Robert L. Hirsch, SAIC, Project Leader
Roger Bezdek, MISI
Robert Wendling, MISI

March, 2005

This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.

The peaking of world oil production presents the U.S. and the world with an unprecedented risk management problem. As peaking is approached, liquid fuel prices and price volatility will increase dramatically, and, without timely mitigation, the economic, social, and political costs will be unprecedented. Viable mitigation options exist on both the supply and demand sides, but to have substantial impact, they must be initiated more than a decade in advance of peaking.

In 2003, the world consumed just under 80 million barrels per day (MM bpd) of oil. U.S. consumption was almost 20 MM bpd, two-thirds of which was in the transportation sector. The U.S. has a fleet of about 210 million automobiles and light trucks (vans, pick-ups, and SUVs). The average age of U.S. automobiles is nine years. Under normal conditions, replacement of only half the automobile fleet will require 10-15 years. The average age of light trucks is seven years. Under normal conditions, replacement of one-half of the stock of light trucks will require 9-14 years. While significant improvements in fuel efficiency are possible in automobiles and light trucks, any affordable approach to upgrading will be inherently time-consuming, requiring more than a decade to achieve significant overall fuel efficiency improvement.

Besides further oil exploration, there are commercial options for increasing world oil supply and for the production of substitute liquid fuels: 1) Improved Oil Recovery (IOR) can marginally increase production from existing reservoirs; one of the largest of the IOR opportunities is Enhanced Oil Recovery (EOR), which can help moderate oil production declines from reservoirs that are past their peak production: 2) Heavy oil / oil sands represents a large resource of lower grade oils, now primarily produced in Canada and Venezuela; those resources are capable of significant production increases;. 3) Coal liquefaction is a wellestablished technique for producing clean substitute fuels from the world’s abundant coal reserves; and finally, 4) Clean substitute fuels can be produced from remotely located natural gas, but exploitation must compete with the world’s growing demand for liquefied natural gas. However, world-scale contributions from these options will require 10-20 years of accelerated effort. Dealing with world oil production peaking will be extremely complex, involve literally trillions of dollars and require many years of intense effort.

To explore these complexities, three alternative mitigation scenarios were analyzed:

  • Scenario I assumed that action is not initiated until peaking occurs.
  • Scenario II assumed that action is initiated 10 years before peaking.
  • Scenario III assumed action is initiated 20 years before peaking.

For this analysis estimates of the possible contributions of each mitigation option were developed, based on an assumed crash program rate of implementation.

Our approach was simplified in order to provide transparency and promote understanding. Our estimates are approximate, but the mitigation envelope that results is believed to be directionally indicative of the realities of such an enormous undertaking. The inescapable conclusion is that more than a decade will be required for the collective contributions to produce results that significantly impact world supply and demand for liquid fuels.

Important observations and conclusions from this study are as follows:

    1. When world oil peaking will occur is not known with certainty. A fundamental problem in predicting oil peaking is the poor quality of and possible political biases in world oil reserves data. Some experts believe peaking may occur soon. This study indicates that “soon” is within 20 years.

    2. The problems associated with world oil production peaking will not be temporary, and past “energy crisis” experience will provide relatively little guidance. The challenge of oil peaking deserves immediate, serious attention, if risks are to be fully understood and mitigation begun on a timely basis.

    3. Oil peaking will create a severe liquid fuels problem for the transportation sector, not an “energy crisis” in the usual sense that term has been used.

    4. Peaking will result in dramatically higher oil prices, which will cause protracted economic hardship in the United States and the world. However, the problems are not insoluble. Timely, aggressive mitigation initiatives addressing both the supply and the demand sides of the issue will be required.

    5. In the developed nations, the problems will be especially serious. In the developing nations peaking problems have the potential to be much worse.

    6. Mitigation will require a minimum of a decade of intense, expensive effort, because the scale of liquid fuels mitigation is inherently extremely large.

    7. While greater end-use efficiency is essential, increased efficiency alone will be neither sufficient nor timely enough to solve the problem. Production of large amounts of substitute liquid fuels will be required. A number of commercial or near-commercial substitute fuel production technologies are currently available for deployment, so the production of vast amounts of substitute liquid fuels is feasible with existing technology.

