In all human history, no substance has so changed economies, social structures, and lifestyles so rapidly, so profoundly, and affected so many people as has oil. Oil brought personal motor transport, intercontinental air travel creating worldwide economic and cultural interchanges, revolutionized agriculture and manufacturing, and lifted much work from the backs of many people.

Oil converted muddy trails into millions of miles of paved roads and continues to maintain them. Oil powers vehicles to transport goods cheaply across great distances. Oil is high-density energy in a most convenient form, which can be taken to remote areas for use and can be stored easily for long periods of time. Oil's versatility in end use has no equal. Oil also is a cause of war.
 

Very importantly, oil is also raw material for myriad products including medicines, paints, and plastics.

Oil and its close companion, natural gas, are the bases for thousands of other petrochemical products, especially chemicals to promote crop growth and to defend crops against insects and diseases. Bartlett (1986) correctly states that modern agriculture is simply a way of converting petroleum into food.
 

The coming of oil has changed the world, presumably much for the better. However, oil is finite, and its inevitable eventual decline and departure will be a signal event in human history. Anticipating the time when this trend begins is the subject of this paper. It seems to be sooner than most people expect. The critical related matter of what alternatives may exist to replace oil also is briefly examined.

Oil has formed in the upper approximately 16,000 ft of the Earth's crust since at least as far back as the Cambrian Period, some 550 million years ago (MYA). It is a rich inheritance of highly concentrated solar derived energy captured by myriad organisms, chiefly algae, and then distilled by geological processes into an energy form that is unequalled by any other energy source in its versatility and convenience in handling. Now, within one human lifetime, one-half of this unique 550 MYA inheritance will have been spent. The remainder will go very fast.

We are not in the Age of oil —the term "Age" is for things of longer duration. We are living in a brief oil interval. The average citizen' pulling into a service station and saying "Fill 'er up" gives little or no thought to where the oil came from or how long that stream of gasoline will continue to be available to flow into the car's tank. There were two brief times in the 1970s when the American public was made acutely aware of what the lack of oil might portend. However, those long gas lines are now a faded memory and, for the youngest third of the U.S. population, there is no such memory at all.

If the public does think briefly about future oil supplies, the question usually asked is, "How long will oil last?" This is the wrong question. Oil will be extracted in some insignificant quantity perhaps 200 years from now. The critical question is: "When does the peak of world oil production occur?"

The importance of the "peak of world oil production" generally is not grasped. However, it is the peak time which is critical. Hubbert (1967) emphasized its significance stating:

In other words, the peak.

The United States already has felt briefly the importance of peak. Precisely, as forecast by Hubbert in 1956 (either ignored, or regarded in gross error by most people at the time), U. S. oil production peaked in 1970. The United States actually had to begin to import oil about 15 years before the peak was reached, as demand already had outstripped production capacity by peak production time. The fact that U. S. production had peaked in 1970, and then began to decline further assured the success of the Arab oil embargo against the United States in 1973, and altered U. S.-Middle East foreign policy.

The United States well past its peak of oil production, now imports more oil than it produces. When world oil production peaks, there will be nowhere else for the world to go for more oil. The problem then will become the harsh reality of distribution of an irreversibly declining resource, rather than dividing more and more oil, as has been the pleasant experience to the present. This is when final competition begins for the last half of world oil reserves. It will be a global struggle. All countries will be involved —the industrialized countries more so than the less developed countries. For the first time, the entire world will be locked in one massive contest for a single resource.

This may be the most important event in human history. In terms of lifestyles, our relatively cheap and abundant food supplies, the manufacture of many things, which depend on the energy of oil, and the distribution of these products, the beginning of the decline of oil production will be a momentous event.

We have forecast the peak of oil production and related data in each of 42 countries, accounting for 98% of 1996 world oil production (Table 1).

From production data, together with a judgmental amalgam of the possible production variables in the 42 countries, we have produced a composite world oil production profile from 1960 to 2040 (Fig. 1). The figure also shows the profile and production peaks .of OPEC and non-OPEC nations, and the significant final cross-over point when OPEC nations (by that time chiefly those of the Persian Gulf region) will be producing one-half and, from then on, more of the world's remaining oil.

