INTRODUCTION

Reserves versus Resources

Our present oil supply comes from reserves discovered many years ago, most of which will be produced and consumed by 1999. New reserves must be continually found to provide a future supply of conventional crude oil for the world. The essence of petroleum exploration is to convert unknown resources to recoverable reserves.

Misuse of two technical terms, reserves and resources, has caused much confusion in the general public; they must be carefully distinguished. Reserves of petroleum are the amount that petroleum engineers know they can produce at a profit from known fields, using known techniques, in a known time. Resources, in contrast, are theoretical estimates of all of the oil and gas that may exist in any given area, most of which are unlikely ever to be converted into reserves.

Resource prediction is not a precise science, because results are not repeatable by other appraisers within an acceptable margin of error. Government planning agencies estimate resources, whereas industrial companies normally appraise only reserves. The total theoretical resources of any nation or region are largely irrelevant to operating oil companies, which are critically concerned with the next drilling prospect's potential reserves. Busy oil men publish little of what they know or plan to do.

The public appears to be interested mainly in the immediate economic availability of oil and gas rather than in geologically postulated petroleum resources. The key issue to the public is not how much oil may exist theoretically, but when it can be produced and what it will cost. Water and many metals can be repeatedly recycled, but once petroleum is burned, it is gone forever. The majority of the public does not appear to be educated on this difference. For all practical purposes the world will have to operate on its known (proven) reserves of oil and gas during the next decade, because of the long lead times required for exploration and development of discoveries. Under such conditions, discussion of potential resources badly misleads the public and politicians, many of whom assume hat reserves and resources are synonymous. Well-intentioned but careless scientists who continue to discuss resources instead of reserves may be one of the causes of our government's lack of realistic energy policies (Figure 1) (McKelvey, 1973).

Short-Term versus Long-Term View

Optimism is always encouraged in petroleum exploration, but it is misleading to imply to any government that oil discoveries are probable where geologic prospects are poor. It is preferable in such cases to hope for the best but plan for the worst. Overoptimism as a policy may be disasterous. Also, the important difference between technical and economic feasibility is often ignored. For example, it is technically feasible to send a person to the moon, but it is not economically feasible to do so.

A long-term view is academic when we are faced with critical short-term deadlines. Time is of the essence during energy shortages when fuel and electricity uses are necessarily curtailed. In general, Americans tend to be short-term oriented, and because of our political system, any time beyond the next election seems to be long-term for many of our politicians. The table below shows the only

End_Page 159------------------------

sources of oil that will be available for various time periods in the future:

Table

Regional Variations in Oil Richness

Oil prospects are not equal in all countries, nor do they stop at national borders. It is commonly overlooked by the public and many economists that some parts of the world are much richer in oil than others. Only a few areas have the high "A-factor" that puts conventional petroleum into subsurface strata. It is statistically possible, but geologically unreasonable, to apply an "average" amount of oil recovery per square mile to an untested basin. The huge resources of the statisticians result when they multiply large alluvium- or water-covered areas by an estimated uniform barrels per square mile factor. Unfortunately, this method is not geologically valid, since some of the small oil-rich areas are far richer than the extensive poor ones. (For example, tiny Kuwait has far more reser es than all of South America and Australia combined.) Oil and gas accumulate under very special circumstances in relatively small traps, leaving the remainder of the earth barren. The world's producing areas amount to <1% of the total surface of the continents and offshore shelves.

The sizes of oil fields in a given basin have roughly log normal distributions, meaning that there are only a few large fields but many tiny ones (Halbouty, 1970; Klemme, 1971b; Ivanhoe; 1976a). The five biggest fields usually contain most of the oil in any basin (Klemme, 1983). Also, the largest fields are the easiest to find because they are the biggest targets and are usually discovered early in the exploration if all of the basin is "politically accessible," (open for leasing). The major producing fields that comprise "big oil" are much more critical to any nation or international company than the very small fields having <5 million bbl recoverable. But every bit helps, particularly to a less developed country (LDC) that imports oil.

Exploration in the super oil-rich countries is much more of a sure thing (by several orders of magnitude) than long-shot exploration in the still unproductive nations. Large countries, such as the United States, the Soviet Union, Canada, Mexico and China, have a much better chance of finding commercial production simply because they have many basins, some of which may be rich in oil, among their extensive barren regions. This is in sharp contrast to the numerous small nations that have only one tiny basin in which to find oil. The few apparent exceptions to this size generalization are the small countries that share part of a rich petroleum province with large neighbors (e.g., Kuwait). Small isolated islands do not usually have any petroleum unless they are located within a producing asin (e.g., Trinidad) (Ivanhoe, 1979).

U.S. versus Foreign Oil Business

Two things are required to develop new oil reserves: geologic prospects and exploratory drilling. This study shows that from 1958 to 1980, the oil discovered per unit of effort in foreign, noncommunist, oil-producing nations averaged 30 times higher than in the 48 conterminous states of the United States. But in 1982 the exploration footage drilled in the conterminous United States totaled ~95 million ft versus 28 million ft for the rest of the noncommunist world combined. This indicates that there are strong political and economic reasons for the international oil industry's preference to drill in the conterminous United States, although the nation is long past its prime oil discovery years (which occurred during the 1930s) (Nehring and Van Driest, 1981). This business preference is ue to the legal peculiarity that the conterminous United States is the only place in the world where farmers usually own the onshore mineral rights and where the politically independent courts are impartial in law suits in resolving the contractual obligations between oil land lessors and lessees. This results in many oil deals and high drilling rates by small independent entrepreneurs who find it difficult to operate elsewhere. Comparable drilling rates and the number of discoveries of tiny fields will never be approached anywhere else in the world (Odell and Rosing, 1975; Grossling, 1976).

