review of conservation efforts, (4) an emergency energy supply plan, and (5) a review of ongoing energy research. The Division of Energy and Power Development (DEPD) has been responsible for the preparation· of both the 1981 and 1982 plans, but in both cases has made extensive use of contractors. The 1982 report was largely written under a $390,000 contract with the national accounting and consulting firm of Booz, Allen & Hamilton. However, the firm's name does not appear on the cover or in the introduction, and we do not know how much of the report's content and format should be attributed to Booz-AJlen, and how much to the DEPD staff. The 1982 report was designed, in its own terms, "to focus existing energy information to support current decision-making needs and to provide a sense of priority across state projects and programs."6 The report is well written, contains few serious errors of fact or obviously faulty analysis, and provides the mandated "existing energy infor- mation" in a convenient format. The "plan" does not fare well in its attempt "to provide a sense of priorities," however. With respect to the really tough social and political issues raised by Susitna and the state's hydropower program generally it largely leaves the field to Acres and Battelle; the Plan's treatment of the Rail belt hydroelectric construction proposals is confi- ned to less than two pages of text. After urging the state to continue planning for Susitna, the report warns that the project's "impact may be to severely limit the consideration of less costly alternative Alaskan based resources such as coal or residual oil." 7 Alaska Energy Planning 10/26/82 Page 7 In other areas, inquding the treatment of Alaska's complex sy$~.~111 of energy subsidies, the authors develop an extensive and unique da~a·:~~llection, but seem reluctant to draw the prescriptive conclusions that clearl:y foll.ow from it. Finally, many of the study's featured findings are pedestrian, for ex~mple the conclusion that "opportunities exist to increase the completeness and accuracy of Alaska's energy data;" 8 Some of these deficiencies are pr?bably the result of the short time in which the state's contractor was required to produce a draft report. In one of their concluding sections the ~uthors seem to recognize these shortcomings, proposing that next year's energy plan focus on developing a "strategic energy planning process." The discus- sion of how that might be accomplished is one of the most interesting in the entire report.9 Alaska Energy Planning 10/26/82 ' Page 8 The Conceptual Framework for Considering Electrical-Generation Alternatives for the Railbelt The Acres report sets out most clearly the conceptual framework shared by all three studies. In the Railbelt, the ke'y issue is to identify the combination of electrical generating facilities that is most likely 'to b¢ the cheapest in the long run. The main choices are: ( l) a two-stage strategy involving Susitna River hydroelectric power; (2) continued reliance mainly on gas-fired combustion turbines (either "simple-cycle" or "combined-eye!~" plants), or on some combination of gas turbines and coal-fired steam plants; (3) and a combination of smaller hydropower facilities with thermal. generation. The crux of the economic comparison between Susitna and thermal generation is the com~ison, over time, of hydro construction costs ana thermal-p ant fuel costs. The chief hydroelectric option, which centers on the Susitna River projects, involves a relatively high front-end capital expense per unit of capacity but very low continuing costs for maintenance and operation. Combustion turbines, on the other hand, are relatively cheap to install per unit of generating capacity, and the cost of the electricity they produce is principally the cost of the natural gas used as fuel. emil- fired steam turbines would be less capital intensive than hydro, and while they would cost considerably more to build per unit of capacity than gas turbines, they might still provide the cheapest base-load power if the price of coal (per unit of electricity generated) were sufficiently ~low that of natural gas. Out of the many issues that are relevant to this choice, the present review focuses on the way the various reports deal with --- a. Future world oil prices; b. Future Railbelt electricity demand; c. Future Railbelt fossil-fuel prices; Alaska Energy Planning 11/03/82 Page 9 d. Construction costs for hydro projects; e. The appropriate interest or discount r~te; and f. Risk and uncertainty regarding these and other issues. The six iss~es fit together as follows: (a) World oil prices will powerfully influence Alaska economic activity, ahd through it electricity demand, by determining state revenues from petroleum royalties and taxes, and thus state spending. Oil prices are also a major influence on Alaska economic development and thereby on electricity demand, byr way of their impact on energy- related private investment ---in oil and gas exploration, coa:I export projects, the Alaska Highway gas pipeline (ANGTS), petrochemicals manufacturing, and the like. World oil prices, moreover, may influence the price_s of natural gas and coal for electrical generation in the Railbelt. Indeed, the Acres and Battelle analyses seem .to treat world oil prices as the crucial force determining fossil-fuel prices in the region. (b) Electricity load growth. Susitna generating capacity would be very "lumpy" as well as capital-intensive; additions would come in multi-billion-dollar packages or not at all. Gas-turbine capacity can be added ih small increments, however, with coal-fired steam turbines and some smalier hydroelectric options falling between the Susitna projects and gas turbines in "lumpiness". If projected power demand and demand-growth rates are high, they can be expected . to liquidate any excess generating capacity rapidly; high load-growth forecasts therefore improve the prospective economic ranking of the Susitna projects, all other things being equal. With low or uncertain load growth, however, the larger hydro projects pose a greater risk than do thermal plants that underutilization Alaska Energy PlanrHng 10/26/82 Page 10 of c apacity would result in high unit costs for elec tricity. Thus, the risk of temporary or permanent overbuilding would be least in a str ategy built around gas-fired combustion turbines, with the risks somewhere in-between for smaller hydroelectric projects and for coal- fired steam generation. (c) Future fuel costs and (d) e~ted construction costs. In the framework described here, the comparison of electricity costs must focus most sharply on the cost of fuel for gas-fired com~tion turbines, and on the original construction cost for the proposed hyd.ro- eJectric plants. The relative cost of electricity from coal-fired steam plants will depend more on capital costs than electricity from gas tur- bines, but more on continuing operating (fuel) expenses than hydro- electric power. Estimates of construction costs and future fuel costs are both subject to great mcertainty ---and the treatment of this tmeertainty (f) below is itself a major issue in any comparison. (e) Discomt rates. Because the Susitna plants would be capital- intensive, long-lived, and take many years to build, the long-term cost of electricity from these projects would be highly sensitive to interest rates. This would be true whether the interest rates in question were the rates the state would have to pay to borrow for Susitna construc- tion, or the rates it could have otherwise earned on money appropriated to build Susitna. The net benefit from the Susitna option is, therefore, most sensitive to the choice of interest rates used to "discount" future costs and benefits. (f) Treatment of risk and uncertaint:y. The various factors that influence future Alaska economic activity and thus electricity demand (including but not limited to future world oil prices); Susitna and other generating-plant construction costs; future Alaska fossil-fuel prices Alaska Energy Planning 11/03/82 Page 11 {which may or may not be closely linked to world oil prices); and future in~erest rates are all unknown today. Important assumptions that the ",' ... ,,, .. analysts plug into their models are thus essentially guesses. These guesses may be informed or ignorant, and insightful or obtuse, but their impact on the final comparison will reflect both the raw values assumed by the analysts, and the way in which the analysts deal with· the risk and uncertainty that surround them. Subsequent sections of this review reveal considerable disagreement with some of the raw values Battelle and Acres have assumed in the studies, and the probabilities they have assigned to these values, but not with Acres' desig11 and execution of the "multivariate risk analysis" used to inte- grate these assumptions. Alaska Energy Planning 10/26/82 Page· 12 Future Oil Prices World oil prices and Alaska state revenues. From the standpoint of Alaska policy-makers, no aspect of the current scene is more contusing than the recent radical change in the the state's official oil- price forecasts, and the forecasts of state revenues that depend directly on oil prices. This change in the oil-price outlook invalidates virtu- ally every important economic and policy conclusfon in the studies reviewed here. · Between June 1980 and January 1982, the Alaska Department of Revenue's quarterly Petroleum Production Revenue Forecast predicted that nominal-dollar ("inflated") oil revenues would increase over the four years beginning on the forecast date at a compound annual rate between 12.2 and 25.8 percent. In its March 1982 Forecast, the Depart- ment's three-year estimate of the expected annual change in state revenues fell abruptly to a negative 0.8 percent.10 Specifically, the "most likely'' projection in the March Forecast was that the weighted average wellhead value for Alaska North Slope {ANS) crude oil in fiscal year (FY) 1983 would be 29 percent lower than in FY-1982, and that world prices would then resume their nominal- dollar increas~, at a compound annual rate of about 7 percent. Not until the beginning of FY-1987 did the Department expect prices to regain FY-1982 levels. With respect to constant -dollar ("real") oil prices, the March Forecast boldly reported "a consensus that oil prices will continue to fal1,"11 and projected declining real oil prices through 1998.12 The authors of the present review agree that world oil markets cannot sustain the level of prices reached in early 1981, and that prices in any given year during the remainder of the century are likely to be considerably lower in real terms than they are today. Alaska Energy Planning 10/26/82 Page 13 There was, however, no consensus on the long-term oil-price outlook that one could prudently rely upon last March, and none exists today. What has happened, instead, is that the near-consens~s which did exist aJ the ·beginning of 1981 has been shatt.~red, namely the assump- tion that ·.th~ long-term trend in oil prices was inexorably upward. Abrupt changes in expectations have not been a pro!J.lem unique to Alaska's official revenue forecasters; a review of the energy literature generall~ confirms that a widespread reevaluation began in late 1981 and early. 1982. Few authorities any longer confidently assume that the energy-price increas~s of 1973-1981 will continue unabated through the . rest of the Century, and an increasing number are suggesting, as the present reviewers have done since 1980,13 that the price rises of the , 1970's may never resume. The crude-oil price slump of late 1981 and early 1982, which few industry and government forecast~rs anticipated, drew attention to the difficulties of predicting energy pric~s, but is also push.ing forecasters into a rnore general reexamination of both the demand and supply of petroleum, and the way in which they determine oil prices in the long run. As late as Septerpber 1980, it was possible for Cambridge economist Nicholas Kaldor to write serlouslx that," ••. OPEC changed everything. By cornering oii it manag~c;i to increase the price four-fold, then double it again, and presumably it could be doubled again, without any really serious impact on consumption." (emphasis added)14 It is now clear th.at the world economy has a much greater capacity and willingness both to substitute other fuels for ~igh:-priced p~troleum and to economize on energy generally than had been widely supposed. Over the past six months, virtually every published authority in the area of petroleum demand has radically altered its expectations regarding future u.s. and world petroleum d~mand.15 Alaska Energy Planning 11/03/82 Page 14 On the supply side, so much excess oil-producing capacity now exists that it is hard to contrive any scenario in which OPEC, Saudi Arabia, or anyone else, can long function as a "price-maker" in world oil markets. To the extent that there is any consensus about world oil markets among the experts today, the managing director of Royal Dutch-Shell summarized it well when he wrote that "we are in for a period of severe and unpredictable discontinuities.1116 ,l 7 An advance draft of Tussing's "Reflections on the End of the OPEC Era", .included as an appendix to this review, takes a backward look at the events that led most foreca- sters in the late 1970s· to expect ever-increasing world oil prices, and the reasons such an outlook seems Wltenable today. Consequences for Alaska. These changes in outlook have extra- ordinary significance for Alaska, because its economy, like that of (say) Kuwait, is a net exporter (seller) of energy. Well over half of Alaska economic activity depends directly or indirectly on crude-oil produc- tion. The largest such influence operates directly through state oil royalties and prodtiction taxes, and if prices continue to faJJ, resulting reductions in state revenues will make it impossible for state spending to continue its new-found role as the main prop and guarantor of Alaska's economy. Table 1 below compares the state's June 1981 and June 1982 forecasts. At the same time the Department of Revenue was reducing its revenue expectations generally, it also decided to emphasize the uncertainty of petroleum-price forecasts, and began highlighting its "30-percent" rather than its ''expected value" series. The different percentage figure indicates the Department's judgment about the probability that actual revenues will be Jess than the figure shown. Alaska Energy Planning 11/03/82 Page 15 Table 1 1982-Dollar Petroleum-Revenue Projections bl the Alaska Department of Revenue, June 1982 vs June 1981 Oil and Gas Revenue (~ Millions) June 1981 Forecast June 1982 Forecast Fiscal ~Expected "30%" Year "Expected Value" Value" Series 1983 4030 2654 2399 1984 4137 2657 2250 1985 4271 2623 2177 1986 4448 ' '2953 2411 1987 4713 3305 2644 1988 4851 3196 2507 1989 4983 3365 2595 1990 4742 3095 2246 1991 4544 2714 1862 1992 4382 2477 1668 1993 3979 2285 1427 1994 3637 2149 1265 1995 3144 1826 1059 1996 2701 1622 936 1997 2289 1608 908 · Alaska Department of Revenue, Petroleum Production Revenue Forecast, Quarterly Report, June 1981, p.l3; June 1982, p. 18, Personal Communication, Charles Logsdon to Erickson. In 1981, state and lo~al government employment directly accoun- ted· for 21 percent of Alaska non-agricultural wage and salary employ- ment. 18 State government expenditures, in turn, were 86 percent financed in FY -1982 by oil production revenue,l9 and Alaska local governments received about two-thirds of their revenues from the state government. State aid to the ~ity and Borough of Juneau, for example, will equal 262 percent of local property tax r~venues in FY- 1983.20 Much of the revenue received by several other local govern- ments in Alaska, moreover, comes f.rom direct taxes on oil industry activity and property. Alaska Energy Planning 11/03/82 . . . Page 16 These illustrations do not begin to encompass the indirect ef- fect of expectations regarding future oil prices on the state econo- my. These expectations largely determine the level of private-sector investment in oil and gas, coal, and other energy-extraction, conver- sion, and transportation ventures; energy-industry service activity; and have an added "multiplier" impact on the Alaska economy 1!! the income flowing from such investment activity into the trade, finance, and service industries. A large, if not precisely measurable, part of the present boom in the Anchorage area reflects private investment com- mitments made in 1979-1981 on the basis of a bullish outlook about future oil prices. This boom is unlikely to survive long once those expectations have been shattered or drastically modified. 21 The situation is quite different for energy-importing states like New York or California, where a radical increase or decrease in energy prices would at the most, over the short run, cause no more than a three or four percent change in the major economic indicators such as employment, gross state product, or disposable income. In these importing regions, the dominant impacts of energy price changes will operate rather diffusely, through the influence of fuel prices on the real incomes of consumers, and through the impact of changed fuel prices on production costs, and thereby on prices, sales, and profits in manufacturing, transportation, and commerce. In Alaska, the potential response is an order of magnitude larger, and is dominated by impacts on employment and population that flow directly from the primary role that petroleum production and state government spending (89 percent supported by petroleum production) play in the regional econoi:ny. Alaska Energy Planning 11/03/82 Page 17 Load Forecasts ·: :· ·the studies reviewed here pay close attention to the usual income and price-factors that affect energy demand, and carefu11y evaluate the impacts of different oil price scenarios on ~he electrical-power costs implicit in various energy development schemes .• But remarkably, they largely ignore the possibility of a major decline in oil revenues, and the direct effects such a decline would have on the Alaska economy . 22 The Battelle reports base their forecasts of energy demand on scenarios and econometric modeling studies generated by ISER in 1980 and 1981, using its Man-in-the-Arctic (MAP) model. Acres' growth scenarios are, in turn, adapted from those of Battelle, and the Battelle rep9rts provide the clearest explanation of the economic-development and state spending assumptions that went into the forecasts of Alaska Railbelt economic activity. Battelle offers five "scenarios", rariging from "low" to "superhigh", and a sixth scenario (tagged "fiscal-crisis") which shows very high spending in the 1980s, followed by a drastic decline in the 1990s. The "moderate-growth" case. Battelle's "moderate" case (which the report defines as having a 50-percent probability of being exceeded) shows population in the Railbelt growing at a compound annual rate of 2.15 percent.23 This scenario is powered by assumptions that state spending will increase from the FY-1981 level (when general-fund appropriations were about eleven thousand dollars per capita) proport~o­ nally with per-capita personal income, that the Alaska Highway gas pipeli~e (ANGTS) will be under construction by 1983,24 that the PacAlaska LNG project will come on line between 1985 and 1987, that a 100,000-barrel per day refinery will be built in Valdez, and that 7 billion barrels of oil will be discovered and developed on federal outer continental shelf (OCS) acreage leased through 1989.25 Alaska Energy Planning 11/Q3/82 Page 18 The Acres study adopts the Battelle load-growth scenarios26 with some modifications, which are not always described in sufficient detail to allow critical examination. Acres summarizes the outcome, how- ever, as follows: Between 1981 and 2010, the mid-range forecast s.uggests that electrical and energy demand will grow at an annual rate of about 3.5 percent, with the high and low range limits at ·about 4.6 percent and 2.8 percent, respectively ••• Under the mid-range forecast, currently scheduled additions are sufficient until1993 to meet rising demand as well as to replace aging unitS which must be retired. Between 1993 and 2010, about 1400 megawatts of capacity must be added to the system If meet additional demand as well as to replace aging units. The "low-growth" case. The BatteUe report states that there is only a 5 to 10-percent chance economic activity will at any point dip below the values projected in the "low-growth" scenario.28 A review of Battelle's assumptions underlying this boundary case show population increasing at a compound rate of 3.4 percent in the 1980-1985 period,29 and constant per capita state spending (base? on the exceedingly high FY -1981 base) and construction of ANG TS, but with a lower level of offshore oil activity and no Valdez refinery or Cook Inlet LNG plant. None of Battelle's moderate-case assumptions, as listed above, now appears likely to materialize. The gas pipeline has, for example, been put on indefinite hold; with the most optimistic outlook, construc- tion could not get under way before 1985 at the earliest. The Valdez refinery _project was scuttled about a year before the Battelle report was deHvered (and never did achieve much credibility among petroleum experts). 30 And Battelle itself has elsewhere virtually written off the LNG project.31 The most stunning discrepancy between Battelle's assumptions and what now seems realistic concerns the state government budget. Even Alaska Energy Planning 11/03/82 Page 19 )~. , , r-· in Battelle's low-growth scenario (given the low-case population growth projections), total state spending in FY-1985 would have to be in the neighborhood of $4.3 billion.J2 However, state appropriations for FY- 1983 were only $2.7 billion, or $6,590 per capita, a fall of 45 percent ti-'o'rri the previous year's peak and about 27 percent less than the low- case figure ·used by Battelle. The Division of Legislative Budget and Audit expects that even this spending reduction will leave the state budget with a deficit of $400 milli~n relative to projected revenues.J3 I As we noted previously, the state's official March 1982 forecast expected real prices to remain below thej_r FY-1982 levels through FY- 1998. Total petroleum-production revenues projected for FY-1985 were $2.8 billion. 