THE WORLD PETROLEUM LIFE-CYCLE:
ENCIRCLING THE PRODUCTION PEAK, by Richard Duncan,
Institute on Energy and Man, Seattle, WA, 1997

If we want to preserve our civilization ... we must find ways of tapping other energy sources than the quickly vanishing fossil fuels. [1]
-- Hans Thirring, 1958

Abstract

This paper presents a new method for predicting petroleum production. The method is used to predict production from 1996 to 2040 for the top 41 oil-bearing nations, seven geographical regions, OPEC versus non-OPEC groups, and the World total. Findings: The FSU and Saudi Arabia will lead the nations. The Middle East commands the regions. OPEC overtakes non-OPEC in 2006 and dominates thenceforth. World production peaks at 28.5 billion barrels in 2005, and then falls to 10.9 billion barrels in 2040—a decline of 62% in 35 years. Details of the new method are given in the Appendix.

Introduction

In 1961, President John F. Kennedy requested a fresh study of US natural resources. The report on energy was directed and written by eminent geologist and geophysicist M. King Hubbert. For that study the earlier petroleum forecasting techniques were no longer appropriate.

Although exceptional progress was made in the above study, the forecasting method that emerged has significant shortcomings. [3.4]

The goals of this paper are fourfold.

1. Introduce an entirely new method to forecast petroleum production.
2. Use it to predict the year that OPEC production will (again) dominate World production.
3. Predict the year that World production will reach its all-time peak.
4. Explore the likely consequences of the shifts and decline.

The new forecasting method is called the Numerate-Empiric Model. [4,5] A discussion appears in the Appendix {Goal 1}. The main text focuses on future petroleum production and its likely social-economic impacts {Goals 2-4}.

We proceed with the petroleum producers.

The Petroleum-Producing Nations

Mother Nature, as it were, played favorites when she parceled out the World's petroleum resources. Table 1 lists the World's main 41 petroleum-producing nations. [6,7] Each nation is modeled separately.

• Column C gives the year that production peaked or is predicted to peak. The USA and Romania peaks are certainly correct because their production has been long falling and their reserves are shrinking fast. However, the peaks of other nations are less certain, e.g., Venezuela and Saudi Arabia.
• Column D gives magnitude of the peaks in billions of barrels per year (Gb/yr). The differences are vast. For example, the FSU peak in 1987 was a whopping 4.62 Gb/yr compared to the Papua New Guinea peak in 1993 of a minuscule 0.04 Gb/yr—a difference of (sic) 12,000%.
• Column E records the production rate for 1995. The differences are staggering. For example, in 1995 there were some 190 nations in the World. The top 41 logged 98.4% (i.e., 24.3/24.7, Table 1) of the total production; the omitted 149 nations—a meager 1.6%.
• Column F gives the production forecasts for 2040. Comparing columns E and F. World production from 1995 to 2040 will fall by 55%.
• Columns G and H give the cumulative production for years 1995 and 2040. Column I gives the Expected Ultimate Recovery (EUR) in billions of barrels (Gb). Column G is historic; columns H and I are forecasts. Comparison of G. H and I indicates the 'stage' (or 'age') of the petroleum life-cycle. Some examples are given following Table 1.

Table 1. Petroleum Production Summary:
Nations and World

* OPEC Member; 'G' means billion; 'b' means barrels; 'yr' means year; 'EUR' means Expected Ultimate Recovery.

Half Gone: Cumulative production of the 41 nations reaches 1005 Gb in 2004, marking the year that half the World's producible oil will be gone.

The USA: By 1995 the US had exhausted 194.4 Gb, i.e., over 75% of its original endowment. That's history. But by 2040, over 98% will be exhausted.

The Bottom Ten: By 2040, ten nations will have depleted virtually all of their oil: Peru 95.0%, Norway 95.1%, Mexico 96.2%, Australia 96.4%, Qatar 96.4%, Syria 96.5%, Algeria 97.0%, USA 98.3%, Romania 98.3%, and Angola 98.8%.

Asian Tigers?: In terms of oil resources, the so-called 'Asian Tigers'—such as China, India and Indonesia—are just pussy-cats in disguise.

Note 1. The forecast of the total World EUR is 2071 Gb (Table 1) -comfortably within the range of 'Expert' estimates. [5]

Let's look at some national examples.