    8. Intervention by governments will be required, because the economic and social implications of oil peaking would otherwise be chaotic. The experiences of the 1970s and 1980s offer important guides as to government actions that are desirable and those that are undesirable, but the process will not be easy.

Mitigating the peaking of world conventional oil production presents a classic risk management problem:

  • Mitigation initiated earlier than required may turn out to be premature, if peaking is long delayed.
  • If peaking is imminent, failure to initiate timely mitigation could be extremely damaging.

Prudent risk management requires the planning and implementation of mitigation well before peaking. Early mitigation will almost certainly be less expensive than delayed mitigation. A unique aspect of the world oil peaking problem is that its timing is uncertain, because of inadequate and potentially biased reserves data from elsewhere around the world. In addition, the onset of peaking may be obscured by the volatile nature of oil prices. Since the potential economic impact of peaking is immense and the uncertainties relating to all facets of the problem are large, detailed quantitative studies to address the uncertainties and to explore mitigation strategies are a critical need.

The purpose of this analysis was to identify the critical issues surrounding the occurrence and mitigation of world oil production peaking. We simplified many of the complexities in an effort to provide a transparent analysis. Nevertheless, our study is neither simple nor brief. We recognize that when oil prices escalate dramatically, there will be demand and economic impacts that will alter our simplified assumptions. Consideration of those feedbacks will be a daunting task but one that should be undertaken.

Our study required that we make a number of assumptions and estimates. We well recognize that in-depth analyses may yield different numbers. Nevertheless, this analysis clearly demonstrates that the key to mitigation of world oil production peaking will be the construction a large number of substitute fuel production facilities, coupled to significant increases in transportation fuel efficiency. The time required to mitigate world oil production peaking is measured on a decade time-scale. Related production facility size is large and capital intensive. How and when governments decide to address these challenges is yet to be determined.

Our focus on existing commercial and near-commercial mitigation technologies illustrates that a number of technologies are currently ready for immediate and extensive implementation. Our analysis was not meant to be limiting. We believe that future research will provide additional mitigation options, some possibly superior to those we considered. Indeed, it would be appropriate to greatly accelerate public and private oil peaking mitigation research. However, the reader must recognize that doing the research required to bring new technologies to commercial readiness takes time under the best of circumstances. Thereafter, more than a decade of intense implementation will be required for world scale impact, because of the inherently large scale of world oil consumption.

In summary, the problem of the peaking of world conventional oil production is unlike any yet faced by modern industrial society. The challenges and uncertainties need to be much better understood. Technologies exist to mitigate the problem. Timely, aggressive risk management will be essential.

Oil is the lifeblood of modern civilization. It fuels the vast majority of the world’s mechanized transportation equipment – Automobiles, trucks, airplanes, trains, ships, farm equipment, the military, etc. Oil is also the primary feedstock for many of the chemicals that are essential to modern life. This study deals with the upcoming physical shortage of world conventional oil — an event that has the potential to inflict disruptions and hardships on the economies of every country. The earth’s endowment of oil is finite and demand for oil continues to increase with time. Accordingly, geologists know that at some future date, conventional oil supply will no longer be capable of satisfying world demand. At that point world conventional oil production will have peaked and begin to decline.

A number of experts project that world production of conventional oil could occur in the relatively near future, as summarized in Table I-1.1 Such projections are fraught with uncertainties because of poor data, political and institutional selfinterest, and other complicating factors. The bottom line is that no one knows with certainty when world oil production will reach a peak,2 but geologists have no doubt that it will happen.

Table I-1. Predictions of World Oil Production Peaking

Projected Date Source of Projection
2006-2007 Bakhitari
2007-2009 Simmons
After 2007 Skrebowski
Before 2009 Deffeyes
Before 2010 Goodstein
Around 2010 Campbell
After 2010 World Energy Council
2010-2020 Laherrere
2016 EIA (Nominal)
After 2020 CERA
2025 or later Shell
No visible Peak Lynch

Our aim in this study is to

  • Summarize the difficulties of oil production forecasting;
  • Identify the fundamentals that show why world oil production peaking is such a unique challenge;
  • Show why mitigation will take a decade or more of intense effort;
  • Examine the potential economic effects of oil peaking;
  • Describe what might be accomplished under three example mitigation scenarios.
  • Stimulate serious discussion of the problem, suggest more definitive studies, and engender interest in timely action to mitigate its impacts.

