EnergyResource s�___.cornucopia or Empty Barrel?t

Peter J. McCabe2

ABSTRACT fuels are driven more by demand than by the geo­ logic abundance of the resource. Over the last 25 yr, considerable debate has con­ Examination of some energy resources with tinued about the future supply of fossil fuel. On well-documented histories leads to two conceptual one side are those who believe we are rapidly models that relate production to price. The closed­ depleting resources and that the resulting short­ market model assumes that there is only one source ages will have a profound impact on society. On of energy available. Although the price initially may the other side are those who see no impending cri­ fall because of economies of scale long term, prices sis because long-term trends are for cheaper prices rise as the energy source is depleted and it despite rising production. The concepts of becomes progressively more expensive to extract. resources and reserves have historically created By contrast, the open-market model assumes that considerable misunderstanding in the minds of there is a variety of available energy sources and many nongeologists. Hubbert-type predictions of that competition among them leads to long-term energy production assume that there is a finitesup­ stable or falling prices. At the moment, the United ply of energy that is measurable; however, esti­ States and the world approximate the open-market mates of resources and reserves are inventories of model, but in the long run the supply of fossil fuel the amounts of a fossil fuel perceived to be avail­ is finite, and prices inevitably will rise unless alter­ able over some future period of time. As those nate energy sources substitute for fossil energy sup­ resources/reserves are depleted over time, addi­ plies; however, there appears little reason to sus­ tional amounts of fossil fuels are inventoried. pect that long-term price trends will rise Throughout most of this century, for example, significantly over the next few decades. crude oil reserves in the United States have repre­ sented a 10-14-yr supply. For the last 50 yr, resource crude oil estimates have represented INTRODUCTION about a 60-70-yr supply for the United States. Division of reserve or resource estimates by cur­ The supply and price of energy over the next rent or projected annual consumption therefore is century will have a profound impact on nations, circular in reasoning and can lead to highly erro­ the energy industry, and the geoscience profes­ neous conclusions. Production histories of fossil sions. Anticipation of future trends determines investment strategies, play portfolios, research directions, career decisions, and government poli­ cies. In this paper, I do not precisely predict the IC>Copyright 1998. The American Association of Petroleum Geologists. All rights reserved. future, but rather I examine historic trends and •Manuscript received September 9, 1997; revised manuscript received clarify the foundations on which one may build April 8, 1998; final acceptance May 28, 1998. one's own predictions. 2U.S. Geological Survey, Federal Center MS 939, Denver, Colorado 80225; e-mail: [email protected] The future availability of fossil energy resources I wish to thank the following individuals: Donald L. Gautier, U.S. has been much debated over the last 25 yr (see, for Geological Survey, who first suggested the warehouse concept of resources to me; Keith W. Shanley, Amoco Production Company, with whom I had example, Polesetsky, 1991; Myers and Simon, 1994). several discussions of recent trends in the petroleum industry; Kevin M. On one side are the "neo-Malthusians; who perceive Pickup, The Coal Authority (U.K.), who provided recent information on British that current trends of energy production are not sus­ coal resources; and Duncan Millard, U.K. Department of Trade and Industry, for providing data on British coal prices. The manuscript benefited tainable, and who foresee that we will soon run out significantly from the reviews of Keith W. Shanley (Amoco Production of energy resources. On the other side are the "cor­ Company), John B. Curtis (Colorado School of Mines), Joseph R. Studlick, nucopians," who believe that there is no scarcity of Jr. (UNOCAL), and the following U.S. Geological Survey reviewers: Donald L. Gautier, James W. Schmoker, Michael D. Lewan, Gordon L. Dolton, and energy resources and that, over time, energy Katherine L. Varnes. This paper is published with approval of the director of resources will become more widely available at the U.S. Geological Survey, but the interpretations and opinions presented are mine, not those of the U.S. Geological Survey, whose scientists have lower costs. These opposing viewpoints on energy diverse opinions on this and most other subjects. resources are part of larger differences in perception

2110 AAPG Bulletin, V. 82, No. 11 (November 1998), P. 2110-2134. McCabe 2111

that are fundamental to many of society's concerns relationship betweenpopulation and resources pub­ and political differences in the 1990s. The neo­ lished at the beginning of the 19th century. Britain, Malthusians, for example, claim that the planet can­ for example, faced major shortages of firewood in not sustain the rapid population growth and econom­ the early 14th and 16th centuries; these shortages ic progress of the last few decades. Unless current resulted in rising firewood prices (Hatcher, 1993). trends are halted or reversed, they anticipate that Worries of running out of petroleum appear to have there will be major famines, shortages of natural begun shortly after commercial production started resources, and environmental degradation that will in 1859. In the United States, at least six oil-shortage be devastating to the human race. Holders of this scares took place in the firsthalf of the 20th century, viewpoint commonly argue for birth control, conser­ with shortages blamed on factors ranging from the vation of resources, and preservation of the environ­ development of the Model T to Britain's supposed ment. They tend to argue for more government regu­ goal of dominating the world petroleum market lation and policy to achieve those goals. By contrast, (Fanning, 1950). The tremendous growth in energy the cornucopians have no such concerns and may production after World War II led to renewed fears believe that the quality of life will be raised as the that fossil fuel supplies would be exhausted, with world population increases. Holders of this view­ consequent detrimental impact on society. Meadows point tend to argue against contraception, to believe et al. (1972), in their report for the Club of Rome, that free markets will ensure an abundant supply of argued that the growth in world population, indus­ resources, and to believe that environmental con­ trial production, use of fertilizers, and use of most cerns have been greatly exaggerated. They argue for natural resources w�s exponential, and that such less government. Much of the neo-Malthusian/cornu­ growth was impossible to sustain. Meadows et al. copian debate on energy resources has been, and (1972) regarded energy resources as fmite and calcu­ inevitably will be, dominated by nongeologists. lated the years at which fossil energy would be Consequently, geologists must communicate their depleted according to three scenarios (Table 1), scientific work in a way that allows others to gain a which have proven to be overly pessimistic; howev­ better appreciation of the nature of energy er, the 1974 Arab oil embargo focused the public's resources. In this paper, I intend to question some attention on energy resources and, for many, gave long-held beliefs of both the neo-Malthusian and cor­ instant credibility to the Club of Rome report. nucopian camps. Perspectives in the paper also may The early 1970s also saw the peak of U.S. crude challenge some common notions held among many oil production. This peak had been predicted by geologists. Hubbert (1956, 1969). Hubbert's predictions were Assessments of fossil energy resources are done based on a model of a cycle of production of finite by many organizations, such as federal geological energy resources (Figure 1), and his methodology surveys, state and provincial surveys, government is discussed at length in following sections of this departments of energy, land use agencies, energy paper. Today, the neo-Malthusian viewpoint is companies, and consultants. Results of these assess­ widely held by society and is strongly pushed by ments are used for such diverse purposes as devis­ the environmental movement that uses phrases ing foreign policy, setting taxation levels, planning such as "sustainable development" and "the plan­ land use, and developing investment strategies. et's carrying capacity," and by advocates of alter­ Assessments therefore are perhaps the most impor­ nate energy sources [e.g., Flavin and Lenssen tant way in which geologists convey information (1994)]. This viewpoint also is held by many scien­ derived from their science to nongeologists who tists [e.g., Bartlett (1986, 1994), Campbell (1997), make critical decisions that affect the general pub­ and Edwards (1997)]. As t.he N ationa! Research lic. Without understanding the basis of the esti­ Council Committee on the Status and Research mates of an assessment, decision-makers are in sig­ Opportunities in the Solid-Earth Sciences claimed nificant danger of misusing the resource numbers. in 1993 (National Research Council Committee on In this paper, I suggest ways in which geologists the Status an� Research Opportunities in the Solid­ might better communicate information on energy Earth Sciences, 1993, p. 297), "It is only a matter resources to nongeologists. of time until global shortages of petroleum resources develop."

RATIONALES FOR NEO-MALTHUSIANAND CORNUCOPIAN VIEWPOINTS Cornucopian Viewpoint

Neo-Malthusian Viewpoint Although many people have worried about ener­ gy resource depletion and argued for conservation, Concern about depletion of energy resources many others have bought large cars, turned up the almost certainly predates Malthus's studies of the air conditioning, and voted against subsidies for 2112 Cornucopia or EmptyBarrel?

Tab le 1. The Club of Rome Forecasts of the Year Global Fossil Fuel Resources Would Be Totally Depleted

Known Reserves Known Reserves (Known Reserves) x 5 1972 Consumption Exponential Consumption Exponential Consumption

Petroleum 2003 1992 2022 Natural gas 2010 1994 2021 Coal 4272 2083 2122

*From Meadows et al. (1972). Forecasts are the estimated year at which fossil fuel resources would be totally depleted under three different scenarios: depletion of known reserves at 1972 consumption rates, consumption growing at the pre-1972 exponential rate with reserves held at 1972 known level, and exponential growth in consumption with reserves at five times those knowri in 1972. A fivefold growth in reserves in the third category apparently was an attempt to illustrate a best-case scenario that was considered highly unlikely.

(A) by a wide range of economists and scientists (see 14 25 papers in Simon, 1995). The main argument made by the cornucopians is that the historical trends are 12 Production ..; v"\ for lower prices for resources through time despite 1- 20 '"' \0 10 0'1 s rising production rates. Plots of the prices of miner­ - ...... _I 15 al and energy resources over- time (see, for exam­ 8 ;gc: ::0 Price I 0 6 ple, the trend for crude oil in Figure 1) show a � I 10 = decline that is steep in the early years of produc­ u 4 / •t:: J � tion, and that becomes flatter as time proceeds p., / 5 2 ../ (Simon et al., 1994; Adelman, 1995; Hausman, 0 1995; Myers et al., 1995). Fluctuations in price may 1880 1900 1920 1940 1960 1980 occur, but historically prices return to a point along a smooth curve. Energy prices over the last 20 yr (B) have been volatile as a result of major events on the

....� � world's political scene, but when these prices are � Ultimate corrected for inflation, oil prices are now at a level c:

. similar to that of the early 1970s (Figure 2). The ....2 u cornucopians argue that because of our imagina­ ::1 "0 tion and creativity, resources will be found to meet £ whatever demand arises. New oil and gas fields are found, production from coal fields continues to expand, methods to extract more oil and gas from reservoirs are developed, and cheaper sources of Time energy replace more expensive sources. Within the geologic community, a cornucopian viewpoint is Figure 1-Two time-series plots that are fundamental to shared by those explorationists who believe there the cornucopian and neo-Malthusian viewpoints of is plenty of oil and gas still to be found, although energy resources. (A) A plot of the price and world pro­ most of them believe that rising prices would help duction of crude oil shows that prices have fa llen over their cause. As Halbouty (1986, p. 10) said, "It has time, despite a tremendous increase in production, sug­ been predicted over and over for decades that the gesting that there is no real shortage of crude oil(from world was running out of oil and gas. Those pes­ Simon et al., 1994). (B) In contrast, the Hubbert curve indicates that a finite resource willbe inevitably deplet­ simistic doomsayers have consistently been proven ed(from Hubbert, 1956). wrong."

The Debate public transport, apparently in the belief that there is no problem; nevertheless, until recently there This section is far too brief to adequately cover

· have been few who publicly advocated the view­ all the rationalebehind. the neo-Malthusian and cor­ point that there is no long-term problem. The main nucopian viewpoints, and the reader is urged to proponent of the cornucopian viewpoint was the consult the cited references; however, know that late Julian L. Simon. His earlier work (Simon, 1980, emotional statements, personal attacks, deductive 1981) received considerable negative criticism reasoning, categorical statements, and other exam­ (Bartlett, 1985), but his viewpoints are now shared ples of deceptive arguments are more common in McCabe 2113

60 70 overly pessimistic because global energy consump­ tion did not grow at an exponential rate and, as is 60 50 Production r discussed in following paragraphs, many new 00 /' � 0'1 50 reserves were found. Will that trend continue? 0'1 40 - ;g According to Bernstein (1996, p. 6), the history of / 40 i:l :a 30 = those who predict future trends is one of "persis­ / 30 tent tension between those who assert that the �0 :3 ·c 20 best decisions are based on quantification and Q., / 20 � numbers, determined by the patterns of the past, 10 10 and those who base their decisions on more sub­ 0 0 jective degrees of belief about the uncertain 1960 1970 1980 1990 2000 future." Ironically, both the cornucopians and neo­ Malthusians claim that the other group relies on gut Figure 2-The price and world production of crude oil feeling rather than having faith in the trends of vari­ have been volatile over the last 25 yr in response to ous data sets. In the next sections we look more major world events. When adjusted for inflation (using closely at the nature of the data sets used by both the consumer price index), today's price is similar to sides in the debate. what it was at the beginning of the 1970s. Production and original price data from Energy Information Administration (1998). THE NATURE OF RESOURCES AND RESOURCE ASSESSMENT many of these papers than is the norm within geo­ What is an energy assessment? Geologists have logic literature. Historically, neo-Malthusians have diverse opinions as to what an assessment is, and been pessimists, often claiming that catastrophe is there are diverse assessment methodologies [see, just around the comer. Gever et al. (1986, p. 5), for for example, the papers in Rice (1986)]. One gener­ example, in explaining the neo-Malthusian perspec­ al consensus is that an assessment involves much tive, pointed to "the current economic turbulence more than just calculating numbers, that it should and confusion as symptoms that humans are be based on a foundation of geologic knowledge, approaching, if not exceeding, their carrying and that it requires an examination of as much data capacity," a claim that seems a little odd when as is feasible within the time frame of the study. viewed a decade later. Cornucopians are opti!nists The calculation of amount of resource is more and claim that history bears them out, but is that properly termed an estimate. For nongeologists, confidence misplaced? Many geologists may sym­ however, there understandably is little or no inter­ pathize with the comment of Gever et al. (1986, est in how the geologists have done the assess­ p. 96): "to realize the eXpectations of Simon and his ment, and there is an assumption that the method­ fellow cornucopians that natural resources will ology is scientifically sound. The only question that continue to become cheaper and more 'abundant' · most nongeologists will ask of an assessment is: in the long run, the net supply of fuel must contin­ How much is there and where is it? In this section, ue to rise and its dollar price must continue to I examine different types of assessment estimates fall-an expectation contradicted by the Hubbert and discuss how they should be interpreted; analysis and by common sense." As scientists, how­ ever, we can put little weight on common sense; the history of science shows that what is consid­ Reserves ered as common sense at one time is often regard­ ed as nonsense by later generations (the ideas of Reserves of a fossil energy are defined as the Agassiz, Darwin, Wegner, and Vail, for example, ini­ identified accumulations that can be extracted tially were regarded by many as contradicting com­ profitably with existing technology under present mon sense). The validity of Hubbert-style analyses economic conditions (McKelvey, 1972; Brobst and will be discussed in following sections of this Pratt, 1973). Although this definition has apparent­ paper. ly satisfied most geologists for many years, there are The amount of fossil fuel on our planet, of aspects of this definition that may strike the nonge­ course, is finite. Does that mean that our civiliza­ ologist as unusual. A basic problem with the con­ tion will run out of fossil fuel at some time in the cept of a reserve is that there has been no uniformi­ future? Is that time imminent? Or will we develop ty or stated policy as to the time period over which alternate sources of energy long before most of the the "existing technology" and "present economic planet's fossil fuels are extracted? In retrospect, the conditions" are anticipated to prevail. An under­ predictions of the 1972 Club of Rome report were standing of the nature of reserves can be gained 2114 Cornucopia or EmptyBarrel?