This graph forecasts the peak of world oil production will occur about the year 2007 —only 7 years away! After this time, with the world production peak at 30.6 Gb a year (the International System term "Gb" is equivalent to the use of the term billion barrels in the United States), world production drops rapidly to 24.6 Gb a year in 2020 and to 11.5 Gb per year in 2040 —a total decline of 62% in just 33 years. Note also the important date when OPEC production becomes the dominant world oil source. This is also at the time of world oil production peak, 2007. This may have significant implications, because at that time the world's remaining oil reserves and production will be mostly under control of the Persian Gulf nations.

Present world oil production is about 25 Gb a year. The 2040 production will be less than one-half the present oil consumption, and will face a demand from a world population which is estimated to be 50 to 70% greater than at present. Compounding world energy demands will be the increasingly industrialized nations (particularly Southeast Asia, China, and India) wanting more energy per capita. China, Southeast Asia, and India, now with some 60% of the world population, are getting motorized wheels. If China used oil on a per capita basis as does the United States, China alone would use more than ten million barrels a day beyond the present entire world oil production.

Although some oil will be produced well beyond 2040, we have not plotted the "tail" of the production curves of individual countries or of the world beyond the year 2040. Production then will be insignificant compared with potential world demand. The happy oil interval will have come to an end for all practical purposes. This is why the question "How long will oil be produced?" is of little consequence.

Graphic illustration of the oil production history since 1960, and projected future production of seven major oil producing areas of the world are shown, respectively, in Figures 2-8. These give a view of the relative abilities of each of these regions to provide for its own oil. Notably, the region with the greatest population, Asia-Pacific (Fig. 8), is the least capable region for providing its own oil.

The basic information shown graphically by Figures 2-8 is tabulated in Table 2.

In earlier years a number of forecasts have been made about the peak time of world oil production. With few data points and many of the potential world oil basins only poorly explored at that time, most of these estimates already have proved to be wrong. Thus, by implication, it may be assumed that current projections also will (hopefully) be wrong. However, the world's potential oil provinces now are well known (Campbell, 1997b). Recently, based on more and better data, further forecasts have been made and, although each estimate is slightly different than the others, most now cluster remarkably, one might say alarmingly, close together on that critical time. These include forecasts by Campbell (1991,1997b), Campbell and Laherrère (1998), Duncan (1997), Fleay (1995), Hatfield (1997), Ivanhoe (1995, 1996, 1997), and MacKenzie (1996). All estimate the peak by 2013, or earlier except for the most optimistic forecast by Edwards (1997) of 2020. There are, however, some divergent views on the time of world oil production peak as noted by Brown (1998), but those cited here are the most recent and, we believe, the most nearly correct.

Figure 2. North America
 

How our figures and estimates are derived is briefly described here.

The oil production profile (all data include natural gas liquids) from 1960 to 2040 for each country in this study is modeled separately by use of a special program (the World Oil Forecasting Program, outlined later). These 42 production profiles have been melded into a single projected world oil production profile (curve 3, Fig. I).

We owe much to the great pioneer theoretical work by Hubbert that defined the oil production life cycle in terms of discovery, production, and reserves, and which he used so successfully in predicting the peak of United States oil production. However, the oil forecasting program and procedure used in this study departs from and goes beyond his forecasting method to try to anticipate and encompass numerous variables related to future oil production, particularly technological advances in exploration and production, not known in Hubbert's time.

The software ("tool") used for the conclusions expressed in this paper, we have termed the "World Oil Forecasting Program" which consists of two distinct, stand-alone models for each nation.

The first model ("N model") is quantitative, using production data and mathematics on a translated coordinate system to produce an intermediate "helper" forecast for each nation. This, the so-called "guide" forecast ("G forecast"), is a purely mechanical prediction of future production. In some examples, the G forecast can provide useful information about the shape of future oil production by providing a lower boundary on the estimated ultimate recovery (EUR) and the probable shape of the future production curve. However, in other circumstances, it is not useful, as in the situation of the OPEC production quota-limited countries. The N model produces the G forecast, the best forecast we are able to make based solely on historic production data, and mathematics. Data are from British Petroleum (1968-1997) and Campbell (1991). Details are in Duncan (1996).