Petroleum Forecasts: FSU, Saudi Arabia, and Norway

Figure 1 graphs the petroleum production of: (1) the Former Soviet Union, (2) Saudi Arabia, and (3) Norway. Each curve illustrates important aspects of the petroleum life-cycle.

Figure 1. Petroleum Life-Cycle Examples:
FSU, Saudi Arabia, and Norway

The horizontal axis in Figure I goes from 1960 to 2040. The vertical axis shows production scaled from 0.0 to 5.0 Gb/yr. Historic data appears from 1960 through 1995. Forecasted values from 1996 through 2040.

Note 2: The full life-cycle curves go from zero production (on the far left, not shown), through one or more peaks, and ultimately back to zero (on the far right, not shown). No exemptions.

• Curve 1, FSU: The FSU illustrates production restrained (since 1990 at least) by institutional collapse. Production was 1.1 Gb/yr in 1960; then rapid growth through 1977; followed by the Everest-like peak of 4.6 Gb/yr in 1987 - a World record! Next the USSR disintegrated and production plummeted to 2.6 in 1995—a fall of 43% in 8 years! But because of the FSU's huge EUR (i.e., 305 Gb), the forecast takes a sharp upward turn in 1997, followed by a temporary high of 3.7 Gb/yr in 2005. Thereafter it nose-dives to 2.0 in 2040.

Note 3: Whatever shape the forecast, the total area under the curve must equal the EUR, e.g., 305 Gb for FSU, Table 1.

• Curve 2, Saudi Arabia: The Saudi data illustrates production restrained by, e.g., national policy and OPEC quotas. Production was 0.5 Gb/yr in 1960; then grew exponentially until 1973; paused during the Arab oil embargo; then climbed to a high of 3.7 Gb/yr in 1980. As oil prices collapsed, Saudi production fell to a new low of 1.5 Gb/yr in 1984; then rebounded to 3.9 in 1995. In recent years production has been near capacity, and huge upfront capital would be necessary to increase it. With this in mind (and internal problems rising!), production is predicted to increase moderately to 3.9 Gb/yr in 2010 -the all-time Saudi peak. Thereafter, depletion (the ultimate restraint) dominates and production plummets to 1.9 Gb/yr in 2040.

Note 4: It's useful to observe how non-measurable knowledge was incorporated into the FSU and Saudi forecasts above.

• Curve 3, Norway: The Norwegian data illustrates (what economists are wont to call) 'free-market, unrestrained production.' This suggests that the lifecycle is quite predictable. Production started in 1972, then grew smoothly to 1.1 Gb/yr in 1995. From there, it rises to a broad peak of 1.3 Gb/yr in 2000. Thence 'gracefully' slumps to 0.2 in 2040.

A main point of the foregoing examples is that most production life-cycles are highly irregular (such as the FSU and Saudi Arabia). It's rare for a nation to have a symmetric curve like that of Norway. The Numerate-Empiric Model, however, handles all shapes with equal ease.

In the next section we explore the World's petroleum regions.

The Seven Petroleum Regions: Giants and Dwarfs

Table 2 classifies the petroleum-producing nations (listed previously) into seven geographic regions and summarizes the key information. The regions are: I - North America, II - South & Central America, III - Europe, IV - Former Soviet Union, V - Middle East, VI - Africa, and VII - Asia & Australasia.

• Column C lists the regional peak year. Note that North America was first to peak in 1984, followed by the FSU in 1987. Thence three regions peak in a flash: Africa and Europe in 2000, and Asia & Australasia in 2001. Next comes South & Central America in 2007. Outlasting all others, the Middle East peaks in 2010.

Table 2. Petroleum-Producing Regions:
Summary and Forecast

Note 5. If the 1987 F5U 'peak' is later topped, then 1987 will simply mark a temporary high (Table 2).

• Column D gives the magnitude of the peak. Note well that the Middle East peak is a towering 11.7 Gb/yr. In contrast, Europe's peak is a lowly 2.4 Gb/yr. A difference of 500%. Jeff and Mutt.
• Columns E and F show that production in all seven regions will be significantly less in 2040 than it was in 1995. Although Middle East and FSU decline by (only!) 22% and 23%, Europe and North America plunge by 82% and 84%, respectively.
• Columns G and H list regional cumulative production for years 1995 and 2040; column I gives the EUR. Taken together, the numbers portend the coming oil crisis. For example, in 2040 the World's oil-producing regions will be depleted as follows: Middle East 84%; FSU 85%; South & Central America 90%; Asia & Australasia 91%; Africa 93%; Europe 93%; and North America 98%. Sputtering on empty.