from Figure 3, which plots U.S. cumulative produc­ 200 �------� tion of crude oil and measured ultimate reserves Measured ultimatereserves (i.e., proved reserves currently booked plus the crude oil produced to date). For the last 80 yr the � United States has essentially identified crude oil reserves (proved) that will be produced over the · � following 10-14 yr. c: .9 The natural reaction that most people have on := 100 learning of a reserve number is to attempt to under­ iii stand that number relative to current consumption rates. If one divides the reserve number by annual consumption, it may appear logical to assume that one has roughly calculated the number of years �lative production before we will run out of the resource, and modifi­ 0 ��----.-r-.-���-r��-.--r-r-� cations may be made by projecting increased or 1920 1930 1940 1950 1960 1970 1980 1990 decreased consumption over time; however, this is a fallacious approach based on circular reasoning. Figure 3--Cumulative production of crude oil and mea­ sured ultimate reserves (cumulative production to date A reserve estimate can be compared to a grocery plus proven reserves)in the United States over time. For warehouse's holding capacity. If one divides the most of this century, the proven reserves measured in capacity to hold a certain product, say baked any year (a) have been sufficient to supply the nation's beans, by the daily consumption of a city, one sees production for the following 10-14 yr (b). Data fro ' how many days the warehouse can supply a city Energy Information Administration (1998) and Oil & ' out of current stock, say 45 days. That does not Gas Journal(1997). iD mean that the citywill face a catastrophe as it runs out of baked beans in 45 days, because the ware­ house itself is constantly being replenished with and Attanasi (1993), Attanasi and Root (1994), and baked beans from canned food manufacturers. A Schmoker and Attanasi (1997). The USGS (U.S. well-designed warehouse will hold enough baked Geological Survey) 1995 national oil and gas beans to meet expected demand for a length of assessment estimated that field growth will add 60 time that has been determined to be optimal in the billion bbl of oil and 322 tcf of natural gas to grocery wholesale business. This concept of reserves over the riext 80 yr (U.S. Geological reserves was articulated by Ezra (1978, p. 82) in Survey National Oil and Gas Resource Assessment describing the British National Coal Board's policy Team, 1995). Field growth is increasingly appar­ on calculating their operating reserves compared ent in worldwide oil and gas fields (Adelman and to recoverable reserves: "There will be a progres­ Lynch, 1997; Dromgoole and Speers, 1997; Smith sive upgrading of recoverable reserves into operat­ and Robinson, 1997). Not all fields show growth in ing reserves,. The intention is to maintain a 'stock' reserves and, although the overall trend is for high­ of at least 50 years of operating reserves at any one er values through time, a substantial number of time." Consciously or not, some time frame over fields are downgraded in size. Factors that include which a reserve is anticipated to be economic is geologic complexity, political pressure, salary incorporated into all reserve estimates. This basical­ incentives, taxation, and the need for investment ly explains why the Club of Rome forecasts (Table capital inf luence field reserve calculations differ­ 1) were erroneous. ently in various parts of the world. A common notion is that addition to proved Much concern has been expressed about the reserves requires the finding of new accumulations decline in U.S. reserves. As Figure 3 shows, ·there of energy. In fact, the vast majority of additions to was a gradual rise in reserves up to 1971, while energy reserves are from areas that have long been production was rising, and there has been a subse­ documented as coal, oil, or gas fields. Figure 4 quent fall. Was the trend in production a function shows the source of annual addition to crude oil of the size of reserves? More likely it was vice versa. reserves in the United States In the last SO yr, only Figure 5 is a plot of reserves divided by annual pro­ 14% of the annual addition to reserves came from duction for 1918-1994. Overall, there has been a new fields. The rest came from step-out drilling steady decline in this ratio. It would be misleading from existing fields (extensions to existing fields), to argue that this trend necessarily portends doom discovery of new pools in existing fields (new for the U.S. oil industry. After all, there was a ten­ reservoirs in existing fields), or from other causes fold increase in annual production over the first 52 that include improved recovery factors (revisions yr of this trend! Rather, the plot suggests that more and adjustments). Rates and effects of field growth rigorous criteria have been used to define U.S. in oil and gas fields have been documented by Root crude oil reserves through time. McCabe 2115

14 Figure 4-Field growth accounts for 86% of the Revisions and adjustments • addition to U.S proved reserves over the last 12 • Extensions to existing fields 50 yr. Over time, revisions and adjustments have New reservoirs in existing fields become proportionally 10 • more important. Note that the large addition to Newfields CJ reserves from the Prudhoe ::E 8 ,.Q Bay field in 1970 was equal c to about 5 yr of annual .sa = fieldgrowth. Data from i:6 6 Energy Information Administration (1998) and on & GasJournal (1997). 4

2

0 1950 1960 1970 1980 1990

§ 16 Resources are defined as the reserves plus all the 14 accumulations of a fossil energy source that may 'g eventually become available-either known accu­ £ 12 mulations that are not economically or technologi­ ] 10 cally recoverable at present, or unknown accumu­ lations, rich or lean, that may be inferred to exist, � : but have not yet been discovered (McKelvey, 1972; 4 Brobst and Pratt, 1973). This definition is more vague than that for reserves. Similar to the amount 2 of gold in seawater, the total amount of any fossil l o��-r���--�r-T fuel present in the Earth's crust (its crustal abun­ 1920 1930 1940 1950 1960 1970 1980 1990 dance) is an astronomic figure that, while of possi­ Figure 5-Plot of U.S. crude oil proved reserves divided ble academic interest, is a number that has no eco­ by annual production for 191�1994. Data from Energy nomic significance. Vast amounts of coal lie in beds Information Administration (1998) and Oil & GasJour- ,/ that are so thin or of such poor quality that it seems nal (1997). highly unlikely that they will ever be mined. Likewise, large volumes of oil and gas exist that most likely will never be exploited, because they assessment is where to slice through the pyramid would be prohibitively expensive to recover. In to encompass only those resources of economic practice, therefore, resource estimates are made significance. The problem is that this is a very only for those accumulations that are seen as subjective decision. Some people are more con­ potentially economic at some time in the future. servative than others in defining what is of poten­ The distribution of fossil fuel can be viewed as a tial economic interest. No one can foresee the pyramid (Figure 6) with a small amount of high­ future. Te chnological advances, for example, may quality resource and with increasingly larger make some resources cheaper to extract, or may amounts of lower grade resource. The costs of change the demand for a fuel. Global economic retrieving the resource increase lower in the pyra­ conditions may increase or decrease the demand mid making a larger amount of the resource avail­ for fossil fuels. able at higher prices. If the total volume of the Figure 7 shows the magnitude of some esti­ pyramid is viewed as the crustal abundance, the mates, which have been published over the last 50 challenge for the geologist in making a resource yr, of the undiscovered, but recoverable, crude oil 2116 Cornucopia or EmptyBarrel?

assessments that were not based on sound geologic c:: 0 studies with an overlay of economic reality. Weeks ·� 4J u e (1959), for example, did not present any evidence CIS :l .b 0 to support his doubling of resource estimates from � Cll � 4J the previous year (Weeks, 1958). The estimates by .... � 0 the USGS between 1961 and 1972 (Zapp, 1961; .... Cll 0 _q� Hendricks, 1965; McKelvey and Duncan, 1965; u :l Theobald et al., 1972) were based on the Zapp bll 0 c:: .... hypothesis, which assumed that the oil to be dis­ 4J .... -� 4J covered per foot of exploratory drilling in a petro­ 2:l u � liferous region would be essentially constant up to c:: - a certain density of drilling. This hypothesis was discredited by Hubbert (1969). Taking those esti­ mates that have scientific and economic founda­ tion, a plot (Figure 8) of U.S. cumulative produc­ tion of crude oil and measured ultimate resources Figure 6--Resourcescan be visualized as a pyramid with (i.e., resources currently estimated plus the crude a small volume of prime resources that are of high qual­ oil produced to date) shows a similar pattern to ity and are easy to extract, and a large volume of that for reserves (Figure 3). Additional resources resources that are of lesser quality and are more diffi­ are added over time to �ompensate for resources cult or expensive to extract. Over time, resources near produced. For the last yr, the United States has the top of the pyramid are depleted, and development of 50 resources farther down the pyramid is initiated. This essentially identified crude oil resources that will figure is basedon a concept ofJ. K. Gray as published by be produced over the following 40-50 yr. Masters (1979). In a visual sense, the assessment cutoff tends to slice lower through the resource pyramid in suc­ resources of the United States At first look, these cessive assessments (Figure 9). The 1995 USGS oil estimates seem highly variable. Note, however, that and gas assessment (U .S. Geological Survey the estimates of over 200 billion bbl were from National Oil and Gas Resource Assessment Team,

600 "'

�c:

V') 500 1.0 � 0'\ V') - 1.0 0'\ a c - "' v1 0'\ u 0'\ 1.0 V') 0'\ 0'\ 0'\ c: - - ::1 N 0'\- 400 0 � r-- :.:. 0'\ < "ii - - "ii .J:J � �.@ .J:J "88 -; (/)� c:: Q V') - ";(I) s r-: 0'\ r-- 00 u .... .8 300 1.0 1.0 0'\ ·- c: r-: 0'\ - 0'\ bllv = 1.0 - 0'\ - 0'\ �.&:l .sa a 00 - 0 0'\ 0 v � V') v -a 00 VIol) 0'\ .... -a 0'\ v - O..s 1.0 - V') [!: ·a 200 00 0'\- -a '

100

Figure 7-A comparison of the magnitude of some mean estimates of undiscovered recoverable crude oil resources of the United States that have been published over the last 50 yr. Estimates are arranged in chronological order from left to right. Other estimates that generally fall within the ranges of these estimates have been made, but they are less well documented inthe public literature. McCabe 2117

I. thus fall in the pyramid belowthe assessment slice in Figure 9. The relationship over time between mea­ 300 G sured ultimate resources and cumulative production E • H • • of crude oil for the world (Figure 10) shows a similar D. trend to that for the United States, with calculated ::c 200 c. resources providing about a 60-70 year supply. .0 = likewise, the size of the perceived ultimate resource :3 of natural gas within the United States has grown iiS over time to compensate for that produced (Dolton 100 et al., 1993; Potential Gas Committee, 1997). As Woods (1993, p. 875) observed, "estimates of ulti­ mate gas recovery should be seen as ... akin to proved reserves, although on a longer time horizon" and esti­ mates "reflect industry perceptions at the time the 1950 1960 1970 1980 1990 estimates were made." Figure 8--Cumulative production of crude oil in the The warehouse concept, therefore, can be applied United States and the estimated ultimate resources to resources as well as reserves. As with reserves, (cumulative production plus estimated remaining re­ geologists have yet to reach some conformity in iden­ sources, including reserves) over the last 50 yr. Letters tifying a temporal framework over which to define indicate the following assessments: A • Weeks (1948); B the economic potent�al of resources. Again, we face • Hubbert (1956); C .. Weeks (1958); D .. Hubbert (1967) the danger of circular reasoning if one divides a (an average of his two estimates); E .. Hubbert (1969); F resource estimate by annual consumption. After the • Miller et a1. (1975); G • Dolton et a1. (1981); H • Mast et resource identified in an assessment has been deplet­ al. (1989); I • U.S. Geological Survey National Oil and Gas Resource Assessment Team (1995) and Minerals ed, there is a vast amount of resource lower down Management Service (1996). the resource pyramid that may become economically viable to recover.

1995), for example, included continuous-type resources such as coalbed methane and tight gas Perceptions Related to Resource and Reserve reservoirs that had previously been considered as Assessments unconventional and largely uneconomic, but which were considered by 1995 as a significant Reserves and resources are subjectively deter­ factor in future U.S. energy production. Field mined; they are energy accumulations that are per­ growth, which also can be considered as accessing ceived to be close to being exploited (reserves) or resources lower in the pyramid, was considered to that are thought likely be exploited in the future be a more important factor in the 1995 assessment (resources). To a significant degree, a reserve or than it had been in earlier assessments. Other resource estimate is a measure of how conservative occurrences of oil and gas, such as oil shales and geologists are in identifying factors that will make gas hydrates, were not considered as economically the resource economic in the future. It is worth­ viable by the 1995 USGS assessment and would while at this point to look at a few examples of

Figure 9-Resource pyramids showing diagrammatically how successive oil and gas assessments tend to slice lower in the pyramid as less easily accessible and poorer quality resources are seen as potentially economic over time. By contrast, the trend is often in the other direction in coal assessments, as more realistic parameters are used to def1ne those resources with economic potential.

Successive assessments 2118 Cornucopia or Empty Barrel?