The guide forecast is just one of many items of information that may be used in the second model portion of the World Oil Forecasting Program.

By definition, "heuristic" denotes a method of solving a problem for which no algorithm exists. It involves trial and error, as in iteration. In this discussion heuristic knowledge indicates "soft," "qualitative," or "judgmental" knowledge. Although judgmental knowledge is lacking in the Numeric model, it is crucial for oil forecasting in the heuristic model ("H model"). The H model provides the user with a powerful interface for oil forecasting, chief of which is a three-curve graph for each nation with years 1960 to 2040 on the x axis, and production on the y axis (Fig. 1. Curve 1 shows the historic data from 1960 through 1996 —a crucial reference for forecasting. Curve 2 shows the guide forecast (previously discussed) and is useful as the lower bound curve. Curves 1 and 2 are important forecasting aids, but they are only the beginning.

Curve 3 also displays the historic data from 1960 through 1996, but this time the data serve as a base for a new and better forecast 1997 through 2040. A so-called graphical input device (GID) makes it easy to enter and run different trial forecasts. After each trial run, a different estimated ultimate recovery (EUR) value is displayed so, after making several runs, the user can select an upper-bound curve for each nation. Thus, now confined by lower and upper curves, and further modified by judgmental input, the user extends the most recent production trend seamlessly into the curve extending through the year 2040, providing what we termed the "judgmental" forecast (J forecast) of future oil production, one nation at a time. Details of the heuristic model are in Duncan (1997).

In our 42-nation study, we also have grouped the nations into seven regions (Figs. 2-8 and Table 2), and made a world summary (Fig. 1 and Table 1), which are the output of the heuristic model.
 

Figure 3. South and Central America

Figure 6. Middle East

Figure 7. Africa

Figure 8. Asia Pacific

This step is qualitative and involves judgmental factors. We recognize that predicting either a nation's or the world's oil production peak is not an exact science. Hennessy (1997) states:

Although qualitative factors are generally more prone to measurement error than quantitative variables, we should not exclude them on that basis alone. How well a model recreates the system's performance —and thereby the model's usefulness— depends on much more than measurement precision.

In our model, we have incorporated, on a judgmental basis, the important variables that we visualize might affect the conclusions.

In general, we have not used reserve estimates in our forecast as we believe these to be less reliable than are the actual production figures. Reserve estimates are subject to political and economic factors. Basing production quotas on reserve estimates, as is done by the OPEC nations, may cause reserves to be over stated. Using oil reserves as collateral for loans as was done by Mexico, also may lead to inflated figures. Government-owned oil companies like to show a gain in reserves each year, and this factor may distort the reserve picture. Laherrère (1995), Campbell (1997a, 1997b), Campbell and Laherrère (1998), Riva (1995), and others have pointed out the risks in using stated reserves by countries, designating some as "political reserves."

Although production figures are subject to some error or deliberate misstatement, we believe that production can be verified more easily than can reserves unseen oil in the ground. In addition, production generally represents the true ability of oil fields to produce, except in instances where production may from time to time be restrained artificially (e.g., OPEC quotas). For these reasons the first part of our world oil forecasting is based chiefly on production history, which also was the basis of Hubbert's successful 1956 forecast of the U. S. 1970 oil peak. However, in this second part of the model, we do employ some reserve figures, which, on a judgmental basis, may differ considerably from those reported by a given country.

Other judgmental elements include ultimate recovery volumes based on what is known about reservoir characteristics, and the effect on production of new technologies (e.g., three- and four-dimensional seismic, directional drilling, COZ and air injection, deeper water, and subsalt drilling capabilities, etc.) (Anderson, 1998). To some extent, these are already factored into present production curves. However, we also have given these some additional weight in the judgmental part of our model, as we project the curves into the future.

The judgmental elements applied to the production curves are an important part of the world oil forecasting program. Purely on the basis of extrapolating production data into the future (the Guide curve), we arrived at a figure of 1303 Gb for the estimated ultimate recovery (EUR) of oil in the 42 nations. However, when the judgmental element is applied, we have a figure of 1996 Gb for the EUR, which we believe to be the more nearly correct. Other variables on which we have commented upon briefly in the following paragraphs may move the peak back. With the judgmental element, we have resolved these into the final world production curve.