34 If non-petroleum revenues add another half billion doliars, the state would have $3.3 billion to spend in FY -1985. The "low-case" population estimate therefore means that $3.3 billion is $6,9 52 per capita, or only about three-fourths of the figure implied by Battelle's "low-case" assumption that state spending will remain at FY- 1981 levels. In short, the median, "most-likely" revenue projections of the state's Division of Petroleum Revenue imply a level of state spending that is far lower than, and inconsistent and incompatible with, even the ·"low-case" boundary conditions assumed in the Battelle analysis of electricity demand and the Acres analysis of Susitna fea~ibility. None of the contractors has used the model to beget a scenario driven by falling real prices for crude oil, but it is clear that oil-price declines of the magnitude implicit in the state's most recent revenue forecasts would, if cranked into the model, result in ·demand-growth projections below those of the most pessimistic cases considered by either Acres or Battelle. We have not calculated the quantitative effect on Alaska's economy of these actual and projected changes in state spending, the Alaska Energy Planning . 11/03/82' P~ge 20 stalemate and possi ble demise of ANGTS, o r the bleak outlook for other large capital projects (including, by the way, Susitna) whose construc- tion lies explicitl y or implicitly behind the l oad forecasts. Clearl:r, however, a realistic, up-t<Hiate view of state revenues and private-sector investment plans implies a "most-likel " future in which electrici demand will be ower than the lowest case postylated by Acres or Battelle •. Lessons for the future. Predictin g future eco no m~c activity is and always has been a tricky business, and Alaska's thin, open economy makes it exceptionall y so. Recent Alaska e l ectrical l oad forecasts have consistently over estim ated the growth of demand ---even without the kind of downturn in Alaska economic activity that is likely to proceed from the current oil-price slump. In 1977, ISER forecast that state electricity consumption would g row between 1974 and 1980 at an ann ual rate of nearly 12 percent, with the lowest possible rat e 8.7 per cent. The actual growth rate was 8.2 percent.35 T he forecasts ISER furni- shed Acres in 1980 i mplied a 4.6 percent annu al growth of Railbelt e lectricity consumption in 1978-1981 in the moderate-growth case, and 4.0 percent in the low-growth case. The actual increase was 3.4 per cent.36 It appears, however, that the forecasting performance of other parties has been even less sat isfac t ory. If there is a lesson to be learned from this experience it is that grave dangers lurk in r elying on a single consultant or institution for insights about the future. The purpose of the Battelle Railbelt Alternatives study was (at least originally) to provide a check on the assumptions and analysis used by the Alaska Power Authority and Acres. One reason the governor's office sponsored and the legisl ature mandated this study was their concern that Acres' stake in Susitna construction might aifect the credibility of its load forecasts and other e lements of its economic and financial analysis. Alaska Energy Pl a nning 11/03/82 Page 21 Events show that the original purpose of an independ~nt study of Rq..i).l?eJt generation alternatives was frustrated when Acres and Battelle ... .. . - were directed to standardize their assumptions on. a number of issues, (e.g., load forecasts) and to rely only on ISER's 1980 economic-develop- ment scenar.ios as the basis for projecting future electricity demand --- probably the most important variable of all in evaluating the wisdom of the massive commitment of present and future resources to a project like Susi tna. Long Term Energy Plan .. The Long Term Energy Plan describes its assumptions concerning the variables that drive the Alaska economy as derived fr~m B_attelle's work.37 However,,. in the case of expected population and employment growth The Long Term Energy Plan respec- tively assumes compound rates of 1.1 and 0.6 percent through the yea_r 2000, figures that are clearly inconsistent with the Battelle projec- tions.38 Unaccountably, the authors of the Plan nevertheless adopt Battelle's Railbelt electricity-demand forecast.39 Alaska Energy Pla!ln.ing 11/03/82 . Page 22 Alaska Fossil-Fuel Availability and Cost The Cook Inlet basin, which contains the bulk of the Railbelt's population and economic activity, is a major natural-gas producing area and contains large deposits of steam coal close to tidewater . Indeed, the bulk of the electricity produced in the Anchorage-Kenai peninsula area today comes from gas-fired combustion turbines, a nd much of the Interior's electricity comes from a coal-fired steam plant at Healy. Thus, the most obvious alternatives to the Susitna project for new electrical generating capacity in the Railbe!t are coal-and gas-fired plants, or a combination of smaller hydro projects with coal-and /or gas- fired plants. The Acres Summary represents both Acres' and Battelle's position with respect to this choice: ... If required generation expansion occurs by continued use of the thermal generating plants, a shift toward increased use of coal will be necessary not only because the Cook Inlet gas reserves may be insufficient to sustain long-term re li- ance upon natural gas in the face of increased demand but also because sharp increases in gas prices will occur in the next decade as old supply contracts expire. The installation of thermal (coal-or gas-fired) plants to meet the demand would offer the consumer no protection against rising costs, since fuel prices will continue to be exposed to inflation and to extraordinary escalation occasioned by world market conditions. 39 Despite the central role that fossil-fuel availabillty a nd cost must play in comparing generation alternativ-es, neither Acres nor Battelle has attempted to assess the probable future cost of fossil fuels for generation in the Railbelt from local coal or gas supply condit ions .• The reports give little if any attention to the incremental cost of gas or coal production in the region, the ownership and regulatory or contrac- tual status of proved and potential reserves, or the other physical, economic, and inst itutional features of the local supply picture for Alaska Energy Planning 11/03/82 Page 23 either fuel. Indeed, the Acres reports and Battelle's Railbelt Alterna- tives reports do not even cite Battelle's own 1982 forecasts of Alaska oil ~nd gas supply and demand for the Alaska Department of Natural Resoo~ces (whose conclusions appear to conf)ict with the assun~ptions used in the rep<)rts reviewed here), or the coal-supply study conducted by Ebasco Services, Inc., under a subcontract to Battelle for the Railbelt alternatives study. Both contractors have, instead, deduced Railbelt coal and natural- gas sUpply costs from (1) recent LNG and coal import prices in Japan, (2) assumptions about future world oil-price movements, and (3) the relative processing and transport costs between the mine ·mouth or gas field and markets in Alaska and Japan. The following passage on the relationship between coal prices and the economics of the Susitna project sets out the implicit pricing model Acres shares with Battelle: • • • Coal mining is assumed to occur at Beluga and an export market for Beluga coal is assumed to exist. To recover the investment in these plants and to account for anticipated major increases in gas costs, as well as inflation and fuel price escalation above the IUfderlying inflation rate, the wholesale price per kWh for electricity will have risen by 1994 to 145 mills (+14.5¢) per kWh and will continue to rise thereafter. It is this trend again:Jt which any pro- ~ hydroelectric development on the SUsftna River ·m~t compete.4D(emphasis added) . The assumption of i•fuel price escalation above the underlying inflation rate" stems directly from the proposition Acres shares with Battelle that the price of coal to new electrical generating plants in I Alaska's Railbelt will reflect its market value in Japan (less trans- . portation costs between Alaska and Japan) and that this market value is in turn, directly determined by world oil prices. The Acres and Battelle . assumptions ~egarding natural-gas prices follow the same logic: The cost of natural gas to electric utilities in the Railbelt will be the Alaska Energy Planning 11/03/82 Page 24 price of LNG landed in Japan (which will move with world oil prices), less liquefaction costs and transportation costs from Alaska. The relevance of "opportunity cost" to Railbelt natural-gas and coal supplr. The concept at the heart of Acres' and Battelle's treatment of both coal and gas prices is opportlllity cost. The opportunity cost of Alaska-produced natural-gas or coal is the highest price their owners could have got for the fuel, even if they actually made a deal at some lower price. Acres succinctly states the principle as follows: ••• if export markets exist for LNG (or coal) at the prices which have been determined in the Report, then it must be assumed that any rational gas (coal) producer in Alaska would select the opportunity to receive the highest price which is offered for the gas (coal). If the gas (coal) producer chooses to sell gas (coal) in Alaska at a lower price instead, then should the Susitna project evaluation be based on this economically inefficient price? If this is so, the project evaluation will not be consistent with generally accepted principles of public investment appraisaL (paren- theses and "coal" added)41 The principle set out in the previous citation is generally a sound one. The cost figure the state ought to use in calculating the net benefits to Alaska of selling royalty oil to an in-state r efiner or petrochemical producer is not the actual transaction price (which state officials may decide to discount for the purpose of fostering industrial development) but, rather, the revenue the state would have got by selling its royalty oil to the highest bidder. Likewise, the cost to Alaskans of electricity from the Susitna project includes all interest the state could otherwise have earned on money appropriated to the project, even if the legislature or the Power Authority decides to set electric rates on the basis of a lower (or even zero) rate of return on that money. Alaska Energy Planning 10/26/82 Page 25 Opportunity cost in this sense measures the true cost of state -~~:·:J.• . . royalty oil sold to an in-state processor and establishes the proper 'Cffscount rate for evaluating investments of surplu"s state funds within Alaska, but Only bt!cause there are ready, arid for practical purposes .. :,~·l :t~;.. ~c. . • -•• Unlimited, eXport markets for royalty crude oil and surplus state money. Thus, h is reasonable to assume that any royalty oil which is not processed within Alaska can and will be sold at world market prices (as limited, of course, by the federal prohibition on exports of crude oil shipped throught the Trans Alaska pipeline). It is also reasonable to assume that any state money that is not s·pent or invested within Alaska can and will be invested elsewhere to get the highest available yield (consistent with the degree of security sought by the state's money managers.) It is not, however, reasonable to suppose that all AlaSka gas or coal that is burned in Alaska would otherwise hav:e been exported at the netback prices assumed by Acres or Battelle. If effective export markets in fact existed for all of Alaska's potential coal and natural gas production at. current Japanese import prices less transport costs, export markets would exist at comparable prices for air of the natural gas that is now shut in or being flared (close to tidewater at least) anywhere in the world. Trillions of cubic feet of gas would not be flared each year in the Middle East, for example, and . . LNG from Cook Inlet would already be flowing into foreign markets to the limit of the region's producing capacity. Likewise, if there were really an effective export market at recent Japanese lmp~rt prices for any coal that is economically feasible to produce in Alaska's Railbelt, there would be little uncertainty about the future development of Beitiga (or Bering River, or other Alaska) coal, while coal prices almost eve~ywhere in the world would have reachep levels comparable with the netback "opportunity cost" that the two reports posit for Alaska coal. Alas~a E·nergy Plan~ing· 10/26/82 Page 26 What the "opportunity-cost" approach of Acres and Battelle ignores is the fact that the world's known reserves of c oal and natural gas, technically capable of development and delivery to Japan at resource costs below that nation's recent import prices, are many times greater than potential Japanese or global markets for coal and LNG at those prices. Since 1974, Japanese consumers (among others) have been willing to pay prices for coal and LNG related to the rising cost of imported oil, but only because the construction of export and import terminals, on-shore transportation facilities, and the specialized bulk carriers could not keep up with the worldwide growth of demand for fuels capable of underpricing OPEC petroleum in the electric-utility and industrial boiler-fuel markets. In 1981-82, however, a combination of falling oil prices, the economic slump, and completion of the first generation of major LNG and coal-shipment projects initiated after the 1973-74 oil crisis, have (1) marked an end to the period of frantic growth in Japanese and worldwide demand for both coal and LNG, and (2) permitted new supplies to c atch up with .and perhaps outstrip that demand. Of the more than twenty new coal-export terminal :developments that were ' under a c tive consideration on the West Coast of North America in early 198!, it now appears that the growth of East Asian demand will economically support at most two such projects between now and the year 2000. There is likewise a growing conviction in industry that LNG- export projects already committed to construction will satisfy Japanese demand for the rest of the century, and that even some of these "committed" projects (the British Columbia LNG venture, for example) may yet be scuttled.42 · Coal prices and oil prices. Acres and Battelle c ombine (1) the peak 1981 spot prices for coal imported to Japan, (2) the assumption Alaska Energy Planning 10/26/82 Page 27 that world oil prices will continue to escalate in real terms; apparently witHout limit, and (3) their "opportunity-cost" netback pricing rriethodo- lo'gy to arrive at the following most-likely constant-dollar price- escalation rates for Beluga Coal (at Anchorage) and Healy coal (at Nen'ana)/13 Table 2. Alaska Coat-Price Asswnptions Acres Battelle Acres Battelle Coal Source Beluga Healy (01/82 cents per mmbtu) 1.43 1.75 1.69 2.43 (annual constant- dollar escalation rate in percent) 2.5 2.7 2.1 2.0 Two weakn~sses in the Acres and Battelle coal-market assump- tions compromise the usefulness of their comparisons between Susitna and coal-fired generation costs: Firstly, if constant-dollar oil prices are not likely to increase at the rates that Acres and Battelle assume throughout their studies, the rationale for similarly-increasing coal prices is mortally damaged. Thus, a more realistic, up-to-date oil-price outlook would call for a reevaluation of the probable cost of coal in Al~ska, and of electricity generated· from coal, just as it calls for reevaluation of most of the important assumptions and conclusions of the two studies. Secondly, even if Acres and Battelle oil-price scenarios had remained plausible, their supposition that Alaska coal-price movements Alaska Energy Plann'ing. 10/26/82 Page 28 will be driven by world oil prices in some straightforward and easily predictable way is unwarranted. The effect of this mechanistic assumption is to evade one o~ the most difficult, yet one of the most crucial, issues in choosing among generation alternatives ---the need to forecast the change in relative prices among various fossil fuels over time. Historical evidence. Acres gives a number of good reasons why production costs alone are not sufficient to predict coal prices at any given time and place. For example, • • • "economic rents" (that is, a price that exceeds produc- tion costs including a normal return on investment) may be eamed by the producers, mine labor and/or governments. -For example, in the interests of supply security, a coal importer may be willing . to pay a price much higher than actual coal production costs. In addition, oil price increases induce increased demand for coal, thus exerting upward pressure on coal prices. 45 The report then remarks that "(h)istorical trends support these observations." "Historically, it has been observed that export prices of coal are highly correlated with oil prices, and that production cost analysis has not predicted accurately the level of coal prices."46 Acres follows with evidence from a number of sources that real coal prices have increased in recent years, presumably at rates exceeding the increase in production costs, and with citations from a number of other authorities who have also predicted or assumed that coal prices will continue to increase at rates comparable with, or tied to, world oil prices. The 'historical data Acres cites do show that rapidly growing coal demand can cause real-price increases, but they do not, in fact, support the notion that coal prices have been tightly linked to oil prices in the recent past. Examine the following three illustrations, for example: Alaska Energy Planning 10/26/82 Page 19 . ·,·· .\.. Coal prices (bituminous, export unit value, FOB u.s. ports) gpew at real annual rates of_ 1.5 percent (1950 to ·-~:.1979) and ~.8 percent (1972 to 1979).'41 In fact, the constant-dollar price of imported crude oil rose at an average rate of 3.9 percent ---almost 2~ times that of the average FOB export value of coal ---between 1950 to 1979, and at an average rate of 29.2 percent ---almost ten times as rapidly as coal prices --- between 1972 and 1979.118 The 1.5-percent average coal-price increase cited by Acres for the longer period was actually less than the average increase in hourly compensation of bituminous coal miners in the U.S. over the same period (1.9 percent). 49 In Alaska, the price of therm~l coal sold to the GVEA utility advanced at real rates of 2.2 percent (1950 to 1978) and 2.3 percent (1'970 to 1978).~0 The contrast between Alaska coal and oil-price trends depends wholly on the years chosen. The constant-dollar price of No. 2 fuel oil in Fairbanks actually decreased at an average annual rate of 0.4 percent between 1970 and 1978. Between 1973 and 1981, however, GVEA's real coal price increased at an average rate of 2.1 percent, while No. 2 fuel oil prices were increasing at an average rate of 15.6 51 percent. In Japan, the average CIF prices of steam coal experi- enced real escalation rates of 6~3 percent per year in the period 1977 to 1981.52 The constant-dollar price of Japanese crude-oil imports, however, increased at an annual rate of 17.6 percent over the same period --- about three times as rapidly as the country's average coal-import . 53 pnce. Table 3 compares the cost of coal and oil at electric generating plants in the United States over the last decade, and is further evidence against the direct relationship between coal and oil prices Alaska Energy Planning 10/26/82 Page 30 Table 3. Cost of Fossil Fuels Delivered To Electrical-Genera!!rnt Plants in the Unitep States 1973-1981 Current Prices GNP 1981 Prices (cents per mmbtu) Oefla-kents per mmbtu~ Resi-tor Resid-Oil-Coal dual 0972 dual Differ- Year Coal Oil = 100) Coal Oil entiat 1973 40.5 78.8 105.8 74.1 144.2 70.1 1974 71.0 191.0 116.0 118.5 318.7 200.2 1975 81.4 201.4 127.2 123.9 306.6 182.7 1976 84.8 195.9 122.7 133.8 309 .o 175.3 1977 94.7 220.4 141.7 129.4 301.1 171.7 1978 111.6 212.3 152.1 142.1 270.3 128.2 1979 122.4 299.7 165.5 143.2 350.6 207.4 1980 135.2 427.9 177.4 147.6 467.0 319.5 1981 153.3 529.0 193.6 153.3 529.0 375.7 Change1 1973-1981 Absolute 79.2 384.8 305.6 Multiple 2.07 3.67 5.36 Armual rate {%) 9.51 17.64 23.35 Source: Monthly Energy Review asserted by Acres and Battelle. The real price that utilities paid for coal did indeed more than double between 1973 and 1981, but at the end of the period, the .fuel-cost advantage of coal-over oil-fired generation (in constant dollars per million btu) was 5.4 times as high as it was at the beginning. Whate ver limited validity there might be in the model behind the A c res and Battelle c oal-price assumptions, it would exist only for spot markets and only for short-term fluctuations. The oil-price surges of the 19 70's, without doubt, stimulated increases in the demand for coal that outran the ability of the industry to open new mines. As a result, excess demand did pull up world coal pric es dramatically --- particularly the s pot prices of steam coal. But Acres ignores the fact Alaska Energy Planning 11/03/82 Page 31 .. ·-~ .. ·; t that slack demand and · overinvestment in coal-producing capacity can jus~, as;. easily cause reductions as surges of demand can cause increases in spot-market coal prices relative to production costs. Moreover th~ Acres report offers the reader no reason to believe ·that ·the excess-demand conditions that existed during the 1970's will continue unbroken for the next tweAty to forty yeq(S. The historic~! record, which stretches back more than two centuries, shows both cyclical and random fluctuations in coal prices, but no evidence of rising rea.l costs over the long term. It is likely the peak of the most recent cycle was reached in 1981; at any rate, nothing in the world supply-demand picture suggests further real coal-price rises in the foreseeable. futtJr~. In the long-nm and on the averya.ge, coal prices must reflect the real resource cost of operung and operating new mines. The real world of coal-purchase contracts. No m<ijor coal-fired electrical generating plant will be planned or built by Alaska utilities or the Alaska Power Authority (or financed with revenue bonds) unless it has secured "dedicated" coal reserves and producing capacity sufficient to supply it with fuel over its economic life, or at least over the period of its long-term financing. A prospective mine operator must, in turn, have a long-term coal-purchase contract from the utilities or the Power Authority in order to get financing for mine development and produc- tion equipment, and working capital. That long-term coal-purchase contract will, in every li{<elihood, be a cost-of-service contract. It will probably provide automatic price ~djustn:'ents for certain production-cost items, and may allow the return to the mine:-owners and/or operators (assuming they are diffe- rent parties) to increas~ in proportion to the Anchorage consumer price index (CPI) or the gross national product deflator. Alaska Energy Plaf\ning 10/26/82 .. ' P~ge 32 It is even less likely that the Alaska Public Utilities Commission (APUC) or the Legislature (in the case of a facility built by the Power Authority) would approve any coal-or electricity-purchase arrangement that was subject to such a floating-price term. There are two reasons for this judgment: (1) The cost-of-service arr!ffiKement described above 1s the way that utilities actually P!!!:~ coal on long-term· cOntract from dedicated mines. Indeed, we haye not been able to locate any instance in which a state public utilities commission approved a utility's long-term coal- purchase contract whose price was tied to the price of oil. Moreover (2) Mine financing will de.es_nd ;Qil the e~tence of long-term coal-purchase contracts. The prepo!!derance of bar~ PQWer in negotiations over such a contract will be on the side of the utilities or the Power Authority, rather than the mine owner. If CEA, for example, offered one or more of the owners of Beluga coal reserves (or any other coal reserve in the Railbelt) a 20-year cost- of-service take-or-pay contract to provide coal for a 200-megawatt generating plant, with a reasonable rate of return to the operator(s) and a reasonable royalty to the owner(s) of the mineral rights, that offer would be much too good to refuse (assuming, of course, that this production volume is sufficient to cover the mine's startup costs). It would indeed be a more attractive business proposition than anything yet offered by potential customers in East Asia. Any opportunity the mine owner(s) may have to export coal to the Far East over the same period will, incidentally, make them more, not Alaska Energy Planning 11/03/82 Page 33 less eager to deal with an Alaska utility on a long-term cost-of-service ba~i~, because of the ability added sales for export would ·give them to capture economies of scale by spreading fixed costs over a larger volume of production. Gas-fired vs Susitna genera~ing costs. The current average cost of Cook Inlet natural gas to Railbelt electric utilities is $0.86 per million btu (mmbtu).54 Simple~cycle gas turbines that can b~ .bought "off the shelf" would produce electricity at capital costs of about $630 pet installed kilowatt. 