2.5 States (Table 2) and the world (Table 1), coal 7 6• . resources have been calculated as much larger than 2.0 oil and gas resources relative to current production 4 5 rates. Is there really that much more coal, or do the :a • • .c underlying bases of the assessments make such a c:: 1.5 0 2 comparison invalid? Most of the undiscovered :.= • resources identified in the oil and gas assessments =2 1.0 E-< probably will play a role in U.S. energy production

• over the next 50 yr, whereas the vast majority of 0.5 estimated coal resources clearly will not. The crite­ ria that have been used by geologists to delineate 0 the economic feasibility of extracting the resource 1950 1960 1970 1980 1990 are much more conservative in the case of oil and Figure tO--Cumulative production of crude oil in the gas than of coal. Compared to conventional oil world and the estimated ultimate resources (cumulative and gas resources, it is difficult to define the production plus estimated remaining resources, includ­ parameters that make a coal economically feasible ing reserves) over the last 50 yr. Numbers indicate the to extract from the ground. In a pictorial sense, it following assessments: 1 =Weeks (1950); 2 .. Hubbert means that the resource pyramid for coal can be (1956); 3 "' Hubbert (1969) (an average of his two esti­ considered as broader and less steep than that for mates); 4 .. Masters et al. (1984); 5 = Masters et al. (1987); oil and gas. Interestingly,_ the same problem arises 6 • Masters et al. (1991); 7 .. Masters et al. (1994). Pro­ whert considering discrete (or conventional) vs. (1998) duction from Energy Information Administration continuous (or unconventional) oil and gas accu­ and Bureau of Mines (1924-1993). mulations. It is relatively easy to defme parameters (field size, depth of water offshore, etc.) that will most likely make a discrete accumulation econom­ how assessment estimates can lead to widely held ically viable, but it is much harder to define the viewpoints that are based more on how the parameters that will determine which continuous reserves and resources are estimated than on any accumulations will be economically viable over real geologic and economic criteria. the next 20-30 yr. The lack of rigorous and uniform criteria makes Much of the perception that there is more coal it difficult to realistically compare reserve and than oil or gas is a function of where the slice has

resource figures between countries, and even more been drawn through the resource pyramids and the · difficult to compare commodities. Generally, the shape of the pyramids themselves. If accumula­ belief is that there are much greater amounts of tions, such as oil shales, deep-basin gas accumula­ coal than of oil or gas. Within both the United tions, and gas hydrates, were included in the total

Table 2. Estimates of U.S. Reservesand Resources fo r Crude Oil, Natural Gas, and Coal*

Supply at Current Production Rate Resource Estimate Assessment Type Referencet (yr)

Crude oil 24.1 billion bbl Proved reserves EIA 10 Crude oil 165 billion bbl Technically recoverable resources USGS&MMS 68 Natural gas 170.5 tcf Proved reserves EIA 9 Natural gas 1412 tcf"* Technically recoverable resources USGS&MMS 75 Coal 474 billion short tons Demonstrated reserve base EIA -470 Coal in 3968 billion short tons Total remaining resources Averitt -3900 lower 48 states Coal in 5600 billion short tons Resources Merritt & Hawley Alaska

*Note that these figures are not comparable because the assessments have been done based on different premises. The "years of supply" column shows the result of a division of the estimate by current production rates, which is a function of the methodology used in the estimate, not the geologic abundance of the fossil fuel. All resource figures include proved reserves. **For comparison purposes, the Potential Gas Committee (1997) estimated 1044 tel of remaining gas resource (proved reserves plus probable, possible, and speculative resources). tElA= Energy Information Administration (1998), USGS & MMS =U.S. Geological Survey National Oil and Gas Resource Assessment Team (1995) and Minerals Management Service (1996), Averitt=Averitt (1975), and Merritt & Hawley=Merritt and Hawley (1986). McCabe 2119

1000 �------�Figure 11-Plot showing 1978 Annual Production and 1994 coal and lignite production from selected regions of the world and a tabulation of 800 the reserves estimated in 1984. The relative size of resources or reserves in different regions is not necessarily an indicator 600 of the proportion of future production that will come from those regions. These estimates of reserves in Europe, the fo rmer 400 Soviet Union, North America, and China were roughly comparable in size, but whereas coal production (histogram) has 200 fal len in Europe and the fo rmer Soviet Union over the last 20 yr, it has risen markedly in North America and China. The 1984 estimates appear to reflect Europe Fonner Soviet Union N. America China �elative rates of production at Reserves 'the time the assessment was made rather than the future xlo3Mt production potential. Reserve Hard coal 101 104 109 99 estimates from Ezra (1984). Production f"lgUres from United Lignite 36 61 86 - Nations (1978-1994). Total 137 165 195 99

oil and gas resource, they would greatly add to its reserve and resource base. In the late 1970s, pro­ magnitude. For example, Smith (1981) estimated duction of coal was roughly the same in four major that the oil shales of the Green River Formation in regions: Europe, the former Soviet Union, North Colorado, Utah, and Wyoming contain 1500 billion America, and China. Proved reserves calculated for bbl of oil. Duncan and Swanson (1965) estimated each region in 1984 also were similar (Figure 11). that there is over 160 trillion bblof oil in shales with­ Over the last 20 yr, however, there has been a dra­ in the United States if one includes the lower grade matic fall in coal production in Europe and in the oil shales. These numbers contrast to the 165 billion former Soviet Union, but a substantial rise in pro­ bbl of remaining technically recoverable crude oil in duction in North America and China. The demand the United States according to the latest estimates of for energy and availability of substitute fuels the USGS and Minerals Management Service (U.S. explain these profoundly different trends. Europe Geological Survey National Oil and Gas Resource may have a lot of coal, but much of it is uneconom­ Assessment Te am, 1995; Minerals Management ic. In 1995, for example, the German government Service, 1996). The amount of methane trapped in subsidized the German hard coal industry $119/ton hydrates worldwide is estimated to be 660,000 tcf produced (Roodman, 1997); for comparison, U.S. (Kvenvolden, 1993), which contrasts to the 9800 tcf hard coal currently sells for about $20/ton. By con­ of remaining conventional natural gas worldwide trast, China has a rapidly expanding economy and (Masters et al., 1994). Arguably, such accumulations somewhat limited oil and gas resources. The tie, or will be exploited long before many of the coal lack thereof, between resources/reserves and pro­ resources. duction histories is discussed for two specific Contrary to what one might anticipate, the pat­ examples in the next section. tern of exploitation of resources over time need Historically, perceptions of the amount of fossil not necessarily bear any relationship to the distri­ fuel resources have had a profound impact on many bution of calculated resource or reserve figures. political and investment decisions. In retrospect, Countries such as Saudi Arabia, Iraq, and Iran have many of those perceptions have been ill founded. In obviously not produced crude oil at their full the 1970s, for example, there was a belief that the potential for the last 25 yr (the years of maximum United States was running out of natural gas (Bupp production for these countries were 1977, 1979, and Schuller, 1983). Natural gas suffered from avail­ and 1974, respectively), despite their enormous ability problems in the regulated interstate market 2120 Cornucopia or EmptyB arr el?

25 Communicating Information on Resources to Nongeologists 20 Geologists generally have not conveyed informa­

.., tion about the distribution and economic signifi­ ¢:: 15 c: cance of energy resources to those outside our � profession in an effective manner. In particular, the 10 concepts of reserves and resources have been � widely misinterpreted by most nongeologists. Throughout the 20th century, the public has been 5 exposed to repeated fears that oil and gas resources will run out within a few years, but overall oil and gas supplies continue to be plentiful. Shortages 1960 1970 1980 1990 2000 and associated high prices have proven to be short-lived. By contrast, the public has been led to Figure 12-Annual production and consumption of nat­ believe that there is a great abundance of coal. In ural gas within the United States from 1950 to 1995. The some areas, however, coal mining has declined or excess of consumption over production has been met almost died out, despite enormous resource and by imports, primarily from Canada. Note how gas con­ reserve figures. Unrealistically high or low resource sumption fell from the mid-1970s to the mid-1980s, largely as a result of a perceived shortage of natural gas. and reserve figures no do':lbt have led to much pub­ Data fromEnergy Information Administration (1998). lic angst and unwise decision making by politicians, government bureaucrats, fmanciers, and company executives. The comucopian viewpoint is based on and, at the same time, reserve and resource esti­ a lack of faith in resource and reserve figures. It is mates were relatively low. As Weaver (1981, p. probably time that geologists abandon the notion 17) wrote in a special National Geographic issue that there is a finite amount of any energy resource on energy, "If unconventional sources should not that is quantifiable in a form that has any societal come through, gas reserves would be gone in significance. another ten years at recent rates of use, without The warehouse comparison to resources and any further additions. And the new finds of the reserves offers an intriguing possibility for an alter­ past decade have averaged only about half of nate method of resource assessment. Although what we consumed." Many policies and invest­ geologists are perennially asked how much of a cer­ ment decisions were based on this perceived tain resource is present in a certain area, the real shortage in natural gas. Provisions of the 1978 question that most nongeologists have is where the Public Utility and Industrial Fuel Act, for exam­ resources are and when will (or could) they be ple, required the conversion of gas- and oil-fired extracted. To better answer that question, it may be generating units to alternate sources if they did more appropriate to identify resources in a tempo­ not meet a stringent cost-benefit test, and required ral framework, as suggested by Lewis (1986). One that utilities not use natural gas after 1990 can predict demand for fossil fuels over the next (Hughes, 1995; Means, 1995). Largely as a result few decades with confidence limits widening of such policies, natural gas consumption in the through time. The challenge for the geologist is to United States fell from a high of 22 tcf in 1972 to a predict where those resources will come from. An low of 17 tcf in 1983 (Figure 12) . Meanwhile, assessment might usefully identify the 10, 20, and coal was seen as abundant, and its use was 50 yr resources, thus identifying the resources encouraged. Coal production in the western most likely to be extracted over lengths of time that United States rose more than 400% between 1973 have some relationship to various types of decision and 1986. The perception of a shortage in natural making. Land-use planners, for example, have a gas has subsequently changed, and restrictions on much longer term perspective than do investment its use have been lifted. By.1996 consumption companies. This approach essentially would blend had rebounded to 22 tcf, although domestic pro­ the concepts of reserves and resources into a con­ duction has risen only to 19 tcf (Figure 12) tinuum. Several advantages are evident with such a because of the competitive advantage of import­ methodology. This methodology would result in ed Canadian natural gas. In 1970 more than 24% assessments of oil, gas, and coal in a similar format of the electricity produced in the United States that would allow a better comparison among the was from natural gas. That percentage fell to 9.5% different resources and would allow more realistic in the early 1990s, but now 37% of the planned evaluations of the geologic factors that will control growth in electrical generation through the year the future evolution of the energy mix. Of course, 2003 will come from natural gas (Hughes, 1995). we cannot precisely predict the demand for each McCabe 2121

4 .------�Figure 13-U.S. crude oilproduc­ tion from 1870 to 1996 and Hub ­ bert' s (1956) predictions of U.S. production. Hubbert made two predictions based on his high 3 and low estimates of remaining resources in 1955. Data on '"" production are from Energy s.J:J hUormationAdministration .J:J ...... (1998), Bureau of Mines 0 2 (1924-1993), and U.S. Cf.l = = Geological Survey(188�1923). 0 J 3 :a"B � � 0.. 1 \0 "' -0'\ ..... 0 v 8 E= 0 0 0 0 0 0 1- 00 0'\ N (") "' 00, 00 00 0'\8 0'\� 0'\ 0'\ 0'\� 0'\ /" ------§N

fossil fuel in the future, and there are many nongeo­ exhausted over a period of time that is determined logical factors that will determine which resources by the size of the resource and the rate of produc­ will be used and over what time period; however, if tion. He suggested that an idealized cycle of pro­ an assessment ranked the potential resource sup­ duction would be bell-shaped when plotted against plies, it might provide a powerful tool for use in time (Figure 1), with a long period of rising pro­ developing scenarios for future energy use. duction followed by a long period of falling produc­ tion because it takes time to assemble and disman­ tle the infrastructure to produce a fossil fuel. DRIVING FORCES BEHIND PRODUCTION Although some production curves do have a shape IDSTORIES that approximates a normal distribution, there is no inherent reason why production should follow M. King Hubbert's�edictions such a pattern. Some production curves show strong asymmetry or polymodal patterns (Figure Hubbert's projections of oil and gas production 14). Hubbert (1956) himself cautioned that produc­ have almost legendary status within the American tion curves may diverge strongly from an idealized geologic community. Ivanhoe (1995, p. 86), for shape, and cited the production of crude oil in example, stated that "The only truly valid scientific Illinois as an example. projection of future oil production yet made was The classic Hubbert curve of energy production that by M. King Hubbert in 1956." Amid consider­ resembles the normal curve of statistical analysis. able controversy, Hubbert anticipated that U.S. oil Scientists and economists are accustomed to data and gas production would peak and subsequently that show a normal distribution, and our familiarity decline. His predicted production curves have a with the bell-shaped curve may subliminally sug­ patternsimilar to actual production curves over the gest some statistical rigor to the Hubbert curve last 40 yr (Figure 13). Hubbert deserves much cred­ analyses. Remember, however, that the Hubbert it for his predictions, which were based on more curves are not normal distributions in the classic realistic criteria than those used by many of his sense, where frequency is plotted against some contemporaries, but the conclusion of some that property of a sampled population. Rather, the Hubbert's methodology and assumptions must Hubbert curvesshow a physical property plotted have been correct because his predictions were against time. In a classic normal distribution, the "accurate" is not logical. This section briefly observations are not dependent on one another, reviews Hubbert's major assumptions as presented whereas in a time-series plot of energy production in his papers (Hubbert 1956, 1967, 1969, 1976, there is clearly a dependency of one year's produc­ 1981; Hubbert and Root, 1981). tion on previous years' production. Hubbert believed that resources are finite, and Hubbert's basic assumption is that the area that all resources are depleted and eventually under a complete production curve, the "ultimate 2122 Cornucopia or Empty Barrel?