There are oil deposits, which are termed unconventional, as they can be made to produce oil using special technologies beyond those employed in usual oil recovery by flowing or pumping, which also includes gas repressuring and water flooding. These resources include oil/tar sands and oil from wells by various enhancing methods, including use of detergents, hot-water injection (Cold Lake area, Alberta), and steam flooding. These resources will increasingly be brought into production as economic and environmental factors permit. We have factored these into our forecast on a judgmental basis, but they seem to change the curve derived from a projection of conventional crude oil production only a little. They simply give a modestly extended life to the low far end of the curve of world oil production.

Fouda (1998) describes the conversion of natural gas to liquid fuel, which now is being implemented chiefly in the Persian Gulf region. This will convert excess gas production to oil where facilities for the transport of natural gas are not conveniently available. However, bringing this process on stream cannot be done in significant quantities in time to affect oil production peak.

George (1998) describes the potential for oil production of the northern Alberta Athabasca oil sands and of the heavy oil deposits of eastern Alberta, which, again, similar to natural gas conversion to oil, only can be brought into increased production slowly. The time delay involved and the modest quantity will not affect significantly the time of world oil production peak. The same applies to the large heavy oil deposits of Venezuela (the "cinturon de la brea"). Conversion of coal to liquid fuel can and has been done, but the scale of coal mining and the related processing facilities needed to provide a significant world liquid fuel supply, are beyond early attainment. Shale oil from oil shale does not seem, after extensive and expensive major industry attempts in western Colorado in the 1980s, to be an economically viable source of oil in appreciable quantities (Youngquist, 1998).

Another consideration is the amount of displacement of oil by renewable energy supplies, which could limit demand and therefore production, thus delaying the time of peak production. These include solar, wind, hydro, nuclear, tidal, and geothermal energy sources. However, an evaluation of how soon and to what degree these energy sources could fill the gap created by the declining oil production indicate that such sources, neither individually nor in total, will significantly replace oil within the next 15 to 20 years and perhaps not even beyond that time, given the many ways and especially the volumes in which we use oil today (Youngquist, 1997). The forecast peak will arrive too soon for alternative energy supplies to be developed in time to affect the oil peak.

The problem of bringing on sufficient quantities of unconventional oil resources to appreciably move the peak of world oil production is illustrated by the calculation by Bartlett (1999) that each additional Gb of oil or oil equivalent added to present projected world supply, would move the peak of oil production just 5 1l2 days further away.

We also recognize that economics plays a part in oil production. However, there is a difference between the attractive oil economics by which oil displaced coal, leaving large quantities of coal in the ground, with the situation now with oil. Alternative energy sources possibly to displace oil do not have the great versatility of end use and the energy/profit ratio that oil had over coal. The favorable economics and the greater versatility in end use that oil has compared with coal, which caused the switch to oil, generally do not exist for alternative energy sources. In spite of eventual rising prices, we believe the much preferred and used energy source will remain oil as long as reasonable quantities exist.

Eventually, price will move oil into its higher end use values. Transportation may be one such higher value. The more than 600 million gasoline and diesel powered cars and trucks in use today cannot be replaced quickly or easily with other types of vehicles. Oil for simply burning in the low end use of space heating is likely to become prohibitive in cost. In spite of price, oil will remain in strong demand. For these reasons we believe that the curves, both regional and worldwide, have considerable validity, and economics will not materially change them.

There is a general citizen confidence that alternative energy sources exist or will be discovered to replace oil. "The scientists will think of something" is a widely expressed pleasant public placebo. However, there is no alternative energy source or combination thereof now known that can completely replace oil in all its many and varied uses, particularly with regard to the concentration of such a large amount of energy in such a convenient, easy to handle form for use in mobile machines (cars, trucks, tractors, airplanes, etc.). The energy source that can compare with 5 gallons of gasoline which may be carried conveniently for hundreds of miles to where it may be put in a car to propel it for 150 miles or more at 60 miles an hour is nowhere in sight. Nor has an alternative energy system been developed which can propel an airplane nonstop for 14 h at 600 miles/h (the present New York to Capetown flight).