5 5 At today's Cook Inlet gas prices, the bus-bar cost (the pdce at the plant) of power from such a facility ·will be in the neighborhood of 4.3 cents per kilowatt-hour.56 The generation 'cost per kilowatt-hour would presumably be even lower for combined-cycle gas plants, and lower yet for gas-fired steam turbines operated at high plant factors (say, 80 percent) in base-load service. The 4.3-cent estimate, however, contrasts impressively with projected Susitna elec- tricity prices of 14 cents per kilowatt-hour (assuming the state subsidi- zed the project with a $2.3 billion appropriation) and 30 . cents (the full cost of power based upon 1 00-percent debt finandng). 57 Given these relative costs, it is reasonable to wonder why the Susitna project is even being considered. The Acres Summary gives the answer in brief: • • • between 1982 and 199 3, many of the long-term contracts now held . by utility companies for very favorably priced . Cook Inlet gas will expire. Not only will major increases in electric energy costs result from the require- ment by local utility companies to purchase gas at m...-ket prices, but also known Cook Inlet gas reserves may have been depleted in the early 1990's to the point that reliance upon na.tural gas as the principal ~uel J:or electri~al energy generat1on would no longer be poss1ble. 8 (emphasts added) Acres and Battelle have thus assumed that the "market prices" of Alaska natural gas will rise dramatically. Acres' "low" fossil-fuel price A~aska E,nergy Planning 11/03/82 Page 34 . case, to which the report assigns a 25-percent probability, projects Cook Inlet natural-gas costs at a constant $3.00 per mmbtu, which is three and one-half times present gas prices. The Acres "medium" ("50- percent probability") fossil-fuel price case assumes that these prices will skyrocket between now and the year 2000 at an average annual real rate of 9 percent, to $4.80 per mmbtu. The "high" case (25-percent probability) projects the trend at 11.2 percent annually, with its end- point at $7.22 per mmbtu, a more than eightfold increase over presen~ gas prices.-'9 (All prices are in constant 1982 dollars.) . . The weaknesses of Battelle and Acres in their analyses of gas prices closely parallel those that fatally compromise their analysis of coal prices. The assumption· of high future prices in both reports flows from two basic postulates, {1) that world oil prices will keep increasing in real terms, apparently without limit, and (2) that Cook Inlet natural- gas and Railbelt coal prices are, or at any rate presently will be, tightly coupled to these rising world oil prices. The Cook Inlet gas-price/world oil-price nexus. Acres assumes that the "opportunity value" of Cook Inlet gas is now $3.00 per m!Ylbtu, 2.5 times more than prices currently paid by Anchorage utilities . This assumption is based on the contractor's analysis of LNG export opportu- nities , LNG processing . and transportation costs, and the btu rela- tions.hip between gas and oil. Sin ce the assumption applies to all three cases ---"l o w", "medium", and "high" ---it is an essential element in the analytical process by which Acres discards natural-gas-fired plants as an economical long-term generation option, and by which the report f inds Susitna to be the preferred option.60 More importantly, there is enough evidence to make at least a plausible c ase that Cook Inlet gas prices will be established largely on the basis of factors local to the region. Proved gas reserves, for example, are far in excess of current demand, sufficient to supply the Alaska Energy Planning 10/26/82 Page 35 local utilities for roughly 75 years at present rates of c~nsumption.61 If the ~l1,.1ll capacity of the existing LNG and ammonia-urea plants is regarded as part of regional demand, the same proved reserves are good for about 23 years at e~isting production rates.62 In other regions of North.r America (Alberta, for example), similar reserve-to-production (r/p) ratios are considered evidence of _a gas glut.63 Even these measures tend to understate the potential gas supply in Cook Inlet. "Proved'' reserves (or "identif~ed economically reco- verable reserves", in the terminology of Alaska's Oil and Gas Conserva- tion Commission) constitute only that fraction of the resource base which producers have had a commercial incentive to explore to the point at which the producible volumes are a near certainty. This kind of exploration is expensive and, absent credible near-term market prospects, there is no reason for the lease owners to spend the money. The known reservoirs of the Cook Inlet basin contain considerably more unproved gas (or "indicated" reserves) than the 3.9 trillion cubic feet (tcf) reported as proved at the end of 1981. Nobody knows the volume of indicated gas in these reservoirs with much confidence (otherwise they would be counted in the "proved" column), but unpub- lished estimates in the industry tend to be in the 5. to 10 tcf range.64 Acres and Battelle disregard the high r/p ratios and the region's additional gas-development potential, on the ground that new demands will arise for Cook Inlet gas in the form of additional LNG -exports to Japan or California. This may indeed turn out to be the case, but it is a proposition not very well supported by the experience of the only new LNG export scheme to be seriously proposed ---the PacAlaska project, whose sponsors in September 1982 announced the project's indefinite postponement. 65 Most importantly, overwhelming evidence has accumulated in the last few months that the final-market value of natural gas is consider- Alaska Energy Planning 11/03/82 Page 36 ably lower, and total demand in either the Lower-48 or Japan much more limited, than the industry believed during the 1970's. The new gas-market outlook has cast a serious cloud over all "supplemental-gas" projects ---ANGTS and synthetics ---as well as LNG.67 At any rate, there are no other proposals to establish new Cook Inlet LNG facilities (or increase the capacity of existing facilities).68 Until serious industry l~terest appears in some such project, it is unrealistic to assume that Cook Inlet gas prices will be dictated by "netback" gas values in export markets.69 The world oil-price assumption once again. Acres' and Battelle's Alaska gas-price model, in which Cook Inlet gas prices reflect world oil pric es, is driven by the contractors' assumption of c ontinuously rising real prices for c rude-oil. In Acres' "most-likely case" the rate is 2.0 or 2.6 percent, depending on which of the Acres reports one reads. 71 The report states that its scenarios for gas-price escalation are expected to "follow closely the crude oil-price scenarios."72 For the years through 1998, however, the Department of Revenue assumes that the real price of Saudi "marker crude" (which underlies and drives the state's official revenue forecasts) will decrease at a compound annual rate of almost one percent. At least for the period through FY-98, the state's recent oil-price forecasts are as incom- patible with the Acres and Battelle assumptions regarding gas-price escalation in the Railbelt as they are with the consultants' forecasts of economic activity and energy demand. Alaska Energy Planning 10/26/82 Page 37 _.. ~~~:\<::. Susitna Construction Costs The Acres base-case estimate for the original cost of the Watana unit is $5,081 per kilowatt, and for the Devil Canyon unit $2,265 per kilowatt. (Several other figures appear for the two units in the Battelle and Acres reports, but the distinctions among them are not crucial here.) These figures contrast with an estimate of only .$636 per kilowatt for a 70-megawatt simple-cycle gas turbine. Costs per kilowatt of cap~dty for combined-cycle plants, co~l-fired steam plants, and some smaller hydroelectric projects cqnsidered for the Railbelt fall between these two extremes. Thus, the Susitna project is by far the most capital-intensive of the major electrical-generation "options considered by Acres and Battelle, and this capital-intensiveness means that the unit cost of Susitna power is much more vulnerable to capital-cost overruns than power from combustion or steam-turbine pl~nts. The authors of this review are not in a position to scrutinize the Acres estimates of Susitna construction costs as such; the way in which Acres deals with cost-overrun and related risks is within the scope of this review, however. Large construction projects have become notori- ous in recent years for costs running far above and sometimes many times higher than the cost estimates on which the owners ---private corporations and public agencies alike ---based their dedsior:t to go ahead. The TAPS and ANGTS experiences, among others, have made Alaskans particularly sensitive to the cost-overrun issue. One obvious question in evaluating and comparing any large P.roject proposal like Susitna is how much credibility anyone can place in the cost estimates and budgets. Public officials and the general public have become increasingly sophisticated about the tendency of project sponsors, their engineers, and their consultants to underesti- mate costs, and to downplay the risk of delay, false starts, and other Alaska Energy Planning 10/26/82 Page 38 causes of overruns. They have become increasingly skeptical about the figures offered by project promoters. Overruns and other cost-estimation errors are not a new pheno- menon,73 however, and there are a few generalizations that the research literature supports with a good deal of confidence74 Errors in construction-cost estimates can be separated, at least conceptually, into two components: variance and bias. Variance is a. measure of the expected departure of the actual cost either up or down from the estimated cost, and bias is the expected departure in one consistent direction (usually upward) from the estimated cost. These measures of error depend in different degrees on (1) the specific features of the project, (2) the general economic environment, and (2) the institutional framework in which the project is carried out. Character of the project. Cost-estimation variance clearly tends to increase with the size of project; novelty in design, location, or construction technique; the time required for conception, authorization, design, construction and shakedown; and the numbe r of licenses, per- mits, and degree of regulatory surveillance. Both m easures of estima- tion error decline with the amount of experience in similar or related construction on the part of the owners, designers, and contractors. Many of these features are intercorrelated: larger projects tend to be custom-designed, more complex, take longer to complete, and involve a greater number of regulations and regulatory entities. The larger the project, moreover, the fewer similar projects the owners, designers, and c ontrac tors are likely to have had experience with. Because of these intercorrelations, the effects of the various factors are nearly impos- sible to distinguish in practice, but --- On their face, the specific featl.D"es of the Susitna project suggest a high risk of cost-estimation error. Alaska Energy Planning 10/26/82 Page 39 Among other things, Susitna would be (a) one of the highest dams in the world, (b) the largest enterprise anywhere, ever, of its particular type, .(c) the highest-latitt.Jde large-scale hydroelectric project in the world and, so far as we can find, the laq~est civil works proj_ect ever att~r:npted above 55th parallel. The Acres study has attacked these project-specific features of the cost-estimation risk from tw9. angles; at the core qf the whole Su~tina feasibility study ~s an impressively designed, comprehensive engineering-type "multivariate" risk analysis. Acres then checked the findings of this analysis a.gainst a comparison of cost-estimates and \ actual costs on a large sample o~ completed federal water projects. In the core analysis, Acres has created a model of the construc- tion process in which the study te~m has identified each major uncertainty, including everything the designers think might go wrong, in the following areas: 8 categories of natural risks (flood, ice, etc.), 2 categories of design-contr~lled risks (seepage, etc.), 6 categories of construction r~s (equipment availability, labor disputes, etc.), 4 categori~s of human ri.$:; (including contractor capability and quality control), and 2 categories of special risks (inclu9ing regulatory delay) The Acres team assigned probabilities to th~ various contingen- cies in ~ach category and then combined these probabilities mathemati- cally, to create a schedule that shows the probabilities of various levels of total cost (and a number of other outcomes of the construction process). With the availability of computers capable of manipulating huge array~ of variables each of whkh is paired to a probability coefficient, this kind of risk analysis has become common in the engineering of Alaska Energy Planning 10/26/82 Page 40 large and complex projects, where it is a powerful tool for helping designers choose the project configuration that minimizes expected costs, subject to certain maximum a cceptable cost, schedule, and operations risks. The general method used by Acres is pr obably the best available technique for comparing expected costs and risk factors among variants of a single project concept, and it is reasonably reliable for projecting which of (say) three proposed configurations of the project would have the lowest expected cost, and which would have the lowest risk of unacceptable cost overruns or operating failure. For a number of theoretical and practical reasons, however, such risk analysis is probably less valid and reliable for comparing widely different technologies, facilities, or strategies, or for generating abso- lute-dollar cost estimates, cost-overrun, or schedule-risk estimates to be used outside the design process, particularly as input to benefit-cost analysis. No matter how competent and objective the analysts are, it is almost impossible for them to escape a certain amount of misplaced specificity,75 subjectivity and over-optimism,16 institutional blind spots,77 and underallowance for non-completion.78 Nevertheless --- The Acres approach to controlling cost-estimation variance stemming from the character of the project is as rigorous as any we have seen. More importantly, perhaps --- We have not found any evidence of bias in the cost- estimation procedure or in the construction-cost aspects of Acres' multivariate risk analysis. In the second phase of its risk analysis the Acres report implicitly recognizes the inevitable subjectivity of any engineering-type risk analysis, and checks its results against a sample of actual project experience. Acres' discussion acknowledges shortcomings in the survey data base ---particularly the fact that the sample is made up of Alaska Energy Planning 10/26/82 Page 41 ,, . :.~ ... projects completed before passage of the National Environmental Policy Act, the Endangered Species Act, and other environmental legis- lation had their present great impact on project schedule and comple- tion. Further examination reveals that the sample offered by Acres has virtually no resemblance to the Susitna projects in type, location, scale, or timing. More comforting historical evidence that overruns can be controlled exists, however, in the construction record of the 5,225 MW Churchill Falls hydroectric project on James Bay in Canada. ChurchHt Falls is probably the closest parallel to Susitna in scale or location anywhere in North America.79 ' The Churchill Falls project1 whose design and con- strtJcti'N' were l}'lanaged by a ·j~lrit 'Y!!!!ture of Acres and a Bechtel subsi<Uar:y, was completed essentially on time and on budget. General economic conditions. It is inevitable that overruns (which correspond to a downward bias in cost estimates) are more frequent and generally larger in periods when inflation is accelerating, and also in periods when environmental, safety, and other kinds of regulation are getting more complex and demanding. The upward push that thes~ factors give to costs is even stronger when inflation proceeds out of (or coincides with) an economic boom, because accelerating re.al economic growth tends to push construction wage rates, building-materials prices, . and other construction costs (including contractor markups) ahead of general inflation (as measured by the Gross National Product deflator or the Consumer Price index). Accelerating inflation also tends to mean nsmg interest ·rates, which result in higher interim financing costs (in utility parlance, "allowance for funds used during construction" or AFUDC), and thus cause final project costs to increase even faster than the wages of construction labor and the cost of building materials (themselves racing ahead of the rest of the economy). As the 1960s and 1970s saw aU of Alaska Energy Planning 11/03/82 Page 42 these c onditions, it is not surprising that the experience of this period fostered a belief that large overruns are the rule rather than the exception in large projects. The economic and regulatory forces that generated the construc- tion-cost overruns of the 1970s have largely run their course. General inflation will probably deeelerate (meaning that escalation rates built into construction-cost estimates will typically be too high rather than too low}. This trend probably means lower nominal interest rates as well and, as a result, pre-construction estimates of interim-financing costs (AFUDC} will tend to be too high. Real economic growth rates are likely to be lower than in the 1960s and early 1970s, moreover, causing the construction-cost indexes to increase less rapidly than general inflation. Finally, the impact of environmental and safety regulation on costs and schedules will tend to be less severe than at present. While we do not anticipate a significant retreat from the goals motivating today's environmental and safety regulation, regulation wiU tend to be more sensitive to cost-effectiveness criteria and on balance less dilatory. For these reasons --- Big cost cost and schedule overruns are not likely to. be the rule over the next decade as they have been over the last. A number of large projects are likely to surprise their sponsors and the public by coming in on time and under budget. Susitna might be one of these. Institutional considerations: The Road to WPPSS. It seems reasonable to expect governmental entities to have weaker incentives for c ost-minimizing design and cost-effective contrac ting than private enterprise, and for regulated private utilities (which profit by enlarging Alaska Energy Planning 10/26/82 Page 43 their "rate base" and are not supposed to profit _from cost-cutting) to hav-e· weaker efficiency incentives than unregulated private enterprises. The existing literature on defense-procurement costs seems to support the first thesis, but we do not know of any systematic comparison reJating the costs of otherwise similar plants owned by different kinds of organization. A casual survey by the reviewers shows mixed results: the publicly-owned Washington Public Power Supply System (WPPSS) is likely to ride its overruns into the biggest financial default of any industry in US history. On the average, publlc entities in the United States have worse schedule and cost performance on nuclear plants than private· utilities. (TVA's reactors ·seemed to be doing better than the / average, however, before construction was stopped, making their effec- tive unit cost of capacity infinite.) But Ontario Hydro's performance in nuclear plant construction (and operation) has· been better than any private utility in the United States (a result that stems at least in part from its choice of a much less troubfesome reactor technology). Clearly, this issue needs further investigation. Nevertheless--- The Susitna proj~ has disturb!ng ~allels with WPPSS. Like WPPSS,. Susitna is sponsored ·by a single-purpose entity (the Alaska Power Authority) whose importance in the world rests primarily on this enterprise; and like WPPSS, that entity has had little or no experience in designing or building works of a sirh ilar kind or scale. The technical case for the project is built to. a large extent on exaggerated projections of future electricity "needs" and on unrealistic projections of alternate-fuel costs. More importantly, however, the decision process is highly political, and a large part of the political constituency for the project is indifferent to its power-generation economics, sound or otherwise. Many advocates see the project mainly in terms of "jobs" Alaska Energy Planning 11/03/82 Page 44 or an undefined thrust for "economic development", in which the perceived benefits increase proportionally with construction costs. And like WPPSS' 11net-billing11 agreements with the Bonneville Power Admi- nistration, at least