estimates to determine the ultimate cumulative production. Consideration of Figures 6, 8, 10, and 14 shows the folly of such approaches. An underlying assumption in the Hubbert analy­ ses is that once a resource is found it will be expe­ ditiously extracted. This perspective was based on the history of the oil industry. Until the mid- 1970s, as each new field was found there usually was a ready market for the oil, as well as the technology to economically develop the resource. Conse­ quently, new fields were rapidly developed. It seemed reasonable to assume that all the oil ever found would be consumed. A different perspective might have been developed from a study of the coal or gas industry. Many coal deposits are well known but have not been developed, and because of the size of the resource, it is far from certain that all the coal will ever be produced. Many gas resources have been known for decades before commercial production began, and much natural gas has been flared for lack of a market. The timing of extraction of a resource is driven by market and political forces and technology availability. The importance of these forces in determining produc­ tion curves is apparent in the following examina­ tion of the history of four fuels for which there is a long-term record of production and price. These Figure 14-Not all production curves approximate a historical perspectives form the basis of a theoreti­ bell-shaped curve, as is evidenced by production curves cal model to explain the relationship between pro­ of crude oil in Ohio and Dlinois. Hubbert (1956) indicat­ duction and price of fossil fuels. ed that although a bell-shaped curve could be used as an idealized model, many production curves are more com­ plex. He used the Dlinois curve to 1955 to demonstrate Coal inEngland from 1550 to 1700 the point. Ironically, he used the Ohio curve to 1955 as an example of a production history that approximated The early use of coal in Britain has recently been hisidealized curve, and regarded Ohio's oil resources as virtually exhausted. Data on production are from Ener­ documented by Hatcher (1993), who is the source

gy Information Administration (1998), Bureau of Mines for the information in this section. Between 1550 (1924--1993), and U.S. Geological Survey(1883-1923). and 1700, there was a dramatic increase in British coal production. Coal began to be transported along the coast and by river to cities, where it was cumulative production," would be equal to the size used for such diverse purposes as domestic heat­ of the resource. A fundamental problem with this ing, making lime, and brewing. Much of the trans­ assumption is that, as was demonstrated in the pre­ ported coal came from northeast England, from vious section, a finite resource cannot realistically where coal shipments grew from around 50,000 be measured. All assessments have a strong subjec­ tons/yr in the mid-16th century to close to 700,000 tive component and, at best, can only identify tons/yr by the end of the 18th century. The reason those resources that appear to have some econom­ for this rapid growth was the increasing price com­ ic potential within the foreseeable future. Hubbert petitiveness of coal relative to the renewable fuels adjusted his own projections in his series of papers that had previously been the primary source of as more data became available. In 1956, for exam­ heating. Deforestation of the English countryside ple, at the time of his initial projections, there had and rapid population growth led to significant been virtually no exploration in Alaska or in most shortages of firewood in the second half of the of the offshore regions of the United States. Several 16th century, as is evidenced by many laws that workers (e.g., Gever et al., 1986; Smith and Lidsky, were enacted to conserve woodlands. The price of 1993; Campbell, 1997; Edwards, 1997) have subse­ firewood rose 400% between 1540 and 1620, com­ quently updated Hubbert's predictions either by fit­ pared to a 300% rise in the cost of living over the ting idealized curves to more up-to-date production same period. By the end of the 16th century, many information, or by using more recent assessment consumers had abandoned firewood as a fuel and �cCabe 2123

were using turves (peat), sedges and reeds, and charcoal for heating. The price of these renewable 300 f, fuels, however, rose at a faster rate than coal Turves (peat) I ""' (Figure 15), and coal eventually gained a substantial price advantage. The rate of rise in the price of coal /'J ' was close to the overall rise in cost of living, and 8 - 250 I \ thus was essentially a constant price throughout II \ � the 17th century. Although firewood and other u renewable fuel provided the vast majority of ;f I SedgW,eod• N 0'1 England's energy supplies in 1550, by the second 11'1 200 )/ I Charcoal half of the 17th century coal was meeting over half u 1\ / u I�---- of the country's fuel requirements. This dramatic ·c ll.. I change took place despite the general view that � / ,/ / I > coal was an inferior fuel that emitted noxious I .i /_ fumes, and that the changeover from wood to coal � 150 � / required conversion of hearths and flues. I / -.//,f I I 7-- - / ---.,/,:

Whale Oil in 19th CenturyAmerica 100

From the 17th to the 19th century, whale oil was used extensively for home illumination, street 1600 1620 1640 1660 1680 1700 lighting, and machine lubrication. The whale oil industry has been documented by Bockstoce Figure 15--T he relative price of fuels supplied to Kings College, Cambridge, from 1592 to 1702. Note how the (1986), Coleman (1995), and Ross (1985), which price of coal rose at a substantially lower rate than that are the primary sources for this section. of renewable fuels. Modified from Hatcher (1993). Throughout the 19th century, the United States dominated the world's whaling industry. Initially, the American fleet harvested the rich whaling discarded the rest of the whale. Although the discov­ grounds of the North Atlantic but, as those areas ery of crude oil occurred at a fortuitous time, it is became depleted in whales, the whalers began likely that other substitute fuels would have been expansion into the South Atlantic in the 1750s, found to replace whale oil as its price rose. For entered the Pacificin the 1790s, and had reached example, in the early 19th century, prices for whale Hawaii by 1819. This renewable resource was oil plummeted in Britain as the country began to proving to be an exhapstible, if not finite, resource. manufacture oil and gas from its rich coal resources. A plot of American whale oil production in the 19th century is a bell-shaped curve, with peak pro­ duction around 1850 (Figure 16). The subsequent Pennsylvania Anthracite fall in production reflects the increasing difficulty that American whalers had in finding whales. For 100 yr, Pennsylvania anthracite was a major Whale oil prices had fluctuated throughout the component of the U.S. ene:rgy mix. The history of 18th century due to the effects of poor whale har­ Pennsylvania anthracite has been documented by vesting seasons and wars, but there was a steady Eavenson (1942) and the Hudson Coal Company rise in prices between 1850 and 1865, reflecting (1932), which are the primary sources for this sec­ increasing demand and growing scarcity (Figure tion. Until 1920, when it was overtaken by petro­ 17). The first discovery of subsurface crude oil leum, anthracite was the second most important occurred in Pennsylvania in 1859, just as a major fossil energy source in the United States, and the shortage in whale oil loomed. Commercial develop­ vast majority was mined in Pennsylvania. Today, ment of crude oil was rapid (Figure 16), and within production has almost ceased (Figure 18). The pro­ less than 10 yr caused a dramatic drop in demand duction curve of Pennsylvania anthracite is perhaps for whale oil and a consequent drop in prices the closest to a full-cycle Hubbert-style curve of any (Figure 17). Some demand for whale oil continued fossil fuel. The only significant variations from a through the rest of the century for a variety of uses, smooth curve are due to coal-miner strikes and the including softening leather and as a lubricant in somewhat lower production during the Depression rope making, but whale oil became a by-product of and the higher production during World War II. whaling as the focus moved to baleen. By the turn Following Hubbert's assumptions, it might be sur­ of the century, the value of whale oil was so low mised that virtually all the anthracite resources of that many whalers extracted only the baleen and Pennsylvania have now been extracted; however, 2124 Cornucopia or Empty Barrel?

"'"' 12 � 1.60 (I) ;::::J 5 .._, 1.40 10 = "' ..9 1.20 0- 'a 0- 4 X c:l 1.00 X 8 "' ... "' 0.80 'i) �II) u 3 Ill] 6 - 0.60 Ill� if - 0 II) II)

Figure 17-Wholesale price of whale oil in the United 0 0 States from 1849 to 1913. From datain Bockstoce (1986). 1800 1820 1840 1860 Prices are shown in actual dollars; there was little infla­ tion over this period. Figure 16--Annual whale oil and crude oil production in the United States between 1800 and 1875. Note the markedly different scales for the two energy sources. Note that the two resources are measured in different Coal Company (1932, p. -378) stated: "Oil and gas ways (a standard barrel of whale oil contained 30 U.S. also compete with anthracite for domestic heating gal compared to the 42 gal in a standard crude oil bar­ purposes, but careful studies have shown that in (1995); rel). Whale oil production is from Coleman the great majority of cases the cost of these fuels is crude oil production is from U.S. Geological Survey considerably higher than anthracite. Oil and gas are (1883-1923). luxury fuels and their main appeal to the consumer is that of convenience." While the anthracite indus­ try worried about imports of anthracite from Arndt et al. (1968) estimated 19 billion tons of Britain and the possibility of Russia dumping remaining identified resources of Pennsylvania anthracite at "prices ruinously low," it was the com­ anthracite, and only 156.3 million tons were pro­ petition from alternate fuels at increasingly com­ duced between 1968 and 1995. The Energy petitive prices that brought the end to anthracite as Information Administration (1996) identified a a significant fuel for domestic heating. reserve base of 7.2 billion short tons as of ]anuary 1, 1993, but production in the 1990s has averaged less than 4 million tons/yr. �ven allowing for con­ British CoalIndustry from 1830 to 1995 siderable waste of resources and reserves in excess of the production numbers, these figures indicate The primary sources for this section are books by that the resource is far from depleted. The history Church (1986), Supple (1987), and Ashworth (1986). of the price of Pennsylvania anthracite (Figure 19) Britain was the world's leading producer of energy also is revealing. One might have expected a steady, for much of the 19th century because of its coal min­ ifnot exponential, rise in price as the resource was ing industry. Before 1870, Britain produced more depleted. Instead, the price has been relatively than 50% of the world's coal, and it was the world's steady when adjusted for inflation. The price in leading producer until 1900. Coal fueled the rapid 1995 was essentially the same as in 1924, near the industrialization of Britain and allowed it to build an peak of production. The overall steady price of empire, but British coal production peaked in the anthracite is in line with other energy prices in the early decades of the 20th century and since hasfallen United States (see Figure 2) and suggests that com­ dramatically (Figure 20). petition with alternate fuels was a critical factor. There is a widespread belief that Britain's coal The production curve for Pennsylvania anthracite resources are now depleted, but the facts are rather reflects the demand for anthracite within the east­ different, and the history of resource appraisals ern United States Anthracite's main use was for (Figure 21) is revealing. The sharp fall in reserve domestic heating for which it was ideal because it estimates between 1915 and 1946 partly reflected produced high heat and little smoke, and burned sterilization of coal resources due to mining. Mine longer than other coals. The expanding urban pop­ plans, especially the old room-and-pillar design, u1ations of the 19th century provided a ready mar­ inevitably leave a substantial amount of coal in the ket for Pennsylvania anthracite. By the early 1930s, ground, and in most cases it wou1d be very costly competition from alternate fuels had begun to bite to later recover any remaining coal from an aban­ into the domestic heating market. The Hudson doned mine. The decline in reserve figures in the McCabe 2125

100 80 � 11"1 0'1 0'1 "' 80 - c 60 .s � c t: 60 0 0 E-o .c Cl.l t: 40 c 0 .c 40 Cl.l ... � QJ Q., � QJ 20 20 (J if 0 0 1840 1860 1880 1900 1920 1940 1960 1980 1900 1920 1940 1960 1980

Figure 18--Annual production of anthracite in Pennsyl­ Figure 19-T he price of Pennsylvania anthracite from vania from 1827 to 1995. Although this production 1900 to 1995 adjusted for inflation. Although the produc­ curve strongly resembles a complete idealized Hubbert tion curve (Figure 18) may suggest that the resource has curve, the resource is far from exhausted. Data from been depleted, the price curve is similar to that for Energy Information Administration (1998), Bureau of crude oil (Figures 1, 2). The high prices in the late Mines (1924-1993), U.S. Geological Survey (1883--1923), 1970s and early 1980s reflect world events (Figure 2) and Eavenson (1942). rather than any geologic factors in Pennsylvania.· The apparent sharp increase in price in 1922-1923 is due to a significant deflation in the cost of living created by a 1965 and 1969 reserve estimates ref lected in­ marked fall in the price of agricultural products that resulted from surpluses following World War I. The creased mechanization of the mines and a view that real price of anthracite and other nonr«;newable much of the earlier estimated coal reserves were no resources remained relatively constant during that longer economically recoverable. In contrast to the period. Data from Energy Information Administration oil and gas industry, where increased technology (1998), Bureau of Mines (1924-1993), and U.S. Geologi­ makes more of a reservoir accessible, mechaniza­ cal Survey (1883-1923). tion of mines tends to make less coal accessible. Not only is the machinery designed for coal beds of a certain thickness, but it is difficult to operate figures of the 1960s assessments was that they large-scale mining equipment in areas with com­ "were directly related to the low priority given to plex geology (e.g. , where coal beds are cut by coal during that period and bore no relation to the faults or channels). Increasing mechanization made reserves actually in the ground." If one substituted it more critical to distinguish those coals that lay "high priority" for "low priority," the same state­ higher in the resource pyramid (see Figure 6), ment could be made of the 1978 assessment. Much because these could be extracted at a much lower lower, and arguably more realistic, reserve figures cost than coals lower in the pyramid. By contrast, in were calculated in the early 1990s, prior to dena­ earlier times, when coal was hand-hewed, the cost tionalization of the British coal industry. The evolu­ differential was not great. The successive assess­ tion of these coal assessments demonstrates that ments from 1915 to 1969 drew the cutoff plane resource assessments can be primarily controlled higher in the resource pyramid. The increase in by perception rather than unbiased objectivity. reserves in 1978 was due in part to an active explo­ In hindsight, the 1978 assessment can be seen as ration program in the 1970s, which identified sever­ a failure to identify coal that had proved to be eco­ al new major accumulations of coal. A significant nomic over the last 20 yr, but it shows that the aspect of this more optimistic assessment, however, marked decline in British coal production is not for was the perceived need for more coal in Britain. want of coal . The 1978 assessment identified Ezra, as chairman of the National Coal Board, was a almost twice as much "recoverable" coal than had prominent advocate for coal who regarded the coal been mined in Britain to that date. As Ezra ·(1978) industry as essential to Britain's future (Ezra, 1978, pointed out, even that figure could have been very 1984). The 1978 assessment suggested that, at substantially increased if it had included coal in 1978 rates of production, the country had reserves thinner seams, coal below 4000 ft depth, and off­ that would last for well over 300 yr (Ezra, 1978). shore coals. Oil and gas exploration in the North The 1978 assessment cutoff was clearly lower in Sea has also identified extensive coals. Carbon­ the resource pyramid than the assessments of the iferous coals extend throughout much of the 1960s, because the need for coal was thought to southern North Sea, and there are also thick coals have increased in the wake of the 1973-197 4 ener­ in the Permian-Triassic, Jurassic, Cretaceous, and gy crisis. Ezra's (1978, p. 82) perception of the low Tertiary (Knight et al., 1996). We could mine 2126 Cornucopia or EmptyBarrel?

