Oil and Gas Resources—Welcome to Uncertainty

There were some less sanguine estimates from those who looked at the ever-declining curves of oil field production and projected rising costs through time. But these more cautious projections appeared to be overshadowed by the upward path of U.S. oil production. As the world's leading oil producer, the nation passed the 1 billion-barrel level in 1929, 2 billion in 1948, 3 billion in 1966, and reached the 3 and one-half billion level in 1970. The perennial optimism of the wildcatter—"Give us an incentive and we'll go find you some oil and gas"—was well supported by over 100 years of production history. The United States seemed a permanent fixture as the world's number one producer of oil and gas.

The undercurrent of concern during the 1960s over declining exploratory activity in the United States elicited little real attention outside of the oil and gas industry itself and a small circle of petroleum specialists. It was easy for others to treat these worries as merely the customary background noises that accompany an industry's efforts to encourage favorable treatment by Congress on taxation, incentives, or protection from foreign competition. However, the major disturbance caused by the Organization of Petroleum Exporting Countries (OPEC) oil embargo in 1973 brought an immediate end to this lack of public attention.

In 1975, a report by the Committee on Resources and the Environment (COMRATE) of the National Academy of Sciences, based on a review of contemporary estimates, stated that, of the original stock of crude oil and natural gas liquids (249 billion barrels), only 113 remained to be discovered. For natural gas 530 trillion cubic feet (of an original 1,227 trillion) remained. [Footnote 1] This marked the end of general optimism both in industry and government about the future U.S. oil and gas resource position. For the public and Congress, whose ears are normally more receptive to good news, it was a shocking revelation to learn that instead of over 400 billion barrels of liquid hydrocarbons there might be much less. To have this unwelcome news appear in the midst of the oil and gas industries' post-embargo clamoring for high prices resulted in both public confusion and distrust. With respect to natural gas, the winter crisis of 1976-77 caused renewed doubts and confusion among the public, the media, and members of government.

In the three years since the COMRATE report, several staff members at Resources for the Future have been looking into questions about oil and gas reserves and resources. It, therefore, seems appropriate at this time to distill from these recent RFF efforts as much understanding about oil and gas resources as possible. We have no intention of producing a new set of resource estimates; there are more than enough of these. Rather, we hope to show why we keep getting different signals about the status of our oil and gas resources. If we can reduce some of the confusion, our efforts will be well rewarded.

Obviously, it will not be possible to explore in these few pages all of the problems in the collection and use of oil and gas statistics. Our attention is directed toward the different methodologies and perspectives employed by the various analysts who produce conflicting estimates.

The following discussion draws heavily upon a number of recent RFF activities, including: a workshop on oil and gas resources sponsored by the National Science Foundation, a study on resource terminology sponsored by the Electric Power Research Institute, a workshop on Maximum Efficient Rate (MER) of oil and gas production sponsored by the U.S. Department of the Interior, and a workshop which reviewed the Federal Energy Administration's National Energy Outlook, 1976 sponsored by the National Science Foundation. In addition, members of the RFF staff have participated on a regular basis in the work of the committees and boards of the National Academy of Sciences, the Gas Policy Advisory Council of the Federal Power Commission, and the oil and gas resource appraisal groups of the American Association of Petroleum Geologists and the U.S. Geological Survey. The contribution to this summary report of Dr. John C. Calhoun, Jr., Vice Chancellor of Texas A&M University who directed the oil and gas resources workshop, is especially acknowledged.

A Matter of Definition and Classification

An oil or gas reservoir is not a subterranean cavern filled with oil and gas which we empty like a huge storage tank. During geological time, varied mixtures of crude oil, natural gas, and salt water were formed and moved about in the interconnected minute pores of certain kinds of rock where they have remained trapped. When a driller's bit penetrates the rock, natural pressures cause a slow migration of the fluids toward the well bore. The well operator may decide initially, or eventually, to give the flow of oil and gas an assist through the application of the sucking action of a pump, or by fracturing the rock around the well, or by injecting water, chemicals, heat, or gases into the rock. To understand this production process, three things must be kept in mind: 1, the flow of fluids through the rock pores is a function of the physical forces at work; 2, the quantity resulting from additional effort gradually diminishes, just as wringing a wet rag produces less and less water; and 3, the only actual measurement that can be made is of the oil and gas produced at the surface—all other information about the reservoir is estimated.

An oil and gas reservoir or pool is basically a hydraulic unit where all the interconnected pores holding the oil and gas in the rock behave as a single fluid system. Theoretically, a well drilled at precisely the right place would be all that would be needed to produce all of the oil or gas the reservoir will ever produce, given sufficient time for the fluids to move through the rock to this one point. In one geographic area, encompassing a few or thousands of acres, there may be a series of reservoirs or individual traps containing oil and gas that are geologically related but not physically interconnected. To find and produce all of the oil and gas requires additional wells, dispersed either vertically or horizontally. A single isolated reservoir or a group of reservoirs related by physical proximity and geological origin are identified as an oil or gas field. Once the oil and gas exploration teams have found a specific bed of rock that contains oil and gas accumulations, they will tend to follow this "play" by drilling down to that bed or zone over an extended area. A discrete geological environment having a large number of oil and gas fields is known as a basin or province. In the United States, there are over 100 basins found clustered in five major regions. Within the basins, there are thousands of fields and many thousands of individual reservoirs. Over 2 million wells have penetrated the earth in the vicinity of these oil and gas traps, and more than 500,000 successful ones are still producing oil and gas. Approximately 10,000 wells are abandoned each year.

Once the physical characteristics of an oil and gas resource system is appreciated, the complexity of the question "how much oil and gas do we have?" becomes more apparent. Any response can be no more than a judgmental estimate. Intelligent communication about oil and gas resources becomes exceedingly difficult unless both the questioner and respondent understand what kind of data they are using. A start in this direction is to use a diagram becoming common in governmental circles (see figure 1).

A complete oil and gas resource diagram is a pictorial representation of all unproduced natural oil and gas hydrocarbons that may exist. We can also visualize a valve attached to the diagram representing the oil and gas wells through which oil and gas is removed from the reservoirs that have already been penetrated by the drill. Beyond this valve there is a conduit which represents the pipelines, tankers, barges, railroad cars, and trucks used to move the oil and gas through processing and on to the final user. It is worth repeating that past production, that is, the quantity already delivered by this system, is our only actual measurement. That quantity of oil and gas is gone forever. References to original oil and gas in place mean the sum of both the remaining oil and gas plus all that has ever been produced.