In Chapter II we describe the basics of oil production, the meaning of world conventional oil production peaking, the challenge of making accurate forecasts, and the effects that higher prices and advanced technology might have on oil production.

Because of the massive scale of oil use around the world, mitigation of oil shortages will be difficult, time consuming, and expensive. In Chapter III we describe the extensive and critical uses of U.S. oil and the long economic and mechanical lifetimes of existing liquid fuel consuming vehicles and equipment.

While it is impossible to predict the impact of world oil production peaking with any certainty, much can be learned from past oil disruptions, particularly the 1973 oil embargo and the 1979 Iranian oil shortage, as discussed in Chapter IV. In Chapter V we describe the developing shortages of U.S. natural gas, shortages that are occurring in spite of assurances of abundant supply provided just a few years ago. The parallels to world oil supply are disconcerting.

In Chapter VI we describe available mitigation options and related implementation issues. We limit our considerations to technologies that are near ready or currently commercially available for immediate deployment. Clearly, accelerated research and development holds promise for other options.

However, the challenge related to extensive near-term oil shortages will require deployment of currently viable technologies, which is our focus.

Oil is a commodity found in over 90 countries, consumed in all countries, and traded on world markets. To illustrate and bracket the range of mitigation options, we developed three illustrative scenarios. Two assume action well in advance of the onset of world oil peaking – in one case, 20 years before peaking and in another case, 10 years in advance. Our third scenario assumes that no action is taken prior to the onset of peaking. Our findings illustrate the magnitude of the problem and the importance of prudent risk management.

Finally, we touch on possible market signals that might foretell the onset of peaking and possible wildcards that might change the timing of world conventional oil production peaking. In conclusion, we frame the challenge of an unknown date for peaking, its potentially extensive economic impacts, and available mitigation options as a matter of risk management and prudent response.

The reader is asked to contemplate three major questions:

  • What are the risks of heavy reliance on optimistic world oil production peaking projections?
  • Must we wait for the onset of oil shortages before actions are taken?
  • What can be done to ensure that prudent mitigation is initiated on a timely basis?


A. Background

Oil was formed by geological processes millions of years ago and is typically found in underground reservoirs of dramatically different sizes, at varying depths, and with widely varying characteristics. The largest oil reservoirs are called “Super Giants,” many of which were discovered in the Middle East. Because of their size and other characteristics, Super Giant reservoirs are generally the easiest to find, the most economic to develop, and the longest lived. The last Super Giant oil reservoirs discovered worldwide were found in 1967 and 1968. Since then, smaller reservoirs of varying sizes have been discovered in what are called “oil prone” locations worldwide — oil is not found everywhere.

Geologists understand that oil is a finite resource in the earth’s crust, and at some future date, world oil production will reach a maximum — a peak — after which production will decline. This logic follows from the well-established fact that the output of individual oil reservoirs rises after discovery, reaches a peak and declines thereafter. Oil reservoirs have lifetimes typically measured in decades, and peak production often occurs roughly a decade or so after discovery. It is important to recognize that oil production peaking is not “running out.” Peaking is a reservoir’s maximum oil production rate, which typically occurs after roughly half of the recoverable oil in a reservoir has been produced. In many ways, what is likely to happen on a world scale is similar to what happens to individual reservoirs, because world production is the sum total of production from many different reservoirs.

Because oil is usually found thousands of feet below the surface and because oil reservoirs normally do not have an obvious surface signature, oil is very difficult to find. Advancing technology has greatly improved the discovery process and reduced exploration failures. Nevertheless, oil exploration is still inexact and expensive.

Once oil has been discovered via an exploratory well, full-scale production requires many more wells across the reservoir to provide multiple paths that facilitate the flow of oil to the surface. This multitude of wells also helps to define the total recoverable oil in a reservoir – its so-called “reserves.”