some of the financing schemes being considered for Susitna (a direct appropriation from the state treasury, or general- obligation bonding) would bypass the need to convince lenders that the project is prudently conceived and designed and will be implemented efficiently. In our judgment, the most serious cost-overrun risks connected with Susitna do not flow from oversights or biases in Acres• engineer- ing-cost estimates. Nor do we believe that big cost overruns will be as prevalent on large custom-engineered construction projects as· they have been since the mid-1960s. The paramount construction-cost risk connected with the Susitna project is not that the Acres cost estimate is too low and, as a result, encourages the state of Alaska to invest in a project that turns out to be Uneconomic. The Acres estimates, in other words, may be on target. The most serious cost risk associated with Susitna is, rather, that --- An Acres cost estimate which is essentially correct may combine with unrealistic forecasts of Railbelt electri- city demand and fossil-fuel prices to encourage an Alaska investment in a project that would be uneconomic even in the absence of construction-cost overruns. This is precisely what is likely to happen if the Alaska Power Authority and the Legislature rely on the present Battelle and Acres reports as the last word on project feasiblity. Ironically, the latter risk begets the risk, from an entirely different source, of large cost overruns relative to the Acres estimate, no matter how good that estimate in itself may be. If the legislature authorizes construction of the Susitna project on the basis of fanciful Alaska· Energy Planning 10/26/82 Page 45 ,; f •·••. ~· economic assumptions, financing the project wil! likely require state gove·rnment to shelter consumers from the true cost of Susitna power by means of a direct capital grant, or to shelter lenders from the true ec<?nomic risks of project investment by means of general-obligation bo,nd financing. These devices, adopted precisely because the 5usitna prQject co.uld not stand on its own feet, would deprive the project of any important constituency for good management and cost control to offset the interests that stand to benefit from overbuilding, goid- plat!ng, and wasteful mismanagement. Non-completiOn risk. In the last few years, construction of a large Aumber of electrical generating plants (mainly nuclear, but a few coal-fired) in the United States has been abandoned, suspended, or stretched out. Most of these plants were, at the time, suffering major technical, regulatory, and/or financing problems, out it is a mistake (albeit a widespread one) to regard these problems as the cause of abandonment or delay. In almost every recent case of .power-plant non- ·c:ompletion in the United States the root cause has been falling load-growth projections. . Systems that had to pare back their. construction schedules in line with revised demand forecasts naturally chose to abandon or delay their most troubled projec~s, while plants that were not justified by prospec- tive load growth naturally became relatively difficult to finance. If the utilities in question really "needed" the added capacity, they would have persevered in trying to build them despite their .technical and regula- tory problems. Likewise, the utilities would have been able to finance most of the troubled plants if lenders had been confident that consumer demand would generate sufficient revenue to pay their costs. Acres concluded that there is only a negligible risk of non- completion stemming from those contingencies Acres explicitly consi- Alaska Energy Planning 11/03/82 Page 46 dered in its analysis of construction risks; we do not dispute that judgment. In recent years at least, no major hydroelectric project in North America has been terminated or indefinitely postponed for engineering reasons after construction began. Hydro facilities are not, however, inherently immune to cancellation or prolonged delays stem- mi ng from some of the troubles that have beset nuclear-plant construc- tion plans, including environmental regulatio n or 1itig3tion.80 There is certainly no reason to believe t hat hydropower is any less vulnerable to non-completion risks that arise from changed economic circumstances or bad pl anning and management than are n uclear and fossil-fueled plants. If the state does abandon Susitna construction in the face of now- unforseen engineering, environmental, regulatory, or financial pro- blems, it will at bottom be because the state and/or the lenders belatedly realized that the demand for Susitna power at its expected cost no l o nger justifies the additional outlays necessary for completion. This, rather than the fickleness of construction-cost estimates, is the real lesson of the WPPSS disaster for Alaskans. Alaska Energy Planning 10 /26/82 Page 47 p· . Iss mailClOg ues ·,Real discoWlt and interest rates. Benefit-cost analysis is the techniq.ue of comparing all of the benefits created by a project with all of its costs, no matter when they occur. In earlier y~a·rs, the end produ€t of such an analysis was usually a benefit/cost ratio: a worth- while. project was one with a ratio greater than 1.0, and the best of a group· of competing projects was the oAe with the highest ratio. More tecef\tly, the concept of net benefits (total benefits less total costs) has become more fashionable: a worthwhile project is one that creates a positive net benefit, and the best project from a group of projects is the one that creates the greatest net benefit. Acres has used a variant of this approach that assumes total benefits to Alaskans of meeting their electric power demand to be the same regardless of how the power is generated. Thus, the system that has the lowest expected cost is the one to be chosen, and under most assumptions, Acres finds that the Susitna project results in lower expected costs than any of its alternatives. All of these benefit-cost approaches require the analysts to choose a discoWlt rate for translating the costs and benefits of various future years to present value at the time the investment decision is \ made. Thus, the stated goal of the Acres analysis is to find the system •. . that has the lowest expected costs as of 1982. The outcome of benefit-cost analyses involving long-lived capital- intensive investments is very sensitive to the choice of dJ.sco.unt rates, therefore making it a consequential and often controversial issue. The various assumptions adopted by Acres, for example, lead to the finding that Susitna is most likely the least-cost alternative if costs and benefits are discounted at a real (inflation-adjusted) rate of 4 percent or less, but' that it is probably not the least-cost alternative if costs and Alaska Energy Planning 10/26/82 Page 48 benefits are discounted at .5 percent or more. A more familiar way of stating this finding is that money invested in Susitna is likely to earn the people of Alaska more than 4 percent and less than 5 percent per year, in excess of the return necessary to offset inflation. After a clear discussion of the various discount-rate concepts ' used in benefit-cost analysis, Acres points out that the proper concept to use depends upon the purpose of the analysis.81 In the case of the Susitna project, the proper discount rate is one that reflects the investment cost to Alaskans of the money sunk into the project; however, this cost, and thus the discount rate to be chosen, also depend on the specific financing arrangements proposed. The appropriate discount rate to use, and thus the decision whether or not a given project is a good investment, depend on the state of the capital market at the time the evaluation is made. And, benefit-cost analysts need to to take inflation into consideration in choosing a discount rate ---or more precisely, the expected rate of inflation over the life of the investment. Acres makes this adjustment by using "real" or "constant- dollar" interest and discount rates. Even after this adjustment, however, real interest rates vary greatly over time. Interest rates on tax-exempt municipal bonds are normally lower than those on taxable securities of the same quality and maturity. Thus, it is conceivable that an economic-feasibility analysis of Susitna would indciate that the project is a good buy for Alaskans if it can be financed with tax-exempt borrowing, but a loser if the Internal Revenue Code does not permit the Alaska Power Authority to sell tax-exempt securities, or if the legislature were to appropriate funds that otherwise could have been invested in high-yielding federal-government or corpo- rate securitie~. To the extent Susitna (or any alternative to Susitna) is to be debt- financed, the appropriate discount rate is the interest rate at which the Alaska Energy Planning 10/26/82 Page 49 money would be borrowed and, to the extent the state proposes to under~rit~ the project with appropriated funds, the proper discount rate is the return the state could otherwise e_arn by investing the same funds. The Acres analysis d~s not explicitly take. in~o acc~u~t tl)e difference between the state's borrowing and lending costs. Nor is the w o ;$t-'' '• • ' '\ 0 • ,.; ' report. clear what Acr:es assumes. ~b~ut the f~.deral income-tax stat~s of Su.si_t _ry_a . debt. It do~s not, at any rate, offer any coherent suppoq for ~ull)ing that the. AI~ Power Authority will ~· able to bcx't:ow in the •. . . tax-exempt bond market. Neither do~s the report deal with the possible effect of changes in real interest ra~es over time. Acres, instead, uses an approach that bypasse~ aU of these issues: After some general observations about. interest-rate history, the report adopts a d~scount rate of 3 per<;:ent on the following grounds: • • • long-term industrial bond rates. nave. averaged abo~.t 2 to 3 percent in the U$.: in real (inflatipn-adj'usted). terms. Forecasts of real interest rates show average values of about 3 percent and 2 respectively. The US · Nuclear R~gulc:itory Commission has also analyzed the choice of discount rates for investment apprai8(ll in the electric utility industry and has recQmmended a 3 percent real rate. Therefore, a real rate of 3 pe~cen.i has been qdopted· as the lxise. case discount and interest rate for the period 1982 to 204o.82 This approach might have been workable ten or even two years ago, but it is unsu~table in today's financial environment, in which current real yields have only the most tenuous connection with historic levels. matters is actual cost of borrowin or not investin elsewhere at the time the financial commitment is made. In August 1982, inflation (in the form of the annualized rate of ch~nge in the gro~s natiol}~l pr9ject deflator) was running below 7 Alaska Energy Planning lQ/26/82 Page 5q / percent, and most forecasters expected even lower rates of inflation in the late 1980's and the 1990's. Long-term municipal bond yields were in the vicinity of 13 percent, and long-term taxable utility and corporate bond yields 14-15 percent. Both nominal interest rates and future inflation rates were generally expected to fall, but the numbers implied that the real interest rate today (current bond yields less expected inflation) was in the 6-8 percent range. In July, 1982, Data Resources Inc., forecast real long-term interest rates rem aining above 4.5 percent through 1992.83 (In late October, bond yields and forecasts of future inflation had both fallen substantially relative to August levels, but the general relationship among them remained about the same.84) At a discotmt rate corresponding to current real interest-rate levels, even the Acres analysis rejects Susit- na.&' Even with the outdated or questionable assumptions (identified elsewhere in this review) that bias the reports' benefit-cost analysis in Susitna's favor, Acres' own analysis implies that the project is not cost- effective in today's capital market. Thus, it would not be prudent for the state to borrow money at today's market rates to finance the pro- ject even with tax-exempt securities, and it would be even less cost- effective to commit permanent-fund or general-fund revenues that might otherwise be invested at today's high yields. When and whether long-term capital markets will r eturn to normal historical patterns, with long-term (taxable) real-interest rates on the order of 3 percent, is one of the big economic puzzles of the age. The real interest rates that have prevailed recently are about the highest on historical record, and they were not foreseen by any school of economic or financial analysts. The two most fashionable expert explanations for high interest rates today are almost totally contradic- Alaska Energy Planning 10/26/82 Page 51 tory. Stated differently, ~ong-term capital markets are in great disorder and nobody really knows why or what "normalcy" now means. A .3-percent disc01mt rate appears much too low to use as a base case . for evaluatiiig the eeonomic merits of electricity-generatiOn alternatives for the Railbelt. Today's capital market is sending a message to the sponsors of big-ticket private investments (like ·ANGTS, for example) as well as to ·the state of Alaska with respect to Susitna that now is not tlie time to lock billion's of dollars into them. If the rest· of Acres' economic assumptions were sound (they are not), we might reasonably hope that a window would open in (say) two years or ·fiVe years, ih which r eal interest rates would have fallen to a level that makes the prqject cost- effedive. Under today's capital market conditions, however --- Acres• own analysis does not justify or supPOrt a state inveStment in Susitna. Alaska Energy Planning 10/26/82 Page .52 / NOTES Introduction and Summary l. State of Alaska, Division of Energy and Power Development, State o_! Alaska Long Term Energy Plan (Juneau: 1982). The "plan" is in two volumes, an Executive Summary (cited hereafter as L TEP Summary), and the main report {cited as L TEP). 2. Acres American Inc. for the Alaska Power Authority, Susitna Hydroelec- tric Pro·ect Feaslbilit Re ort (Anchorage: 1982). The report encom- passes a "Summary Report" cited hereafter as Acres2 Summary), the main report (cited as Acres), and 73 "reference reports." Reference report R-72, "Task 11: Reference Report: Economic, Marketing, and Financial Evaluation," is cited as Acres, Task 11. Reference report R- 73, "Task 11.03, Close-Out Report, Susitna Risk Analysis," is cited as Acres, T~k 11.03. 3. Battelle Pacific Northwest Labratories for the Office of the Governor, State of Alaska, Railbe}t .Electric Power Alternatives Stud : Evaluation of Railbelt Eleetric Energy Plans Richland: February 1982 , cited hereafter as Battelle. A subsidiary report, "Fossil Fuel Availbility and Price Forecasts" (Richland: March 1982), is cited as Battelle, Fossil ~- 4. The present reviewers were not involved in ISER's production of Railbelt economic scenarios or load-forecasts for Battelle (and indirectly for Acres). ~ackgr;ound to the ~tudies 5. These programs cost the state over $490 million in FY 1982 (LTg.P, p IV- 6). The 1982 legislature elimi nated some programs, including the energy audit subsidies, but the costs of others, such as "power cost assistance," can be expected to increase. If the aggregate costs of these programs stay constant in nominal terms, the energy subsi<Jy programs. , will account for roughly 18 percent of the $2.7 billion appropriated for FY 1983. For a description of each of the state's 2'0 subsidy programs; see .hill_, Exhibit IV-4 (following p IV-6). .·• 6 . LTEP, p 1. 7. ~' p IV-11. Alaska Energy Planning 11/03/82 Page 53 TheThr~Stud!es . ,. ·: : ·-··.; ' . ·'\·.!::.·•' _.,-.. _1 .:..-!·.~Lf __ ,,.,·, ,; . .': c·, · ~~1 ·.~·n::~ <'->LusJ ·A \/i ~:~~\.fi .. ,~c .. [.?~ .•. . . -~: '• .. i: . 8. LTEP·, p IV-4. 9. LTEP.;·.P.P IV-15-IV-22. Future .Oil Prices 10. Div.ision of Petroleum Revenue, Alaska Department of Revenue, Petro- leum Production Revenue Fore<;~st: Quarterly Report. In June 1982 the De,pattment of Revenue's oil-price forecasts were revised slightly upward. In September, however, the Department issued a forecast which in the long term is even more pessimistic (regarding state oil revenues) than the March report. (Anchorage: June 1980 through September 1982), cited hereafter as Pet. Rev. Forecast. 11. Pet~ Rev. Forecast, March 1982, p 5. 12. Charles Logsdon, Chief Petroleum Economist, Alaska Department of Revenue, personal communication to Erickson, July 18, 1982. 13. Cf. Tussing's remarks to ·the quarterly meetings of the US Department of Commerce Economic Advisory Board during 1979 and 1980; A R Tussing, "The 1981 Oil Price Outlook" in The Economic Outlook for .!ill, University of Michigan, November 1980; A R Tussing, "Will Oil Prices Keep Rising? Maybe Not", Anchorage paily News, February 14, 1981; Jon Matthews, "State Revenues Likely Will Drop: Oil Prices are Heading Down", Anchor:age Daily News, May 29, 1981; A R Tussing, "Alpetco's Collapse Has Lessons for Budget Planners", Anchorage Daily ~' May 23, 1981; Bob Shallit, "State Faces Revenue Loss, Econo- mists Say", Anchorage Daily News, January 15, 1982; "Alaska Cuts Forecasts", January 18, 1982; Erickson & Associates for the Alaska State Legislature, The World Oil Market and Alaska S.tate Revenues: A Fifteen Month Forecast (Juneau: March 1982). Also see the appendix to this review, "Reflections on the End of the OPEC Era". 14. "The Energy Issues" in T Barker and V Brailovsky, Oil or Industry, edited proceedings of a Conference on Policy Issues in Energy Self-Sufficient Economies at Different Stages of Industrialization held at Oaxaca, Mexico in September 1980. (London: Academic Press, 1981 , p 3.) 15. The following examples are typical of the lot, as far as the reviewers can tell. In 1977 Exxon forecast that US consumption of petroleum liquids would ·be 20.3 million barrels per day in ]980, but actual consumption was only 16.3 million. And two years ago, Data Resources Inc. (DRI), an authority upon which Acres and Battelle have relied for fuel-price forecasts, was predicting 20 years of 4-percent annual oil- consumption growth in Europe. However, DRI now expects the cont i- Alaska Energy Planning 11/03/82 Page 54 nent's consumption to be below its 1979 peak during the rest of this century. (Reported in P~tro!eum Intelligenc~ Weekly, 28 June 1982, p 7 .) Also s~e the appendix to this review, "Reflections on the End of the OPEC Era" for further elaboration of the world "flight from oil." 16. D. de Bruyne, as quoted in Petroleum InteHigen<;:e Weekly, 14 June 1982, p 9. 17. The published forecasts of government agencie~, and the big consulting firms seem to be among the last to recognize the changed outlook --- just as they were among the last to recognize that the high prices established in 1973-74 would be with us for a while. In response to the review draft of this paper, Acres offered the following list of "major forecasts of oil-price trends": Source Data Resources Inc. International Energy Agency US DOE Energy Inform. Adm. Canada: Energy Mines & Res. Ontario Hydro Energy Modeling Forum Average of ten models Dr. F. Fesharaki Date of Forecast Summer 1982 Spring 1982 Spring 1982 Summer 1982 Spring 1982 February 1982 Spring 1982 Forecast Trend ~percent) +2.8 -0.5 to +2.0 above +3.0 +1.7 +1.8 +1.9 to +5.3 +1.7 From Alaska Power Authority, Susitna Hxdroelectric Project, "Commentary on 'Alaska Energy Planning Studies' (A R Tussing and G K Erickson)", prepared by Acres American Inc., September 7, 1982. While the Acres compilation does not reveal the time span of the various projections, it is not unrepresentative of the kind of forecasters polled, even in late 1982. Significantly, however, this list does not contain a single petroleum-producing company, financial institition, or agency of an oil-exporting political entity, while four (half) of the forecasts come from go_vernmental entities that have a powerful institutional stake (as does the Alaska Power Authority) in perpetuating the belief that real oil prices will continue to rise. . None of the recent forecasts the reviewers have seen, prepared for internal use by major petroleum producers (corporations or ~ov~rn­ ments), expects sustained growth in the real price of oil during th~ .. r;est of this century. All of them now assume continued constariJz-.9o11~ price declines through at least 1985. At least two major integrated oil companies that now assume for their own planning pur-PQ~~~ .. ~na;:t . th~.: Alaska Energy Planning 11/03/82 Page 55 • -••• -r ·' .· long-term oil-price trend will .fall somewhere between a level nominal- dollar and a level constant-4ollar trajectOry, released public forecasts ;in 19'82 that still show long-term re~l-price increases for oil. More signi!icant than what the tompan.ies are saying, · however, is what they are doing. The . dramatic change in industry's long-term oil-price expecta- tions o·v~r the last two years is; impossible to ignore: The companies have ~'ctittled every major ~nsubsidized synthetic-fuels project in North America, and most of the subsiqized ones to boot. Oil companies are cutting !>ack massively on drilling programs, but the most telling indiGator ·of the drastic chang~ fn their expectation is the fact that they are ·now paying only about half as much .per barrel for proved petroleum reserves as they were paying in late 1980. (On the last point, see B F -' PiCchi; "The Valuation of US Petroleum Reserves: Exploding the Myths" h1 Salomon Brothers Inc Stock Research/Industry Analysis, Octob'er i5, 19,82.) 18. Alaska Department of Labor, Alaska Planning Information: 1982, p 26. ' .' 19. Alaska Department of Revenue, Revenue Sour:ces, June 1982, p 7. ' 20. Calculated from dat~ contained in City and Borough of Juneau, "Notice to Taxpayers," May 1982. 21. Recent preliminary work by one of the reviewers on the actual phasing of state expendi.tures suggests that there may be a significant lag (as much as two or three years) between the time that revenues and appropriations began to fall (the interval between the FY-1982 and FY- 1983 budgets) and the beginning of a decline in actual. state expendi- tures. For example, state appropriations for FY-1981 ---when revenues and appropriations were still increasing rapidly ---were $3.1 billion, yet our preliminary data on actual spending (based ·o·n warrants redeemed) show that the state actually spent only· $1.8 billion in that fiscal year. Load Foreccl5ts 22. Acres Task 11, pp 18-30. Acres "tes.ts" the possibility of a positive correlation between oil prices and Alaska energy demand; however, as we show below, the "low load forecast" used in the Acres sensitivity analysis is unreasonably high given present expectations concerning ·state revenue, state spending, and major resource-development pro- jects. 