The productivity capacity of the United States is the amount of oil and gas that can be produced from existing wells during a specified period of time under specified operating conditions. The totality of physical oil and gas in the earth but not yet produced from the continental crust to a depth of perhaps 50,000 feet is sometimes called the oil and gas resource base. There are four kinds of oil and gas found in this resource base. The first kind (segment A in figure 1) consists of oil and gas which has already been found and is considered producible under present prices using current technology. These quantities are customarily known as reserves. The immediately producible portion of these reserves—the oil and gas that will flow from wells in developed reservoirs, the quantity of which can be estimated with considerable accuracy—is classified as proved reserves. The balance, or unproved reserves, has been discovered but cannot be estimated with as great accuracy and may require additional drilling and development (see figure 2).

In segment B we find oil and gas that has been discovered but in the judgment of the operators cannot be produced under current prices with existing technology. These quantities are known as subeconomic resources. There are two kinds of subeconomic resources. First, the unrecoverable, high-cost portion of oil and gas currently left behind in producing reservoirs. Second, oil and gas in other reservoirs that have been found but are not now producing or have been abandoned because they would cost too much to produce due to size or other problems.

Segment C of the resources diagram encompasses the oil and gas that remains to be discovered. Exploratory drilling has not proceeded to a point where there is physical evidence of the actual presence of this oil and gas. There is only expectation, and estimates of undiscovered oil and gas are based solely upon geologic and engineering extrapolation. This requires the use of geological and geophysical data rather than using physical data based upon the actual existence of the oil and gas. It is possible to subdivide undiscovered resources into economic and subeconomic quantities, but to do so requires the analyst to make some sort of assumption about prices and technology conditions. Present prices and technology are frequently assumed despite the fact that the oil and gas, when actually discovered, will be produced under future conditions of price and technology.

The final portion is segment D—other occurrences. This includes any oil and gas left behind that is not expected under any future circumstances to be worth the effort or cost of production, as well as deposits which are considered too small to either find or produce if found. Finally, this category is a convenient place to account for other forms of oil and gas hydrocarbons about which either little is known, or production technology is so immature that economic and technologic judgments cannot be made, even though large quantities may be involved.

Estimation of Reserves

As the drill bit penetrates a rock reservoir for the first time and finds oil and gas, the first questions asked are how much has been found and can it be produced economically. The initial well provides limited information about the rock strata that have been penetrated and nothing about strata that are below the bottom of the well. Once a layer of rock containing oil and gas has been found, the approximate thickness of the bed at that point—anything from a few to hundreds of feet—is known. A core of rock is usually taken from the bed.

Electrical and other measurements are taken inside the hole. All of these data provide information about the porosity and permeability of the rock and the amounts and kinds of fluids it contains. If the reservoir seems to justify production on the basis of this preliminary information, then the drilling equipment is removed, production pipe is put in place, and the well prepared for production. The initial flow of a new well provides information about the production rate, pressure, and other physical data.

At this point, a preliminary judgment on how much oil and gas have been found can be made based on: the flow from a single well; a rock sample a few inches in diameter of a multi acre reservoir; a map of the surface geology; and a seismic "shadow picture" of the structure holding the oil and gas thousands of feet below the surface. Obviously, this first estimate cannot be very precise. Yet based on this one well and past experience with the kind of reservoir that appears to have been found, the engineer makes a judgment. This estimate may range from the least amount of oil and gas that appears to have been found to the outer limit of what the reservoir might ultimately produce if the buried structure is entirely filled with oil and gas.

The scientific guesswork about a reservoir hundreds of acres in size is useful but extremely crude. It is akin to going to an unfamiliar supermarket on a foggy night and trying to estimate the total amount of asphalt used in paving the parking lot, with no other data than a cubic inch sample of the blacktop used. How uncertain these judgments about reserves can be was illustrated in a study published by the National Academy of Sciences in 1976.

The study concerned the amount of gas reserves under lease in certain fields in the Gulf of Mexico. Previous estimates had been made by the technical staff of the Federal Power Commission (FPC), but there was disagreement about their accuracy. The Academy suggested that two consulting firms, experienced in the Gulf fields, should make independent estimates using the same geological survey data that was used by the FPC. This was done for a random sample of nineteen (out of a total 168) leases. All the estimates by these firms proved to be lower than those of the FPC staff, but a comparison between the estimates made by the two firms was more interesting. For one lease, the difference between their estimates was only 10 percent. But for nine of the leases, the upper estimate was from two to eleven times higher than the lower estimate.

Even before a well is drilled, companies will appraise the potential of a new region to help them determine whether or not exploratory wells are worth drilling (see figure 3). Once a well has been successful in finding oil and gas, two new estimates can be made: first, an estimate of the minimum amount (the proved reserve) that seems to be producible by that well and, second, a less certain estimate of what might be the ultimate potential of the entire field. As more wells are drilled and additional production data are gathered, the proved reserve estimate may be revised up or down. The expectation of ultimate production can also change upward or downward, usually over a much wider range than that of the proved reserve figure—several multiples are common. For a typical field it takes approximately five or six years before the proved reserve estimate of remaining oil plus past production begins to approach a true estimate of ultimate production. In other words, it takes a number of years before there emerges a reasonably accurate estimate of what has been discovered in toto in a reservoir. The exact amount of producible oil or gas is not known until the field is permanently abandoned and that oil or gas has been measured as past production.

In addition to a company's estimates of proved and ultimate, other appraisals may be made by a producing company during the life of a field for various purposes. Estimates based on well logs and other data are commonly used by banks for making loans. Information is also released to the press about the importance of new discoveries. The Securities and Exchange Commission expects that companies will release information to stockholders about their holdings and expectations. And, finally, the many kinds of information required by government agencies lead to a number of estimates being provided by a variety of federal offices. Considering the array of purposes for which reserve estimates are made and the constant revision of most of these through time it is not surprising that various reserve reports may appear to be in conflict.

The proved reserves of oil and gas represent only a small portion of the total oil and gas resource base that remains unrecovered in the United States. Yet, these proved reserve data sometimes receive more attention, and in recent years have prompted more controversy than the more significant resource estimates of undiscovered oil and gas. [Footnote 2] For crude oil, proved reserves represent the stock of immediately producible oil from existing wells. The oil producer knows that the amount of oil or gas that can flow in a given year from producing wells is physically linked to the number of wells available and the quantity of oil and gas still remaining in the reservoir. Thus, proved reserves for many years have been the industry's empirical indicator of current capability, not a measure of the total supply of oil and gas left for the future. Equating proved reserves with "years of supply" is particularly misleading.