When free companies explore with their own money there is always a compromise between the gross local business terms (including politics, laws, royalties, taxes, and infrastructure) and the geological merit of an area. The overall foreign investment restraints must be tougher than in the conterminous United States or else the companies would be looking for oil overseas instead.

Countries lacking petroleum delude themselves if they hope to entice small U.S. independent oil companies to drill in their basins. Foreign nations must realize that they are in competition with all other countries, including the United States for the international oil companies' exploration funds.

Fig. 1. Illustration of the differences between resources and reserves: selling versus buying and the view point of the geologist versus the engineer, both of which are needed. (From Ivanhoe, 1980e.)

End_Page 160------------------------

Non-oil-producing LDCs often set their asking terms to oil companies at the high levels used in oil-rich countries, which are economically unrealistic for the nation's lacking in oil. These less fortunate countries may have to offer extremely favorable terms to stimulate interest by independents in their unattractive prospects.

Oil-producing nations have similar problems for any untested basins. Many public servants and politicians ignore or forget that oil and gas companies operating on their own money are not designed to be scientific institutions or semipublic services. Independent commercial oil companies are profit-oriented groups that are in business to earn money for their shareholders. The goal is money, not oil. They will risk their funds in exploration only in places where they can hope to find oil and gas that they can sell at a profit: no profit, no exploration. They can and will go elsewhere or will instead, diversify into any other business that has a better profit potential.

Less developed countries are not necessarily non-oil-producers. All of the 13 members of the Organization of Petroleum Exporting Countries (OPEC) and Mexico are oil-rich LDCs, whereas many industrial nations are lacking in oil (e.g., Japan, Sweden, Belgium, and South Africa). Mexico provides a good example of effective exploration by a big government oil company (Pemex) in an oil-rich LDC. It is the presence of major petroleum reserves, however, that matters, not that the country has a national petroleum company. Many LDCs do not acknowledge that merely setting up a government oil company and getting World Bank loans will not ensure them of any discoveries if their country has no oil (Meyerhoff, 1976).

Modern Petroleum Exploration Technology

Modern petroleum exploration is a very efficient process, and political and leasing delays often consume more time than the necessary geophysical work and wildcat drilling. Maturity of a basin's oil prospects comes wicker than it did in the early history of oil production. Digital seismic surveys have decreased the average finding time required to discover 100% of the reserves in a new basin's five largest fields after the geology is understood and the first large field found from 14 years in the 1940s to 6 years in the 1970s (if all of the basin is politically accessible) (Klemme, 1983). Modern digital seismic surveys give a much more accurate delineation of structure and stratigraphy for the estimated ultimate recovery (EUR) than did older methods, particularly offshore. Large strat graphic trap fields now seem to be possibilities rather than probabilities (Halbouty, 1982). If large commercial-sized stratigraphic oil fields (like East Texas), however, are common, geologists would have discovered more of them in the highly drilled conterminous United States or in other areas via digital seismic "bright-spots" during the last 25 years.

Since the 1960s, digital seismic techniques have reached such a stage of development that a virgin area covered with water (e.g., North Sea) or with permafrost (e.g., northern Alaska) can be easily surveyed and its poor areas eliminated, its best land leased, and its discovery wells accurately located. Modern geophysical and drilling techniques are very effective in finding oil in such areas and thus are just as valid in eliminating poor areas in other regions. There are practically no virgin areas left, and there is no place where an unknown basin the size of the Persian Gulf could go undetected. The recently discovered giant offshore Mexican oil fields all lie within the previously known Gulf of Mexico basin (Ivanhoe, 1979, 1980a, b). The world has now reached a semimature (i.e., "m ddle aged") stage of petroleum exploration.

Methods of Estimating Undiscovered Oil Resources

Estimates of undiscovered petroleum resources are usually made by one of three basic methods or combinations of them, which we can term Economic, Geologic, or Engineering. Each of these approaches has its problems and must be carefully reviewed to determine its degree of reliability. There is no sure way to predict the future.

Semantics are critical. It is often overlooked that there are few provable facts that are 100% certain. A geologic "fact," for example, may be 80% certain and 20% assumption (common sense). Common sense, however, is not a true fact in the scientific sense, but a product of our education and experience. If we change the assumptions, then we will correspondingly alter the final "facts." Thus, we must always separate the provable facts from the assumptions (estimates) to see whether changing the latter influences any conclusions.

Economic Estimates

The inherent assumption of economic estimates is that the petroleum supply is infinite and that all basic problems are those of distribution, which can be controlled by money. The need to actually discover oil and gas before it can be produced is not considered in the estimations. In the past, petroleum geologists found so much oil and gas that much of the public now takes discoveries for granted. Most laymen and many geologists are unaware of how the world's petroleum prospects have decreased since 1970 (Ivanhoe, 1984a,b,c).

Geologic Estimates

Geologic estimates of petroleum resources attempt to predict where the oil and gas are and how much may still exist. A critical shortcoming of such predictions is that the timing (i.e., between now and eternity) of discoveries is not indicated. Results are usually displayed on a cumulative probability graph. The amount of petroleum (zero to infinity) is graphed against the cumulative probability of its occurrence (0-100%) (Figure 2). A 75% cumulative probability on such graphs means that there is a 75% chance that there is more and a 25% chance that there is less than the indicated volume. Low, median, and high estimates (95%, 50%, and 5% probabilities) are often published. Some companies refine their predictions by discounting for the likelihood (e.g., 30%) of any commercial oil bein found.