23. Calculated from data in Battelle, p A-6 (Table A-3). Alaska Energy Planning 11/03/82 Page 56 24. Battelle,. p A.2 (Table A-7). 25. Battelle, p A.2 (Table A-1). According to 0 S Goldsmith, the ISER and Battelle moderate case assumed that per-capita state spending moved proportionally with per-capita personal income. (Personal communica- tion, September 6, 1982) '26. Acres, Task 11, pp 18-50. According to Acres, the State directed Acres to use the Battelle load forecasts. (Alaska Power Authority, Susitna Hydroelectric Project, "Commentary on 'Alaska Energy Planning Stu- dies' (A R Tussing and G K Erickson)", prepared by Acres American Inc., . September 7, 1982, p 2). 27. Acres Summary, p 5. 28. Battelle, p 3.12 (note a). 29. Calculated from data in Battelle, p A-7 (Table A-4). 30. A R Tussing & L S Kramer, Hydrocarbons Proc;essing (Anchorage: ISER, . 1981). 31. Battelle, Fossil Fuel, p 2.7. 32. This calculation assumes a 1982 population of 410,000, and the 3.4- percent annual rate of. population increase in Battelle's forecast. Real dollars are converted to nominal dollars using the seven-percent annual inflation rate assumed by the Department of Revenue (note 13, supra). 33. Pet. Rev. Quart. Forecast, March 1982; Milt Barker, Memorandum to Representative Thelma Buchholdt, 9 June 1982. 34.-Pet. Rev. Quart. Forecast, March 1982, p 13. 35. Arion R Tussing and Associates, Inc., Introduction to Electric Power Supply Planning (Anchorage, 1_980), pp 32-33, 90-91. 36. University of Alaska, Institute of Social and Economic Research, ;'Alaska Electric Power Requirements, A Review and Projection," in Alaska Review of Business and Economic Conditions, June 1977, pp 1, 15~ The · actual growth rate is calculated from data in Alaska Power __ Adrp_ini-~_,., stration, El~ctric Power Statistics, 1960~1980, August 198-.l,;p }9..~; .. · ,: .. ~:· · ·· ' ' • ;. _;: (..,, .._,o.,._.,,,:.,.,r ::: : ~:;~:n1~6c;:~:~::~On, 0 S Goldsmith to Tussing, July 2~,,p,s~,=_t :~' :' ,;; :'· Alaska Energy Planning 11/03/82 Page 57 . \ ~:.:~~*·.:i~f: . ·-~ ··:;;::- .• : l •• - 39. L TEP, p 1-6. As we noted above, Battelle's "low-case" assumption for Railbelt population growth is 1.8 percent annually through :2000 . An- o~ner Battelle study gives a "low-case" projection of statewide popula- tion in the year 2000 that implies an annual growth rate (1980-2000) of 1.7 percent. (Michael J Scott, et al, Historical and Projected Oil and .Gas Consumption. Battelle Pacific Northwest Laborat9ries for t.he Division of Minerals and Energy Management, Alaska D.epartment of Natural Resources, January 1982, p D.lf) lf.O. L TEP~ Exhibit I-1 0 {fo!lowing p I-7). It is not clear, however, which of the many Battelle "cases" was used; the 1980-2000 Railbelt electric energy demand growth rate is listed as 3.5 percent. Alaska fossil-Fuel Availability and Costs .39. Acres Summary, p 5. 40. Acres Summary, p 9. 4-I. Al~ska Power Authority, Susitna Hydroelectric Project, "Commentary on 'Alaska Energy Planning Studies' (A R Tussing and G K Erickson)", prepared by Acres American Inc., September 7, 1982, p 9. 42. ART A, Inc. "Outlook for Proposed Coal-Export Terminals on the West . Coast .o.f North America" (unpublished proprietary report, October '1982), "Full Tanks Make Japan a Tough Customer for LNG", in The Asian Wall Street Journal, September 21, 1982. 43. Battelle, p 2.2; Acres Task 11, Table 18.2 .• 3. 44. omitted 45. Acres Task 11,. p 18-8. 46 .1oc. cit. 47. op. cit., p 18-9. 48. Minerals Yearbook, 1980. 49. US Department of Labor, Bureau of Labor Statistics, Weekly and Hourly Compensation of Production Workers in the Bituminous Coal Industry. 50. Acres Task 11, loc. cit. Alaska Energy Planning 10/26/82 Page 58 .51. Fuel oil prices: 19.50 and 1970, US Department of Labor, Bureau of Labor Statistics, Western Regional Office (San Francisco), telephone commu- nication; coal prices: 1973 and 1981, GVEA public information office, telephone communication; other data from Fairbanks North Star Bo- rough, Community Research Center, The Energy Report. 52 . Acres Task 11, loc. cit • .53. Japan, Commodity Trade Statistics • .54 . Battelle, p 2 .2 (Table 2.1) • .5.5 . Acres, p 2-4. 56. This estimate assumes a heat rate (btu/kilowatt-hour ratio) of 10,000, an interest rate of 1.5 percent, a .50-percent load factor, 5 mills per kilowatt-hour operation and maintenance costs, and present Cook Inlet gas prices. (loc. cit.) The cost estimate would probably be somewhat high even if the Acres-Battelle gas-price assumptions were accepted, as heat rates substantially lower are possible; levelized prices comparable to those used to project Susitna power costs would require the use of "real" {infl ation-adjusted interest rates) of 3 percent {6 to 8 percent would probably be more realistic in today's market), and 5 mills considerably exceeds CEA's current 0 & M costs for gas turbines. 57. Acres, Summary, p .50. 58 . op. cit., p 9 . 59. All of the data used here are derived or calcul ated from Acres, Task 11, Table 18.2.3. The Anchorage natural-gas utility has offered to purchase gas from c urr e ntly shut-in fields on the Ken ai peninsula at a price in the neighborhood of $2.2.5 per Mcf, escalating with the producer-price lndex. The offer was refused but, in the absence of any visible alternative markets, the refusal could be construed as violating the 11d1ligent deveJopment11 provision of the state oil and gas leases. 60. In particular, see the sensitivity of the "net benefits" of Susitna to $0.65 MMBtu change in the price of coal. Acres, Summary, p 46. 61. Battelle, Fossil Fuel, p 2 .29, 2.30 (Tables 2.9, 2 .10). 62 . Assuming no change in the consumption of the LNG and chemical plants. See Battelle, Fossil Fuel, p 2 .12. 63. US Department of Energy, The Current State of The Gas Market, 1982. Alaska Energy Planning 10/26/82 Page .59 64. The US Geological Survey's latest estimates of undiscovered natural gas in the Cook Inlet area are as follows: Low Mean High . {F.95) (F.5) <F.o5> Onshore Associated-Dissolved neg. .2 .6 Non-Associated 1.1 3.3 7.2 Offshore Associated-Dissolved 0 .4 2.2 Non-Associated 0 1.3 5.9 TOTAL* 3.0 5.2 12.4 * Totals assume resources in the four categories are inde- pendent of one another. soUrce: US Geological Survey·, Circular 860, Estimates of Undiscovered Recoverable Conventional Resources of Oil and Gas in the United States 0981), pp 76-,78. Industry personnel in Alaska seem to be somewhat more optimis- tic about the Cook Inlet gas resource than is the USGS. A typical industry view of the Cook Inlet gas-supply situation (the most lucid expression of that view the reviewers have encountered) came from a veteran oil-company geologist: Sure, there's gas shows all over the place. How much? Hell, who knows? But I'll tell you one thing. Ain't nobody goin' to drill much of it till some of that shut-in stuff starts movin'. Similar views are expressed by officials of Pacific Gas and Electric Company. (William L. Cole, personal communication to Erick- son, February 6, 1981). See also, another Battelle study: Michael J. Scott, et al, Historical and Projected Oil and Gas Consumption (Battelle Pacif~c . Northwest Laboratories for tne Division of Minerals and Energy Management, Alaska Department of Natural Resources, Jan·uary 1982) 65. At the end of September 1982, the California Public Utilities Commis- sior1 gave the sponsors of the LNG proje~t 60 days to prove. its economic viability. (Inside FERC, October 11, 1982) The companies ---Pacific Gas and Electric, and Pacific Lighting ---responded immediately with ah announcement putting off construction indefinitely: Alaska Energy Planning 10/26/82 Page 60 ••• the NGPA has made available substantial added supplies of domestic natural gas, according to Bill Wood, gas compa- ny senior vice presiCent and president of Pacific Lighting Gas Supply Co. As a result, Wood said, the company has announced that an affiliate has concluded it will not proceed at this time with construction of a liquefied natural gas (L~G) receiving terminal some 40 miles northwest of Santa Barbara, as well as an LNG liquefaction plant in south Alaska ••• Alluding to the current adequacy of natural gas sup- plies, Wood said: We continue to believe the importation of LNG is an extremely valuable energy source for California. Further we believe that within two or three years we will be much better able to determine when these LNG facilities should be put into operation. (Southern Caltfomia Gas Company press release, October 4, 1982.) 66. omitted. 67. On the export market, see, for example, "Full Tanks Make Japan a Tough Customer for LNG", in The Asian Wall Street Journal, September 21, 1982. For US market conditions see, inter alia, A R Tussing and C C Barlow, "The Rise and Fall of Regulation in the Natural Gas Industry," Public Utilities Fortnightly, March 4, 1982. (An expanded version is to appear as "The Future is Now", in the October 1982 Energy Journal.) Also, Tussing and Barlow, "A Survival Strategy for Gas Pipelines in the Post-OPEC Era", address to the Annual Meeting of the Interstate Natural Gas Association of America, September 27, 1982. (To appear in Public Utilities Fortnightly in January 198.3.) See Business Week's cover story, "Gas Pipeliners: Pric~d into a No-Growth Future, Massacre in the Marketplace", August 2, 1982, pp 44-47. 68. We understand, but have not verified, that current LNG plant siting and safety standards essentially preclude construction of any new LNG facilities on the east side of Cook Inlet. 69. Cook Inlet vs North Slope gas prices. The Acres and Battelle treatment· of North Slope gas is curious at best. For the purpose of determining Alaska economic activity and thus electricity demand, all scenarios assume that ANGTS will in fact be built (thus imparting an upward bias to population, gross state product, and electricity-load forecasts). When they are considering fuel supplies for electrical generation, however, the project is treated as doubtful. The Ac res Summary is typical; after suggesting that "Cook Inlet gas reserves may have been depleted in the early 1990s to the point that rellance upon natural gas Alaska Energy Planning 10/26/82 Page 61 as the principal fuel for electrical generation may no longer be possible," Acres states that The availability during the study period of North Slope gas near Fairbanks remains uncertain. Even if the gas pipeline is built, however, the gas price is projected to be higher than that for the Cook Inlet resource. (Acr.es Summary, p 9; emphasis added) · Acres and Battelle are not only inconsistent in their treatment of the ANGTS construction outlook, but show a double standard in their gas-pricing assumptions. They bpth use a "netback opportu_nity-cost" model to project Cook Inlet gas prices (which r~sult in very high price forecasts), but they reject that approach in forecasting Alask~ prices of North Slope gas (where it would have implied low gas prices for genera- ting .plants in Interior Alaska). In the latter case, Battelle obt<;lins a level 1982-dQllar price of $5.92 per million btu at Fairbanks by adding the pipeline-transportation cost to the statutory wellpead price fo.r Prudt:loe Bay gas, despite the fact that such a price .plus transportation charges beyond Fairbanks would surely m~ke the gas unmarketable if! the .Lower 48. (Indeed, a 1982-dollar price of $5.92 at the lower 48 "tajlgate of ANGTS would probably price North Slope gas out of the market.) Thus, the contrac- tors have assumed as a ce~ty that ANGTS will be built on schedule where this assumption imparts an upward bias to load forecasts, and thereby a bias in favor of Susitna, while they treat ANGTS as doubtful when it undermines the crucial assumption that the Railbelt will have run out of natural gas by the 1990's. For Cook Inlet and North Slope gas, they have used quite <;fifferent pricing models, and in each case they have also chosen the one that results in relatively high gas prices and which are, therefore, most favorable to the economics of the Susitna project. (Battelle, p 2.5.) 70. omitted 71. Acres, Task 11, Table 18.1.1; Acres, Summary, p '!6 (Plate 24). 72. Acres, Task 11, p 18-7. In fact, the gas prices used do not follow these scenarios, but are instead substantially higher. See Acres, Task 11, p 18-7. For even more confusion, see Table 18.2.3, whiCh implies a constant -gas price of $3.00 per Mcf through 1990, and then a sudden ·spurt at a compound rate of 4.7 percent annually. Alaska Energy Planning 10/26/82 Page 62 Cost-Overrun Risks 73. For a review of notorious cost overruns beginning in the 19th century, see Myron Kaplan, "Keeping Engineering Within Budget", in Tec hnology Review, January 1976. 74. Our review included: US Library of Congress, Congressional Research Service, "Cost Escalation in Selected Major Construction Projects," an unpublished s urvey for A R Tussing, chief economist, U S Senate Committee on Interior and Insular Affairs, May 22, 1975; M M Hut- schmidt and Jaques Ge'rin, "Systematic Errors in Cost Estimates for Public Investment Projects," in Julius Ma rgolis (ed), The Analysis of Public Output (New York: Columbia University Press, 1970); Robert H. Havemann, The Economic Performance of Public Investments: An Ex- Post Evaluation of Water Resources Investments (Baltimore: Johns Hopkins Universi ty Press for Resources for the Future In c ., 1972) (This volume is the source of the samaple of federal water projects cited by Acres); Walter J. Mead et al, Trans rtation o f Natural Gas from the Arctic (Washington: American Enterprise Institute, 1977, pp 8 -94; and E W Merrow et at, A Review of Cost Estimation in New Tec hnologies (Santa Monica C A: The Rand Corporation, 1979). 7 5. Misplaced specificity. The cost and schedule figures examined in an engineering-type risk analysis apply only to a specified projec t with a specified design. But almost every major construction project gets redesigned in major respects, both after the risk analysis is completed and before construction begins, and during construction. Thus, sponsors of the Trans Alaska oil pipeline (TAPS) can in good conscience deny that the difference between their original $900 million estimate and the final $9 billion construction cost was all, or even mostly, an "overrun". Likewise, the pipeline engineers who carried out the first risk analysis (which showed a 95-percent probability that the pipeline would come in at less than $1.4 billion), do not believe that their work was incompetent or misleading, because the pipeline that was ultimately built was a very different thing physically from what these engi neers originally had in mind. What would have been improper and misleading, would have been to place great weight on these estimates of costs and risk parameters, (1) in comparing the pipeline with the icebreaking-tanker tec hnology that Exxon was then testing, or (2) in assessing the economic feasibility of developing and producing the Prudhoe Bay reserves. To the credit of the TAPS sponsors, they never attempted to use their risk analysis in this way. Alaska Energy Pla nning 10/26/82 Page 63 The Susitna hydroelectric project has already gone through seve- ral conceptual changes and considerable escalation in real-cost esti- mates under the Corps of Engineers and Acres. The design examined in the present Acres Feasibility Study is surely not the . final design, we have been unable to determine from the current Acres reports how, and at what cost, the truly final design is expected to differ from the concept now under consideration. 76. Subjectivity and Overoptimism. Several kinds of subjectivity tend to flaw' engineering risk analyses as a source for point-estimates of expec-ted costs, or probability distributions· around those point-esti- mates. Firstly and utterly fundamentally, risk analysts build into their model~ only those contingencies that they can foresee. The really major risks are., well, surprises ••• Secondly, the probabilities assigned to most risk factors are franklY. arbitrary and subjective. While a few probability factors (stream-flow conditions, industrial ac<;;idents, etc.) may reflect physical or actuarial analysis, most of the individual figures that go into the calculations are blind guesses, and many of t.hem are inevitably inexpert and .uninformed guesses. Finally~ these subjective 'features of risk analyses leave them vulnerable to designer-and sponsor overoptimism. Only the project's own engineering design team is likely to know enough about it to carry out a competent risk analysis and, as we pointed out above, they often conduct similar analyses in the course of their own design work. Such teams are, however, almost always "believers" in the project analyzed and in its technology (be it hydro or nuclear power, space travel, or whatever); and their organization usually has a material interest in low cost and risk estimates. They are, therefore, among the least likely parties to anticipate all , unpleasant surprises, or to assign sufficient weight to those that they do identify. 77. Institutional blind spots. Engineering-type risk an~lyses almost invari- ably exclude risks arising from certain institutional causes. These include changes in sponsor or public perceptions regarding the need for the project; design error and mismanagement; changes. in laws or regullations; political controversies and lawsuits; changed capital-mar- ket conditions, etc. Together, these factors have played a large part in the cost-overruns, completion delay, and abandonment of "megapro- jects•• in recent years. Among these factors, the AGres report appears to consider only "contractor competence" and "regulatory delay". One famous instance of project failure is the Washington Public Power Supply System (WPPSS) nuclear-plant construction program. Its history included planning and management failures, design errors and Alaska Energy Planning 10/26/82 Page 64 frequent redesign, incompetent or dishonest contractors, regulatory delays, unforseen interest-rate increases, ratepayer protests that cul- minated in a voter initiative to restrict WPPSS' borrowing power, repudiation of purchase obligations by some utilities, and a flock of lawsuits, any one of which may prevent further bond sales. Very few (if any) of these factors were contemplated in the risk analyses that WPPSS presented to prospective lenders and to Washing- ton State's Energy Facilities Site Evaluation Council (EFSEC). None of them, however, was as important in creating the impending disaster than the collapse of the load-growth forecasts prepar~d for or by the Northwest's utilities. Year a~ter year, demand growth fell below the minimum values the utilities had forecast in previous years. (Is there something familiar here?) After billions of dollars had been spent on the WPPSS plants, it turned out that the power from them wasn't needed and couldn't be sold ---even if the system had been able to overcome all its other problems. Almost all of the generating projects in the United States that have r ecently been, or will soon be terminated or indefinitely delayed o r stretched out, have faced a combination of changing designs, esca- lating cost estimates, higher interest rates, regulatory problems, and falling demand forecasts. In most cases however, the last element was the most effective cause, as it undermined both the rationale for the projects, and the capacity of the utilities to finance them. 78. Underallowance for non-completion. The "expected final cost" figures generated by risk analyses almost invariably assume that the project in question will be completed. The Acres Susitna study is no exception, as Acres found the non-completion risks flowing from natural and cons- truction factors, regulatory problems, or cost-overruns that cause mid- construction abandonment, to be "negligible" whether taken in isolation or together. A true "expected-cost" figure for use in a benefit-cost analysis must incorporate the cost of "dry holes", however ---plants that are abandoned in mid-construction or which, even if nominally completed, cannot get an operating license or go on stream. Suppose the expected cost of a finished plant is 120 percent of budget, but that there is a 10 percent probability construction will be abandoned half way to the budget figure, a 5 percent chance it would be abandoned or not operate after the entire budget had been expended, and a 2 percent chance that it wou ld be a washout after 120 percent of budget had been spent. The expected cost of a unit of added capacity would then be 135 percent of the budget figure, rather than 120 percent. Alaska Energy Planning 10/26/82 Page 65 .· ~ ,, . t: . ,;. ~ i ... ., .-.. ~-··· 79. See the Appendix to Alaska Power Authority, Susitna Hydroelectric Project, "Commentary on 'Alaska Energy Planning Studies' (A R Tussing and G K Erickson)", prepared by Acres American Inc., September 7, 1982. Also, "Hydropower Broker Robert Byrd, Quebec's Master Ener- gizer of James Bay, in Engineering News ReCord, February 12, 1981. 80. The Tennessee Valley Authority's Tellico Dam is notorious because of the. -"snail-darter" case that held up filling of the reservoir for more then two years after the dam was completed. The reason this is the only example that comes to mind is that most of the major hydro sites in the US Lower 48 had already been developed before the plague of cost-overruns, litigation, and regulatory and economic troubles began to overwhelm the nuclear power industry. There are other recent' instances of controversies that precluded construction of hydropower projects, however ---the New River project in Virgina and North Carolina, for example. An interesting aspect of this case is tha~, by the time that the federal courts finally authorized operation of the pr:-oject, new studies by TV A showed that it would be uneconomic to install turbines and fill tl)e reservoir eV'en after the dam itseH had been completed. Tellico had, in the meantime, generated such a political constituency that Congress ordered the TVA to go ahead with it. Financing Issues 81. Acres Task ll, p 18-3/4 82. ibid. 83. Business Week, July 27, 1982. 84. October 14-15 yields for new utility bonds were 12~-14 percent; Aa industrials 12.1 percent; US treasury bonds (1994-99) 10·.1 percent, and the 20-bond Bond Buyer index of municipals 9~ percent. Long-term inflation expectations, however, had fall~n to the· 2-5-percent range. For a summary of current market expectations, see Kenneth H Bacon, "Price Risk; Fed, Forecasters Differ on Inflation Outlook; Firms Have to Gamble. Companies Counting on Rate Staying 5% May Suffer Reserve Sees Further Fall. The Trauma of Disinflation" in the October 14, 1982 Wall Street Journal. 