Since proved reserves plus past production are normally less than we might reasonably expect to produce from known oil and gas fields, do we have any estimates of this undeveloped and less certain, part of our discovered oil and gas reserves? Unfortunately, there are no regular government or industry-wide efforts to report on what is known as indicated or inferred oil and gas. [Footnote 3] The American Petroleum Institute (API) does report on additional reserves of oil that could be produced from secondary recovery projects but that are not yet fully evaluated at the time of the proved reserve estimations. An industry-sponsored effort, called the Potential Gas Committee, has included this portion of the gas reserves in its occasional reports on gas resources. The Federal Energy Administration in its 1975 survey of operators had hoped to go beyond merely proved reserves, but its final report only included oil from secondary and tertiary projects not the less certain oil and gas quantities. Currently, the new Department of Energy is again considering how to define and request data on oil and gas reserves that are not reported as proved.

The U.S. Geological Survey in its 1975 Circular 725 relied upon the use of a statistical ratio devised by M. King Hubbert to account for indicated reserves. This ratio is based on the historical relationship of the amount of oil and gas that has been added through extensions and revisions to proved reserves during the typical life of reservoirs. The relationship shows that approximately 80 percent more oil and gas will be produced from known fields than is currently being reported as proved. Although proved reserves data are considered to be accurate within plus or minus 20 percent, this refers to the oil and gas expected to flow from existing wells. On the average, almost twice as much oil and gas will ultimately be found in these fields once their true size or limits become fully identified. This is not an intentional understating or hiding of reserves, but merely a reflection that the definition of "proved" limits the estimators to the drilled portion of the field.

One must be aware that when estimates go beyond proved, accuracy deteriorates rapidly, with errors of perhaps 50 percent or more for indicated reserves (mostly oil and gas resulting from further development of the reservoir) and amounting to perhaps several multiples for inferred reserves (oil and gas resulting from the discovery of additional reservoirs within the same field). Many U.S. professionals and Canadian government specialists would prefer, because of the uncertainty, to consider inferred oil and gas as not actually discovered.

To obtain more information on what lies beyond proved reserves in known fields involves a reservoir-by-reservoir examination of considerable magnitude. Considering the range of judgment involved and the unavoidable approximations, the added information obtained may not be worth the cost of acquiring it. In addition, there are problems of handling proprietary data, which, in any event, would be diverse, constantly changing, and of unknown quality and usefulness. It would require combining the personal judgment of the ultimate potential for field A made by a geologist from Company X with estimates for field B from Company Y's geologist, through all of the thousands of fields in the United States. In the final analysis, to know that indicated and inferred U.S. reserves are considered to be 2.4 times our proved reserves instead of 1.8 times has little significance in determining our policies with respect to oil and gas. The more important questions are found in the categories of subeconomic and undiscovered oil and gas resources. It is upon these quantities that our energy future depends most heavily in the medium term.

Estimation of Subeconomic Resources

Subeconomic resources include all of the oil and gas in known reservoirs that is not producible by present technology at present prices, but may become producible in the future with improved technology or higher prices. It should be noted that a simple downward movement in prices or other incentives can cause some reserves to be reclassified, at least temporarily, as subeconomic resources.

As a field is developed, following the drilling of a discovery well, the producer adopts a plan which he hopes will extract all of the oil from the reservoir that can be produced. He then hopes to sell the oil at an anticipated price that will return as much or more than the additional cost of producing it. He hopes that the aggregate return from all wells, over and above the direct costs of production, will pay him not only for production costs but also for the costs of exploration, dry holes, and abandoned wells. To minimize the duration of his exposure to uncertainty, he hopes the pace of production will permit a quick return of his initial investment.

The investment decision in oil production is a balancing of the total quantity to be recovered, the various costs, the rate at which the oil will be recovered, and the price at which it can be sold. Once the decision is made on how many wells to drill and what recovery technology to use, that is, natural flow, pumping, water flooding, injection of steam, or other methods, the amount of oil that will be recovered and the rate at which it will reach the surface are limited within a fairly narrow range. To change that plan, additional investments must be made in drilling additional wells or in altering the production methodology being used. Such a change in the production scheme will be adopted only if the faster production or greater recovery can justify the extra cost.

Thus, an increase in the price of oil or gas may not be adequate to change the plan for the operation of a field already being developed. The only effect of higher prices in that event may be to permit the reservoir to decline to a lower level of daily output per well before it is abandoned because of low oil flow or gas pressure. This additional quantity of oil and gas produced in the later life of a reservoir may only involve a 1 or 2 percent increase in the ultimate recovery because most of the oil or gas left behind is entrapped in the reservoir and could only be recovered by the use of a different technology. However, higher prices which occur before a production plan is fully implemented in a new field can lead not only to later abandonment but to higher recoveries because the prices can be reflected in a timely investment in a modified development scheme.

The appearance of significant improvements in production technology or markedly higher prices can justify modifying the way new reservoirs as well as fully developed fields are being operated. Even an abandoned reservoir can be reopened, although this is less likely because of the expense. It should be noted, however, that new methods of enhancing the recovery of oil and gas should not be viewed as applicable to all kinds of reservoirs. How successfully a new technology can be employed is determined by the kind of structure and natural energies in the reservoir, as well as the kind of rock that is found in it.

There has not been much experience in estimating the size of the national subeconomic resources of oil and gas because in the past the opportunity to discover new and plentiful oil and gas resources has always seemed more attractive to industry. Even for known fields, the estimation of subeconomic resources is complex. First, the estimator must face uncertainty about new and perhaps untested technology. Second, there is need to deal with the effect of price on production using established technology, as well as what price is required to make new technology commercially feasible. Third, there is a lack of information on exactly how much oil or gas is left in the reservoir to be recovered by new technology. Finally, if the data are to have meaning, there is need to deal with the problem of the time over which these prices and technology can be assumed to occur.

It is perhaps surprising that the amount of oil left behind in a reservoir is uncertain. Depending upon the kind of reservoir and the years during which it was exploited, oil recovery from the initial development plan used can vary anywhere from 10 percent to 80 percent of the oil estimated to have been in place originally. In some cases, reservoirs have been reworked with a secondary production technology long after primary methods were begun. More recently, developed fields tend to be exploited by several integrated methods. Since oil in place, recoverable reserves, and a reservoir's recovery factor are all parts of the same equation, it is apparent that estimates for two of them allows the derivation of the third. Thus, if greater production from a reservoir is achieved than originally [Footnote 4] expected, one is never sure whether the cause is more oil in place, more reserves, or a higher recovery factor.

There is some indication that the overall recovery factor for oil in the United States did not improve very much during the 1960-75 period for several reasons. U.S. production shifted from regions with naturally higher recovery potential to areas with poorer recovery potential. Early estimates of the quantity of oil to be recovered by primary recovery techniques were probably overstated or, conversely, the oil in place may have been understated. Finally, there is a tendency to use a standard recovery factor in relating future production expectations to oil in the reservoir. Each of these tendencies could contribute to the assumption that the U.S. recovery factor has remained near 30 percent for many years. Realistically, it must be concluded that the true national recovery of oil is an unknown percentage.