Moreover—In 1995 most industrial nations relied heavily on imported oil, e.g., USA 47%; Japan 94%. By contrast, the same year Middle East nations exported 82% of their oil. However, by 2040 the Middle East (itself) will have depleted 84% of its producible oil. With that situation looming, Will Middle East nations still agree to export their oil to the West? If so, under what conditions? ....

Figure 2 (top next page) shows the production curves of three key oil-producing regions: the Middle East, North America, and Europe.

Figure 2. Regional Petroleum Life-Cycle Examples:
Middle East, North America, and Europe

• Middle East production (curve 1) was 1.9 Gb/yr in 1960; then grew exponentially until the Arab oil embargo of 1973; dipped slightly in 1974; next jumped to its historic high of 8.0 Gb/yr in 1976. Thereafter (in concert with falling oil prices), production plummeted to 4.0 in 1984; thence a strong rebound to 7.3 in 1995. Post-1995 production continues its strong upward trend until the all-time peak at 11.7 Gb/yr in 2010. Production remains robust until 2018—then takes a dizzying, FSU-like dive to 5.3 in 2040.
• North America production (curve 2) was 3.2 Gb/yr in 1960; 5.0 in 1973; then took a temporary dip to 4.5 in 1976. Thereafter (boosted by Alaska, Canada and Mexico), North America reached its (certain!) all-time peak of 5.6 Gb/yr in 1984. Thereafter production slipped to 5.0 in 1995. Long past its prime, North America falls to 0.7 in 2040—a mere 13% of its 1984 peak. Not to be missed is the smoldering production discrepancy from 1987 through 1995 between the Middle East and North America (curves 1 and 2, Figure 2).
• Europe production (curve 3) was insignificant until 1975 when large oil fields were tapped in the North Sea. From that point, production climbed to 0.3 Gb/yr in 1976; grew steadily to 1.4 in 1986; leveled until 1990; thereupon spurted to 2.2 in 1995. This upward trend continues until the all-time peak is reached at 2.4 Gb/yr in 2000. Thence production declines to 0.4 in 2040.

Note 6: In Figure 2, observe that North America and Europe curves are highly symmetric. And after 1984, the Middle East too.

The areas under the curves of Figure 2 clearly show that the Middle East is the World's towering Oil-Goliath (curve 1). Per Table 2, the Middle East's EUR is more than twice that of North America (curve 2), and some nine times that of Europe (curve 3).

Oily politics in view?

OPEC versus Non-OPEC: The Gap Ahead

The twelve member nations of the Organization of Petroleum Exporting Countries (OPEC) are identified by asterisks (*) in Table 1 (previous). Six are in the Middle East; four in Africa; and one each in South America and Asia & Australasia.

The question is- not whether, but - When will OPEC dominate World petroleum production? This looming event, 'Cross-Over Point 3' (Figure 3), is crucial because it's sure to have social, economic, political, military, and (yes) religious consequences. Figure 3 shows the OPEC and nonOPEC curves. Imagine oil production as a marathon race.

Figure 3. OPEC versus Non-OPEC Petroleum
Production: 1960-2040

• In 1960, non-OPEC production led OPEC by a 'comfortable' 1.5 Gb/yr, but the gap was closing fast (Figure 3).
• Then in 1970, and for the first time, OPEC passed non-OPEC; this event is labeled 'Cross-Over Point 1.' It set the stage for an OPEC power play.
• It came suddenly in 1973—the 'Arab oil embargo.' The USA was a prime target.
• Thereafter, as OPEC lagged, non-OPEC kicked into the lead in 1977—labeled 'Cross-Over Point 2.''
• Following 1977, OPEC stumbled while non-OPEC sprinted. By 1984 non-OPEC commanded a 7.9 Gb/yr lead over OPEC (i.e., 123%). Economists and politicians (and other prodigals) shouted, "The oil crisis is over!"
• But by 1995 OPEC had closed the gap to 4.2 Gb/yr (i.e., 41%) and was coming on strong.
• Then with surging strength, OPEC overtakes non-OPEC in 2006, Cross-Over Point 3 {Goal 2}. Thenceforth OPEC dominates World production (shaded area, Figure 3).
• OPEC reaches its peak at 15.0 Gb/yr in 2009, and -although staggering itself - OPEC nevertheless manages to widen its lead to 5.3 Gb/yr in 2026.
• Nearing the end, OPEC still maintains a commanding lead of 2.3 Gb/yr in 2040.