300.------. 200�------, -- Ultimate reserve 180 estimates "' 160 = 0\ 200 � 140 �(J X ·.s � 120 � ::E 100 �(J ·c:: � 100 � 80 � � 60 40 Cumulative production 20 0 0 0 ...... _ C'l -.:!' 00 0\ 0\ 0\55 0\ - - - §C'l 0 0 0 \Q 00 00 00 80\ Figure 20-Production of coal in the United Kingdom - from 1830 to 1996. The dominant energy source in the world for much of the 19th century, British coal produc· Figure 21-Plot of cumulative production and ultimate tion has steadily declined in the 20th century, but not reserve estimates of British coal from 1830 to the pres­ for lack of coal in the ground. Data from Bureau of ent. Production data from Church (1986) and data in Mines (1924-1993), U.S. Geological Survey (1883-1923), Bureau of Mines and U.S. Geological Survey reports. Church (1986), and data supplied by U.K. Department of Reserveestimates from Royal Commission on the Coal Trade and Industry. Years with abnormally low produc· Industry (1925), Ezra (1978), and John T. Boyd Compa­ tion correspond to miners' strikes. ny (1993).

much of the deeper, thinner, and offshore coal so that by 1967 transport was a negligible market with the technology available today, but obviously for coal. Gas produced from coal was a significant such ventures would be very costly and could not component of British energy use for over a century. produce a product that was price-competitive in This gas was used for street and domestic lighting, today's market. Britain, therefore, has abundant heating, and cooking. In 1965, however, natural gas amounts of coal remaining in the ground. How was discovered under the North Sea, and within 7 much of that coal should be categorized as yr virtually all gas appliances had been converted resources or reserves is a matter of perception of to burn the new fuel. Natural gas also made it fis­ what will be economic in the future and over what cally attractive for many consumers to convert time frame. their domestic heating from coal to gas. Another An understanding of the controls on the rise and important market for British coal was in exports. fall in British coal production can be gained by Although Britain had been the world's leading examining the way in which the coal was and is exporter of energy during the 19th century, that consumed (Figure 22). Over 90% of the coal is now market rapidly disappeared as other countries used for electrical power generation, although its exploited their own coal resources or substituted use in that sector has recently fallen because of other cheaper energy sources. The bell-shaped competition from natural gas. The primary use for curve of British coal production (Figure 20) thus most of the last 170 yr, however, has been for can be viewed as the sum of the various consump­ industrial purposes, and the rise and falls of these tion curves (Figure 22). The rise in production was consumption curves reflect the increasing industri­ a response to growing demand for energy, whereas alization of Britain up to World War I and the subse­ the fall was a product of a drop in demand for quent deindustrialization. At the beginning of the energy for some purposes and substitution of 20th century, coal fueled all mechanized trans­ cheaper energy for other uses. Of the consump­ portation. By World Wa r II, most ships had convert­ tion curves, only that for exports and industry ed to diesel, and the railways converted to diesel or other than iron and steel approximates a symmetri­ electric traction in the late 1950s and early 1960s, cal bell-shaped curve. Ifthe coal industry had been

Figure 22-Consumption of British coal by sector. The rise and fall of the British coal industry (Figure 20) has been driven by the demand fo r British coal rather than any geologic limits on production. The decline in consumption in some sectors has been dramatic over a short period of time when an alternate cheaper fuel became available. Data fr om Church (1986), Supple (1987), Ashworth (1986), British Geological Survey (1992), and data supplied by U.K. Department of Trade and Industry. Million Metric Tons Million Metric Tons Million Metric Tons Million Metric Tons ...... VI ...... IV ...... IV Vl 0 IV 0 0 0 0 0 0 0 0 0 0 0 0 I I I 0 - I I I I I (j � ...... 1840 0 1840-l \. 12.18 40 ::s lMO ...... c:n� P' 0 ::s K "'tj If 1860 1860 1860-l \ I �· 1860 0 0.. c:n :4 en ..... 0 1880 0 1880 1880 1880 ......

1900 1900 1900 1900

1920 1920 1920 1920

1940 1940 1940 1940

1960 1960 1960 1960

1980 1980 1980 1980

2000 2000 2000 2000

Million Metric Tons Million Metric Tons Million Metric Tons Million Metric Tons 0\ ...... Vl IV .f:>. IV .f:>. IV .f:>. 0\ 00 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 � tr1 0 0 P' 1840 0 1840 � 1840� 10'18 40 c:n ::s (") 0 0s ..... c:n P' s...... 1860 � 1860 ... . 1860 1860-f In (") - e. ::s (")§" 0...... c:::: -· til 1880 c:: 1880-l 1880 1880 0 c:n I�::s 0 ::s q 0 '< 1900-l\ 1900 1900 1900 I�0 . ::s 1920 19201 \ I 1920� � I 19201 � I a:: 1940 n 1940-l ""'- I 19401 "\_ I 1940 -i / I 1 ..c:- I r'l � 1960 1960 I A' I � l9W � j 1980 1980 198 1980 N '""' N 2000 ·� 2000�- y------2000 � 2000 ..... 2128 Cornucopia or Empty Barrel?

18�------, The open-market model (Figure 24) assumes that

0 16 there is a variety of available energy sources and 0 builds on the concept of the resource pyramid 14 -; (Figure 6). New energy sources usually are devel­ 12 �0\ oped to meet a local or niche demand. As the new - 8 10 energy source is developed, its costs tend to fall ·s. 8 because of technological innovation, economies of scale, and the development of efficient systems to ] 6 ti transport the fuel to market. With lower costs, the :I :o' new energy source may become price-competitive < with the existing energy source (or sources). If the lower prices are sustained, the new energy source eventually would replace all or most of the market share of the preexisting energy source. Substitution is especially likely to take place if the preexisting Figure 23-The price of British coal from 1830 to 1996. energy source is becoming scarcer or more expen­ The price is normalized so that 1990 price "" 100, and sive to extract. The low cost of the new fuel may prices from 1900 to 1996 have been adjusted to account itself stimulate increased demand for energy, which for inflation based on the consumer price index (there was relatively little inflation in the 19th century). Com· will further increase production. Eventually, how­ pare this trend to that for U.S. coal as typified by Penn­ ever, the energy source becomes victim to a sylvania anthracite (Figure 19). Data from Ashworth renewed cycle of substitution as an alternate (1986), Church (1986), Supple (1987), and datasupplied source is developed. The open-market model, by U.K. Department of Trade and Industry. therefore, may result in a bell-shaped energy pro­ duction curve where the decline in production is unrelated to depletion of the resource. The decline based on a single market (e.g., transport or domes­ part of the curve may be steeper than the rising tic heating), its production curve would have been part of the curve because consumers tend to rapid­ strongly asymmetrical. ly switch to cheaper alternative fuels. The real price of British coal rose steadily The closed-market model (Figure 24) assumes throughout much of the 19th and 20th centuries at that there is only one source of energy available, what appeared to be an exponential rate until the and that its supply is finite. It is essentially the 1980s, when the price began to fall rapidly (Figure model followed in Hubbert-type analyses. The ini­ 23). The rise in price was presumably because of the tial development of the resource may result in insular nature of the British energy economy until falling prices for the reasons previously presented. the latter part of the 20th century. In 1965 coal still Prices may then remain flat for a time, if the total accounted for more than 90% of energy consump­ consumed is a small fraction of the total resource, tion exclusive of the transport sector. Until the dis-:. but the price inevitably will rise as the prime covery of North Sea gas and oil, coal was essentially resources are extracted and it becomes necessary the only fossil fuel domestically produced. Blockades to extract more inaccessible or lower grade during both world wars had a profound impact on resources (i.e., moving lower down the resource British perspectives, and the idea that Britain should pyramid). As the resource is depleted, the price be as energy independent as possible was a notion rises exponentially. The production curve may be that prevailed until the Thatcher era. By the early bell-shaped if the market can initially bear the ris­ 1980s, imported coal was cheaper than that pro­ ing prices, but inevitably demand and production duced domestically, and today the United Kingdom will fall as prices rise quickly and the resource is imports about 20% of the coal it consumes. The rise depleted. The decline part of the curve may be pro­ in British coal prices took place, therefore, in a mar­ longed if the rising prices effectively ration usage ket where there was little in the way of substitute for high-priority or essential needs. fuels available. The open- and closed-market models are idealis­ tic, and most, ifnot all, energy production histories show some elements of both models. Most produc­ Long-Term Production and Price Trends tion histories, however, appear to have stronger affmities to the open-market model. The history of The foregoing production histories provide a use­ British coal production does resemble to the ful background to two theoretical models proposed closed-market model because of the insular nature here that explain the relationships between produc­ of the country and past government policies. The tion and price over time. One model is for an open open-market model explains the long-term trend market, and the other model is for a closed market. for steady or falling prices of energy in the United McCabe 2129

Open Market Closed Market

\ Price of substitute \ energy source ,__.-- Economies of \ scale allow " falling prices Depletion results "' in higher prices ' energy source

------

Time Time

(1) (1) � � � � Cheaper � = /; � = Depletion of 0 U,wer pric substitutes 0 ...... resource � ..... u stimulate result in u ::I ::I limits demand decreased "'tj "'tj production demand �e �e

Time Time

Figure 24-Idealized relationship between price and production of an energy resource in an open market (left) and a closed market (right). The substitution of one energy resource for another due to lower prices (open market) is historically shown in thesubstitution of renewable fuelsby coal in 17thcentury Britain (Figure 15) and of whale oil by crude oil in theUnited States (Figures 16, 17). The presenceof an open market explains the steady price of Penn­ sylvania anthracite (Figure 19), whereas the closed market explains the dramatic rise in price of British coal until the mid-1970s (Figure 23), when the British energy market began to more closely resemble the open-market model, although both fuels show a classic bell-shaped production history (Figures 18, 20).

States, despite rising demand over time. Coal substi­ and U.S. crude oil, that already had reached a relative­ tuted for firewood; petroleum substituted for ly mature level of development. Ifthe United States whale oil fo r illumination purposes, and eventually did not now import significant amounts of crude oil, for coal fo r transportation; gas substituted fo r coal the price fo r domestically produced oil might be in domestic heating; and coal and natural gas cur­ much higher because of the costs that would be asso­ rently compete fo r the electrical generation sector. ciated with increased drilling programs and extrac­ The last few decades have seen the rapid globaliza­ tion fromnoncon ventional sources, such as oil shales tion of energy production. Figure 25 shows the dra­ (i.e., extracting oil from further down the resource matic rise in world production of crude oil and nat­ pyramid). On the other hand, the necessity of devel­ ural gas, a production that was, until the 1950s and oping alternate sources of energy might well have 1960s, dominated by the United States. A similar led to much lower prices for some substitute fuels. increase has occurred in world coal production. Numerous factors made this rapid expansion possi­ ble, including technological advances in offshore THE FUTURE OF THE OIL AND GAS INDUSTRY drilling, much-improved transportation capabili­ ties, relative political stability, and the increased Time-series graphs commonly are used by those wealth of the world's population. This increase in predicting the futureof energy production and prices. world energy production provided fuels that could Cornucopians believe in regression to the mean substitute fo r energy sources, such as British coal fo r energy prices, whereas the neo-Malthusians 2130 Cornucopia or EmptyBarre l?

70 and neo-Malthusians shows, however, that even yes­ Crude oil terday's normality is poorly understood. Hopefully, 60 the discussion of the nature of resources and reserves, and the production histories presented in 50 this paper, shed some new light on the topic. In light of the data and historical perspectives presented in >. "' 40 this paper, it is worthwhile to reexamine some wide­

;g.c ly held beliefs and opinions about the future of the .c 30 energy industry. This reexamination is done by dis­ cussing four statements that are typical of many that 20 have been made over the last decade.

10 "A Sharp Rise inEnergy Prices Is Inevitable 0 Because We Are Rapidly Depleting Energy 0 0 0 0 0 0 0 0 0 0 ..... C"' 11"1 \0 r-- 00 0\ Resources"-True or False? 0\ 0\ 0\ 0\ 0\ ..... �- ..... 0\ ...... 0\ .....0\ 80 History indicates that in most cases energy prices Natural gas in real qollars, similar to other commodity prices, fall 70 over the long term despite rising production; howev­ er, we also have examples of sustained rising prices 60 in markets with a limited availability of substitutable .. 50 energy sources. The reduction in trade barriers and globalization of trade, fo r the most part, have created $¢:: = 40 open-market settings for energy in most of the major � countries of the world. The critical question is :c 30 E-o whether the trend of substitution of energy sources 20 U.S.A. will continue indefinitely, allowing energy prices to remain steady or even fall. If the world remains 10 dependent mainly on fo ssil fuel, then clearly it willbe necessary to exploit resources farther down the 0 resource pyramid through time. Economies of scale 0 0 0 0 11"1 \0 r-- 00 0\ and lower prices due to technological advancements 0\ 0\ 0\ 0\ 0\ ...... eventually may be outweighed by the higher costs associated with extracting lower grade resources. Figure 25--Until the mid-1950s, over halfof the world's crude oil was produced in the United States, and over The planet itself is a closed fo ssil energy market, and half the world's natural gas was produced inthe United sooner or later one might expect it to follow the States before the early 1970s. The rapid increase in closed-market model of exponentially rising prices. world production of fossil fuels and the removal of There appears to be little sign yet that prices have trade barriers have provided an open-market setting started a consistent rise. In my opinion, long-term where competition from imports has helped keep price trends will continue to be level, or even fall prices of domestically produced fuels steady. Data from when adjusted fo r inflation, because the size of fo ssil Energy Information Administration (1998). energy accumulations is so vast and because substitu­ tion of energy sources is likely to continue. For crude oil this suggests long-term prices in the range of $13 worryover exponential plots of energy consumption to $16 per barrel at 1998 prices. and have faith in Hubbert's bell-shaped curves of Inevitable fluctuations in price will be related energy production. Risk-management studies, how­ to short-term imbalances between supply and ever, strongly suggest that it is dangerous to predict demand due to factors such as wars, boycotts, the future from the pattern of any time-series trend. weather conditions, and the political instability of As Bernstein (1996, p. 172) said, "yesterday's normal­ some countries. Such price changes probably are ity may be supplanted today by a new normality that not sustainable fo r more than 3 or 4 yr. High prices we know nothing about." We can only speculate provide an economic incentive to develop new about future changes in such important areas as eco­ energy sources, or to develop methods of using nomic growth, global politics, and technological energy more efficiently. In the long term, develop­ innovations that will surely shape the future of the ment of resources farther down the resource pyra­ fo ssil fu el industry and the need fo r energy mid, or the development of alternate (nonfossil) resources. The debate between the cornucopians energy resources, provides a cap on energy prices McCabe 2131