The current interest in enhancing petroleum recovery by injecting heat, CO2, or chemicals has led to more vigorous examination of subeconomic resources than ever before. Our major oil regions have been examined in terms of the amenability of the various kinds of reservoirs to newer methods for increasing recovery. Although some optimistic suggestions have been made that U.S. recovery could be increased from an assumed 32 percent to ultimately 60 percent, more modest near-term goals are now being set for the upgrading of some of our subeconomic resources to reserves. These suggest an overall increase of perhaps 5 to 8 percentage points in the U.S. recovery factor may be possible.

The uncertainty of how much subeconomic oil and gas may be produced is illustrated by the recent report of the National Petroleum Council (NPC). The additional oil from enhanced recovery, according to the NPC, could be as little as 7 billion barrels, under a price assumption of $10 per barrel (1976), but this would increase to 24 billion barrels at $25 per barrel. The effect on the rate of annual production would also vary. At the higher price level, US oil production could be 3.5 million barrels per day greater in 1995. The uncertainty in the estimates is reflected in the judgment that the higher 24 billion barrel amount is merely the central value of an estimate ranging from as little as 12 or as much as 33 billion barrels. In contrast, another study viewed the outer limit of enhanced recovery at 76 billion barrels at $15 per barrel (1974). Despite the fact that enhanced recovery deals with "discovered" oil in known fields, this does not narrow the range of uncertainty. Technological and economic forecasting of recovery is a source of frustration equal to that of estimating the undiscovered (Table 2).

Estimation of Undiscovered Resources

If the United States would suddenly cease drilling its customary 25 to 50 thousand new wells each year and would be content with merely producing what it could from existing wells, production would decline progressively by approximately 12 percent per year. After some forty years, production would fall to approximately one percent of present production. Since reservoirs do not cease production abruptly, some wells might still be producing a few barrels per day after one hundred years.

Unlike manufacturing, or some kinds of mining operations, the capacity to produce petroleum is not a constant. To avoid a decline in national production, there must be continuous drilling and development. The process of continual annual replacement of what we have produced is heavily dependent upon the magnitude of our undiscovered oil and gas resources.

The potential size of these resources is usually evaluated in one of three ways. There is the geologic or volumetric approach, which attempts to make a direct estimate of the quantity of oil and gas remaining to be discovered and recovered. No attempt is made to show when or if these resources will be produced. The second approach is that of the engineer-manager who makes projections of the drilling, discovery, and production process. These future production profiles implicitly suggest the amount of recoverable oil and gas that is left in the ground. The third methodology is that of the economist who uses the equations in his model to suggest what future supply can be achieved by the oil and gas producers as they respond to price changes. Like the engineer-manager, the econometrician may indicate the quantity of remaining oil and gas resources in his model implicitly rather than explicitly.

The Geological Survey recognizes that Circular 725, while a major advancement, was a first effort and must be used with considerable care. Only the portion of the report concerning "undiscovered, economic" resources of crude oil and natural gas are totally new estimates. All other numbers presented were based on other sources of statistics or were derived by simple ratios. Thus the measured and indicated additional reserves are taken from the reports of the American Petroleum Institute and the American Gas Association. The inferred reserves of oil and gas are based not on an evaluation of fields and basins but on the historical trend of extensions and revisions of proved reserves through time. The subeconomic resources are based on simple ratios using two assumptions—that the average U.S. crude oil recovery might reach 60 percent and natural gas 90 percent at some unspecified time in the future.

The additional data provided by Circular 725 has been useful, but it presents problems for many of its users. If the undiscovered oil and gas resources are shown as a probability range^5 what does one use if one needs a single number? Unfortunately, many seemed to prefer to use the lower limit. In using these data, it has been common to overlook the fact that the subjective judgments behind the estimates were based upon price and technological conditions that prevailed prior to the Arab embargo and the quadrupling of the world price of oil. This leads to the question of how much current prices might alter the estimates. It is believed by many that a recalculation would not make a large difference because the estimates are dominated by large fields which were economic before 1973.

The Survey is now searching out the answers to a number of questions. Can the reserves portion of the estimate, which received modest attention in the first effort, be improved? In estimating subeconomic resources, there is the important question concerning the realism of ever reaching an overall 60 percent recovery. This may now be better understood because of the extensive work just completed in examining enhanced recovery. Since the availability of actual experience by Survey geologists for every oil and gas "play" was naturally somewhat limited, there is an interest in how to tap a broader spectrum of judgment. There is a need to have future appraisals encompass more data on: size distribution of undiscovered fields, depths, reservoir types, and basins found in deep water offshore. These data may be the key to determining actually how "price sensitive" are oil and gas resource estimates. Until these answers are forthcoming, the user of the Survey's Circular 725 must remain fully aware that these are subjective views by a group of government geologists concerning the recoverable amount of pre-1973 "commercial" oil left in the unexplored portions of U.S. basins onshore or to a depth of 200 meters offshore. This recoverable amount is only a portion of the total physical quantity of oil and gas in place remaining beneath the surface in the United States.

Many government or industry projections of future production are not primarily designed to deal with the ultimate size of our oil and gas resources. Nonetheless, they still may foster a public belief that resources are adequate in size to meet the projected goals. In addition, there may be only minimal attention paid to the price required to elicit the necessary investment. And whatever that price may be, the accompanying alterations in the demand for oil and gas, given that price, may not be addressed at all.

One specialized form of engineering projection that has been employed by many authors over the years has been to use a production-history profile that will follow the standard pattern observed when minerals or fuels are produced from a finite deposit. The classical configuration is a bell-shaped curve showing an upward sweep, a peaking of production, followed by a decline to final depletion. Not only a specific deposit, but a state, a region, or a nation, as an aggregate of many deposits, often appears to follow this pathway.

If one assumes that this is the general behavior of production, then it becomes possible to estimate when the peak will occur, the quantity that will be produced, and how it will be distributed over time. Customarily this is done by rather simple curve-fitting techniques using an appropriate mathematical formula. One of the most popular production-history curves has been the logistic growth curve, since it is both bell-shaped and symmetrical when fitted to annual data. It can also be used as a long attenuated "S" showing how cumulative production will approach asymptotically a line representing the maximum recoverable resources. In reviewing forecasts between 1948 and the mid-1950s, RFF's Sam Schurr and Bruce Netschert found at least six authors using this kind of approach who expected U.S. petroleum production to peak by 1970.