85. Acres Summary, pp 24-25. Alaska Energy Planning 10/26/82 Page 66 REFLECTIONS ON THE END OF THE OPEC ERA * By Arion R. Tussing Institute of Social and Economic Research U ni versi ty of Alaska Copyright 1982 Oil Prices and Alaska's Economy Alaska's labor force, population, and personal income have been increasing faster than those of any other state for almost a decade. Very little of this growth would have happened without the two great surges in world oil prices that most people identify with the rise of OPEC ---the Organization of Petroleum Exporting Countries. The five-fold price increase for Middle Eastern c rude oil in 1973- 74, and the resulting three-fold increase in the market value of US domestic crude oil, made it economically feasible to develop the Prudhoe Bay field and complete the Trans-Alaska oil pipeline (TAPS). Even if real oil prices had remained at their 1975-78 levels, Prudhoe Bay oil royalties and taxes would have made Alaska the richest state government per capita throughout the 1980s, and the spending and respending of these revenues in Alaska would have combined with continuing private investment in energy resources to put Alaska among t he faster-growing states in income and employment. World oil prices took off once more in 1979-80, however. Real prices at the Persian Gulf "only" tripled this time, but the stability of transportation charges between the North Slope and Lower-48 refine- ries caused the price of Prudhoe Bay oil ---and consequently Alaska's royalty and severance-tax collections ---to grow almost five-fold. At their peak in 1981, oil revenues were flowing into the state treasury at an annual rate of more than $10,000 per capita. The effect of this bonanza on state spending for government operations, transfer pay- ments, loan programs, and public works was as awesome as it was predictable. In 1980 the Legislature abolished personal income taxes, and in 1982 it voted to distribute a "Permanent Fund dividend" check of $1,000 to each resident. By the second quarter of 1981, when c rude-oil prices began their present downward trend, most private and governmental planning in the state had come to reflect the assumption that oil prices and, with them, the prices of other fossil fuels would keep rising without limit. Disagreement about the long-term oil-price outlook was largely con- fined to the question whether price increases would average two, three, or five percent in excess of general inflation. Alaska Energy Planning 10/26/82 Page 67 . '"' ~-· The prospect of ever-ns1ng oil prices not only implied that Prudhoe Bay would generate more and more revenue for the state, and th~~ ,!J:le spending and respending of that revenue wou]d provide more and more public-and private sector jobs. It also promised to sustain a high level of exploration activity on state lands and the federal Outer Continental Shelf (OCS) and, most likely, a series of major new discov.eries. Rising oil prices convinced sponsors of the proposed Alas.ka gas pipeline that they could· just about ignore the difficulty of marke- ting North Slope gas as a constraint on the project's economic feasibi- lity'. An ever-increasing real price of oil on world markets seemed. to promise Alaska a petrochemicals boom . based on the growing cost- advantage North Slope natural-gas liquids (NGLs) would h~ve over oil- based petrochemical feedstocks used elsewhere; it promised _develop- ment of Alaska's coal for export, as well. Higher and higher prices for oil meant, moreover, that natural gas and coal would be too valuable to use for generating electric power in Alaska; this outlook has been the main rationale for planning a multi-billion dollar hydroelectric genera- ting plant on the Susltna River, and led many Legislators to believe that the state would be able to finance the Susitna project with a direct appropriation from the General Fund. The price of oil is thus the biggest single outside influence on Alaska's economy, and in 1982, tmeerta.i.nty about the oil-pri~e outlook has become the biggest single source of uncertainty about the state's economic future. The possibility that the oil-price boom may be over has profound consequences, which it has thus far been easier to ignore in Alaska than in other energy-exporting areas. Since the upward movement of oil prices ended in 1981, economic distress has already overtaken many of the top oil-expo,rtlng countries. By the end of 1982, all but two OPEC nations are likely to be in a deficit fiscal and foreign-exchange position. In less than a year, a record boom has given way to grave depression in the deep-gas areas of Oklahoma, among oilfield-service contractors, and for the financial institutions that specialized in backing them. All but a handful of synthetic-fuels projects have now been abandoned, and the future of those is doubtful. Several forces join to to perpetuate a petroleum-driven boom longer in Alaska than in other petroleum-exporting regions. Most crucial has been the fact that the actual spending of state money appropriated in 1980 and 1981 has continued to increa~e month by month well into 1982. Another factor has been the momentum of North Slope oil exploration and development programs, some of which would be viable at any world oil-price level higher than (say) $10 per barrel . A third element in the present boom is a host of Anchorage commercial Alaska Energy Planning 10/26/82 Page 68 building projects for which the key decisions and commitments were made before 1981. Finally, the continuing increase in the amount of state money actually flowing into the ecol')omy has combined with the flow of oil-industry investments to perpetuate a mood of high optimism among private investors in Alaska's non-energy industries, particularly residential and commercial real-estate development. The 19Sl Turnirig Point This -paper is intended to put the OPEC oil-price rises of the 1970s and the more recent oil-price decline into historkal perspective. Alaskans can then weigh the chance that energy prices will resume their upward climb, giving new momentum to economic growth in the state, against the prospect that the OPEC era is indeed over --- meaning that Alaska, along with other energy-exporting regions, must accept a radical and mostly downward adjustment in its development expectations. At the end of 1981, oil prices were suffering their greatest decline in half a century, but many industry executives and forecasters ---even outside Alaska ---cautioned that the attendant "glut" was a temporary phenomenon. Despite many signals to the contrary, that opinion has persisted well into 1982, with Occidental's Chairman Armand Hammer and others predicting $100-per-barrel oil in 10 years. The scarcity mentality that fed the price leaps of 1974-and 1979, plus a belief that unpredictable Middle Easterners control the world oil market through OPEC gave suer: assumptions nearly axiomatic status during the past decade. But the exporting nations' boast that "oil in the ground is a better investment than money in the bank" is turning out to be so much wishful thinking. The truth is that neither an end to the recession, nor OPEC attempts at production quotas, nor continued wars in the Middle East will long be able to shore up a sagging crude-oil market. It is, indeed, because oil prices climbed so rapidly and so high in the 1970's that they are now almost certain to fall and keep falling ---perhaps as steeply and as far as they rose. T oday's prices are still higher than markets can tolerate, and the forces that led to the enormous price hikes of the past decade work just as· effectively in reverse. While real (constant-dollar) crude-oil prices are unlikely to rise again to their 1980-81 levels any time in this century, forecasting prices for any specific future year is a nearly hopeless task. The fall in crude-oil prices that began last year will some day give way to another price rise, another decline, and so on. For crude-oil markets are inherently cyclical and, except during a unique period of almost four decades when the State of Texas dominated both U.S. and world crude- oil markets, oil-price fluctuations have been large and frequent. His- Alaska Energy Planning 10/26/82 Page 69 ·,, tory shOws us no long-term oil-price trendS, but only a series of cycles of uneven duration and amplitude. The era of OPEC's opportunistic price-gouging is over, but no other entity is in sight with the power to move oil prices in any consistent direction or to stabilize them at any given level. Market Control by the Texas Railroad Commission (1935-i972).2 To understand OPEC's helplessness in today's .crude oil-market, it is useful to review how the market operated before OPEC came ~o power, and how the Texas Railroad Commission (TRC) managed to exercise control for nearly forty years. The TRC's rule emerged in the mid-1930's from circumstances quite different from those that nurtured OPEC in the 1960's and 1970's. In the era between 1859, the year Colonel Drake first discover.ed oil in Pennsylvania, and the Great Depression, crude-oil markets everywhere were dominated by events in the United States, where one black-gold rush after another unleashed an oversupply and sent prices plummeting. Growing oil demand rapidly restored prices after most of these crises, as petroleum captured markets that were previously ~eld by whale oil, gas, or coal, and as the automobile population swelled. . The Yates field in Texas, for example, was first tapped in 1926. It was the biggest field yet found, and over its first year of production, average crude-oil prices in the United States fell 24 percent. Prices recovered quickly, but in 1930, th~ beginning of the Depressi.on coinci- ded with discovery of the even larger East Texas field. Oil literally ran in the creeks, and prices fell to 10 cents per barrel. Much of this market chaos resulted from the common-law "rule of capture". The principle that nobody owned oil until it was brought to the su.rface generated frenzied competition among drillers . to lift as much oil as they could from each newly-discovered pool before their neighbors got it. The East Texas drilling rush ended in 1931 only when the governor sent . the National Guard into the field to shut down production. The next year, a bitterly divided Texas legislature granted :the TRC authority to limit output from individual wells in the interest of conservation and market order. Under "market-demand proration- ing", refiners told the TRC how much oil they wanted to buy each month, and the Commission parceled out the "allowable" share of this demand to each well. This system assured every Texas producer _(l buyer fqr at least some of his oil, no matter how much excess producing capacity other producers held. The TRC's ability to stabilize the rnar.ket was bolstered by market-demand prorationing in several other st~tes including Louisiana, the number.:.two US producer. Under state regulation, physical shortages and surpluses both became a thing of the Alaska Energy Planning 10/26/82 Page 70 past, "conservation" replaced physical waste, and the violent short-term fluctuations of crude-oil prices ended. Even more important were a series of federal actions backstop- ping TRC authority. In the 1935 Connally "hot-oil" law, Congress made it a federal crime to ship oil produced in violation of state conservation orders. After the War, the executive branch acted to prevent uncon- trolled imports of low-cost foreign c;ude oil from undermining the states' control of US oil supplies. For a while, the handful of largely US-based oil companies :that controlled the oil reserves of the Middle East and the Caribbean had cooperated successfully in limiting petro- leum production from their foreign concessions to just about the amount demanded by their own foreign refineries. Nevertheless, by 19lt8, the huge low-cost oil reserves overseas had become enough of a threat to Texas regulation that the Truman administration started assigning "voluntary" import quotas to the com- panies. In 19 58, after independents like Hunt and Occidental developed enormous new reserves in Libya, President Eisenhower established a mandatory oil-import program (MOIP). The MOIP gave each US refiner the right to import some lower-priced foreign oil, but it enabled the TRC and other state conservation authorities to continlf setting the total volume of crude oil suppliP.d to the domestic market. Critics of market-demand prorationing and import quotas, inclu- ding the 1970 Cabinet Task Force On Oil Import Control headed by George Shultz, saw the. combination mainly as an arrangement that kept US prices artificially high and perpetuated wasteful excess capacity. It did, indeed, shut.in much of the nation's lowest-cost oilat the same time that it created an incentive to d~elop domestic fields that would not have been viable in a free market. But precisely because it sheltered surplus producing capacity, Texas prorationing functioned as the ba- lance wheel of the world oil market for four decades! Texas was the "price-maker": The TRC determined the price all domestic producers received for their oil and it was at the same time the most powerful single force in the world crude-oil market. Other producers were relegated to the passive role of "price-takers" who could always sell as much or as little oil as they wished, once they accepted the price structure established by Texas. The TRC could play the price-maker's role because it had control over sufficient spare producing capacity to supply refiners with aJl the Texas oil they wanted at the established price, offsetting any rise in demand prompted by boom conditions or long-term economic growth, or any drop in output by other producers at home or abroad. The Commission also had the power to enforce production cuts if necessary to prevent a surplus from appearing when oil demand waned or produc- tion outside Texas increased. Alaska Energy Planning 10/26/82 Page 71 The Texas system worked for nearly 40 years, througlh World War II, the US recessions of the 19 50s, and several supply disruptions caused by .Middle Eastern political upheavals in the 19 50s and '60s. Oil supplies were seriously curtailed in 19 52-54, for example, following Mossadeq's nationalization of the Iranian oil concessions, during the Suez crisis of 1956, and again during the 1967 Yom Kippur War. US and world crude- oil prices remained relatively stable, however, as the majors produced more oil in the unaffected Persian Gulf countries, whi~ the TRC and other state commissions increased production at home. From 19 35 to 1972, roughly the period of the TRC's domination of the world market, the average annual change in real crude-oil prices in the United States was only 4 percent, plus or minus, a stark contrast to the average annual change of 21 percent between 1871 and 19 35. One reason for the TRC's success was that it did not exploit its market power opportunistically. The 19 52, 19 56, and 1967 Middle Eastern conflicts offered Texas producers (and the state of Texas, a major royalty owner) a chance for huge short-run profits, as they did the big international oil companies. But each time, the TRC and the majors opted for long-term stability, forestalling the kind of consumer panic that generated the price run-ups of 1974 and 1979 after the TRC had lost control. Once domestic production reached full capacity in 1972, the US government had no choice (politically, at least) other than to do away with import controls, leaving consumers exposed to whatever upheavals might occur ln the oil-exporting countries. Meanwhile, nationalization of the major oil companies' overseas concessions, plus the growing influence of independents (including national oil companies like those of France, Italy, and Brazil), had stripped the majors of their ability to balance supply and demand outside of North America. A supply curtailment by the Arab oil producers, which would have hardly caused a ripple in oil prices ten years or even two years earlier, transformed world energy markets and, for a few years at least, handed control of those markets to OPEC. Panic Pricing in 197.3-74 and 1979. OPEC's spectacular successes in the 1970's were due more to market psychology than to anyone's direct manipulation of crude-oil supplies. OPEC per se did not engineer either of the decade~s great price leaps; they came instead out of consumer panics that spread through the spot market after the 1973-74 Arab embargo and the 1978 Iranian revolution. In both cases, OPEC merely voted, after the panic had run its course, to establish the prevailing spot prices as the base prices for all crude-oil sales. Alaska Energy Planning 10/26/82 Page 72 Crisis psychology was thus the key to the short-term oil market behavior that ratcheted prices upwards. The physical shortfall that provoked the panic of 1973-1974 was proportionally no greater than the shortfalls (or excess supplies) that the industry faced periodically because of unusual weather or the business cycle, and there was no reduction in output at all immediately before the 1979 pric e spiral. In neither case was the actual shortage greater than the sum of (1) the oil then being consumed by electrical-generating and manufacturing plants that had the capacity to use other fuels, (2) the standby or underutilized oil-producing capacity of US and uninvolved foreign producers, and (3) the inventory c ushions that industry ordinarily would have drawn down in order to prevent market turmoil. The price fly-ups, rather, began both times with a handful of large buyers who believed that the "shortage" was real, and who were thus willing to pay almost anything. This crisis mentality had a powerfully perverse effect on the market: instead of restraining demand, soaring spot prices gave the shortage credibility and helped propagate the panic to every class of cons umer, so that demand actually increased. Much of the apparent supply deficiency was caused by hoarding, the most memo- rable example of which was fashion of topping-off gasoline tanks daily in private automobiles. This practice alone probably c reated demand for an additional 600 million barrels of gasollnefn the United States --- equal to about three months of refinery output. However socially irrational that hoarding may have been, it seemed quite reasonable at the time from an individual company or consumer standpoint. In 1973-74, both the Arab producers and the Western media were insisting that the embargo and production cutbacks were in fact harming the consuming countries. Congress had passed an Emergency Petroleum Allocation Act; President Nixon had declared an "energy emergency" and had begun allocating crude-oil and petroleum produc ts. No one knew how long the apparent shortage would last or how high prices might go before it was over; hence it made sense for anyone with a preferred position in the allocation scheme to take every drop of price-controlled gasoline or fuel oil allowed under the rules, regardless of current need, and for every other consumer to buy as much at the prevailing price as he could store. Motorists, households, and businesses all sought to build up and maintain high inventories in c ase things "really got bad" later, while producers, refiners, and others expected to profit from holding products for resale at higher prices in the future. All of these anticipations of course validated themselves: supplies did get tighter and prices continued to rise. Alaska Energy Planning 10/26/82 Page 73 1Jle Role of Spot Prices.8 Oil producers and refiners usually try to plan their physical operations and to budget their purchase outlays and revenues well in advance. For this reason, the great bulk of the world's crude oil moves in "captive" channels from producing companies to their own refinery affiliates or on relatively long-term contracts between producers .and refiners. "Spot" transactions ---sales of a single tanker-load or less ---usually account for only a few percent of world supply, but they are an indispensable part of the total market because they allow any company or government to dispose of a temporary oversupply or fill a temporary shortfall. A general surplus or shortage equal to only (say) three percent of total world demand may .thus show up as a surplus or shortage amounting to fifty or one-hundred percent of normal spot- market demand. As a result, spot prices tend to fluctuate daify and seasonally, and to range widely above and below "posted" or contract price levels, which typically change slowly and infrequently. Changes in crude-oil spot prices occasionally herald deep-seated market changes, but more often they are only exaggerated reflections of unexpected weather or business conditions, the buildup or drawdown of inventories, or political events. After some such contingency has caused spot prices to diverge sharply from contract prices, the spot market normally returns to a relatively narrow band of prices in the vicinity of previous contract-price levels. What was special about the OPEC-dominated markets of the the 1970's is that they twice failed to respond in this normal way. After the panics of 1973-74 and 1979, spot prices did not fall back to pre-crisis levels; instead, contract prices rose ---by OPEC decree ---to the peak values to which the panic had carried spot prices. This feat was OPEC's great triumph which, ironically, is now begetting its downfall. The Power of Saudi Arabia and OPEC. The TRC determined prices by actively manipulating the aggre- gate supply of crude oil; as a state agency, it had the power to enforce its orders on the many thousands of Texas producers regardless of their conflicting individual interests and viewpoints. OPEC, on the other hand, has never had any authority over the diverse and sometimes warring sovereignties that make up its membership. Nevertheless, in the early 1970's, once the surplus capacity of Texas and Louisiana had disappeared, Saudi Arabia by itself conceivably could have taken over the TRC's balance-wheel role. Its potential authority came from a combination of the world's largest reserves of conventional crude oil with a population on the order of only five million. The Saudis have thus had the same mix of assets, at least Alaska Energy Planning 10/26/82 Page 74 theoretically, that earlier gave the TRC its power ---the ability to increase or decrease output over a wide range. Proved and indicated reserves in Saudi Arabia number in the hundreds of billions of barrels ---how many hundreds, no one knows or much cares, because it has never been worthwhile to carry out the intensivefxploration and development work needed to get an accurate estimate. The known reserves are, in any event, so large and so easy to develop that it would take Saudi Arabia only a few months to double its exports from the current level of less than 6 million barrels per day. (After all, production was almost 11 million barrels per day only one year ago.) With two or three years for drllling of development wells and