Of course, questions still remain about the World situation in 2006. For example: Will the OPEC nations still exist nine years hence? Who will control the flow of oil therein? Anticipating another 'Arab oil embargo,' will the USA take 'preemptive action'? .... ?

Next—the BIG, b-r-o-a-d picture.

The World Petroleum Life-Cycle: Options Closing

Figure 4 shows World petroleum production from 1960 to 2040 (curve 1, top line). For comparison, the OPEC and non-OPEC curves are repeated here (curves 2 and 3).

Figure 4. World, OPEC, and non-OPEC Petroleum
Production Profiles: 1960-2040

• Historic Data: In 1960 World production reached 7.9 Gb/yr (curve 1). Strong growth continued to 1973; then (after a brief hiatus) it surged to 21.6 in 1976. Thence production slumped to a temporary low of 20.3 in 1982. Thereafter World production grew to 24.4 Gb/yr in 1995.

• The Future: From 1996 onward, production increases briefly, then eases off to the World all-time peak at 28.5 Gb/yr in 2005 {Goal 3}. Decline begins gradually, then avalanches in 2009 as both OPEC and non-OPEC go into steep-slide. By 2040 World production has fallen to 10.9 Gb/yr—down 62% in 35 years.

It's not that other energy sources aren't feasible. It's that none will be on-line by 2005. And continuing our present course, none by 2050 either.

Not appreciating the danger and therefore failing to act against it, may be the biggest danger we faced
-- J. Richard Gott, 1996

It's time to wrap things up.

Summary and Conclusions

A new method is presented for predicting petroleum production—called the 'Numerate Empiric Model.' It is, in my experience, more flexible and more accurate than earlier methods. All forecasts are based on both quantitative and nonmeasurable knowledge. The N-E Model produced the unique production forecasts shown in Figures 1-4 and 7. Details are given in the Appendix.

Separate models were tailored for each of the World's top 41 oil-producing nations. Key findings are summarized in Table 1. Three national examples are shown in Figure 1: FSU, Saudi Arabia and Norway. Although the production patterns are quite different—all increase, peak, thence decrease. No exceptions. The FSU and Saudi Arabia dominate future production.

Next, oil production is categorized by seven geographical regions (e.g., North America, Europe, Middle East, etc.). Results are summarized in Table 2. Note that the Middle East's Expected Ultimate Recovery (EUR) is more than twice that of North America, i.e., 763 to 368 billion barrels. Further, the production in all regions is significantly less in 2040 than it was in 1995. Figure 2 shows three regional examples: Middle East, North American and Europe. The Middle East, of course, is the World's Petroleum Prodigy.

Figure 3 compares OPEC and non-OPEC production. OPEC production first surpassed non-OPEC in 1970, thence (following the infamous 'Arab oil embargo') fell woefully behind by 1984. At that point, OPEC again surged and by 1995 the gap was closing quick. Thereafter, non-OPEC fatigues, while OPEC continues its torrid pace. OPEC inexorably passes non-OPEC in 2006, labeled 'Cross-Over-Point 3.' From there on it's OPEC all the way (shaded area, Figure 3).

The 41 national forecasts are summed to give the World petroleum profile from 1960 to 2040, Figure 4. The World all-time peak occurs at 28.5 billion barrels per year in 2005. Thereafter, World production plummets to 10.9 billion barrels per year in 2040—a fall of 62% in 35 years. The OPEC Cross-Over Point in 2006 is shown for comparison.

The inexorable dynamics of World oil production will have human consequences in the extreme. Briefly put: By 2005 the future of oil has come and gone—but still no new energy source in sight. Could Hans Thirring be right?

APPENDIX:

The Numerate-Empiric Model

Recall Goal 1 of this paper: 'Introduce an entirely new method to predict petroleum production.' The method is called the Numerate-Empiric ('N-E') Model. Experience over the past year demonstrates that the N-E Model greatly improves the flexibility and accuracy of petroleum forecasts as compared to earlier methods. Moreover, it is fully consistent with (but very different from) the flagship work of M. K. Hubbert. [2, 3] Three steps are involved. Suggestion: Glance ahead at Figures 5-7; then come back to Step 1, below.