Among this group, perhaps the greatest amount of attention in recent years has been directed toward the work of M. King Hubbert. Relying heavily on the logistic curve and a family of various statistical series to track the behavior of U.S. oil production, his analysis is both extensive and detailed. Among the family of interrelated curves that he uses are: cumulative proved discoveries, cumulative production, proved reserves, annual production, annual increases and decreases in proved reserves, and discoveries of oil per foot drilled versus total footage drilled. The first three of these, with the appropriate fitted curves, are shown in figure 6.

The logistic curve provides a particularly good fit to any historical series that is approaching or has already reached a plateau. But one must be careful not to assume that the fitting of the curve to the several variations of the same historical series in some way confirms the validity of its use. Despite the vigor with which Hubbert examines past behavior and projects the various patterns of U.S. oil discovery and production into the future, this does not necessarily indicate that the logistic curve is a more reliable predictor of the future than any other curve that might have been chosen.
The use of mathematical formulas to project trends forward provides an aura of precision and objectivity. However, the process of fitting and projecting is a more subjective process than it might appear. The choice of the type of curve to be used preordains in a general way what the future will look like. Then the analyst must exercise further judgment as to the time period to be used and how the curve is fitted to the data. Judgments at this stage of the analysis are particularly critical, because the manner in which the final years are fitted affects the steepness of the expected production decline.

In addition to his decision to use the logistic curve rather than another, Hubbert's case for future decline is bolstered by his assumption that a declining amount of oil will be found per foot drilled. This is a persuasive position to take if one expects that the largest and near-surface fields have been found first. Hubbert supports this hypothesis by another projection involving again the choice of proper data and a mathematical formula° to project the trend of oil found per foot drilled. In this extrapolation he shows that despite extensive drilling the quantity of oil found in the future becomes relatively insignificant and his gloomy expectations for the future are further substantiated.

In mature, densely drilled areas, such as the onshore areas of the lower 48 states, one might be persuaded that it is unlikely that there are many "surprises" left. It does appear unlikely that there are a number of Prudhoe Bays cleverly hidden by nature along the Gulf Coast or in the Rocky Mountains. For these areas it seems to be a question not of new peaks in production but of the duration of the current level of production and the nature of the subsequent decline. But in those areas where drilling has been infrequent—natural gas at greater depths along the Gulf or oil and gas in the Arctic or farther offshore— the future is not as clearly defined. For these regions, most analysts, including Hubbert, do not trust unalterable formulas but turn back to more traditional geologic speculation about the quality of the targets that may be found in these unexplored volumes of rock

Production-history profiles provide for many a sense of "rightness" because they follow the classical pattern of mineral production. Yet they leave a number of questions unanswered. Is it fair to assume that the historical interrelations between the many factors that affect oil exploration and production will continue unchanged into the future? If our production of oil and gas has peaked, has that fact been induced by external institutional factors rather than by a limitation in our hydrocarbon endowment? Might not our technical ingenuity or the new frontiers of exploration once again produce a major surprise? If secondary peaks have been observed in the past for states and nations, why not for the United States?

There are no certain answers to these questions. But for the United States, M. King Hubbert has reminded us of one unavoidable truth—we are not debating whether there will be decline, merely when and how. This does not deter the exploration optimist from reminding us how we drilled for many years in Texas, Saudi Arabia, and the Arctic with little or no success.

Econometric models have their own special link to the past. The response, or elasticity, of oil and gas supply to price must be judged in large measure in terms of historical data, despite the realization that in each future year we will deal with a different segment of the original resource. Future resources may very well differ in character and, as a consequence, in cost from those discovered in the past. Economic behavior patterns of operations conducted on vast federal leases in 1,000 feet of water are not the same as those encountered in the private farmlands of Kansas. Nor will the response of gas supply to a doubling in price (in constant dollars) be the same when it starts at 10 cents per MCF as when it starts at $1. A reason to question further the future validity of past experience is to recall that much of the past was characterized by smaller movements in the price of oil and gas relative to other prices, and that for the most part this was downward not upward.

Since so many analyses have come to depend upon it, the further work of the USGS has become extremely critical. Currently, the Survey is hoping to refine its presentation of probability data on undiscovered oil and gas resources so that the full range of potential resources within the hypothetical and speculative categories is more apparent. This will allow for an appreciation that beyond the 5 and 95 percent probability boundaries there still remain possibilities for either zero finds or major discoveries for which past experience has not prepared us. Recent interest on the part of the National Petroleum Council and other groups in enhanced recovery will now permit the Survey to be somewhat more specific about the magnitude of subeconomic, discovered resources. In addition, the presentation of data on indicated and inferred reserves (reserves beyond proved) in known fields may be expanded by the Survey.

A number of agencies have joined forces with the Survey in this effort to determine how additional information in economics and technology could be combined with the essentially geological data from the Survey's Resource Appraisal Group. This Inter-Agency Study Group on Oil and Gas Supply is examining not only the amount of the oil and gas resources present but their distribution with respect to size, geography, and depth. An attempt will then be made to define the level of exploratory drilling activity required to find these deposits. This can then be followed with studies that deal with drilling, production, and finding costs, which incorporate considerations of reservoir depth, water depth, and other geological characteristics.

Currently the group is in a testing phase, using three pilot areas to determine the feasibility of performing these various tasks. It is not expected that the current effort will attempt to deal immediately with the time distribution of future oil and gas supplies. But it is hoped that a better appreciation of the magnitude of oil and gas resources at various costs will be obtained, as well as an understanding of the exploratory effort that will be involved.

Other Occurrences

Other occurrences are frequently the source of possible deception about the size of the nation's usable oil and gas supplies. Billions of barrels of oil in low-quality shale, gas locked in impervious shales and sandstones, methane found in coal beds or dissolved in brines under great pressure at depths of 15,000 feet are all a part of our physical resource base. They can and should be accounted for in any total resource inventory, but they cannot and should not be considered comparable to reserves or subeconomic resources. The likelihood of their soon becoming producible under present or near-future prices and technology is small enough that their importance for present generations is uncertain. Thus considerable caution must be taken to avoid giving them too much leverage in current decisions. After fifty years of effort and anticipation, the first commercial barrel of U.S. shale oil has not yet been produced. To be deceived by a too hasty reliance on methane dissolved in the waters of the Gulf of Mexico would be foolish indeed.

Although the oil shales of the West have become the classic example of a "just-around-the-corner" resource, we must somehow account for such a vast quantity of hydrocarbons. Many oil and gas resource appraisals do not include the oil shales because they are restricted to conventional crude oil, natural gas, and natural gas liquids produced from wells. Other analysts do not include them because they are not economically producible at the present time. If, however, a complete accounting is desired, then it is appropriate to at least identify these as other occurrences or noneconomic resources which are currently not produced and are likely to be significantly more costly than other forms of energy now being used. Whether quantification is attempted depends upon the purposes of the inventory.