construction of gathering lines and terminals, exports probably could rise to something like 18 million barrels per day. Indeed, before the 1973 embargo, the big oil companies (the Aramco shareholders) that controlled the Saudi concession were planning for production on the order of 20 million barrels per day by 1976. Saudi market power rests on the ability to c urtail as well as to increase production. The countrys small population has permitted Saudi Arabia to reduce its output by almost half over the last year, from 10.6 million barrels per day in August 1981 to about 5.5 million in May 1982, without suffering a fiscal or foreign-exc hange crisis. Throughout the 1970's, therefore, Saudi Arabia, with or without the cooperation of other OPEC nations, had much of the wherewithal to stabilize the market in just the same way as the Texas Railroad Commission once did . The OPEC Mystjque. It was a worldwide obsession with scarcity, rather than deliberate management of total world supplies, that underpinned the OPEC mystique and locked in the high prices OPEC decreed in 1974 ~nd 1979 after the direct causes of panic had vanished. The doctrine of imminent resource exhaustion was embraced in the 1970's by a broad spectrum of parties who had e ntirely different world-views and diffe- rent ends. ' Environmentalists hoped to slow the wasteful plundering of the earth's riches; oil companies were seeking to ward off price controls and attacks on their tax preferences; alternative-energy entrepreneurs sought business; politicians found in the energy crisis a moral equivalent of war; civil servants made it the rationale for massive expansion of their agencies and intervention into almost everything; and an army of academics, consultant s and journalists waxed r ich and famous by studying, interpreting, or .advocating national energy policies. Each group wanted to believe, or at least to persuade others, that "the wolf is really here" ---that OPEC's prices might have risen too abruptly for Alaska Energy Planning 10/26/82 Page 75 comfort but that, in the last analysis, those prices only e xpressed the dictates of geological necessity. ''Oil in the Gromd is Better than Money in the Bank." In this intellectup.l climate, each price ~ncrease, regardless of its proximate cause, helped convince the oH-exporting nations that "oil in the ground is a better investment than money in the bank." This doctrine could remain valid for just as long as most producers believed in it, because it encouraged them to hold oil off the market in the faith that its value would be much higher in the future. It was ·therefore t~5 most effective and durable weapon in OPEC's ideological arsenal. Although the organization had no enforcement machinery, and did not even attempt a prorationing scheme until 1982, its members did reduce production when preservation of the price gains of 1974 and 1979 required it. When the OPEC nations cut back their total exports in 1974 and again in 1979, it is important to note that they· made the required cuts individually. They did so without coordination or urging by OPEC, because they had more money than they needed at the moment, and because they believed that their oil would be worth more later. Even the 1973-74 price rise had been so immense that most OPEC countries ha~ subs tantial financial surpluses; only Algeria, Ecuador and Indonesia were in deficit. In 1975 OPEC as a whole had a $59-billion (or 14- percent) surplus of export revenues over import expenditures. Several countries understandably concluded that their economies couldn't ab- sorb further increases in oil income without generating intolerable inflation and social unrest. It also seemed obvious to the producers that oil prices would continue to advance at a higher rate than their surplus cash would yield in risk-free financial instruments. Thus, Kuwait, Libya, and Venezuela together reduced their exports by 4 million barrels per day after the 1974 price rises. Saudi Arabia abandoned Aramco's 20 million barrel- per-day target in 1974, and cut production sharply in January 1979 and then again in April ---ostensibly to offset an imminent oil glut, which was in fact an aftermath of the price panic that followed the overthrow of the Shah. In neither price rise did O.PEC as such haye any role in initiating or orchestrating the curtailments. Thus, a short-lived belief in acute scarcity twice created a .real scarcity that caused spot prices to soar. A belief that chronic energy shortages would engender a permanent seller's market then led produ- cers and consumers alike to interpret an unusual and otherwise transi- tory market phenomenon as obedience to holy writ. The self-fulfilling doctrine that oil in the ground was the world's best investment not only encouraged OPEC officially to adopt spot market prices generated by Alaska Energy Planning 10/26/82 Page 76 consumer panic; it also enabled those prices to stick. In 1973-74J the world price of crude oil (measured at the Persian Gulf) increased five- fold in real terms, and then in 1979-80 it again tripled (a vivid contrast to the decades of tranquility under the TRC). The End of the OPEC Era. OPEC's hold over world energy markets in the 1970's was no less real because lt was mainly psychological. However, the cartel's mystique is far more fragile than the earlier market power of Texas, which stemmed from the TRC's direct control over production volumes. Today, few of the material requisites for further OPEC success remain. Its share of the world oil market has fallen from 55 percent in 1974 to 31 percent in August 1982; and Saudi Arabia's share is already less than the share Texas held as late as the mid-1960's • Some recognition of these shifting realities began to strike the Saudi leadership only after two deliberate production cuts in 1979 had locked in a series of huge price inc reases voted by OPEC. Saudi Arabia's actions have now become more-or-less consistent with the professions of the kingdom's oil minister, Sheikh Yamani, who had long given lip-service to the cause of moderation and market order. Expli- citly invoking the memory of the TRC, Yamani claims to have engineered the 1980-81 "oil glut" ---increasing production from less than 8 million barrels per day to almost 11 million, specifically in order to bring prices down to $34 per barrel and to persuade his OPEC partners they should join a prorationing scheme under Saudi leadership. In 1982, after succeeding too well, perhaps, the Saudis have abruptly reversed course, now cutting their exports by nearly half in an attempt to support the $34 price. But Saudi Arabia appears to have been too late in discovering the market power it alone possessed. While the TRC held c rude-oil prices in the United States above short-term free-market levels, it still kept them low enough to encourage ever-increasing oll consumption and stave off the development of alternative energy sources. The Saudis, however, wittingly or unwittingly had a key role in both OPEC price hikes of the 1970's, unleashing inexorable and profound reactions from both producers and consumers, which today threatens to make OPEC oil a dispensable commodity. Contrary to a near-consensus of industry, government, and the academic/consulting community during the 1970's, crude-oil demand does respond ---s lowly but massively ---to price changes. In the long run, higher prices have a profound effect on oil supply too, but the relationship is too complex to pursue in detail here. In any event, non- OPEC output has grown rapidly and will continue to grow: Production from the North Sea, Alaska, and Mexico, for example, increased by 4 Alaska Energy Planning 10/26/82 Page 77 million barrels per day between 1977 and 'early 1982, and Mexico's .exports ---driven now by economic necessity ---could increase another 3, 5, or more millions of barrels per day before 1990. Most clearly and most importantly, however, high oil prices ·are shrinking oil demand both by reducing total energy consumption and by making coal, natural gas, nuclear power, and other energy sources more attractive. The Flight from Oil. After a modest dip in 1974, total world oil consumption resumed its growth, and finally turned down only in 1979. This experience reinforced the impression that oil demand was insensitive to price changes, misleading economists as well as industry executives ahd government officials in both oil-producing and oil-consuming countries . An absolute decline in US oil consumption was first visible in 1979; the rest of the industrialized world followed a year later. In retrospective, it is remarkable how many were unable to see what was happening. Exxon in 1977 forecast that US consumption of petroleum liquids would be 20.3 million barrels per day"in 1980. In 1979, Shell predicted 18.6 million barrels per day consumption in 1980, and both the Oil and Gas Journal and the Independent Petroleum Association of America forecast 18.4 million. As late as mid-1980, Shell had only revised its published estimate downwards to 17.2 million barrels per day, while the Independent Petroleum Association of America had come down to 17.4. At year-end in 1980, however, average consumption for the year stood at only 16.3 million barrels per day. The drop in total oil use over the last three years and the experts' tardiness in recognizing the trend of consumption stem from profound changes in the structure of world energy demand, which have actually been under way since 1974. From 1960 to 1973, oil prices were Iow and declining in real terms. As a result, absolute oil consump.tion in the industrialized countries grew at an annual rate of 7.6 percent. Japan led this growth with an 18-percent average over the thirteen-year period. After 1974, however, the quadrupled crude-oil price led to a gradual leveling off of demand for oil everywhere. Total oil consump- tion in the industrialized countries dropped slightly in 1974-75; growth resumed between 1975 and 1979 at an annual rate of about f percent, but this partial recovery only concealed the fundamental shift that had taken place in the world's energy-use patterns. More telling than gross consumption figures is the change in oil use. per unit of economic activity, or "gross domestic product" ---the oil/GOP ratio. After rising at an annual rate of 1.3 percent from 1960 to 1973, the oil/GOP ratio for the major industrialized countries showed a 1.5-percent annual decline between 1973 and 1979. The 1979 Alaska Energy Planning 10/26/82 Pag~ 78 upheaval initiated an even more decisive and long-lasting shift away from oil, reflecting both an immediate reaction to the second OPEC price surge and the delayed but cumulating response to the price increases of the early 1970's. From 1979 to 1981, oil consumption in the industrialized countries fell 7 percent per year, and the oil/GOP ratio fell at an annual rate of 8 percent. Since the l atter measure represents the amount of oil used per ...Ut of economic activity rather than an absolute figure, its fall implies that an end to the present recession will not be the panacea that much of the energy industry and many analysts still seem to anticipate. The die has been cast. It is unlikely that individual homeowners will tear the i nsulation out of their houses if the price of home heating oil drops, or scrap their new fuel-efficient automobiles in response to lower real prices for gasoline. Nor will the construction and automobile industries abandon their new energy-efficient designs. Those who attribute the oil glut and the current "softness" of oil markets primarily to the world r ecession forget that economic recovery will mean a more rapid replacement of existing vehicles, industrial machinery, and buildings with models designed since 1974 in response to high energy prices. Except in a couple of OPEC countries, no new base-load electrical generating plants fired by oil, or large-scale oil-fired boilers of any sort, have been built since the mid-1970's; over the past decade, industry has been relentlessly converting existing oil-burning equipment to coal, natural gas, and other energy sources. Because changes in the world's fuel-use patterns are generally embodied in long-lived capital- intensive investments such as buildings, transportation equipment, and industrial machinery, the extended period it has taken for the 1974 price rises to produce an absolute decline in oil consumption only reflect the time reql;lired to replace these assets. This long lag in adjusting the world's capital stock to changed ener:-gy-supply conditions also suggests that the all-time high oil prices of 1974-1982 will influence consumption patterns for many more years, even if world oil prices now fall as rapidly and as far as they rose in the 1'70's. The truism that the world's petroleum resource is finite thus does not mean that oil prices will continue to rise. The world has no demand for crude oil as such, but only for the heat, motive power, and chemical building bl ocks it provides, and only for so long as it is the cheapest source of these goods. No matter how scarce natural petroleum liquids become, their prices can not rise and remain above the cost at which each of these wants can be dispensed with or served in some other way. It should be fairly obvious now that predictions of $100 per barrel oil are ludicrous. At $15 per barrel, oil was already more expensive than coal in most of the world, and had consequently priced itself out of electrical-generation and other large-scale stationary-heat and boiler- Alaska Energy Planning 10/26/82 Page 79 . ·~ . . . ' . fuel markets. At well under $50 per b_arrel, given a few years for market and infrastructure development, liqu_id petroleum products wou.ld have become marginal even as transportation fuels, increasingly reptaced by some combination of compress.ed and lique ~ied hydrocarbon gases and alcohols. A world that is already fleeing from oil at $32 per barrel would hardly have any use for it at two or three times that price. This dynamic does not bode well for OPEC, or for the ability of Saud~ Arabia or anyone else to manipulate· or st.abilize the market. On.iy when'·ctemand faHs is the power of a price-maker tested. Can the OPEC nations, many of whom are deeply in debt, now afford to cut back production as they must? On this point too, the OPEC of 1982 is as different from its predecessor, the Texas Railroad Commission, as it is from the OPEC of the mid-1970's. In comparing OPEC with the TRC, it is essential to remem~er that the Commission's power developed during the Depres- sion, and that its institutions were designed expressly to manage a chronic excess of producing capacity. Once that excess was gone, the TRC became impotent and economically irrelevant. OPEC, in contrast, showed it~ muscle · under totally opposite conditions. It twice seized upon a brief moment of consumer panic; convinced itself and consumers alike that a permanent world oil scarcity existed, and for a while reaped the benefits of a seller's market even after the ·foundations of that market had vanished. There is little prospect that OPEC can function effectively in a chronic buyer's market, especially in the face of the organization's current internal dissensions and the precarious financial situation of its members. At its March 1982 meeting, the group made its f irst serious attempts at TRC-style prorationing. The experiment was an instant failure, with at leeif2 three members brazenly exceeding their quotas from the beginning. By July 1982, Ira!l was selling 1.0 million barrels per day above its quota, Nigeria .3 million, and Libya .25 million. Venezuela ---the sole advocate of OPEC prorationing before 1980 --- had threatened to start selling more than its assigned 1.5 million barrels per day if the other countries didn't get in line, and in August, the S~udis, ·who had already reduced their output by 45 percent h1 the hope . of supporting the $34 marker price, were also hi~_!ing that they would increase their exports if the cheating didn't stop. Even holding the line at today's production level is not enough to bolster OPEC's flagging power, as world oil consumption continues to shrink and the production of non-members ---especially Mexico --- continues to grow. The organization as a whole must somehow manage to reduce production even further if present prices are to hold. Yet its member-nations individually face internal problems and pressur es that urge them in just the opposite d.irection. Alaska Energy Planning 10/26/1,2 Page SO The greatest source of downward pressure on prices is the shaky financial condition of the exporting countries, a drastic turnaround from the situation of 197 5. Since 1973, the OPEC nations' spending for imports has risen at an average annual rate of 30 percent, because of ambitious industrialization plans in every one of them and extravagant purchases of military hardware in many. Already, the combination of declining oil demand and rapidly rising expenditures has resulted in trade deficits for all but three OPEC members. Unless oil production or prices increase sharply, every member, including Saudi Arabia, could slip into deficit by the end of 1982. These deficits, exacer bated by the continuing Iran-Iraqi war, are already beginning to take their toll as the most hard-pressed countries, in search of revenues to pay for today's imports, produce as much oil as . they can sell. Moreover, with declining or even stable prices and real (inflation-adjusted) interest rates at their highest level in history, the slogan that oil in the ground is a better investment than money in the bank is obsolete even for countries that don't have an immediate cash- flow or foreign-exchange deficit. In the 1980's, it is hard for even a cash-surplus oil-exporter to avoid recognizing that "oil in the ground is a non-earning asset", which ought to be cashed out so the proceeds can be invested in high-yielding financial instruments. This doctrine is just as true and may be just as self-fulfilling today as was the opposite notion in 1975 or 1979. To put OPEC's weakness into further perspective, consider the following: *In August 1982, world crude-oil production was about 54 million barrels per day. Out of this total, the Saudi share was about 5.5 million, or 10 percent; all of OPEC was producing about 17 million barrels per day, or 31 percent of world supply. If new production in non-OPEC countries plus further declines in con- sumption were to equal only 10 percent of present world demand, OPEC's members would have to reduce their own production by 32 percent in order to defend any chosen price level. Saudi Arabia, which has already reduced its exports by 4-5 percent over the last year, can not and will not acc ommodate muc h of this burden, as a 10 percent shift in world supply or demand would be just about equal to the country's current export volume. Further growth in non-OPEC production and a further fall in world consumption are not only plausible but nearly inevitable. Thus Saudi Arabia's reign as world price-maker is ending virtually as soon as it began. *Conservation, fuel-switching and recession caused the non- c ommunist world's ..>il consumption to fall by 7.5 million barrels Alaska Energy Planning 10/26/82 Page 81 • per day between 1979 and mid-1982. If consumption declined by only half as much over the next two years, OPEC's output would have to fall by ali amount equal to the combined production of Kuwait, Libya, Algeria and · Indonesia in August 1982, or by 68 percent of current Saudi output. · · *Crude oil production from Alaska, Mexico and the North Sea increased by more than 4 million barrels per day between 1977 and 1981. If all non-OPEC producers were to increase their output by another 4 million, OPEC could maintain control of prices only if its members could cut production by at least the equivalent of 73 percent of August 1982 Saudi Arabian production. *Production from Iran, the world's former number-two oil exporter, has fallen 4 million barrels per day from its 1974 peak. The former number three exporter, Iraq, is producing 2.6 million barrels per day less than the peak it reached in 1978. If the war between these countries should end and they returned to the market with their 1978 sales volumes, other OPEC countries would have to curtail production by an amount equal to 90 percent of August 19 82 Saudi output. *Finally, if by chance the last three developments all took place, and if OPEC hoped to sustain world prices at current levels, it would have to find places to cut production by least 12.7 million barrels per day ---75 percent of the organization's current output, or 231 percent of August 1982 Saudi production. The range of conditions within which OPEC, Saudi Ar:-abia, or anyone can continue to dictate or even defend the level of world oil prices is thus extremely narrow. The reckless opportunism that held sway in the 1970's is now taking its toll. Long-term changes in supply and demand adverse to -OPEC's interests have been under way ever slnce the cartel's first big coup in 1974. As these changes become visible to everybody, the mystique that has been OPEC's chief source of power will vanish along with forecasts of hundred-dollar oil. The world market will soon be, if it is not already, out of anyone's control. What Have We Learned? A big new disturbance in world oil markets could push prices either up or down. It is still conceivable, if only barely so, that a sharp economic upturn and an exceptionally cold winter could combine with tbe right kind of Middle Eastern political crisis, and send prices soaring for a third time to levels significantly above those reached in 1980-81. The probabilities, however, weigh heavily on the other side. There is a huge overhang of excess producing capacity in the oil- exporting countries. Several of them are in big fiscal distress; Mexico Alaska Energy Planning 10/26/82 Page 82 in particular has both the ability and a desperate need to increase oil exports. Meanwhile, the price-induced flight from oil is still gathering momentum that will not be spent for years no matter what happens to oil prices today. AU of these forces together, not to mention a worldwide econo- mic slump that is far from over, add up to irresistible downward pressure on oil prices. Prices must eventually go down, and they must go down substantially. The serious questions are whether they will descend smoothly or chaotically, and how deep they will go. There is still a sliver of a chance that prices could firm for weeks or months, or even ---given the unlikely coincidence of events decribed above --- increase once more. ·aut a market collapse this year or next has a far bigger likelihood, a collapse every bit as spectacular as the two price eruptions of the 19 70s. Looking back across the years of OPEC and Energy Crises to the relative tranquillity of the TRC era and beyond, there are a several lessons for the future. 