Geologic estimates can take several basic forms, including volumetric appraisals, geologic consensus, and exploration plays.

Volumetric appraisals are scientific estimates of a region's potential resources (not reserves). The basic procedure is to calculate the basin's area or volume and then multiply it by some assumed geologic factor to get the total oil and gas that might be in the region. Many assumptions are involved in the process (Mast et al., 1980).

End_Page 161------------------------

Geologic consensus appraisals are simply the averages of the opinions of various experts about the undiscovered resources of an area (see other paper by Ivanhoe, this volume). The quality of a consensus can vary tremendously and critically depends on the experts' geologic knowledge about the area being studied. A minimum of six independent opinions, from both qualified petroleum geologists and engineers, are probably required to produce a valid geologic consensus. Each expert's opinions should be shown and identified on the final graphs. Individual appraisals for any basin are always suspect because there is no way to establish an expert's degree of optimism until his appraisal is compared with those of a group of his peers. Petroleum engineers are usually less optimistic than geologi ts, but both are much more conservative in areas where they have been personally humbled by the actual local problems of finding oil than in foreign areas where they lack first-hand knowledge (Ivanhoe, 1984f) (Figure 2).

A current scientific consensus is that the world will ultimately recover about 2,000 billion bbl of conventional crude oil (Hubbert, 1979; Wood, 1979; Nehring, 1980; Grenon, 1982). This figure, however, is probably slightly high because it was estimated by committees of scientists who had no way of being familiar with all of the world's onshore and offshore petroleum potential. The degree of optimism tends to be in inverse relationship to local knowledge.

The third form of geologic estimates, exploration plays (sets of drilling prospects), can only be realistically appraised if considerable amounts of seismic and geologic expertise have been built up in a given basin to provide a substantial amount of hindsight. An exploration staff develops several plays for a basin (e.g., the Santa Barbara Channel) for which such geologic parameters as structure, stratigraphy, depth, and size are reasonably well known. An estimate of the possible size of the postulated prospects is then made. This gross volumetric estimate may then be reduced by the assumed petroleum probability (e.g., 30%) of any commercial oil being found. (Operating oil company staffs are much more conservative in predicting probable commerciality than are government committees wh are rarely humbled by dry holes.) An entire region (e.g., the North Sea) can be approached in the same manner (Roadifer, 1975; Miller, 1982; Baker et al., 1984). Credibility dwindles, however, as hindsight decreases. At times it seems that all of the geologic effort involved in calculating and discounting the size of postulated plays in virgin areas may be no more reliable than simply scaling the qualitative reserve potential from graphs of maximum field expected versus minimum economic field (Ivanhoe, 1976a, 1979; Klemme, 1983; Baker et al., 1984).

Engineering Estimates

Engineering estimates of future oil and gas discoveries are much more realistic and conservative than economic or geologic appraisals. They attempt to tell us where, how much, and approximately when petroleum may be found and produced. Such appraisals are repeatable within acceptable variations and are usually made by petroleum engineers with extensive industrial experience. Some of the exploration play appraisals mentioned above approach being engineering estimates if they forecast when the oil can be produced.

M. K. Hubbert is the dean of engineering appraisers. While working at Shell Oil Company 30 years ago, he predicted by projecting finding rates that the United States would reach its production peak in about 1970 and discoveries would decrease thereafter (Hubbert, 1950, 1956). This prophecy was largely rejected by the oil industry because it ran against the conventional optimistic outlook of the time. Hubbert was proven correct, however, when the oil production peaked exactly as he predicted. There is no longer any scientific question of whether, but only when and how, our globe's oil age will effectively end during the first half of the twenty-first century (Hubbert, 1982; Ivanhoe, 1984e).

Hubbert's ideas were adapted to provide the basis for a series of engineering appraisals of most oil-producing countries by the U.S. Department of Energy/Energy Information Administrations/Foreign Energy Supply Assessment Program (FESAP). The small group of practical FESAP petroleum engineers and geologists patiently assembles and synthesizes the known data on reserves and oil

Fig. 2. Graphs showing cumulative probability versus estimated ultimate recovery of (a) U.S. oil (after Nehring, 1981), and (b) a typical producing basin in Colombia. The latter is a typical curve of a geological consensus. Individual opinions are distinguished by different symbols to show the range of optimism. X indicates the computed average. (From Ivanhoe, 1984f.)

End_Page 162------------------------

and gas production of all major petroleum countries so that all this information will be available to the U.S. government when the next energy crisis strikes. Their practical engineering reports, with an indefinite shelflife, will be invaluable at that time (Dietzman et al., 1979, 1982, 1983a, b, c, d, 1984a, b). (The FESAP reports are gradually superseding earlier preliminary assessments by the USGS (see Masters et al., 1981-1984).

Another important but simple engineering method is the oil discovery index (DI) rate method, which is discussed in the following section.

FOREIGN OIL DISCOVERY INDEX RATE METHOD

The Hubbert and FESAP engineering estimate methods are both quite mathematical and require knowledgeable technicians to make, update, and interpret the computations. Consequently, a simpler method was needed for routine planning purposes. I modified the old discovery index (DI) rate method (Lahee, 1946; Moody, 1978; Haun, 1981) for Occidental Exploration and Production Company (Oxy), to quantify the effectiveness of the past petroleum exploration of different nations. Projections of future oil and gas production in noncommunist countries were not the original objectives, but evolved as the study progressed (Ivanhoe, 1983, 1984, a, b, c).