because these fuelswill substitute fo r what we now energy should have a symmetrical bell-shaped curve. regard as conventional energy sources. In addition Many production curves have complex patterns (see to short-term periods of high prices, there may be for example, Figure 14), although some do times when prices are substantially below those of approximate to a bell-shaped curve (Figures 18, the present. It is worth remembering that four of the 20). Production histories, in large part, are deter­ countries with large oil reserves, Russia, Iran, Iraq, mined by demand for the fuel, and consumers sub­ and Saudi Arabia, are currently producing crude oil stitute one fuel for another over time as prices at well below their capacity; however, low prices fluctuate. Substitution can occur over a short time are not sustainable either. Many operations would period. The history of British coal consumption become unprofitable at low prices and, as they leave (Figure 22) shows how rapidly some of these substi­ the market, the overall supply would decrease, thus tutions can take place. The production of those fuels raising prices to close to their current level. The key that are primarily used by one consumer sector can to survival in the energy patch fo r individuals or be particularly affected by substitution. At present, companies is to be adaptable and to have plans to about 65% of petroleum in the United States is con­ capitalize on periods of high prices and plans to sur­ sumed by the transportation sector. If other fuels vive periods of low prices. become cheaper to use in vehicles, it could cause a dramatic fall in demand for crude oil. Coal is even more dependent on a single sector: About 88% of "Those Countries with the Most Oil Reserves coal is used by electric utilities. Natural gas has a Will Control the Wo rld's OilMarket in the more diversified consumer base. The presumed Coming Decades"-True or False? decline in U.S. crude oil production will be deter­ mined by the future availability of cheaper fuels fo r Estimates of reserves and resources are percep­ transportation, either imported crude oil or alter­ tions of quantities of fo ssil energy that will be nate fuels such as natural gas or electricity. extracted in the future. That does not mean that they will be extracted. Other resources, not includ­ ed in the estimates, may be extracted well before "North America Is a Mature Area inRegard those reserves and resources are exhausted. Figure to Petroleum Exploration and No Large Oil or 11 shows that there has been no strong correlation Gas Fields Remain To Be Discovered"- between the size of a region's coal reserves and the True or False? region's subsequent coal production. Some coun­ tries, particularly in the Middle East, have likewise Compared to much of the world, North America yet to realize their full potential fo r crude oil pro­ has been extensively explored. Prospects fo r finding duction as indicated by reserve and resource esti­ giant oil and gas fields may appear to be much mates. Coal reserves may represent thousands of brighter overseas; however, plenty of oil and gas years' supply at current production rates, but remains to be discovered on this continent. The unless alternate (nonfossil fuel) energy sources uppermost part of the resource pyramid has been remain economically uncompetitive, which seems found and exploited, but thereremains a large part highly unlikely, the vast majority of that coal seems of the lower pyramid that may be exploited in the destined to remain in the ground. Likewise, some future. The oil shales of the United States and the oil part of the world's estimated conventional oil and sands of Canada could be considered as the lower gas resources may well never be extracted, as part of the crude oil pyramid: resources that dwarf cheaper alternatives become available. With steady conventional resources in size, but which are rela­ or falling prices, it makes little sense to curtail pro­ tively expensive to extract. However, plenty of the duction now in the hope of reaping higher rewards more "conventional" resources that lie higher in the later. As in the past, those countries that produce pyramid ha� yet to be fo und. Large fields continue fo ssil energy, rather than those countries that have to be discovf;red, but in places that require more fo ssil energy resources, will determine the world's sophisticatecke<:hnical skills for exploration. The energy market. deep-water and subsalt plays in the Gulf of Mexico are examples of where a combination of new tech­ nologies and lower costs of exploration and produc­ "Although a Decline inU.S. Crude Oil tion have provided new exploration opportunities. Production Is Inevitable, That Decline Will The subsalt play is estimated to contain 1.2 billion Take Place Over Many Decades"-True or bbl of recoverable oil and 1 5 tcf of recoverable gas False? (Montgomery and Moore, 1997). Scoping volumes for the deep-water Gulf of Mexico are in the 15 bil­

· There is no inherent reason why a curve that lion bbl of oil equivalent range, with announced dis­ plots the history of production of a type of fo ssil covery volumes to date of more than 2 billion bbl 2132 Cornucopia or EmptyBarr el?

(Lawrence, 1994). Shell's Mars discovery alone any one time, the price of extraction increases as accounts fo r 700 million bbl of recoverable one goes lower in the pyramid, but over time reserves of oil (Mahaffie, 1994). Deep-basin gas advances in technology tend to decrease the costs deposits are a case where new technologies and of extracting all fuels. the ability of geologists to think beyond the norm Within the next few decades it is unlikely that have created new opportunities in what were we will begin to significantly diminish the resource thought to be well-explored basins. The Elmworth pyramid. The abundance of fo ssil energy and the discovery in Alberta, fo r example, contains 17 tcf possibilities of energy substitution make it likely of natural gas (Masters, 1984). Large discoveries are that long-term price trends are likely to rise at rates always appreciated, but it is important to remem­ no faster than the rate of inflation. Historically, cor­ ber that in the United States, field growth over a porate management and government planners 4-5 yr period adds crude oil equivalent to the often have been swayed by prevailing trends. When amount discovered in the Prudhoe Bay field (Figure energy prices rise sharply they may adopt a neo­ 4). Oil and gas that will comprise fu ture field Malthusian outlook, and when prices fall they may growth can be considered as being lower in the take a comucopian viewpoint. Such reactions have pyramid; specifically oil and gas that is in the more caused major swings in investment strategies, poli­ subtle traps, the more co�partmentalized reser­ cies, and staffing. A more reasonable approach voirs,. or in less mobile fo r¢.s. The keys to contin­ should take into account the nature of the energy ued growth of reserves indude cost-containment pyramid and the concepts of energy substitution. measures, such as business alliances, and the con­ With a longer term vision, decision-makers could tinued development of new prod\}ctiontechnolo­ stay the course despite short-term swings in prices. gies. The North American oil industry remains amazingly robust considering its long history. Its future health depends on the industry's continued REFERENCES CITED

ability to bring the region's oil and gas resources to Adelman, M. A., 1995, Trends in the price and supply of oil, in J. L. market at a price that is lower than that of compet­ Simon, ed., The state of humanity: Oxford, Blackwell, ing imports or alternative energy sources. This p. 287-293. Adelman, M. A., and M. C. Lynch, 1997, Fixed view of resource challenge will require substantial technical innova­ , tions and the ability of geologists and their man­ limits creates undue pessimism: Oil & Gas Journal, v. 95, no. 14, p. 56-60. agers to look beyond the norm. Arndt, H. H., P. Averitt, J. Dowd, D. J. Frendzel, and P. A. Gallo, 1968, Mineral resources of the Appalachian region-coal: U.S. Geological Survey Professional Paper 580, p. 102-133. CONCLUSIONS Ashworth, W., 1986, The history of the British coal industry, vol­ ume 5, 1946-1982: the nationalized industry: Oxford, Clarendon, 710 p. The cornucopians are correct in claiming that Attanasi, E. D., and D. H. Root, 1994, The enigma of oil and gas there is a lot more fo ssil fuelthan many resource or field growth: AAPG Bulletin, v. 78, p. 321-332. reserve estimates imply. Hubbert-style analysis, Averitt, P., 1975, Coal resources of the United States: U.S. invoked by many neo-Malthusians as a demonstra­ Geological Survey Bulletin 1412, 131 p. Bartlett, A. A., 1985, Book review: The Ultimate Resource by J. L. tion of how quickly energy resources may be Simon: American Journal of Physics, v. 53, p. 282-285. depleted, does not hold up to scrutiny of its basic Bartlett, A. A., 1986, Sustained availability, a management program assumptions. Because new resources become viable for nonrenewable resources: American Journal of Physics, to extract as prices rise or, more usually, as new v. 54, p. 398-402. Bartlett, A. A. , 1994, Reflections on sustainability, population technologies are developed, it is not possible to growth, and the environment: Population and Environment, quantify a finite amount of energy resource that will v. 16, p. 5-35. be available in the future. History also suggests that Bernstein, P. L., 1996, Against the gods: the remarkable story of many of the resources that we anticipate as being risk: New York, Wiley, 383 p. economically viable in the future probably will Bockstoce, J. R., 1986, Whales, ice, and men: the history of whal­ ing in the western Arctic: Seattle, University of Washington never be extracted. The shape of a production Press, 400 p. curve is as much controlled by the demand fo r the British Geological Survey, 1992, Energy minerals; United energy source as it is by its availability. The resource Kingdom Minerals Yearbook 1992: British Geological Survey, pyramid concept provides a useful framework fo r p. 2-1-2-12. Brobst, D. A., and W. P. Pratt, 1973, United States mineral understanding the nature of energy resources, and resources: U.S. Geological Survey Professional Paper 820, has the advantage of not suggesting a finite 772 p. resource. The pyramid also demonstrates why sub­ Bupp, I. C., andF. Schuller, 1983, Natural gas: conflicts and com­ jective criteria, used by geologists to decide which promise, in R. Stobaugh and D. Yergin, eds., 1983, Energy future (3d ed.): New York, Vintage Books, Random House, resources to assess, influence the size of their p ..69-99. assessment estimates. The pyramid illustrates the Bureau of Mines, 1924-1993, Mineral Resources of the United concept that vast resources are available and that, at States, annual editions 1924-1931, and Minerals Yearbook, McCabe 2133

annual editions 1932-1993: U.S. Bureau of Mines. Hudson Coal Company, 1932, The story of anthracite: Scranton, Campbell, C. J., 1997, The coming oil crisis: Multi-Science International Textbook Press, 425 p. Publishing Company and Petroconsultants, Brentwood, U.K., Hughes, W. R., 1995, Future of gas as an electric generating fuel, 210 p. in R. E. Willett, ed., The 1996 natural gas yearbook: New York, Church, R., 1986, The history of the British coal industry, volume Wiley, p. 275-302. 3, 1830-1913: Victorian pre-eminence: Oxford, Clarendon, Ivanhoe, L. F., 1995, Future world oil supplies: there is a finite 831 p. limit: World Oil, October 1995, v. 216, p. 77-88. Coleman,}. 1., 1995, The American whale oil industry: a look back John T. Boyd Company, 1993, Independent analysis, 21 closure

\ to the future of the American petroleum industry: Non- review collieries, British Coal Corporation, United Kingdom: renewable Resources, v. 4, p. 273-288. London, Her Majesty's Stationery Office, 130 p. \� olton G. L., et al., 1981, Estimates of undiscovered recoverable Knight,}. 1., B. J. Shevlin, D. C. Edgar, and P. Dolan, 1996, Coal -conventional resources of oil and gas in the United States: U.S. thickness distributions on the UK continental shelf, in R. Geological Survey Circular 860,87 p. Gayer and I. Harris, eds., Coalbed methane and coal geology: Dolton, G. 1., D. L. Gautier, R. F. Mast, and D. H. Root, 1993, U.S. Geological Society of London Special Publication 109, Geological Survey estimates of natural-gas energy resources, in p. 43-57. D. G. Howell, ed., The future of energy gases: U.S. Geological Kvenvolden, K. A., 1993, A primer on gas hydrates, in D. G. Survey Professional Paper 1570, p. 495-506. Howell, ed., The future of energy gases: U.S. Geological Survey Dromgoole, P., and R. Speers, 1997, Geoscore: a method for quan­ Professional Paper 1570, p. 279-291. tifying uncertainty in field reserve estimates: Petroleum Lawrence, D. T., 1994, Turbidite technical challenges in the deep­ Geoscience, v. 3, p. 1-12. water Gulf of Mexico, in P. Weimer, A. H. Bouma, and B. F. Duncan, D. C., and V. E. Swanson, 1965, Organic-rich shale of the Perkins, eds., Submarine fans and turbidite systems: Gulf Coast United States and world land areas: U.S. Geological Survey Section SEPM Fifteenth Annual Research Conference, Circular 523, 30 p. p. 217-220. Eavenson, H. N., 1942, The first century and a quarter of American Lewis, C. }. , 1986, Are our oil and gas resource assessments realis­ coal industry: Baltimore, Waverly Press, 701 p. tic?, in D. D. Rice, ed., Oil and gas assessment-methods and Edwards, J. D., 1997, Crude oil and alternate energy production applications: AAPG Studies in Geology 21, p. 195-202. fo recasts for the twenty-first century: the end of the hydrocar­ Mahaffie, M. J., 1994, Reservoir classification for turbidite inter­ bon era: AAPG Bulletin, v. 81, p. 1292-1305. vals at the Mars Discovery, Mississippi Canyon 807, Gulf of �nergy Information Administration, 1998, Historical data on fos­ Mexico, in P. Weimer, A. H. Bouma, and B. F. Perkins, eds., sil energy production, Accessed Submarine fans and turbidite systems: Gulf Coast Section SEPM March 15, 1998. Foundation Fifteenth Annual Research Conference, Ezra, D., 1978, Coal and energy: the need to exploit the world's p. 233-244. most abundant fo ssil fuel: NewYork , Wiley, 182 p. Mast, R. F., G. L. Dolton, R. A. Crovelli, D. H. Root, E. D. Attanasi, Ezra, D., 1984, Coal supplies, in T. L. Shaw, D. E. Lennard, and P. E. Martin, L. W. Cooke, G. B. Carpenter, W. C. Pecora, and P. M. S. Jones, eds., Policy and development of energy M. B. Rose, 1989, Estimates of undiscovered conventional oil resources: New York, Wiley, p. 109-125. and gas resources in the United States-a part of the nation's Fanning, L. M., 1950, A case history of oil-shortage scares, in L. M. energy endowment: U.S. Geological Survey and Minerals Fanning, ed., Our oil resources: New York, McGraw-Hill, Management Service, 44 p. p. 306-406. Masters, C. D., D. H. Root, and W. D. Dietzman, 1984, Distribution Flavin, C., and N. Lenssen, 1994, Power surge: guide to the com­ and quantitative assessment of world crude oil reserves and ing energy revolution: New York, Norton, 382 p. resources: Proceedings of the 11th World Petroleum Congress, Gever, J., R. Kaufmann, D. Skole, and C. Vorosmarty, 1986, v. 2, p. 229-237. Beyond oil: the threat to food and fuel in the coming decades: Masters, C. D., E. D. Attanasi, W. D. Dietzman, R. F. Meyer, R. W. Cambridge, Massachusetts, Ballinger, 304 p. Mitchell, and D. H. Root, 1987, World resources of crude oil, Halbouty, M. T., 1986, Basins and new fr ontiers: an overview, in natural gas, natural bitumen, and shale oil: Proceedings of the M. T. Halbouty, ed., Future petroleum provinces of the world: 12th World Petroleum Congress, v. 5, p. 3-27. AAPG Memoir 40, p. 1-10. Masters, C. D., D. H. Root, and E. D. Attanasi, 1991, World Hatcher, }., 1993, The history of the British coal industry, volume resources of crude oil and natural gas: Proceedings of the 13th 1, before 1700: towards the age of coal: Oxford, Clarendon, World Petroleum Congress, p. 51-64. 624 p. Masters, C. D., E. D. Attanasi, and D. H. Root, 1994, World ,/llausman, W. J., 1995, Long-term trends in energy prices, in }. L. petroleum assessment and analysis: Proceedings of the 14th Simon, ed., The state of humanity: Oxford, Blackwell, World Petroleum Congress, v. 5, p. 529-541. p. 280-286. Masters, J. A., 1979, Deep-basin gas trap, western Canada: AAPG Hendricks, T. A., 1965, Resources of oil, gas, and natural-gas liq­ Bulletin, v. 63, p. 152-181. uids in the United States and the world: U.S. Geological Survey Masters, }. A., ed., 1984, Elmworth-a case study of a deep-basin Circular 522, 20 p. gas field: AAPG Memoir 38, 316 p. Hubbert, M. K. , 1956, Nuclear energy and the fo ssil fuels: Shell McKelvey, V. E., 1972, Mineral resource estimates and public poli­ Development Company Publication 95, 40 p. cy: American Scientist, v. 60, p. 32-40. Hubbert, M. K., 1967, Degree of advancement of petroleum explo­ McKelvey, V. E., and D. C. Duncan, 1965, United States and world ration in United States: AAPG Bulletin, v. 51, p. 2207-2227. resources of energy: American Chemical Society Symposium Hubbert, M. K., 1969, Energy resources, National Academy of on Fuel and Energy Economics, v. 9, p. 1-17. Sciences-National Research Council, Resources and Man: San Meadows, D. H., D. L. Meadows,}. Randers, and W. W. Behrens ill, Francisco, Freeman, p. 157-242. 1972, The limits to growth: New York, Universe Books, 205 p. Hubbert, M. K., 1976, Outlook for fuelreserves , in D. N. Lapedes, Means, R. C., 1995, To market, to market: competition, deregula­ ed., McGraw-Hill Encyclopedia of Energy: New York, McGraw­ tion, and jurisdiction, in R. E. Willett, ed., The 1996 natural gas Hill, p. 11-23. yearbook: New York, Wiley, p. 11-60. Hubbert, M. K., 1981, The world's evolving energy system: Merritt, R. D., and C. C. Hawley, 1986, Map of Alaska's coal American Journal of Physics, v. 49, p. 1007- 1020. resources: Alaska Division of Geological and Geophysical Hubbert, M. K., and D. H. Root, 1981, Outlook for fuelreserves, in Surveys Open-File Report 84-24, 64p. S. Parker, ed., Encyclopedia of energy (2d ed.): New York, Miller, B. M., H. L. Thomsen, G. L. Dolton, A. B. Coury, T. A. McGraw-Hill, p. 45-56. Hendricks, F. E. Lennartz, R. B. Powers, E. G. Sable, and K. L. 2134 Cornucopia or Empty Barrel?