Hydrocarbons occur in many forms in nature. Just as there are many types of coal (anthracite, bituminous, or lignite) there are heavy oils, tar sands, and kerogens which will not flow to drilled wells. This requires the extraction of the material either through the use of heat and chemicals or physically mining the rock so that it can be processed above ground. Since these are sedimentary deposits, they can be vast in extent but highly variable in recoverable energy content. In effect, they are low-grade deposits requiring expensive processing. As such they must be considered as either subeconomic or probably non-exploitable in any period of time that is of significance to present generations.

A number of largely unexploited sources of methane, the most abundant of the natural gases, have also attracted considerable attention. Among these are natural gas in dense sandstones of the West and the Devonian shales of the East' where the rock is relatively impermeable and does not allow the gas to flow freely to a well. As a consequence, the drilling of a well in these formations is not often rewarded with a great quantity of producible gas or a high daily rate of production. Methane in coal is well known as a hazard to mining and is actually recoverable by drilling holes in the coal bed in advance of mining. Another recent discovery has been of the presence of methane in underground salt water found at considerable depth in the Gulf Coast area. The gas is held in the water by the great pressures that exist at the depth.

Relatively simple calculations of the volume of oil shale in the Piceance basin, tarlike substances in Utah, and methane in coal beds or other geological settings yield vast quantities of energy that physically exist. However, like exotic rocks on the moon, the fact of their existence should not be confused with economic and technological accessibility.

In the other occurrence classification, there is also that portion of conventional oil and gas that we do not expect to recover. Similar to low-grade oil shale, it would be physically possible to produce this oil and gas at great cost. One could literally mine an oil reservoir and produce all of the oil, or let a gas well produce until there was no more pressure left. Obviously, long before this, it would be far more sensible to use some other source of energy.

Thus, those portions of our oil and gas resources unlikely to ever be recoverable can be accounted for among the other occurrences.

Productive Capacity

It is virtually impossible to determine how much oil and gas can be produced in any given year solely on the basis of knowing the quantity of proved reserves of oil and gas. If the nation finds itself lining up at gas stations or shutting down factories because adequate pressure cannot be maintained in all the utility mains during cold weather, the immediate supply problem is the productive capability of the delivery system, not proved reserves or undiscovered resources. (7)

Over the years little study has been directed toward understanding the limits of the delivery system upon which we depend to move energy from the well to the burner tip. For the fossil fuel group, we have only the American Petroleum Institute's (API) estimate of productive capacity. This is the maximum daily rate of production which could be attained under specific condition on March 31 of any given year. It would require ninety days to reach, starting January 1, and is based upon existing wells, well equipment, and surface facilities. The estimate provides for no reduction in ultimate recovery, and environmental damage or other hazards are not accepted.

Obviously, it is useful to have such a measure of our capability. It is important, however, to be aware that the API definition of national productive capacity does not imply anything about the sustainability of this capacity over any specified period of time beyond the ninety days. Beyond March 31 of the year of estimate, the productive capability would begin to decline. This particular measure of productive capacity does not encompass our capability, or lack of it, for storage, transportation, and processing facilities to handle the oil once produced.

The gas shortage of the winter of 1976-77 was a combination of system limits and the declining deliverability of gas from existing wells. Through emergency measures, such as denying some customers gas and shifting gas between systems on an ad hoc basis, the nation coped with the situation. This does not alter the fact that the gas deliverability from wells in 1977-78 will be different from what it was in the past winter. It is not generally appreciated that ad hoc plans that worked once may not work as well in another heating season. The fact remains that our oil and gas systems involve thousands of enterprises delivering fuel to millions of households and commercial customers. Exactly how that system works and what its capability might be, given the declining production curves of oil and gas, is only partially understood. As of midwinter, the weather and the amount of natural gas in storage did provide some optimism for the 1977-78 heating season.

The inability to discriminate between reserves and deliverability is the source of considerable confusion in the reporting on the oil and gas situation in the United States. References to the estimated total reserves in a field or resources in a new region are equated with annual production or requirements. Since only approximately 10 to 15 percent of the reserves of a field can be produced in a single year, if connected to a delivery system, billions of barrels of oil reserves or trillions of cubic feet of gas translate into a much smaller amount of oil and gas available even in the early peak years.

To understand oil and gas supply requires more than a realization that estimating reserves is an inexact process; it also involves an appreciation of the limited capability of a well to produce oil and gas upon demand. The time required to explore, find, drill development wells, lay pipe, and provide process facilities is a further restraint on translating reserves into production. To this year's energy consumer the only supply that counts is deliverable oil and gas, not reserves or resources. If flow is inadequate, periods of five to ten years or more and considerable investment will be required to alter it in any significant way. Considerably little attention to the limits of this process, and perhaps less to reserves, seems warranted.

It is understandable that the productive capacity of the vast oil and gas producing industry and its downstream facilities presents problems in terms of measurement. One would expect, in contrast, that the capability of a known producing reservoir would be a reasonably precise number. This has taken on a new importance in recent years, as questions have been asked about whether or not producers holding federal leases have been producing oil and gas as diligently as they might if the price for oil and gas were higher.

This particular question has revived interest in a measurement that appeared during World War II called maximum efficient rate (MER). In concept, MER can mean the theoretical physical capability of a reservoir to produce oil and gas over time as a hydraulic unit. To exceed this rate may reduce the amount of oil and gas recovered. In practice, MER has come to mean the maximum rate, to terms of barrels or cubic feet per day, a reservoir can produce "efficiently" and "economically" from a fixed number of wells under actual operating and market conditions. In effect, this is the design capacity of the reservoir development plan, and reflects what the operator feels is economically justifiable. Producing the existing wells too rapidly could cause oil and gas to be lost in the reservoirs. Drilling additional wells could increase the flow from a reservoir without harm and might even increase the amount ultimately produced, but the decision then rests on whether the extra expense of the well can be justified by the economic advantage to be gained by producing more oil or gas or producing it faster. That particular judgement may depend on whether you are the producer or the federal government leaseholder.

Even the term efficiency provides its own share of problems. To the economist, efficiency will tend to be interpreted as determining how recovery should be distributed over time to arrive at a maximum value for the oil and gas produced regardless of the physical recovery. The reservoir engineer's technical efficiency will be to achieve maximum physical recovery at the lowest cost under current market conditions. The government administrator may be interested in a production rate that provides a maximum quantity of oil and gas to the public when it needs it, commensurate with a reasonable return to the producer and acceptable payments to the Treasury of bonuses and royalties. For a given reservoir, the annual production under these three criteria will not necessarily be the same.