Discovery Index Rationale

The rationale of the DI method is very simple. The technical effectiveness of oil exploration (new recoverable oil per unit of exploratory effort) and general oil richness of any country can be measured by the annual oil DI rate which was defined for foreign countries (Ivanhoe, 1983) as the annual number of barrels of new recoverable reserves added per foot of exploration drilling. The new recoverable oil added is equivalent to the annual production plus the recoverable reserves at the end of the year, minus the reserves at the first of the year.

Natural gas was not included in the DI because foreign gas is not the economic equivalent of easily transported oil, and foreign gas statistics are usually incomplete and untrustworthy. Communist nations were not covered because of the lack of reliable drilling statistics.

A country's DI rate is basically dependent on geology, not drilling. It is the discovery rate that matters. If the geologic conditions for finding oil are so poor that the DI rate is zero, then no amount of drilling will discover any oil. If twice as many exploratory wells are drilled in a producing basin in a given year, these wells will probably discover about twice as much oil, which would result in approximately the same DI rate for that year. Inevitably, the quality of prospects decreases as time goes by because the biggest fields (largest targets) are discovered first, which reduces the DI rate in future years.

The DI graphs summarize the oil industry's worldwide discovery record over many years, which is the true state-of-the-art of petroleum exploration. Figure 3 shows DIs for the conterminous United States on a nonlogarithmic graph. Semilogarithmic graphs were used for all other countries and regions to cover the range of DI values in Figures 4, 5, and 6. Extrapolation of the historical DI record will tell us roughly when and how much future oil may be discovered in any region. This is in contrast to the annual AAPG statistics on North American drilling activities that since 1946 have reported the number of new-field wildcat wells that have discovered "significant" reserves, which they define as new fields containing > 1 million bbl of oil. The AAPG's qualitative field sizes (A, B, C, , E, and F) do not actually tell us specifically how much oil was found in a given year and neglect the giant (>500 million bbl) and supergiant (>5,000 million bbl) fields that are most important (Ivanhoe, 1980a; Nehring, 1981). The AAPG method was designed for the United States, and cannot be used in many foreign countries where detailed statistics about new-field wildcat wells are not available. My foreign DI includes primary and secondary reserves added by pool extensions, whereas the AAPG method does not. No reliable foreign statistics before 1945 exist.

The old DI method probably had not been used more recently by others because the foreign data were only ~80% complete or accurate; the other 20%, which were biased by political, economic, and technological factors, required critical analysis and reconciliation by one objective person to produce consistent and usable statistics. The final results on my semilogarithmic DI graphs are not significantly affected by errors in the foreign DI computation procedure. The DI results are substantially repeatable by others. Consequently, the simple DI method will most likely provide realistic oil discovery rate projections for countries and regions during the next 0-15 years, with an estimated accuracy of about ± 20%. This is better than for most other forecasting methods used in short-and ong-term planning purposes (Grenon, 1982).

DI Methodology

The oil potential of different nations and regions can be compared from their DIs. Discovery indexes for foreign countries are more difficult to calculate than for the United States, because information on past drilling activities and reserve estimates are incomplete and scattered. Nevertheless, enough data exist in commonly available publications (which can be tabulated by year and country) to provide the basis for calculation of subquantitative (logarithmic) DIs.

Statistics on annual production and drilling are actual facts, whereas those for reserves are only estimates that must be critically audited. My reserve audit procedure is similar to balancing old checking accounts; that is, one starts at the last balance and works backward to find and correct obvious discrepancies. The latest (1981) reserves were presumed to be reliable (because they have had the most time to self-correct), and they provide an up-to-date base for future DI studies.

There is always a delay between the discovery year and the date that a new oil field's reserves are finally declared (booked). This delay, however, is not critical when comparing long-term trends between countries as long as the same computation method is used for all nations in a DI study. Several significant economic factors, such as financing, infrastructure, and delineation drillings, are incorporated when delayed reserves are booked that are not

End_Page 163------------------------

Fig. 3. A nonlogarithmic plot of oil discovery indexes (DI) versus years for the conterminous United States, showing significant political, economic, and technical developments. (From Ivanhoe, 1983.)

End_Page 164------------------------

Fig. 4. A semilogarithmic plot of regional composite oil discovery indexes (DI) versus years for North America, western Europe, Latin America, and Perisan Gulf. The empirical average DI decline rate is - 7% per year. (From Ivanhoe, 1983.)

End_Page 165------------------------

Fig. 5. Semilogarithmic plot of composite oil discovery indexes (DI) versus years for Africa showing three DI decline periods. (From Ivanhoe, 1984b.)

End_Page 166------------------------

Fig. 6. Semilogarithmic plot of composite oil discovery indexes (DI) versus years for the Far East showing three distinct DI decline periods. Several nations had Soviet advisors during the period 1945-1981, consequently their DI statistics may be too optimistic. (From Ivanhoe, 1984b.)

End_Page 167------------------------

available the year fields are first discovered. Such delays provide a much better idea of the EUR than does a single discovery well (Ivanhoe, 1984d).