Varnes, 1975, Geological estimates of undiscovered recover­ Simon, J. L., G. Weinrauch, and S. Moore, 1994, The reserves of

able oil and gas resources in the United· States: U.S. Geological extracted resources: historical data: Nonrenewable Resources, V SurveyCircular 725, 78 p. v. 3, p. 325-340. Minerals Management Service, 1996, An assessment of the undiscov­ Smith, A. L., and B. J. Lidsky, 1993, King Hubbert's analysis revisit­ ered hydrocarbon potential of the nation's outer continental shelf: ed: update of the lower 48 oil and gas resource base: The Minerals Management Service OCS Report MMS 96-0034, 40 p. Leading Edge, v. 12, p. 1082-1086. Montgomery, S. L., 11nd D. C. Moore, 1997, Subsalt play, Gulf of Smith, J. W., 1981, Oil shale resources of the United States: Mexico: a review� AAPG Bulletin, v. 81, p. 87 1-896. Mineral and Energy Resources, Colorado School of Mines, Myers, J. G., S. Moor�, and }. L. Simon, 1995, Trends in availability v. 23, no. 6, p. 1-20. of non-fuel mineritlsJn J. L. Simon, ed., The state of humanity: Smith, N. J., and G. H. Robinson, 1997, Technology pushes Oxford, Blackwell, p. 303-312. reserves 'crunch' date back in time: Oil & Gas Journal, v. 95, Myers, N., and J. L. Simon, 1994, Scarcity or abundance? A debate no. 14, p. 43-50. on the environment: New York, Norton, 254 p. Supple, B., 1987, The history of the British coal industry, volume National Research Council Committee on the Status and 4, 1913-1946: the political economy of decline: Oxford, Research Opportunities in the Solid-Earth Sciences, 1993, Clarendon, 733 p. Solid-earth sciences and society: Washington, D.C., National Theobald, P. K., S. P. Schweinfurth, and D. C. Duncan, 1972, Academy Press, 346 p. Energy resources of the United States: U.S. Geological Survey /on & Gas Journal, 1997, Energy statistics sourcebook; Tulsa, Circular 650, 27 p. Pennwell, 582 p. and CD-ROM. United Nations (1978-1994), Energy statistics yearbook, annual V' Polesetsky, M., ed., 1991, Global resources: opposing viewpoints: editions: New York, United Nations. San Diego, Greenhaven Press, 288 p. U.S. Geological Survey, 1883-1923, Mineral Resources of the Potential Gas Committee, 1997, Potential supply of natural gas in United States, annualeditions: U.S. Geological Survey. V the United States: Potential Gas Agency, Colorado School of U.S. Geological Survey National Oil and Gas Resource Assessment Mines, Golden, Colorado, 130 p. Team, 1995, 1995 national �sessment of United States oil and '­ Rice, D. D., ed., 1986, Oil and gas assessment-methods and appli­ gas resource: U.S. Geological Survey Circular 1118, 20 p. cations: AAPG Studies in Geology 21, 267 p. Weaver, K. F., 1981, Our energy predicament: Energy, National Roodman, D. M., 1997, Reforming subsidies, in L. Starke, ed., State Geographic Special Report, February 1981, p. 2-23. of the world 1997: a Worldwatch Institute report on progress Weeks, L. G., 1948, Highlights on 1947 developments in fo reign toward a sustainable society: New York, Norton, p. 132-150. petroleum fields: AAPG Bulletin, v. 32, p. 1093-1160. Root, D. H., and E. D. Attanasi, 1993, A primer in field-growth esti­ Weeks, L. G., 1950, Discussion of "Estimates of undiscovered mation, in D. G. Howell, ed., The future of energy gases: U.S. petroleum reserves by A. I. Levorsen": Proceedings of the Geological Survey Professional Paper 1570, p. 547-554. United Nations Scientific Conference on the Conservation and Ross, W. G., 1985, Arctic whalers, icy seas: Toronto, Irwin Utilization of Resources, 1949, v. 1, p. 107- 110. Publishing, 263 p. Weeks, L. G., 1958, Fuel reserves of the fu ture: AAPG Bulletin, Royal Commission on the Coal Industry, 1925, Report of the Royal v. 42, p. 431-438. Commission on the coal industry (1925), volume 1: His Weeks, L. G., 1959, Where will energy come from in 2059?: Majesty's Stationery Office, London, 294 p. Petroleum Engineer, v. 31, p. A24-A3 1. Schmoker, J. W., andE. D. Attanasi, 1997, Reserve growth important Woods, T. J., 1993, How ultimate is ultimate gas recovery?, in to U. S. gas supply: Oil & GasJournal, v. 95, no. 4, p. 95-96. D. G. Howell, ed., The future of energy gases: U.S. Geological Simon, J. L., 1980, Resources, population, environment: an over­ Survey Professional Paper 1570, p. 869-875. supply of false bad news: Science, v. 208, p. 1431-1437. Zapp, A. D., 1961, World petroleum resources, in U.S. Geological Simon, }. L., 1981, The ultimate resource: Princeton, Princeton Survey, Domestic and world resources of fo ssil fuels, radioac­ University Press, 415 p. tive minerals, and geothermal energy: Report prepared for the Simon, J. L., ed., 1995, The state of humanity: Oxford, Blackwell, Natural Resources Subcommittee of the Federal Science 694 p. Council, 6 p. and 1 table.

ABOUT TilEOR AUTII

Peter J. McCabe

Pete McCabe obtained a B.Sc. degree in geology from the Uni­ versity of Hull, United Kingdom (UK), and a Ph.D. in geology from the University of Keele, UK. He worked at McMaster University, the University of Nebraska, Exxon Production Research Company, and the Alberta Research Council before joining the U.S. Geological Survey in 1987. McCabe's research interests include sequence stratigraphy, coal geology, oil and gas resources of the Asia Pacific region, and assess­ ment of fossil energy resources. a series of supply curves for recoverable oil and gas in Energy resources­ the United States. cornucopia or empty barrel?: Reply HUBBERT-STYLE ANALYSIS Contrary to Duncan's claims, I did not write that the Peter J. McCabe* bell-shaped Hubbert curves are normal probability dis­ tributions or that the curves show no dependency of one year's production on previous production history. In fact, I aimed to dispel a common notion that the I welcome the opportunity to reply to the discussions bell-shaped curves of Hubbert are normal probability by Campbell, Duncan, and Laherrere, three of distributions. To quote from my article, "Hubbert the listed experts of Oil Crisis dot-com , a group that lobbies governments quency is plotted against some property of a sampled and industry to warn of a perceived impending short.:. population. Rather, the Hubbert curves show a physi­ age of crude oil. Their disc:Ussions include numerous cal property plotted against time ....in a time-series points; I will address their comments collectively in plot of energy production there is a clear dependency this response by major theme. of one year's production on a previous years' produc­ tion" (McCabe, 1998, p. 2121). Production curves are graphs of mathematical functions that have two vari­ ECONOMIC CONTROLS ON OIL ables: date and production (the dependant variable). PRODUCTION They are not statistical distributions plotting the fre­ quency of a single variable. Consequently, production Campbell claims that the production of �onventional curves cannot be normal distributions. Laherrere also oil is "relatively insensitive to price, investnient, and fails to grasp this elementary point where he discusses technology" and Laherrere claims that, whereas the "statistical theory" and applies the Central Limit The­ availability of coal is dependant on price, with lower orem (a. theorem of statistical distributions) to produc­ concentrations becoming available at higher prices,this tion curves. is not the case with oil. These are surprising suggestions Unfortunately, Duncan seems to have missed the that defy economic logic, and there is abundant evi­ major arguments against Hubbert-style analysis. These dence to the•contrary. Many small United States oil do not center at all on the mathematical shape of the fields that had been profitable at $"15-20/bbl in the curve . The key point is that this type of analysis as­ mid-1990s were shut in when the price fell to close to sumes thatwe can know the quantity of resource avail­ $10/bbl Qr less in the late 1990s. The low oil prices of able and that all of that resource will be used. The 1999 led to maj or cu�backs in exploration and devel­ historical data pr�sented in my article, however, show opment in many parts of the world, including the (1) that assessments of the total quantity of resource North Sea, and layoffof many employees. The amount available change (iri.crease or decrease) over time and of oil available for production is closely tied to prices, (2) that a production cycle of a resource can be com­ although it may take several years to increase produc­ pleted without depletion of all the resource. tion substantially after a price rise. In an elegant dem­ The production history of a fuel is driven by the onstration that the future availability of conventional interaction of demand and supply. Many have tried to oil resources is tied to price, Attanasi (1998) developed develop mathematical models that attempt to predict the outcome of the interaction of supply and demand

Copyright Cl2001. The American Association of Petroleum Geologists. All rights · throughtime for a varietyof markets. Perhaps the best . reserved. known are models that aimed to predict the futureper­ *U.S. Geological Survey, Federal Center MS939, Denver, Colorado 80225; fo rmance of the stock market. These efforts have all [email protected]. been doomed to failure in the long term because of I wish to thank Tom Ahlbrandt and Jim Schmoker of the U.S. Geological Survey for unanticipated changes, particularly in demand. Devel­ reviev.�ing this reply and making many useful suggestions. I also thank Bob Crovelli .of the U.S. Geological Survey for his insights on mathematics and statistics. opment of new technologies, wars, human ingenuity, Manuscript received August 15, 2000; final acceptance November 25, 2000. and the unpredictable nature of human behavior are

MPG BULLETIN, V. 85, NO. 6 (JUNE 2001), PP. 1093-1097 1093 among the many factors that may change demand "recC?verable resources" and "reserves" are terms that over time in ways that cannot be anticipated are defined as much on the basis of economics and mathematically. technology as on geology. They are dynamic numbers that change through time in response to changing eco­ nomic factors, technological advances, and increased RESOURCES AND RESERVES geologic knowledge. Resources and reserves are not fixed numb ers as Laherrere seems to believe. Laherrere' s main criticism is that I do not define re­ serves and resources; however, the two terms are de­ fined and discussed at length in my article (McCabe, WHAT IS "CONVENTIO NAL"? 1998, p. 2113-2115). The definitions used are those of McKelvey (1972) and Brobst and Pratt (1973) . La­ Campbell claims that world discovery rates of "con­ herrere has his own different definitions of the terms ventional oil" are plummeting but greatly exaggerates and then criticizes my article based on his use of the this by not including oil from deep-water fields. This terms. His definitions are clearly different fromthe way would be akin to someone in 1935 claiming that oil these terms are normally used. Laherrere uses the term . fo und at depths below 3000 ft, in wells drilled by the "resource" fo r ahy hydrocarbon in the ground, but, as new rotary rigs, was unconventional, or sorheone in pointed out in my article, the term is normally used 1960 claiming that offshore production was unconven­ ollly for that part of the global fossil fuel endowment tional oil. If 4 billion bbl of deep-water oil are being that is thought to have the potential to be recovered fo und every year it can hardly be classified as uncon­ one day (sometimes referred to as "recoverable re­ ventional anymore. Many examples exist of resources sources") . In fact, "resources" are defined by the U.S. that were previously considered uneconomic, such as Bureau of Mines and U.S. Geological Survey (1976) as basin-centered gas, coalbed methane, and heavy oils accumulations that are in such a fo rm that economic that are now critical components of the mix of pro­ extraction is currently or potentially feasible. The re­ duced fossil fuels. Because new technologies permit vi­ coverable resource base increases through time (figures able economic extraction of a wider array of resources 8, 10 in McCabe [1998]) with the exp ansion ofknowl­ over time, "conventional" and "unconventional" are edge and the development of new technologies. La­ subjective terms that depend on one's perspective. Ir­ herrere uses "reserve" for that part of the resource that respective of terminology, it is the total amount of oil will be recovered one day. The ter:m "reserves," how­ discoveries that should be of concern. Given that real ever, covers only that part of the known resourcethat prices have been falling for most of the last 20 yr and has the potential to be recovered within the foresee­ that present reserves could last for decades, it could be able future under current economic" and technologic argued that discovery rates have been surprisi�glyhigh conditions (ge.nerally 10-20 yrbut significantly longer over the last decade. Additionally, the announced new for coal) . As figure 3 in my article (McCabe, 1998) discoveries are not grown. Assuming modest reserve shows, thereserve number for United States crude oil growth, discovery rates of the last decade may well throughout the 20th century was the amount of oil have matched production. that was recovered within approximately the following 12 yr. Part of the resource base is converted to reserves on an ongoing basis as they are discovered or become RESERVE GROWTH economically viable. It is also possible for hydrocarbons to be removed from the reserve category if develop­ Reserve growth is well documented in the United ments prove the hydrocarbons are not likelyto be re­ States and is heavily dependent on new technology. covered within the foreseeable future. For example, 24 The concept of reserve growth in fields outside the tcf of natural gas was removed from the reserves for United States is a contentious issue with Campbell, Alaska in 1988 when the prospects fo r development of who considers growth of P5o reserves throughout the gas from the North Slope were diminished (EIA, world to be very minor. If estimates considered to be 1998). Likewise, the dramatic downsizing of the re­ P50 were in fact the median estimates, then reserve serve figure for British coal from 1920 to 1990 was a growth would indeed be very small. Analysis of world result of a decreased prospect for development-not a field-size data for the period from 1981 to 1996 by the decrease of coal in the ground. The point is that both U.S. Geological Survey (Klett et al., 2000), however,