B. Oil Reserves

The concept of reserves is generally not well understood. “Reserves” is an estimate of the amount of oil in a reservoir that can be extracted at an assumed cost. Thus, a higher oil price outlook often means that more oil can be produced, but geology places an upper limit on price-dependent reserves growth; in well managed oil fields, it is often 10-20 percent more than what is available at lower prices.

Reserves estimates are revised periodically as a reservoir is developed and new information provides a basis for refinement. Reserves estimation is a matter of gauging how much extractable oil resides in complex rock formations that exist typically one to three miles below the surface of the ground, using inherently limited information. Reserves estimation is a bit like a blindfolded person trying to judge what the whole elephant looks like from touching it in just a few places. It is not like counting cars in a parking lot, where all the cars are in full view. Specialists who estimate reserves use an array of methodologies and a great deal of judgment. Thus, different estimators might calculate different reserves from the same data. Sometimes politics or self-interest influences reserves estimates, e.g., an oil reservoir owner may want a higher estimate in order to attract outside investment or to influence other producers.

Reserves and production should not be confused. Reserves estimates are but one factor in estimating future oil production from a given reservoir. Other factors include production history, understanding of local geology, available technology, oil prices, etc. An oil field can have large estimated reserves, but if the field is past its maximum production, the remaining reserves will be produced at a declining rate. This concept is important because satisfying increasing oil demand not only requires continuing to produce older oil reservoirs with their declining production, it also requires finding new ones, capable of producing sufficient quantities of oil to both compensate for shrinking production from older fields and to provide the increases demanded by the market.

C. Production Peaking

World oil demand is expected to grow 50 percent by 2025.4 To meet that demand, ever-larger volumes of oil will have to be produced. Since oil production from individual reservoirs grows to a peak and then declines, new reservoirs must be continually discovered and brought into production to compensate for the depletion of older reservoirs. If large quantities of new oil are not discovered and brought into production somewhere in the world, then world oil production will no longer satisfy demand. That point is called the peaking of world conventional oil production.

When world oil production peaks, there will still be large reserves remaining. Peaking means that the rate of world oil production cannot increase; it also means that production will thereafter decrease with time.

The peaking of world oil production has been a matter of speculation from the beginning of the modern oil era in the mid 1800s. In the early days, little was known about petroleum geology, so predictions of peaking were no more than guesses without basis. Over time, geological understanding improved dramatically and guessing gave way to more informed projections, although the knowledge base involves numerous uncertainties even today.

Past predictions typically fixed peaking in the succeeding 10-20 year period. Most such predictions were wrong, which does not negate that peaking will someday occur. Obviously, we cannot know if recent forecasts are wrong until predicted dates of peaking pass without incident.

With a history of failed forecasts, why revisit the issue now? The reasons are as follows: 1. Extensive drilling for oil and gas has provided a massive worldwide database; current geological knowledge is much more extensive than in years past, i.e., we have the knowledge to make much better estimates than previously.

2. Seismic and other exploration technologies have advanced dramatically in recent decades, greatly improving our ability to discover new oil reservoirs. Nevertheless, the oil reserves discovered per exploratory well began dropping worldwide over a decade ago. We are finding less and less oil in spite of vigorous efforts, suggesting that nature may not have much more to provide.

3. Many credible analysts have recently become much more pessimistic about the possibility of finding the huge new reserves needed to meet growing world demand.

4. Even the most optimistic forecasts suggest that world oil peaking will occur in less than 25 years.

5. The peaking of world oil production could create enormous economic disruption, as only glimpsed during the 1973 oil embargo and the 1979 Iranian oil cut-off.

Accordingly, there are compelling reasons for in-depth, unbiased reconsideration.

D. Types of Oil

Oil is classified as “Conventional” and “Unconventional.” Conventional oil is typically the highest quality, lightest oil, which flows from underground reservoirs with comparative ease. Unconventional oils are heavy, often tar-like. They are not readily recovered since production typically requires a great deal of capital investment and supplemental energy in various forms. For that reason, most current world oil production is conventional oil.5 (Unconventional oil production will be discussed in Chapter VI).