It is an axiom that recoverable reserves plus total production can never decrease with time, so the true increase in reserves can never be a negative number in any year. Thus, proceeding back in time in 1-year increments from the 1981 tabulations, the reported reserves are adjusted each year as required to balance the accounting. Major corrections are made at erratic plus or minus DI "spike" years, which are easy to identify on a long-term tabulation. A common problem occurs when fields finally go into production: at these times the more conservative petroleum engineers take over for the optimistic geologists and politicians and begin to report "recoverable" reserves rather than "oil-in-place." The application of a realistic 10-40% oil recovery factor abruptly reduces the reserves and the DIs for that year. If the DI is ever negative for a given year, it is invalid and must be corrected by adjusting the reported reserves at the last major positive DI spike. Table 1 gives an example of the oil reserve adjustments for one country (Greece). Fallow years without reserve additions between discoveries can be handled for a few years by averaging the new reserves per total exploratory footage over all of the barren period. Such averaging can also be used for any rare cases in which reserves have been backdated to the year of discovery or in which drilling statistics are estimated. This smoothing can be recognized on the DI graphs as a horizontal line for several years.

The final tabulation and computation procedure can be outlined by these steps:

Table

Foreign gas without an established market is not the commercial equivalent of easily transported oil and belongs in the economically inaccessible "inactive reserve" (resource) category (e.g., Australia northwest shelf, offshore Trinidad, and Arctic Alaska). It costs from three to four times more to transport 1 Btu of natural gas than it does 1 Btu of crude oil in the conterminous United States (Megill, 1984). Lack of long-term trust from one government to the next is a critical obstacle for the international financing and use of natural gas. Consequently, gas is not included in the DI computations. The often-used factor of 6 million cu ft per barrel of oil equivalent (BOE) is the calorific, not the economic (usability) conversion of their values. Natural gas liquids (NGL) that are gas by-products should be assigned to the "inactive" gas reserves (resources) until they actually come on production. Gas hydrates in deep ocean basins are not even in the inactive reserve or resource categories.

A higher DI occurs for any year in which local exploratory drilling was under-reported. This is common for those nations (e.g., Cuba and India) that had Russian advisors who are reluctant to account for all of their exploration efforts. The Soviets reluctantly admit to drilling dry holes, thus unsuccessful wildcat become "scientific/stratigraphic tests" and accordingly are not listed (if any wells are reported at all). Consequently, DI rates are always suspect and may be overly optimistic for any country during its guidance by Communist advisors. No communist country was included in the DI study because of the incomplete drilling statistics.

Exploratory footage missing in any AAPG table can be estimated for many nations from the text of the annual

Table 1. Example of oil reserve adjustments for discovery index computations for Greece.

End_Page 168------------------------

AAPG reports. In those rare cases where exploration drilling footage is completely omitted, other references (such as Oil and Gas Journal) or estimates can be used to provide realistic numbers. Cumulative drilling footage is often used (rather than years) in studies of North American discovery trends, but such refinements were not justified for this post-1945 foreign project.

Alaska and the 48 conterminous states of the United States were treated as separate regions in this DI study. Failure to separate Alaska from the conterminous United States skews the exploration statistics for both. Alaskan operating problems, lack of infrastructure, government-owned mineral lands, and other factors are much more typical of major oil companies' foreign operations in the LDCs than of the small independents in the rest of the United States. Furthermore, Alaskan oil is overseas oil; it is shipped to the conterminous United States by tankers from Valdez, which are subject to interruption by terrorists (Ivanhoe, 1983, 1984d).

A significant trend cannot be established in 2 years. Isolated drops in exploration drilling and DIs are often due to temporary national political and/or economic factors (e.g., Iraq/Iran war) rather than a major change in the geologic merit of a country. High and low DI spike years on DI graphs are not particularly meaningful. (An analogy is the daily weather versus the annual climate.) Erratic DI spikes are often the result of irregular time lags between the dates of new discoveries and the time when their reserves are finally reported. If enough countries are combined, many DI spikes are eliminated, but one must adjust all numbers for the individual nations before they are added together.

Each nation must be evaluated separately, not only as a part of a continent or region. Politics and laws, rather than geology, determine where and when oil companies may work (e.g., People's Republic of China). Two countries sharing common oil basins may have very different exploration politics, problems, and success (e.g., northeastern Mexico versus southern Texas).

Graphs of Annual Exploratory Drilling and New Oil Reserves

Figures 7 and 8 summarize the basic information collected and analyzed during the DI studies. These graphs summarize the actual (adjusted) quantities involved, rather than the DI rates that were derived from them (Ivanhoe, 1984c,d). Figure 7 shows the total exploratory drilling in millions of feet, and Figure 8 gives the total (adjusted) new recoverable oil reserves booked per year from 1937 to 1982. Dates of relevant political and technological developments are indicated.

Both of the figures show curves of three separate areas: (1) the conterminous United States (excluding Alaska); (2) the remaining noncommunist ("free world") countries (including Alaska and the Persian Gulf nations); and (3) the Persian Gulf nations.

This three-way breakdown helps put the noncommunist countries' oil supply into proper perspective. There are two abnormal regions that tend to skew all of these statistics: (1) the conterminous United States with its large amount of drilling and low discovery rates, and (2) the Persian Gulf with its small amount of drilling and huge oil reserves. Statistical details from either of these unusual regions may have little direct bearing on the rest of the world.