The distinguished British scientist and statesman, Sir Crispin Tickell (1994), has expressed the same opinion regarding oil alternatives and the future stating, ". . .we have done remarkably little to reduce our dependence on a fuel [oil] which is a limited resource, and for which there is no comprehensive substitute in prospect."

We do not presume to have the precise date of the peak of world oil production. Campbell (1997a), in discussing oil reserve estimates, stated, "All numbers are wrong: that much we know. The question is: how wrong?" The same thought can be applied to the peak time of world oil production. Our estimate may be wrong, but we do not believe it is far wrong, and that the critical date of peak world oil production is in sight. Even allowing for the highly unlikely event of the discovery of IS more "super-giant" fields, our happy but brief oil "interval" we have had the good fortune to enjoy, a bright almost instant flash even in human terms, shortly will become history.

We use a special map, termed a "scatter diagram," to track our progress in establishing the peak. Each forecast pinpoints a milestone along the route. The first forecast put the peak in year 2005, with production at 28.50 Gb (Duncan, 1997). The second forecast, the subject of this paper, puts the peak at 2007, with that year's production at 30.64 Gb. Successive points are marked on the map and connected by lines, each line, on average, encircling closer to the peak. By repeating this effort annually we expect to forecast correctly the time of the peak within the next few years.

However, the exact date will not be known for certain until some time after that year arrives, when it can be viewed in retrospect with the downward trend well established. We expect that the peak at the time will not be recognized easily and rather than a sharp peak, there will be a gradual leveling out of production, and the peak will be obscured in a fairly flat, plateau-like curve. Our figures portray this shape of the peak in predicting the annual Gb of oil production for the year 2001.at 29.75; 2003, 30.41; 2005, 30.59; 2007, 30.64; 2009, 30.40; 2011, 29.52; and 2013, 28.32. The peak may be more evident by a sharp rise in price as demand begins to exceed supply at prior price levels.

An outgrowth of our study has been an estimate of the amount of world oil which ultimately will be recovered. The most recent discussion of the future of world oil production is an outstanding article by Campbell and Laherrère (1998) with which we are in substantial agreement. These authors estimate the probable ultimate world oil recovery at about 1.80 trillion barrels. By using a somewhat different approach, our estimate is 2.04 trillion barrels. However, Campbell and Laherrère do not include natural gas liquids in their estimate. We do. If one assumes 8% additional oil from natural gas liquids, then the Campbell/Laherrère figure is about 1.94 trillion barrels compared with our figure of 2.04 trillion —in remarkably close agreement. Both studies show, as Ivanhoe (1995) also has noted, that there is a now-visible limit to world oil supply and we are approaching the halfway point of its consumption.

The important point of these on-going studies is that by all reasonable estimates, the peak of world oil production will be reached within the lifetimes of most people living today. We should be concerned now.

It is difficult to get the attention of either policy makers or the public, in general, concerning something, which will happen 10 or 15 years hence. However, here are simple facts, which should surely gain attention. The world now consumes about 25 Gb of oil a year. However we are finding, the world around, less than 5.5 Gb annually (Campbell, 1997c). Campbell and Laherrere (1998) state, "In the 1990s oil companies have discovered an average of seven Gb a year; last year they drained more than three times as much."

These authors also note that the world oil discovery rate peaked in the early 1960s and they add, "About 80 percent of the oil produced today flows from fields that were found before 1973, and the great majority of them are declining."

Any company which is selling its product faster it is being replaced is going out of business. The world is going out of the oil business. Political leaders, business planners, and social policy makers take note. Heed the words of Aldous Huxley: "Facts do not cease to exist because they are ignored."

The World Oil Forecasting Program is available at Website: < www.halcyon.com/duncanrc/ >. All programs are free (both Macintosh and Windows versions), including the latest run-time version of Stella 5.0. Users can study our forecasts, or make their own forecasts for 42 nations, seven regions, and the world. All historical data and the guide forecast are automatic. No programming is needed.