E. Oil Resources6

Consider the world resource of conventional oil. In the past, higher prices led to increased estimates of conventional oil reserves worldwide. However, this pricereserves relationship has its limits, because oil is found in discrete packages (reservoirs) as opposed to the varying concentrations characteristic of many minerals. Thus, at some price, world reserves of recoverable conventional oil will reach a maximum because of geological fundamentals. Beyond that point, insufficient additional conventional oil will be recoverable at any realistic price. This is a geological fact that is often misunderstood by people accustomed to dealing with hard minerals, whose geology is fundamentally different. This misunderstanding often clouds rational discussion of oil peaking.

Future world recoverable reserves are the sum of the oil remaining in existing reservoirs plus the reserves to be added by future oil discoveries. Future oil production will be the sum of production from older reservoirs in decline, newer reservoirs from which production is increasing, and yet-to-be discovered reservoirs.

Because oil prices have been relatively high for the past decade, oil companies have conducted extensive exploration over that period, but their results have been disappointing. If recent trends hold, there is little reason to expect that exploration success will dramatically improve in the future. This situation is evident in Figure II-1, which shows the difference between annual world oil reserves additions minus annual consumption.7 The image is one of a world moving from a long period in which reserves additions were much greater than consumption, to an era in which annual additions are falling increasingly short of annual consumption. This is but one of a number of trends that suggest the world is fast approaching the inevitable peaking of conventional world oil production.

F. Impact of Higher Prices and New Technology

Conventional oil has been the mainstay of modern civilization for more than a century, because it is most easily brought to the surface from deep underground reservoirs, and it is the most easily refined into finished fuels. The U.S. was endowed with huge reserves of petroleum, which underpinned U.S. economic

growth in the early and mid twentieth century. However, U.S. oil resources, like those in the world, are finite, and growing U.S. demand resulted in the peaking of U.S. oil production in the Lower 48 states in the early 1970s. With relatively minor exceptions, U.S. Lower 48 oil production has been in continuing decline ever since. Because U.S. demand for petroleum products continued to increase, the U.S. became an oil importer. Today, the U.S. depends on foreign sources for almost 60 percent of its needs, and future U.S. imports are projected to rise to 70 percent of demand by 2025.8

Over the past 50 years, exploration for and production of petroleum has been an increasingly more technological enterprise, benefiting from more sophisticated engineering capabilities, advanced geological understanding, improved instrumentation, greatly expanded computing power, more durable materials, etc. Today’s technology allows oil reservoirs to be more readily discovered and better understood sooner than heretofore. Accordingly, reservoirs can be produced more rapidly, which provides significant economic advantages to the operators but also hastens peaking and depletion.

Some economists expect higher oil prices and improved technologies to continue to provide ever-increasing oil production for the foreseeable future. Most geologists disagree because they do not believe that there are many huge new oil reservoirs left to be found. Accordingly, geologists and other observers believe that supply will eventually fall short of growing world demand – and result in the peaking of world conventional oil production.

To gain some insight into the effects of higher oil prices and improved technology on oil production, let us briefly examine related impacts in the U.S. Lower 48 states. This region is a useful surrogate for the world, because it was one of the world’s richest, most geologically varied, and most productive up until 1970, when production peaked and started into decline. While the U.S. is the best available surrogate, it should be remembered that the decline rate in US production was in part impacted by the availability of large volumes of relatively low cost oil from the Middle East.

Figure II-2 shows EIA data for Lower 48 oil production,9 to which trend lines have been added that will aid our scenarios analysis later in the report. The trend lines show a relatively symmetric, triangular pattern. For reference, four notable petroleum market events are noted in the figure: the 1973 OPEC oil embargo, the 1979 Iranian oil crisis, the 1986 oil price collapse, and the 1991 Iraq war.