1. Worldwide scarcity and rising real resource costs had little or no direct res · ill f or the worldwide energy-price uphea of the 1970s. The earths's lmown resources still include plenty of crude oil that could be developed and produced at resource costs (capital, material, and labor costs) well below 1973 real prices. Considering these resources alone, there is enough low-cost oil left to satisfy the world's current rate of consumption for several decades. 2. In the absence of an effective price-maker like the Texas Railroad Commission, crude-oil markets are inherent- ly cydical. Oil demand is highly responsive ("elastic") to price changes, but this response is very slow, because fuel-use patterns are embodied in capital goods whose turnover is measured in decades: buildings, trans- portation equipment, industrial machinery, and production processes. For the same reason, demand is exceedingly inelastic to price changes in the short run. This contrast between short-and long-term price- responsiveness inevitably fosters cyclical price behavior. In the 1970s short-term price-inelasticity spawned a steep cyclical upswing ~.fter years of artifically-maintained stability, and in 1981, a high long-term price-elasticity finally began to show itself in the beginning of a downswing. If there is no "surge-tank" or "damping" mechanism comparable to market-demand prorationing, moreover--- Alaska Energy Planning 10/26/82 Page 83 < .: ~ ... 3. Market structure and psychology can exaggerate an .. episodic oil-price fluctuation, up or down, far out of propor- tion to the original supply-demand imbalance that trigger~ it. Inventory accumulation or liquidation, the financial position of major producers, and consumer panic can all cause markets to behave perversely over a "short-run" that can last for several years. In a mockery of the "normal'~ supply-demand map, an oil-price rise can for a while create an incentive to build inventories, and sustained price rises can encourage the withholding of production. A price reduction, likewise, can provoke liquidation of inventories and the expansion of output. As a result --- 4. A small excess of demand or supply, real or imagined, can send the market soaring or plummeting far beyond the price level that ultimately could have brought it back into balance. Thus, there is no stable equilibrium toward which an unregulated petroleum market unfailingly "hunts" once it is disturbed. The uphea- vals of the 1970s, which carried prices well above any level that could be sustained in the long run, have now set the stage for a descent far below the range of sustainable prices. 5. No cartel or regulatory system could have held world oil prices at the low levels of the early 1970s, and none . can do so in the future. Before 1973, state regulators in the United States and the cartel of international companies maintained prices that were above the shortest-term "market-clearing" levels, but which were still so low that oil totally dominated transportation-fuel markets (even capturing rail- roads that had earlier been powered by electricity generated from coal) and, except in a small corner of the United States, virtually swept coal from the world's markets for industrial boiler fuels and organic- chemical feedstocks. These prices were, at the same time, too low to perpetuate the surplus oil-producing capacity in the United States, to which the state regulators owed their market power. Though the world's stock of very low-cost oil was still immense, the loss of spare capacity in the United States concentrated the power to increase or curtail production rapidly in a handful of economically underdeveloped and politically turbulent countries. The problem, therefore, was not a permanent worldwide scarcity of "cheap" oil, but rather, the absence of short-term oil-demand flexibility, together with the disappearance of the short-run supply Alaska Energy Planning 10/26/82 Page 84 flexibility that had previously been exercised by governmental and private institutions that had were consciously striving for market order. In these circumstances, a relatively small curtailment of sales by a few producers openly aiming at market disruption could and did trigger an upward explosion of prices. If world oil prices now fell to pre-1973 levels (in constant-dollar terms) once more, a world-wide "energy crisis" would be with us again sooner or later. Likewise, however --- 6. No group of producers could long hold yorld oil prices at the high levels of the early 1980s, and it .is wilikely anyone will ever be able to do so. Today's prices are not viable because they are well on the wq.y tQ pricing oil out of both the industrial fuels market and the marke~ for petrochemical feedstocks. If prolonged, today's prices would even begin to erode oil's monopoly in transportation-fuels markets. Oil at $30 and up has, therefore, guaranteed the emergence of excess producing capacity not in just one or a handful of political entities (Texas and Lousiana, for example, or Saudi Arabia, Kuwait, and Abu Dhabi), but all over the globe. 7. Market stab~ ii( price requires the supply- demand balai\Ce to 00\ytly and in the normal direction to any price d!ange, and prices to respond prompt- ly and in the normal direction to any d!ange in the supply- demand balance. If the world is to avoid repeated violent swings in oil prices, market arrangements must be such that a small rise in oil prices can cause either a sizeable increase in effective oil supply or a sizeable decrease in oil consumption, or both. A small drop in prices must, likewise, be able to induce a prompt reduction in supply or increase in consumption. 8. Short-term supply-side adjustments that foster :;i?ce stabili,j rather than instability require a TRG-style _ice-maker _:_but none is now in siiflt. Any supplier or group of suppliers that hopes to regulate the market must have the ability and the will to swing world oil production upward to satisfy any surge of demand or supply interruption, or (more importantly now) to swing it downward in order to make room for a surge of supply or slump in demand. The system run by the TRC underpriced away its power to increase output whenever it was needed. OPEC as such never had either the will or the capacity to take responsibility, and Saudi Arabia ---out of greed ?r timidit>:, we ~ay never know ---blew its chance. It has now overpraced away Its ability Alaska Energy Planning 10/26/82 Page 85 -·-: .. •.; to . reduce production sufficiently to support world prices at present leveis or, most likely, at any level. . .. ·~ .. The only phiusible new candidates for ·price-maker may be PEMEX (Mexico's state oil company) and the US Strategic Petroleum Reserve. Even if one of them maneuvers itself into the right strategic spot in the world. market, however, there is only the barest chance that domestic polftks in the United States or Mexico would perm.it either insti tution to move quickly, independently, and responsibly eno.ugh to serve as the world's oil-supply balance wheel. 9. The only price level that even a supply-side pr:ice:. maker can maintain for long is one that foSters demand-side stability as well. 10 . Specifically, the range of sustainable oil prices is limited to those prices at which oil, coal, and gas are effective and close competitors in the world's markets for electrical-generation fuels, industrial boiler and stationary heating fuels, and petrochemical feedstocks. If the price of oil remains within a range where oil, naturai gas, and coal effectively compete for industrial sales in North America, Europe, and East Asia, many of the world's large energy-consumers will !ind it worthwhile to install dual or multi-fuel capacity, ~xpressly in order to take advantages of small shifts in relative prices. T~e ability of a large consuming sector to switch fuels rapidly in response to changes in relative fuel prices or availability would preempt the perverse market behavior that has permitted small market shoc)<s to explode into global crises. Multi-fuel consumers would simply let go of enough oil in a tight market, and absorb enough additional 9i1 in a slack market, to avoid even the illusion of a physical shortage or surplus. The greater this demand-side flexibility becomes, the more modest will be .the world's need for a supply-side price-maker like the TRC, the less onerous will be the price-maker's task if one is still needed, and the less damage an incompetent or irresponsible price-inaker will be able to cause. Unlike pre-1973 prices, the $10-to-$18 price range is high enough to cover the cost of mining and transporting coal, and of burning it in an environmentally acceptable fashion, almost but not quite everywhere Alaska Energy Planning 10/26/82 Page 86 in the world. These prices are also high. enough to justify shipping liquefied natural gas (LNG) from any low-cost gas-producing area near tidewater to almost any port in the world, and to justify building transcontinental natural-gas pipelines (though maybe not the Alaska or Yamal pipelines). Prices in this range would still leave oil holding a significant fraction of the markets for electric-utility and industrial boiler fuels and for petrochemical feedstocks. Any price excursion outside of this range, however, would still carry the threat of steep price f luctuations further away from, or substantially overshooting, any attainable equilibri um. History offers some empirical support for the viability of a long- term world oil price in the $10-to-$18 range. Over 110 years of crude- oil price records in the United States, the average price in 1982 dollars has been almost exactly $13 per barrel and, despite an average constant-dollar price fluctuation of more than 20 percent per year, no long-term trend can be detected. {The average 1982-dollar price between 1871 and 1925 was $12.96 per barrel, and the average price between 1926 and 1980 was $13.0q per barrel.) Thus, the safest guess as to the average crude-oil price over (say) the next 25 years may be about $13 per barrel in 1982 dollars. However--- 12. These generalizations do not warrant a forecast of a $13 price, or any other specific price at any specific fut\a"e time. In the absence of a secure mechanism for getting world oil prices into this range and keeping them there for several years, and in the absence of a competent and responsible successor to the Texas Railroad Commission, the prospect is for wide and unforeseeable fluctuations in world oil prices, like those that occurred before the TRC took control ln the mid-1930s. The managing director of Royal Dutch Shell, D. de Bruyne, summarized the new outlook well when he wrote ~t "we are in for a period of severe and unpredictable discontinuities." The most ambitious forecast we dare make with any confidence is that --- without some new market-ordering mechanism, which is not now in sight --- 13. World oil prices will fluctuate both randomly and cyclically. In any given future year, however, the most likely price will be far below l979-1982levels. In summary, there is no basis in geology, resource-economics or history for predicting a never-ending increase in the real price of oil. Private investments and governmental institutions founded on that proposition are sure losers. Alaska Energy Planning 10/26/82 Page 87 NOTES *Connie· C Barlow, Mi c helle Celarier, Gregg K Erickson, and Samuel A Van Vactor all helped develop the concepts in this Raper and clarify their exp~ess.lon; M A Adelman provided useful corrections and s4ggestions. 1. '"OPEC Will Survive•, Oilman Hammer Says." Associated Press story in S~attle Post-Intelligencer, July 12, 1982, p B8. 2. For the history of oil conservation and the rise of the Texas Railroad Commission, see Wallace F Lovejoy and Paul T Homan, Economic Aspects of Oil Conservation Regulation, (Baltimore: The Johns Hopkins University Press, 1967), pp 33-57, and Stephen L McDonald, Petroleum Conservation in the United States: An Economic Anal sis, (Baltimore: The Johns Hopkins University Press, 1971 , pp 29-55. 3. From the J.uly 1882 Scientfic American: The history of the discoveries in the Pennsylvania oil fields has been one of a series of disapppintments to the producers. From 1866 to 1872 the price per barrel averaged from $4 to $5, and the producers were making money rapidly. Then the field in Butler County was struck, and from that day to · this the production has been greater than the consumption. Then came the Bullion pool with its 2,000- and 3,000-barrel wells, which forced the price down to $1.50. This field was soon exhausted, and better times for the producers were at hand when the Bradford field, the largest in extent ever known, was opened. Then Bradford began to decline and again a silver lining was seen, but again disiappointment came. In May of last year the first well was struck in Allegany County, New York, and a new field was opened that soon more than made up for the decline. Then was the great ''646" well struck, and with it followed disaster to the owners of wells generally, and lower-priced oil than since the summer of 1874, when for a time it sold for 45 cents a barrel. Where the next field will be is only a matter of conjecture. 4. For the his~ory of import controls, including the influence of the Texas Railroa9 Commission on import policy; see M A Adelman, The World Petroleum Market (Baltimore: The Johns Hopkins University Press, 1972), pp 150-154. Alaska Energy Planni~g 10/26/82 Page 88 5. Lovejoy and Homan, op cit, pp 265-285, and the Cabinet Task Force on Oil Import Control, The Oil Im rt uestion (Washington: US Govern- ment Printing Office, 1970. pp ·24, 121, 216, 242-246. 6. Adelman, op cit. For a history of oil prices .during the en~re era, see pp 131-191. 7 •. M A Adelman, "Coping with Supply Insecurity," The Energy Journal, Oc tober 1982, pp 1-16. 8. On the role and operation of spot markets, see Paul H Frankel, Topical Problems (London: Petroleum Economics, Ltd), July/ August 1973, p xx, January/February 1976, p iv and June 1979, p xvii-xviii. 9. Estimated proved reserves as of January 1, 1982 can be found in Robert J Enright, "Worldwide Report," Oil and Gas Journal, December 28, 1981, p 86. The Journal lists Saudi Arabia's proved reserves as 164.6 billion. John Blair in The . Control of Oil (New York: Vantage, 1978), pp 18-19, quotes Yamanl as saying that "Saudi Arabia's 'true reserves' are more than two and a half times the 'ultra conservative numbers' at which 'proved reserves' were being carri-ed." In 1972, James Akins, then US Ambassador to Saudi Arabia, told Senator Mike Gravel and one of the authors that Saudi Arabia's reserves were "realistically" at least 700 billion barrels and "probably closer to a trillion." At a Central Intelligence Agency briefing one of us attended in 1975, an Agency spokesman gave almost the same estimates (likely from a common source) of the ultimate reserves in the known fields in Saudi Arabia. He added that Iraq's reserves were probably "almost as big". For our present purposes it doesn't matter which of these reserve estimates is the most realistic ---even the most conservative of them implies that Saudi Arabia is physically capable of producing consider- ably more than 20 million barrels per day without any new disc overies. 10. See, for example, the remarks of Jahangir Amuzegar, Iranian Ambassa- dor-at-Large and sometime petroleum minister, at a 1975 Salomon Brothers conference in London (World Petroleum: The Economics of Current Pricing and Supply Policies. London: Salbro Press 1976, p 30): Notwithstanding Western calculations and projections to the contrary, OPEC members believe that their oil reserves underground will be worth more in the future c ompared to the present ---even with accumulated returns on the invested revenues. Adelman makes a plausible case that this notion is economically fallacious and, by implication, cynical and deliberately misleading. (M Alaska Energy Planning 10/26/82 Page 89 A Adelman, "OPEC as a : Cartel" in Griffin and Teece, OPEC Behavior and World Oil Prices, pp 38-53) ··''· '>Jiowever, Adelman's argument that oil reserves never earned as much as financial investments rests on discount rates that reflect the . short life-expectancies of governments in Third World oil-exporting countries. This approach ignores the ideological content of national policy .in· such countries. Economic policy in most OPEC nations is either made by nationalist bureaucrats who view their nations as something different from the present government, or by heads of state who believe that their persons are identical with the nation, which is itself immortal. Either case results in lower discount rates and longer amortization periods than Adelman assumes for a non-ideological world. Adelman's analysis ·also virtually dismisses the specific ideological role played by the concept of oil-reserves as a long-term investment. OPEC spokesmen were doubtless sincere when they insisted that the asset-value of their resources was appreciating at a higher rate than the real rate of return on risk-free financial iiwestments. My ground for accepting such professions at face value stems both from personal contact with high-placed and lowly' Believers, a11d from the fact that it was in the oil-exporters~ interest that they and their customers both believe their motives were something more honorable than greed, and their production-scheduling was built on something more substantial than simple opportunism • At the same 1975 conference one of us directly addressed Amuzegar's 1975 argument and anticipated Adelman's 1981 argument: To Karl Marx, who gave us the concept, "ideology" was a body of doctrine that provided a religious, moral, or scientific cloak to self-interest. Ideology is in the first place a political weapon: if they believe in it, its sponsors can · d~w from it moral -fervor and confidence of success. And· ari. eff'ective ideology can also captivate or neutralize its adversaries. Believing one's own propaganda uncritically (or that of one's opponents) has, however, Zed to some remarkable foolishness, as various Crusades from the Middle Ages to. Vietnam have shown. I suppose that my message today is not to take OPEC's rhetoric too seriously, nor the opposing rhetoric • • • The conservationist element in OPEC doctrine deseMJes more serious attention (than its profession of solidarity with the poor and exploited everywhere), particularly as it is a notion shared by a rich, industrialized non-OPEC oil producer like Canada and by a variety of environmentalists and Malthu- sian doomsayers in all of the rich countries. The common theme of all these parties is that mankind ought to keep its Alaska Energy Planning 10/26/82 Page 90 c heapest energy source, petroleum, in the ground because it will be more valuable in the future than it is today. This proposition cannot be dismissed out of hand. There may well be some producing country or countries with reserves of only ten to fifteen times current produc tion, without the hope of major new discoveries and with only limited opportunities for productive investment at home. Such a country could reasonably estimate the so-called user cost of its petroleum ---the present value of future production given up by producing today---to be as great or greater than the current world price. Such a country might reasonably believe, in other words, that its oil could appre- ciate in value over the average life of its reserves at a higher rate than the rate of earnings on risk-free foreign investment. Or it may believe that the risks ---market and political ---of all foreign investments are so great that they make speculation in oil inventories at home a more prudent investment. I am not certain there is such a country ---but that C'OWltry surely is not Iran or Venezuela, whose ability to absorb foreign exchange in profitable domestic investment ventures is insatiable, nor is it Saudi Arabia, whose potential reserves are so huge that the present value of a barrel of oil not produced today is truly negligible. No, to each of these countries, limiting production is rational not because its oil will be more valuable in the future but because less production means higher prices today. Conservation, however, sounds more noble in the producers' own ears than maximization of monopoly profit, and it appeals to a fashionable intellec tual current in the rich consuming nations. The conservationist rhetoric is, therefore, a particularly effective ideological weapon of the cartel. (emphasis added)(A R Tussing, discussant, comments on speeches by Chief M 0 Feyide, Secretary-General of OPEC; Amuzegar; T 0 Enders, US Assistant Secretary of State for Economic Affairs; Adelman; and P T Frankel, Director of Petroleum Economics Ltd., op cit, pp 41-44) We need not be overly skeptical about the OPEC nations' belief in a doctrine that helped enrich them, when the same doctrine was believed by so many statesmen and scholars (including the majority of "energy economists") in Europe and America, who used it to rationalize policies that helped impoverish their own nations. 12. Youssef M Ibrahim, "Saudi Role in OPEC Under Siege" in The Wall Street Journal, July 21, 1982, p 33. Alaska Energy Planning 10/26/82 Page 91 13. "Oil Nation Warns its Partners", Associated Press story in the Seattle Post-Intelligencer, July 8, 1982, p B9. On Saudi Arabia's threats, se Platt's Oilgram, J,tdy 1.3, 1982, p lA. 1/f. D. de Bruyne, quoted in Petrol~um .Intelligence Weekly, June llf, 1982, p 8. Alaska Energy Planning 10/26/82 . Page 92 MEMORANDUM . State of Alaska TO: Commiss i oners. and Spec i ~l Assistants mes M. Souby, Division of Pol and Pl anning Office of the Governor DATE: FILE NO: TELEPHONE NO: SUBJECT : November 24, 1982 465 -3577 Alaska Energy Planning Studies The proposed Susitna hydroe l ectric project has been t he subject of extensive study over the past few yea r s . Recent reports by Acres American ,Incorporated and Battelle Pacific Northwest Labo r a- tories conclude that construction of ·the Susitna dams would probably yi eld significant economic benefit over the long term compared with the best electric generation alternatives. Due t.o the magnitude of the issue and the complexity of the documents issued by Acres and Batte l le~ this Division engaged the Inst i tute of Soc i al and Economic Research at the University of Alaska to review the studies and to comment on the· strengths and weaknesses of the arguments underlying their conclusions . The review was performed by Arlon Tussing and Gregg Erickson , and is presented in fina l form i n the enclosed report . As you may recall, th i s report received considerab l e attention i n the press two months ago when circulated in draft . 02-001 A( Rev.l0/79)