Exploratory Drilling

It is apparent by visual inspection of the three curves on Figure 7 that the conterminous United States is still the oil industry's favorite country for exploration drilling. This is because the farmers usually own the onshore mineral rights, thus small oil deals are feasible for independent entrepreneurs. In the period from 1973 to 1982, the exploratory drilling footage in the conterminous United States ranged from 2.1 to 3.0 (average, 2.5) times that of all of the rest of the noncommunist world combined, and 85 to 155 (average, 120) times that of the Persian Gulf nations. Before the 1973 Arab oil embargo, the exploration spread was even greater, reaching a maximum of 7 times at the first U.S. drilling peak in 1956, which did not occur in the rest of the noncommunist world. After the 1973 and 1979 oil price shocks, the U.S. exploratory drilling doubled to a peak in 1982, with a corresponding but less pronounced increase elsewhere. As might be expected, the quality of the oil prospects and finding rates decreased during these drilling surges.

About one-third of the rest of the noncommunist world's exploration drilling effort is in Canada. A change in Canada's federal energy policies reduced exploration tax subsidies after 1980, which resulted in a corresponding decline in drilling.

New Recoverable Oil Reserves

The DeGolyer and MacNaughton (D&M) 1945-1982 engineering reports do not show which (new or old) oil fields are responsible for any reserves, nor what time lag may be involved in the bookings (see Figure 8). This is a basic difference between the D&M reserves, which include annual changes resulting from improved recoveries, and the AAPG procedures, which qualitatively attempt to back-date, after 3-and 6-year intervals, any reserve changes for specific fields per AAPG field size categories. In the long run, the final amount of oil discovered must be the same regardless of whether the D&M or the AAPG procedure is used. The quantitative D&M engineering reserves, however, are more useful for economic planning purposes than the AAPG qualitative field sizes. No attempt was ma e in this study to correlate any D&M reserves with specific fields.

DISCOVERY INDEX: RESULTS AND DISCUSSION

General

Two distinct foreign categories exist: countries with commercial oil discoveries and those without. In 1981, there were 150 noncommunist countries and 18 communist countries. This study shows that 74 countries produced commercial oil, including 31 exporters, 13 of which are in OPEC. For the 94 less prospective nations where commercial oil has not yet been found, the DI rate is zero.

End_Page 169------------------------

Fig. 7. Exploratory drilling of noncommunist world from 1937 to 1983. The three curves distinguish the conterminous United States, the remaining noncommunist countries including Alaska, and the Persian Gulf. About one-third of the noncommunist world's drilling (excluding the United States) is in Canada. (From American Association of Petroleum Geologists, 1945-1983.)

End_Page 170------------------------

Fig. 8. New recoverable oil reserves booked per year between 1937 and 1982. The three curves distinguish the conterminous United States, the remaining noncommunist countries including Alaska, and the Persian Gulf. (From DeGolyer & MacNaughton, 1945-1982.)

End_Page 171------------------------

The potential reserves of undiscovered fields determine how attractive any country's remaining oil prospects may be, as well as the justified exploration investments, risks, and rewards. On a world scale, any nation's prospects decrease logarithmically in the following manner:

                    Oil discovery index (DI) rate
World grade   (bbl/exploration ft)

Stupendous          >10,000
Excellent           1,000-10,000
Good                100-1,000
Fair                10-100
Poor                1-10

The computed geologic potential of any country represents the "average" grade to date including all past dry holes, and it may not necessarily indicate the quality of any remaining individual drilling prospects. The average remaining prospects in the conterminous United States (1982 DI = 19), however, are approaching the "poor" (DI = 10) category. The oil prospects here are being steadily depleted (Figures 3 and 9) (Ivanhoe, 1983, 1984a, c,e).

Digital seismic and common depth point (CDP) technologies were developed in the United States during the late 1950s and by 1965 were commercially available everywhere. This was the most important exploration breakthrough since the introduction of onshore analog reflection seismic surveys in the early 1930s, because digital seismic surveys made oil prospecting effective at sea and through permafrost. Digital seismic surveys have now been run across every politically accessible continental shelf in the world. Offshore China is the most recent area to be surveyed, thus the U.S.S.R. is the only remaining "nondigital seismic" nation. In many countries (e.g. United Kingdom, Mexico, etc.), there was a pronounced upward jump (digital shift) in the DIs when digital seismic technology opened up virgin offshore basins in the late 1960s and the new reserves were reported soon after 1970 (Figure 4). The digital shift is often very distinct if it is offshore, as in the United Kingdom, but it is sometimes less spectacular for countries, such as the United States and Libya, where digital seismic surveys were first introduced to onshore areas that already had considerable oil production. In the latter cases, the digital shift may appear as a flat or climbing plateau rather than as two offset, steeply declining DI trends (Figure 9). A possible exception to this two-phase digital shift is in the Persian Gulf region where the decline has been fairly steady since 1956, partly because many offshore reserve increases there resulted from extensions of onshore fields, in other words, by devel pment rather than exploration drilling (Figure 4).

Three recognizable "DI periods" can be identified on most national and regional DI graphs: (1) onshore DI decline; pre-1960s; (2) onshore/offshore DI increase due to digital seismic surveys; 1960-70; and (3) offshore DI decline; post-1970. (Figs. 5, 6, and 9). The third (postdigital offshore) DI decline period is almost 15 years old and is well advanced in all regions of the noncommunist world. Oil discovery rates are definitely decreasing everywhere (Ivanhoe, 1983, 1984a,b,c,d).