Figure II-3 shows Lower 48 historical oil production with oil prices and technology trends added. In constant dollars, oil prices increased by roughly a factor of three in 1973-74 and another factor of two in 1979-80. The modest production up-ticks in the mid 1980s and early 1990s are likely responses to the 1973 and 1979 oil price spikes, both of which spurred a major increase in U.S exploration and production investments. The delays in production response are inherent to the implementation of large-scale oil field investments. The fact that the production up-ticks were moderate was due to the absence of attractive exploration and production opportunities, because of geological realities. Beyond oil price increases, the 1980s and 1990s were a golden age of oil technology development, including practical 3-D seismic, economic horizontal drilling, and dramatically improved geological understanding. Nevertheless, Figure II-3 shows, Lower 48 production still trended downward, showing pronounced response to either price or technology. In light of this experience, there is good reason to expect that an analogous situation will exist worldwide after world oil production peaks: Higher prices and improved technology unlikely to yield dramatically higher conventional oil production.10

G. Projections of the Peaking of World oil Production

Projections of future world oil production will be the sum total of 1) output from all of the world’s then existing producing oil reservoirs, which will be in various stages of development, and 2) all the yet-to-be discovered reservoirs in their various states of development. This is an extremely complex summation problem, because of the variability and possible biases in publicly available data. In practice, estimators use various approximations to predict future world oil production. The remarkable complexity of the problem can easily lead to incorrect conclusions, either positive or negative.

Various individuals and groups have used available information and geological estimates to develop projections for when world oil production might peak. A sampling of recent projections is shown in Table II-1.


Robert L. Hirsch, SAIC, Project Leader
Roger Bezdek, MISI
Robert Wendling, MISI

March, 2005

1 A more detailed list is given in the following chapter in Table II-2.

2 In this study we interchangeably refer to the peaking of world conventional oil production as “oil peaking” or simply as “peaking.”

3 Portions of this chapter are taken from Hirsch, R.L. “Six Major Factors in Energy Planning”. U.S. Department of Energy. National Energy Technology Laboratory. March 2004.

4 U.S. Department of Energy, Energy Information Administration, International Energy Outlook – 2004, April 2004.

5 U.S. Department of Energy, Energy Information Administration, International Energy Outlook – 2004, April 2004.

6 Total oil in place is called the “resource.” However, only a part of the resource can be produced, because of geological complexities and economic limitations. That which is realistically recoverable is called “reserves,” which varies within limits depending on oil prices.

7 Aleklett, K. & Campbell, C.J. “The Peak and Decline of World Oil and Gas Production”. Uppsala University, Sweden. ASPO web site. 2003.

8 U.S. Department of Energy, Energy Information Administration, International Energy Outlook – 2004, April 2004.

9 U.S. Department of Energy, Energy Information Administration, Long Term World Oil Supply, April 18, 2000.

10 The US Lower 48 experience occurred over a long period characterized at different times by production controls (Texas Railroad Commission), price and allocation controls (1970s), free market prices (since 1981), wild price swings, etc., as well as higher prices and advancing technology. Nevertheless, production peaked and moved into a relatively constant rate of decline.

11 Bakhtiari, A.M.S. “World Oil Production Capacity Model Suggests Output Peak by 2006-07.” OGJ. April 26, 2004.

12 Simmons, M.R. ASPO Workshop. May 26, 2003.

13 Skrebowski, C. “Oil Field Mega Projects – 2004.” Petroleum Review. January 2004.

14 Deffeyes, K.S. Hubbert’s Peak-The Impending World Oil Shortage. Princeton University Press. 2003.

15 Goodstein, D. Out of Gas – The End of the Age of Oil. W.W. Norton. 2004

16 Campbell, C.J. “Industry Urged to Watch for Regular Oil Production Peaks, Depletion Signals.” OGJ. July 14, 2003.

17 Drivers of the Energy Scene. World Energy Council. 2003.

18 Laherrere, J. Seminar Center of Energy Conversion. Zurich. May 7, 2003

19 DOE EIA. “Long Term World Oil Supply.” April 18, 2000. See Appendix I for discussion.

20 Jackson, P. et al. “Triple Witching Hour for Oil Arrives Early in 2004 – But, As Yet, No Real Witches.” CERA Alert. April 7, 2004.

21 Davis, G. “Meeting Future Energy Needs.” The Bridge. National Academies Press. Summer 2003.

22 Lynch, M.C. “Petroleum Resources Pessimism Debunked in Hubbert Model and Hubbert Modelers’ Assessment.” Oil and Gas Journal, July 14, 2003.