A difficult aspect of using MER is that it is not a constant. Since production of a well or reservoir is from a declining reserve, the producible quantity actually decreases from day to day. In practice, MER is determined periodically for most federal leases. It should be remembered that a reservoir may include a large number of wells and that the maximum efficient rate of production applies to the whole reservoir and not to any individual well. In the past, MER was considered a conservation technique to prevent wasting the natural energies of the reservoir by moving the various fluids so rapidly that oil or gas is left behind entrapped in the reservoir. Even so, in emergencies such as World War II, it is sometimes considered in the nation's best interest to produce oil or gas at a rate that actually causes some loss in ultimate production.

Offshore operations from costly platforms have introduced new dimensions to production and transport that have increased the difficulty of determining what is both physically and economically possible to accomplish. As a result, MER, which in the past has been primarily a state conservation regulatory tool, creates a number of problems when it is used as a measure of diligence in exploiting federal leases. The federal government continues to look for a means whereby it can properly monitor operator performance in terms of the national interest without harming the entitlement of the leaseholder to serve and to protect his own interests as a private enterprise.

Conclusions

The RFF staff has now engaged in over two years of studying, discussing and explaining the uncertainties of oil and resource estimation. That has led not to better numbers but to perhaps a better understanding of what the existing data can or cannot do for us. By and large, we find that most examinations of oil and gas resources reflect in part the professional background of the estimator but most importantly the purpose for which the work is designed. Many estimates that have been published provide total future quantities of oil and gas that may be produced rather than supply in the economic sense or rates of production over future time. All too frequently, these totals may be translated into years of remaining oil and gas by dividing them by the current or some other assumed rate of production. This leads to the too-simple conclusion that we may be out of oil and gas at the end of that number of years.

Published estimates of total future producible hydrocarbon fluids provide the public with a narrow view of future oil and gas supply. This is compounded by the fact that the public does not know how to interpret the figures. As one RFF workshop participant noted, "the difficulty [in publishing estimates] was the problem and confusion in the public's mind of what all these numbers mean. It has just been an absolute mess. They [the public] have taken undiscovered resources and related them to reserves, and this was not really our intent at all. Suddenly, we find ourselves quoted in the most peculiar ways. And much to our embarrassment."

To the economist, it is important to know how much and for what period of time oil and gas production rates might be increased by a change in the economic structure of the industry or in the cost—price ratios. Or how sensitive future oil and gas production rates are to changes in technology. Answers to these kinds of questions are not contained in the usual published estimates of future oil and gas reserves. As a participant in one RFF workshop said, "Not a single technique, approach, publication, or anything, has yet adequately dealt with what the economists would call the supply schedule. Somehow we have got to get some indication of what different levels of future supply would be available at different cost—price relationships."

In any given year the production of U.S. crude oil encompasses production from a reservoir that has been newly discovered, along with the production from reservoirs ten to fifty years of age. The important point to note is that the future rate of supply will be a composite of the rates from both old and new reservoirs.

The experts who have appeared in the RFF workshops are agreed that the rates of oil and gas production cannot be increased indefinitely. At some point, the rates must inevitably peak and then continue to decline until the hydrocarbon resources of the earth are exhausted. What the experts cannot agree upon is whether the annual rates of oil and gas supply can still be increased, and over what period of time, before the final decline begins. The extremes are illustrated by proposing different extensions into the future of the past oil production rate curve (see figure 7). The conservative view states that most of our giant fields have been found; the maximum annual rate has already been reached; and that, henceforth, there will be nothing but decline (see Curve A). An optimistic view might be that the annual oil production rate may still rise in response to exploration and new technology to some future maximum from whence it would decline until all reservoirs were exhausted (see Curve B). A moderate view would be a future annual production rate curve somewhat between these two extremes (see Curve C).

Whatever extension is predicted for the future rate of oil supply, the area under the resulting profile of the future can be no greater than the total amount of oil which one estimates can ultimately be recovered. Thus, the estimator of the conservative situation (Curve A) not only envisions a declining rate of production but also a limited amount of total production yet to be achieved. The optimistic estimator (Curve B) sees not only an increased annual rate of supply but also a larger volume of oil yet to be produced. Whether the estimator approaches the problem in terms of rates or total future production, the results of the estimate must be internally consistent with respect to the relationship between rates and cumulative production.

Many of the published estimates of future oil and gas supplies have provided a value for the total supply through use of the traditional "volumetric" method. No matter how polished and sophisticated the details may be, the volumetric method still contains two basic perceptions: (1) that the occurrence of hydrocarbons in an unexplored geological region and the parameters that are associated with its occurrence will probably bear a relationship to previously explored regions, and (2) that the searching efficiency for finding oil in new areas will likely resemble what it has been in the past in older regions.

Any resources estimate for future oil and gas production in an unknown region may convey a misleading impression to the nonprofessional. The only thing an estimator can say with absolute accuracy is that he does not know whether there is oil or gas in a given region until wells are drilled to find out. The history of past estimates is rife with situations in which either little oil was found in areas where there was a high expectancy or much oil was found in areas where there was a low expectancy. Thus the U.S. Geological Survey's probabilistic 2 to 4 billion barrels of undiscovered recoverable oil off the eastern U.S. Atlantic shore may be used by many as a "guaranteed" number. It may become the basis of a political decision either to explore or ignore the area. An individual particularly concerned about environmental protection might be inclined to conclude that this amount of oil does not justify taking the risk of polluting the environment to go after it. The fact is that there may be no oil at all off the Atlantic shore or there may be many times the estimated USGS figure.

If one prefers to approach the future by considering the production rate curve rather than total future production, then the immediate problem is the lack of an established theoretical basis for predicting the future shape of that curve. It appears to be somewhat unwarranted to assume that the curve would be symmetrical. On the contrary, there is reason to suggest that the rising leg of the curve is dominated by one physical and economic process, that is, the discovery of new reservoirs, while the decline will be dominated by a different physical and economic process, that is, the depletion of discovered reservoirs.

The nature of oil and gas reservoirs is such that the highest rates of production occur in the early life of the reservoir. The flow factors of a reservoir taken together generally mean that more than half the production from a particular reservoir will occur after the maximum rate of production is reached. If it is possible to supplement the natural producing energy of the reservoir or to apply technology that will make more of the reservoir oil accessible to production, the history of the reservoir may show additional production-rate peaks after the first one has been reached. This was the case in Pennsylvania and in Illinois.