The DI rates in noncommunist countries peaked in 1969. All 150 of these nations, including the 90 non-oil-producers are now politically accessible and have had at least reconnaissance digital seismic surveys run across their continental shelves. We can now drill to any prospective depth on land or in any ice-free ocean. The best structures were drilled and tested during the last 15 years. No new major petroleum provinces are expected unless the deep ocean areas away from the slopes produce sizable commercial oil reserves, which would happen far in the future, if ever. Effective exploration in arctic Alaska with its permafrost was also delayed until digital seismic technology arrived. The National Petroleum Reserve-Alaska (NPR-A) has been surveyed and drilled by the federal government ith poor results. Some parts of the government-owned mineral lands of the conterminous United States and Alaska, however, are still not politically accessible and thus are untested.

No major technological breakthroughs are in sight, although all technologies are being constantly improved to enable us to find smaller and smaller traps and to recover more oil by enhanced oil recoveries (EOR) from those fields that are already known. Governmental geologic research jobs often increase as field sizes go down. National political and military considerations will prolong the search for local oil in many nations long after economic considerations indicate that it should be stopped. Normal technical improvements are already included in projections of the DI rates. What we need are new virgin areas to explore that might contain giant oil fields. Unfortunately none are known. The only completely untested regions of the globe are now those in deep ocean basins and under pack ce. Deep-water, open-ocean completions far from shelf areas may never compete economically with alternate fuel sources (Rozendal, this volume). The stormy, icepack-covered, deep oceans of the Arctic or Antarctica are no more likely prospects than the deep-water basins right next to the U.S. markets. (St. John, 1980; Ivanhoe, 1980e, 1981a,b).

Discovery Index: Decline Rate

A significant trend that came out of this study is the average DI decline rate of 7% per year (Figs. 4, 5, 6, and 9). This decline rate is equivalent to a drop of one order of magnitude per 30 years. It can be used to help project DI trends into the future as oil provinces are depleted. This 7% per year DI decline rate was derived empirically from the DI graphs of the six major oil-producing regions (Figs. 4, 5, and 6), which include North America (1940-1955 and 1967-1977), Western Europe (1958-1969), Latin America (1948-1970), Persian Gulf (1950-1979), Africa (1967-1981), and the Far East (1970-1981). This is a conservative DI decline rate, and a case can be made for a more rapid DI decline for several important regions including the Persian Gulf. After 30 years at a rate of 7% per y ar, the amount of new oil that will be discovered per unit of drilling effort in the noncommunist world may be only 10% of the current rate. The economic and political consequences of such an oil discovery decline are very alarming for the world's people (Ivanhoe, 1983, 1984a,b,c,d).

All effective new petroleum exploration or production techniques have been quickly introduced everywhere in the noncommunist world since World War II. Consequently, all

End_Page 172------------------------

Fig. 9. Semilogarithmic plot of oil discovery indexes (DI) and projected discoveries for the conterminous United States and the remaining noncommunist countries. Composite DI numbers versus years for 1945-1981. Projected oil discoveries for 1982-2000 are given at 1980 exploratory drilling rates: 19 billion bbl for the conterminous United States and 291 billion bbl of recoverable oil for the remaining noncommunist world. Empirical high (and low) dashed lines bracket the trends. Graph for noncommunist countries shows three periods: (1) is erratically high due to several new supergiant oil fields in the Persian Gulf region; (2) includes new onshore and offshore basins in Africa, Nigeria, Libya, western Europe (North Sea), and Saudi Arabia; and (3) includes positive spike from Mexican offshore fields and decreases resulting from the Iraq/Iran war. (From Ivanhoe, 1984a).

End_Page 173------------------------

nations'oil discovery rates show similar technology related surges with little time lag, and their DI curves are roughly subparallel, differing only in the richness of oil in that nation's basins.

Discovery Index Costs and Investment Risks

The DIs and graphs can be used to roughly compare annual exploration costs per effort between countries. To do so, drilling expenses must be comparable and the total cost must include not only the contracted drilling price of the wells, but also all direct and indirect exploration expenses such as overhead, land, legal, bonuses, accounting, seismic surveys, geology, engineering, roads, and well logging. The cost per barrel found is infinite before the first oil discovery (Ivanhoe, 1984d).

The average 1958-1980 spread between the subparallel conterminous United States and noncommunist world DI curves indicates that the oil industry has consistently discounted gross oversea oil prospects versus U.S. long-term profit potentials by ~ 30 times (Figure 9).

WORLD PRODUCTION TRENDS

Unfortunately, the world's petroleum prospects (excluding communist countries) are much worse than they were 10 years ago, but the good news is many people have made a major conservation effort and have accordingly postponed the time of the next major energy crisis (Figure 10). The question is not whether but rather when the next crunch will occur (Hubbert, 1979; Ivanhoe, 1984e). Major new reserves to replace the globe's oil production are simply not being found. Only three giant (> 500 million bbl) oil fields were discovered worldwide from 1979 to 1983. In 30 years (at -7% per year) the oil discovery rate will be only 10% of the present total. The current world "oil glut" will be as fleeting as a desert community's "water glut" caused by one winter storm. (Some glut The total U.S. 1983 expenditure for imported oil was about $55 billion or 90% of our international trade deficit of $61 billion.) The world's production oversupply is a short-term imbalance caused by a sudden drop in total oil consumption resulting from long overdue conservation in the United States and elsewhere. (The U.S. usage of crude oil, natural gas liquids, and their products decreased from 19.0 to 15.2 million bbl per day between 1979 and 1983, which is analogous to dieting from 190 to 152 lbs in 4 years. A European-type tax on gasoline would reduce our usage even more.) Additional conservation will be increasingly difficult to accomplish, thus long-term consumption is expected to eventually stabilize and to balance production. At this point the world price of crude oil will quickly rise again- to the public's surprise and dismay. (Figures 9 and 10).