1 094 Discussions and Replies showed that both OPEC and non-OPEC nations had that have appeared in peer-revie:wed publications with greater increases in supposed P50 than might have been discussion of methodology and supporting evidence. expected fromdocumented United States rates of field growth. It seems highly unlikely, therefore, that we now know the P50 of any fields, except possibly those LIQUID, SOLID, OR GAS-HYDROCARBON that are close to the end of their life. Application of PRODUCTION IS DRIVEN BY MARKET new technologies in futuredecades will surely push the · FORCES estimates upward. It is our inability to foresee tech­ nologies that are developed in future decades that re­ Campbell and Laherrere object to the idea that lessons quires us to increase our estimates of P50 over time. may be learnedfrom looking at the production curves of coal. No matter which hydrocarbon (coal, oil, or natural gas), the availability at different price� depends

U.S. GEOLOGICAL SURVEY ASSESSMENTS · on several factors, such as distance to market, recovery costs (drilling and mining costs related to depth and The latest U.S. Geological Survey world petroleum as­ location), and quality, not just size of the accumula­ sessment (US. Geological Survey World Energy As­ tions. The giant gas fields of the Northwest Shelf of sessment Team, 2000) suggests that 2120 billion bbl Australia, the giant oil accumulations of the Caspian of recover:able conventional oil remain in the world and East Siberian basin, and the· extensive thick coal (exclusive of the United States), compared to the ap­ accumulations of Alaska are all examples of hydrocar­ proximately 539 billion bbl that have been produced bons that have difficulty getting to market because of to date. If the future peak in world oil production is location. In contrast, the small oil and gas accumula­ approximately at a time when half the world's endow­ tions of the J sandstone in the Denver basin, Colorado, ment has been produced, then the peak should lie well and the relatively thin coals of Appalachia, for exam­ past the imminent peaking of production that Camp­ ple, are economically viable because of their proximity bell forecasts. Campbell has criticized the U.S. Geo­ to markets. Both Campbell and Laherrere say that oil logical Survey numbers as being too large, but rather is "either there in abundance or not there at all," but than making a careful critique of the geological data, the fact is that in any petroleum basin there is a distri­ discovery rate information, historical reserve growth bution of fieldsizes and smaller fields are far more com­ data, and methodology used in the assessment, Camp­ mon than large fields (Gautier et al., 199Sj U.S. Geo­ bell chose to dismiss the numbers as a plot by the logical SurveyWorld Energy Assessment Team, 2000). United States government to undermine OPEC's con­ Accumulations that are currently too small to be eco­ fidence (Campbell, 2000). nomic are generally not recorded as fields butas shows.

•' Likewise, in most coal basins there are a few thick coal beds and many thinner seams, including many that are DIFFERE·N T TYPES OF ASSESSMENTS so thin that they are ignored in an assessment of re­ sources. Oil-water contacts are sharp but are certainly Laherrere criticizes the fact that I included only some no sharper than coal-mudstone contacts. Whatever the of the historical estimates of undiscovered oil in my geologic factors controlling accumulation or the physi­ figure 8 (McCabe, 1998). To include all the estimates cal state of the hydrocarbon, oil, natural gas, and coal would be to mix apples and oranges. The estimates are all commodities after they are extracted from the used in figure 8 were all based on detailed geologic ground, and the market forces of supply and demand appraisals of the various basins in the United States. that determine their price and the amount produced The other estimates were based on a statistical tech- are similar to other nonrenewable natural resources . nique that essentially ignored geologic conditions. De­ (Brown and Wolk, 2000) . tails of the rationale as to which assessments were in­ cluded in figure 8 were discussed at some length (McCabe, 1998, p.2116) and reference made to the BRITISH COAL PRODUCTIO.N extensive discussion on this topic by Hubbert (1969). There have, of course, been many other estimates of Laherrere calculates an ultimate reserve figure for undiscovered resources in the United States, but my British coal from cumulative production to date plus figure 7 (McCabe, 1998, p. 2116) includes all those the current reserve figure. He then uses that figure to

McCABE 1095 retrodict (that is, inferring a past state of affairs from profound impact those mistaken forecasts had on so­ present observational data) British coal production. ciety in the 1970s. !!'his reasoning is circular in nature. A Hubbert-style The authors of thediscussions have published pre­ prediction in the early 20th century would have dra­ dictions of a bleak future. In his "Olduvai theory," matically missed the mark in predicting future coal Duncan says that industrial civilization "peaked" about production in Britain because the very high reserve 1978 and will end about 2025 (Duncan, 1997). He figures of that period turned outto include much coal portrays humans as living in a cavemanlike existence that was uneconomic to mine as economic and tech­ by 3000 A.D. Campbell says that we are close to the nologic conditions changed with time. With the com­ midpoint in world oil production. His book portrays plexities of retrodiction, Laherrere finds · it necessary an apocalyptic scene for 2025 with the end of consum­ to model the 'British coal production curve by invok­ erism (Campbell, 1997, p.l14). He forecasts, "The ing a series of Hubbert curves. One curve he·explains World will become a very different place witha smaller as representing a peak in iron and steel use in 1958 population. The transition will be difficult, and for and a second as related to high prices of 1981. Ex­ some catastrophic" with "major economic and political amination of how the coal was used (figure 22 in discontinuity" globally and "great suffering" (Camp­ McCabe [1998]) shows that the final production re­ bell, 199 7, p. 177). Afterthe collapse of civilization as flects a much more complex interplay of various mar­ we know it, his book suggests that humans \Viii be liv­ ket sectors. The minor peak in the late 1950s reflected ing in harmony with nature, in a Garden-of-Eden set­ an economic resurgence afterthe Second World War, ting, for the period of 2050 to 2500 (Campbell, 1997, but coal started to lose its markets for domestic heat­ p. ll5). ing, railroads, gas production, and iron and steel pro­ These types of predictions seem to come from a duction after 1958, largely because cheaper alterna­ narrow perception that oil is the only form of energy tives became available. The use of coal for iron and available to civilization. Oil only became the dominant steel in the 1950s was no larger than it had been since energy source in about 1950. Today about 60% of the 1870, except for periods of lower production during global energy supply comes from other sources, in­ the Great Depression and the Second World War. cluding coal, natural gas, nuclear, and renewables. The The minor peak in the early 1980s is related solely to future energy mix will evolve. These changes may be the increased use of coal for electrical generation at driven by a shortage or increased abundance of one or that time. Although prices peaked in the early 1980s, more fuels, but it is perhaps more likely that, as in the it was the culmination of a long-term price increase past, human ingenuity will create new markets and that began around 1850, not something that started there will be substitution of fuels in various market around 1960 as the Laherrere plot suggests. Also per­ sectors. The current development of fuel cells, for ex­ tinent to not� is that the minor peaks of 1958 and ample, may well create a dramatic decrease in oil con­ 1961 were not related to separate sources of supply, sumption and perhaps a considerable incre�se in nat­ as Laherrere's multi-Hubbert analysis suggests, but in­ ural gas consumption. Considering such a scenario stead reflected increased productionfrom existing coal likely, Sheik Yamani, who once warned the world of mines. the dangers of depleting oil reserves, recently said about fuelcells: "I can tell you with a degree of confi­ dence that afterfive years therewill be a sharp drop in PREDICTING THE FUTURE the price of oil . . ..Thirty years from now there will be a huge amount of oil-and no buyers" (Brandreth, Laherrere points out that forecasts by Halbouty, 2000). Perhaps Yamani's predictions will prove incor­ Moody, and Masters were somewhat offmark but in­ rect, but they emphasize a critical aspect of my article: terestingly does not write about the incorrect and more that future production will be determined as much by recent forecasts of his colleague Campbell (Campbell demand as by supply. [1991 ]; see also discussion by Thomasson [2000]). In Laherrere objects to the term '"neo-Malthusian." 1999, actual production in the United States was al­ The term is widespread in the literature and is used to most twice, and in Canada and the United Kingdom discriminate Malthusians who closely follow Malthus's about six times, what Campbell had predicted for that beliefs from those who have more modern concepts of year in 1991. A discussion of the Club of Rome fore­ why there will be a future shortage of resources. L F. casts was included in my article only because of the "Buzz" Ivanhoe prefers the term "Cassandra" (Brown,

'1096 Discussions and Replies 2000), but unlike Cassandra, some people believe their methodologies, and data analyses are the essence of the predictions, and today's energy pessimists were not scientific method. granted the gift of prophecy by the gods.

REFERENCES CITED HUBBERT'S LEGACY Attanasi, E. D., 1998, Economics and the 1995 national assessment ' Although Campbell, Duncan, and Laherrere invoke his of United States oil and gas resources: U.S. Geological Survey Circular 1145, 35 p. legacy, I suspect that ifM. KingHubbert were still alive Brandreth, G., 2000, Farewell to riches of the earth: Tele�aph, he would identify more with the ideas in my article no. 1857, 25 June. than he would with their viewpoints. When one reads Brobst, D. A., and W. P. Pratt, 1973, United States mineral re­ sources: U.S. Geological Survey Professional Paper 820, 772 p. his articles (for example, Hubbert [1969]) the image Brown, D., 2000, Bulls and bears duel over supply: AAPG Explorer, left is not of an end-member neo-Malthusian. Hubbert v. 21, no. 5, p. 12-15. was a technocrat who believed that technology could Brown, S. P. A., and D. Wolk, 2000, Natural resource scarcity and technological change: Federal ReserveBank of Dallas Economic answer many of the world's problems and discussed and Financial Review First Quarter 2000, p. 2-1 3. how future technological advances might influence his Campbell, C. J., 1991, The golden century of oil 1950-2050: Dor­ predictions. He would have been excited about (rather drecht, Kluwer Academic Publishers, 345 p. than dismissive of) many of the advances in oil explo­ Campbell, C. J., 1997, The coming oil crisis: Brentwood, United Kingdom, Multi-Science Publishing and Petroconsultants, ration and production over the last decade and would 210 p. have considered their implications for predicting fu­ Campbell, C. J., 2000, Misleading USGS report, Accessed July 14, 2000. the total energy mix, not as a fuelin isolation, and he Duncan, R. C., 1997, The Olduvai theory of industrial civiliza­ considered what fuels might replace oil when its pro­ tion, Ac­ duction peaked. Furthermore, he did not see oil as cessed July 14, 2000. EIA (Energy Information Administration), 1998, U.S. crudeoil, nat­ unique-he applied his concepts on resource depletion ural gas, and natural gas liquids reserves 1998 annual report, equally to coal, natural gas, and oil. He also saw the Accessed July 14, 2000. Gautier, D. L., G. L. Dolton, K. I. Takahashi, and K. L. Varnes, at alternate scenarios and did not believe that an end 1995, 1995 national assessment of United States oil and gas of oil production would necessarily lead to a collapse resources-results, methodology, and supporting data: U.S. of civilization. In short, Hubbert's present-day disci­ Geological Survey Digital Data Series DDS-30, 1 CD-ROM.. Hubbert, M. K., 1969, Energy resources: National Academy of ples are giving him a bad name. Sciences-National Research Council, Resources and man: San Francisco, Freeman, p. 157-24 2. Klett, T. R., R. R. Charpentier, J. W. Schmoker, and E. D. Attanasi, oil and gas field sizes (abs.): CONC.i:;USION 2000, Predicting changes in world AAPG Annual Convention, Program with Abstracts, CD­ ROM. In his discussion, Laherrere calls me a comucopian McCabe, P. J., 1998, Energy resources-cornucopia or empty bar­ (which evidently means that, in his opinion, my data rel?: AAPG Bulletin, v. 82, p. 2110-2134. McKelvey, V. E., 1972, Mineralresource estimates and publicpolicy: are not as good as his!). Campbell, however, accuses American Scientist, v. 60, p. 32-40. me of dismissing comucopian and neo-Malthusian Thomasson, M. R., 2000, Petroleum geology: is there a future?: ideas alike. No doubt my article did not satisfy the ar­ AAPG Explorer, v. 21, no. 5, p. 3-10. U.S. Bureau of Mines and U.S. Geological Survey, 1976, Principles dent neo-Malthusians or cornucopians, but hopefully of the mineral resource classification system of the U.S. Bureau the article provides data on the nature of energy as­ of Mines and U.S. Geological Survey: U.S. Geological Survey sessments and the controls on energy production that Bulletin 1450-A, 5 p. U.S. Geological Survey World Energy Assessment Team, 2000, U.S. can be considered by those who have an open mind. Geological Survey world petroleum assessment 2000-descrip­ Optimists and pessimists alike should welcome new tion and results: U.S. Geological Survey Digital Data Series data and different perspectives, whether they support DDS-60, 4 CD-ROMs. their previous beliefs or not. Testing of hypotheses,

McCABE 1097