The best current estimate is that, with present technology and prices, an average of between 30 and 40 percent of the oil known to exist in discovered reservoirs will have been produced at the time the reservoir is considered commercially exhausted. This is an average to be interpreted as we understand expectancy, that is, some people die younger and some live longer. The amount that can be taken from a particular reservoir is dependent upon the nature of the oil itself and the nature of the reservoir. There are known oil reservoirs from which the ultimate production will be as much as 80 percent of the oil contained in the reservoir. In other instances, the amount of oil that can be produced with present technology and prices may be as low as 10 or 15 percent of the oil in the reservoir. If technology improvement or a price rise permitted an abrupt change in recovery factors, a late production rate peak might show in the oil supply curve.

Another approach to examining the future of oil and gas is an examination of the rate at which exploratory drilling finds new oil and gas reservoirs. This approach does not necessarily depend upon prior estimate of whether oil or gas is present. It assumes that, if oil and gas are present, they will be found. The approach requires, however, an estimate of the efficiency for finding in the future. The published graphs which show the manner in which the amount of new oil found in the past related to the total exploratory footage drilled indicate that the finding rate has been decreasing. The reasons generally given for why less oil is being found per foot of exploratory hole drilled include:

1. We have already found the big reservoirs and those near the surface. Future reservoirs will be found at deeper horizons.
2. The most desirable geological regions have been explored and drilled.
3. The geological areas remaining to be be drilled are more inaccessible and more expensive; for example, on the continental shelf.

Whatever the reasons may be for an expected decline in search efficiency, it is relatively easy to see that if one extrapolates this decline into the future, the contribution to production rates due to finding new reservoirs will diminish. Consequently, one would conclude that the maximum production rate probably has been reached, but not all the experts agree that the search efficiency must decline. Both technology and economics could affect the trend.

Disagreement resulting from various methods of estimating future oil and gas supply revolves around whether the volume of oil and gas remaining to be found, or our productive capability, is the primary limiting factor on the domestic production rate in the immediate years ahead. The National Petroleum Council concluded from its inquiry that at least until 1985 the amount remaining to be found is not the limiting factor. NPC visualizes that annual production can increase with appropriate attention to the drilling rates, finding rates, improvements in recovery factor, and economic adjustments.

The straightforward production-history approach of M. King Hubbert and others, which is appealing to many, does appear to be very useful in telling us what is likely to happen in the near future, if we continue doing things more or less the way we have been. It implies that production, drilling, and so on are insensitive to economics and policies. Barring an almost total interruption in exploration, such curves do seem to provide us with what should be our minimum expectation for future U.S. oil and gas output. However, this is not to suggest that any of the other kinds of appraisals are totally free of ties to past reserve and in-place figures and historical, economic, and technological factors.

There is greater satisfaction with recent estimates of future oil and gas supplies because the numbers appear to be converging. Instead of difference in orders of magnitude, two estimates may be within 10 or 25 percent of one another. In part, this merely reflects a greater consistency in methodology and assumptions than previously. Any comfort derived from this apparent consensus can be false. Although two estimators may now agree, even if they have used different methods, this does not necessarily mean that they are both right.

Finally, it is important to emphasize that all oil and gas resource estimates by the many analyses both public and private are dependent upon the same sets of numbers as starting points. Beyond this, there is no right methodology, and estimates are sophisticated guesses at best. All experts are agreed that the usable oil and gas hydrocarbon resources are probably sufficiently limited that the maximum annual rate of production and the decline until reserves are exhausted are events that will fall within a few decades, not much beyond that. The peak in the United States may have already been reached. Yet one must not minimize the importance of capturing the remaining one-half or one-third of our oil and gas.

So basically we are dealing with forecasts of annual production rates for two or three decades, and we must get an idea of the impacts of economic and chronological changes on these rates. Cost data should be assembled so that it will be possible to analyze better the responses to economic change. Attention must be directed to the effect of technological change on increasing the recovery factor of existing reservoirs and the lead times needed to accomplish this.

This work will be aided if we can eliminate some of the past disagreements of estimators that stemmed from a lack of consistency in defining recovery factors and other concepts employed. If estimators are agreed on anything, it is that the definitions of terms must be examined closely and more standard definitions accepted for future resolution, if not of the supply question, then at least of why the estimators disagree in fact. Such a resolution would be an important step toward substantive agreement.

This is not as easy as it may seem. Even a seemingly simple term "total oil and gas in place" changes in meaning due to changes in information and the economic or technological perception of the analyst. Gas in tight sands or heavy oils would not have been encompassed within the definition of that term a few years ago. Yet some output from these sources has now joined the production stream.

The realization that there are no measurements in oil and gas resource appraisals is important to impress upon everyone. Even in discovered reservoirs we do not measure the oil in place—it is estimated. Reserves are an estimated value derived from a prior estimate of the oil in place, taking into account economics and technology. If the oil-in-place estimate changes, so will that of the reserves. Reserves and resources are equal to the estimated oil in place multiplied by an assumed recovery factor, substantially less than 100 percent for oil, less the amount of cumulative production. Nothing could be simpler yet so uncertain. There are no hidden formulas for predicting the end of the finite supply of oil and gas in the United States or the world. The definitive study of future oil and gas supply and how it may be altered by economic and technological parameters that have not yet emerged still remains to be done. It is of little comfort that the final, reliable, appraisal of the oil and gas resources of the United States will prove to be historic rather than predictive.

Footnote 1 (M. King Hubbert, whose work received considerable attention in the press, was among the COMRATE participants. His estimate was reported as 72 billion barrels of oil and 540 trillion cubic feet of gas.)

Footnote 2 (Despite the apparent clarity of the generalized concept of proved reserves the actual estimation requires certain judgements to be made concerning the amount of extrapolation to be used, whether to include oil and gas from known reservoirs not actually being produced, or how to account for oil resulting from secondary stimulation. As a consequence, there are at least nine "official" definitions from various professional, industrial, or government agencies. Each one leads to some variation of the estimates made.)

Footnote 3 (The terms probable and possible are also used to describe discovered reserves that cannot qualify as proved or measured.)

Footnote 4 (Natural gas, because of its physical characteristics, normally has a high recovery factor—on the order of 75 percent. Subeconomic resources and opportunities for enhanced recovery of gas are thus limited for the most part to the fracturing of dense, low-permeability reservoir rock.)

Footnote 6 (In this case a line representing a constant percentage of decrease each year is used.)

Footnote 5 (The Survey does provide an average of the most likely and upper and lower estimates. However, the statistical usefulness of this average is uncertain and it only appears in the tables.)

Footnote 7 (Productive capacity is the more common expression in the oil industry. The gas utility industry's immediate capability is referred to as the "deliverability" of the gas.)

From time to time, RFF will publish a special issue of Resources that focuses on a single, timely topic. This, the first such issue, written by RFF Fellow John J. Schanz, Jr., embodies the results of a series of workshops.