The continual addition of new petroleum provinces is critical to maintain the world's long-range oil supplies. The world total will drop as soon as no significant new basins (or nations) can be brought into production. The question now is where the international oil industry can turn to find new virgin provinces. The non-U.S. regions seem to be our best hope. The Persian Gulf is in a class by itself--no other areas in the world comes close (Figures 4, 7, and 8).

The communist nations were not included in this study because of a lack of statistical information. We know, however, that the small socialist nations have prospected for and produce some oil for many years and that they are oil importers today. Only giant Soviet Union and China are net petroleum exporters and will undoubtedly discover more fields in their extensive virgin or semivirgin onshore and offshore basins. The introduction of long-delayed modern digital seismic oil exploration in newly politically accessible communist nations such as China should help keep up the world's discoveries and production. Let us hope that these nations will add substantially to the free world supplies (Ivanhoe, 1984e).

World's Reserve to Production Ratios

At the world's present semimature stage of exploration, the current reserve to: production (R/P) ratios indicate where future oil discoveries are most likely. The 1982 R/P ratios for the globe's four major petroleum, economic, and political regions (Ivanhoe, 1984e) are as shown below

Region                   R/P

Conterminous United States      8 years
Middle East                    75
Remaining noncommunist world   27
Communist nations              20

Total world                    34 years

DISCOVERY INDEX PROJECTIONS OF OIL TO BE DISCOVERED

All projections are limited by optimistic or pessimistic assumptions, which can be best verified on graphs. The DI graphs summarize the noncommunist world's historical discovery rate since 1945. Recognizable discovery rate trends place limits on the probable order of magnitude of oil that may be discovered until the end of the next decade (1982-2000 inclusive).

The information on the regional graphs shown in Figures 3, 4, 5, and 6 was recombined into two summary curves in Figure 9, one showing the conterminous United States and the other, the remaining noncommunist countires. The estimates for oil discoveries through the year 2000 are based on extrapolations of the post-1970 DI curves, which have both been declining at roughly the normal rate of 7% per year. Exploratory drilling for future years was assumed to be about that of the near-record year of 1980, which was 70 million ft in the conterminous United States and 30 million ft for the rest of the noncommunist world. More or less drilling should find proportionally more or less oil. The annual DI, when multiplied by the estimated exploratory footage, gives the amount of new oil that may b found each year. The sum of the projected annual discoveries provides the gross amount of oil that can be expected through the year 2000. High (and low) projections, to bracket the top (and bottom) of the "band" of the discovery trend, are both indicated in Figure 9.

End_Page 174------------------------

The total low and high projections of new oil reserves to be discovered or developed during the period from 1982 to 2000 are 12 and 19 billion bbl for the conterminous United States and 162 and 291 billion bbl for the remaining noncommunist countries. In summary, it appears that conventional crude oil discoveries in the rest of the noncommunist world may keep pace with production through the year 2000, but that U.S. discoveries (excluding Alaska) definitely will not. In 1982, the United States consumed 5.7 billion bbl of oil, natural gas liquids, and

Fig. 10. Total U.S. oil supply. Total U.S. production and imports for 1940-1983 including Alaska. Total U.S. production includes conterminous U.S. and Alaskan crude oil and natural gas liquids; imports include crude oil and its products. The cost of 1983 imports is calculated as 5 million bbl per day times 365 days at $30 per bbl equals $55 billion. This was 90% of the 1983 U.S. trade deficit of %61 billion. The major decrease in total oil consumption and imports between 1980 and 1983 was due to conservation. Note Mexico's minor share of U.S. imports. (From Ivanhoe, 1980d.)

End_Page 175------------------------

refined products (3.8 billion bbl produced here and 1.9 billion bbl imported). We will consume over 100 billion bbl of oil during the next 18 years at the 1982 rate. But my high projection of U.S. oil discoveries is for only 19 billion bbl of new oil (slightly over 1 billion bbl per year) during this period. By 1999, the conventional crude oil finding rate of the lower conterminous United States may be only one-third of today's rate, or about 0.5 billion bbl per year.

CONCLUSIONS

My final oil projections are regrettably lower than most estimates made by others. This discrepancy is mainly due to the lack of time limits on most oil resource studies. My high U.S. estimate through the year 2000, however, is close to the 90% cumulative probability (i.e., low ultimate production) number of Nehring's (1981) U.S. report (my estimate is 19 billion bbl versus Nehring's 21 billion bbl). Comparing the rest of the noncommunist world's estimate with Nehring's (1982) world paper gives similar magnitudes (my estimate is 291 billion bbl versus Nehring's 475 billion bbl). My studies project that our entire globe's ultimate conventional crude oil recovery may be close to 1,700 billion bbl, which falls within the range of 1,600-2,000 billion bbl predicted by Nehring (Nehring, 198 ; Ivanhoe, 1984e). My figure is calculated from the following (all values in billion bbl):

Table

I reluctantly agree with Hubbert's prediction that the globe's oil age will effectively and during the lifetimes of many alive today. By 1999 we will have spent most of our national petroleum heritage by public consent, and the end of our oil age will be in sight. We will thereafter become one of the oil-poor nations. The social and political consequences of this foreseeable decline in our country's economic base will be staggering for the entire world, particularly when combined with the concurrent population explosion in the LDCs (Fox, 1984, Ivanhoe, 1984e).