• Step 1, Numerate: This step is quantitative. Start with the production data. Then use the Translated Coordinate System (TCS) approach and regression analysis to squeeze the maximum useful information out of the data. [4] This step can be done with any standard Worksheet, such as Microsoft Excel. [9] It provides the values for the Hubbert ('H') Model parameters. But note carefully: (1) the H Model must reference the TCS, and (2) most important, the H Model forecast (e.g., curve 1, Figure 7) serves only as a prototype for the N-E Model user. It is NOT the N-E Model forecast.

• Step 2, Empiric: This step is qualitative - an 'Expert System' technique. Use the important non-measurable knowledge available (e.g., reserve and EUR estimates, the most recent production trend, national policy, production restraints [such OPEC 'quotas'], the 1992 Climate Treaty, etc.), to produce a seamless and consistent production forecast (e.g., from 1996 to 2040). Make several simulation 'runs,' modifying the Expert forecast each time. The 'best' run (your choice) becomes the N-E Model forecast (my choice was curve 2, Figure 7). This step is greatly facilitated by simulation languages such as STELLA. [10]

Note 7: As for production restraints - Asking the oil industry about global warming is like asking the cigarette industry about lung cancer. For instance - Long after the English, French and Spanish have switched from raising sheep to raising camels. the oil industry will still claim, "It's just a theory."

• Step 3, Interpretation: In this step you double-check the N-E Model and strive to understand the production forecast. Does the forecast smoothly adjoin the most recent production trend? Does the area under the curve fall within the range of 'Expert' EUR estimates? What insight can be gained from the H Model forecast (e.g., curve 1, Figure 7)? Further, you'll want to identify the likely consequences of the petroleum forecast (e.g., social, political and economic impacts).

• Iteration: Whenever new production data becomes available, repeat Steps 1-3.

Note 8: Of the 41 nations studied, the N-E Model forecast is usually significantly higher than the H Model forecast. For instance, compare curve 2 with curve 1, Figure 7.

Words, words, words. Let's have an example.

The Numerate-Empiric Model: An Example

[ Because of the great difficulty of putting this example onto a web page, I have elected not to. If you want this section you may request a hard copy of this paper from Richard Duncan at duncanrc@halcyon.com, or download the actual models at http://www.halcyon.com/duncanrc/ -- JH]

References

[1] Thirring, H. (1958) Energy for Man: From Windmills to Nuclear Power. Harper & Row. New York.

[2] Hubbert, M. K. (1982) Techniques of prediction as applied to the production of oil and gas. Oil and Gas Supply Modeling (ed. Gass, S. I.), 16-141. NBS Special Publication 631, Washington, D. C.

[3] Hubbert, M. K. (1962) Energy resources: A report to the Committee on Natural Resources. National Academy of Sciences, Natl. Research Council, Pub. 1000-D.

[4] Duncan, R. C. (1996) The Mexican petroleum 'play' in two 'acts': Taking hold of oil production data. System Dynamics Conference Proceedings, System Dynamics Society, Cambridge, MA.

[5] Duncan, R. C. (1997) Fossil fuel prospects for the twenty-first century. In Solar Power Satellites: The Emerging Energy Option. Glaser, P. E., et al. (eds.). E. Horwood, New York.

[6] BP (1981-1996) BP Statistical Review of World Energy. British Petroleum Company, London.

[7] Campbell, C. J. (1991) The Golden Century of Oil: The Depletion of a Resource. Kluwer Academic Publishers, Dordrecht, Germany.

[8] Leslie, J. (1996) Thc End of the World. Introduction by J. R. Gott. Routledge, London.

[9] MSC (1995) Execl 4.0 for Macintosh. Microsoft Corporation, Redmond, WA.

[10] HPS (1996) STELLA Technical Documentation. High Performance Systems, Inc. Hanover, NH.

[11] Riva, J. P. (1995) World oil production after year 2000: Business as usual or crisis? Congressional Research Service, Library of Congress, Washington, DC.

The second issue of the World Petroleum Forecast (i.e., with data through 1996) will be available about 1 September 1997. Complementary copies are available on request from the Institute on Energy and Man: duncanrc@halcyon.com

Table of Contents  Please send me your comments.