Bill Totten's Weblog

Thursday, December 09, 2004

Hubbert's Peak: The Impending World Oil Shortage

by Kenneth S Deffeyes

Chapter 1 of his book by the same name (Princeton University Press, 2001)

Global oil production will probably reach a peak sometime during this decade. After the peak, the world's production of crude oil will fall, never to rise again. The world will not run out of energy, but developing alternative energy sources on a large scale will take at least ten years. The slowdown in oil production may already be beginning; the current price fluctuations for crude oil and natural gas may be the preamble to a major crisis.

In 1956, the geologist M King Hubbert predicted that US oil production would peak in the early 1970s. <1> Almost everyone, inside and outside the oil industry, rejected Hubbert's analysis. The controversy raged until 1970, when the US production of crude oil started
to fall. Hubbert was right.

[M King Hubbert, 1903-1989, was an American geophysicist who made important contributions to understanding fluid flow and the strength and behavior of rock bodies. Hubbert was at the Shell research lab in Houston when he made his original estimates of future oil production; he continued the work at the US Geological Survey.]

Around 1995, several analysts began applying Hubbert's method to world oil production, and most of them estimate that the peak year for world oil will be between 2004 and 2008. These analyses were reported in some of the most widely circulated sources: Nature, Science, and Scientific American. <2> None of our political leaders seem to be paying attention. If the predictions are correct there will be enormous effects on the world economy. Even the poorest nations need fuel to run irrigation pumps. The industrialized nations will be bidding against one another for the dwindling oil supply. The good news is that we will put less carbon dioxide into the atmosphere. The bad news is that my pickup truck has a 25-gallon tank.

The experts are making their 2004-8 predictions by building on Hubbert's pioneering work. Hubbert made his 1956 prediction at a meeting of the American Petroleum Institute in San Antonio, where he predicted that US oil production would peak in the early 1970s. He said later that the Shell Oil head office was on the phone right down to the last five minutes before the talk, asking Hubbert to withdraw his prediction. Hubbert had an exceedingly combative personality, and he went through with his announcement.

I went to work in 1958 at the Shell research lab in Houston, where Hubbert was the star of the show. He had extensive scientific accomplishments in addition to his oil prediction. His belligerence during technical arguments gave rise to a saying around the lab, "That Hubbert is a bastard, but at least he's our bastard". Luckily, I got off to a good start with Hubbert; he remained a good friend for the rest of his life.

Critics had many different reasons for rejecting Hubbert's oil prediction. Some were simply emotional; the oil business was highly profitable, and many people did not want to hear that the party would soon be over. A deeper reason was that many false prophets had appeared before. From 1900 onward, several of these people had divided the then known US oil reserves by the annual rate of production. (Barrels of reserves divided by barrels per year gives an answer in years.) The typical answer was ten years. Each of these forecasters started screaming that the US petroleum industry would die in ten years. They cried "wolf" During each ensuing ten years, more oil reserves were added, and the industry actually grew instead of drying up. In 1956, many critics thought that Hubbert was yet another false prophet. Up through 1970, those who were following the story divided into pro-Hubbert and anti-Hubbert factions. One pro-Hubbert publication had the wonderful title "This Time the Wol!
f Really Is at the Door". <3>

Hubbert's 1956 analysis tried out two different educated guesses for the amount of US oil that would eventually be discovered and produced by conventional means: 150 billion and 200 billion barrels. He then made plausible estimates of future oil production rates for each of the two guesses. Even the more optimistic estimate, 200 billion barrels, led to a predicted peak of US oil production in the early 1970s. The actual peak year turned out to be 1970.

Today, we can do something similar for world oil production. One educated guess of ultimate world recovery, 1.8 trillion barrels, comes from a 1997 country-by-country evaluation by Colin J Campbell, an independent oil-industry consultant. <4> In 1982, Hubbert's last
published paper contained a world estimate of 2.1 trillion barrels. <5> Hubbert's 1956 method leads to a peak year of 2001 for the 1.8-trillion-barrel estimate and a peak year of 2003 or 2004 for 2.1 trillion barrels. The prediction based on 1.8 trillion barrels makes a better match to the most recent ten years of world production.

In 1962, I became concerned that the US oil business might not be healthy by the time I was scheduled to retire. I was in no mood to move to Libya. My reaction was to get a photocopy of Hubbert's raw numbers; I made my own analysis using different mathematics. In my analysis, and in Hubbert's, the domestic oil industry would be down to half its peak size by 1998. Fortunately, universities were expanding rapidly in the post-Sputnik era, and I had no trouble moving into academe.

Hubbert's prediction was fully confirmed in the spring of 1971. The announcement was made publicly, but it was almost an encoded message. The San Francisco Chronicle contained this one-sentence item: "The Texas Railroad Commission announced a 100 percent allowable for next month". I went home and said, "Old Hubbert was right". It still strikes me as odd that understanding the newspaper item required knowing that the Texas Railroad Commission, many years earlier, had been assigned the task of matching oil production to demand. In essence, it was a government-sanctioned cartel. Texas oil production so dominated the industry that regulating each Texas oil well to a percentage of its capacity was enough to maintain oil prices. The Organization of Petroleum Exporting Countries (OPEC) was modeled after the Texas Railroad Commission. <6> Just substitute Saudi Arabia for Texas.

With Texas, and every other state, producing at full capacity from 1971 onward, the United States had no way to increase production in an emergency. During the first Middle East oil crisis in 1967, it was possible to open up the valves in Ward and Winkler Counties in west
Texas and partially make up for lost imports. Since 1971, we have been dependent on OPEC.

After his prediction was confirmed, Hubbert became something of a folk hero for conservationists. In contrast to the hundreds of millions of years it took for the world's oil endowment to accumulate, most of the oil is being produced in 100 years. The short bump of oil exploitation on the geologic time line became known as "Hubbert's peak".

In Chapter 7, I explain how Hubbert used oil production and oil reserves to predict the future. We scientists don't like to admit it, but we often guess at the answer and then gather up some numbers to support the guess. A certain level of honesty is required; if the numbers do not justify my guess, I don't fake the numbers. I generate another guess. Hubbert's oil prediction was just barely within the envelope of acceptable scientific methods. It was as much an inspired guess as it was hard-core science.

This cautionary note is needed here: in the late 1980s there were huge and abrupt increases in the announced oil reserves for several OPEC nations. <7> Oil reserves are a vital ingredient in Hubbert's analysis. Earlier, each OPEC nation was assigned a share of the oil market based on the country's annual production capacity. OPEC changed the rule in the 1980s to consider also the oil reserves of each country. Most OPEC countries promptly increased their reserve estimates. These increases are not necessarily wrong; they are not necessarily fraudulent. "Reserves" exist in the eye of the beholder.

Oil reserves are defined as future production, using existing technolology, from wells that have already been drilled (not to be confused with the US "strategic petroleum reserve", which is a storage facility for oil that has already been produced). Typically, young petroleum engineers unconsciously tend to underestimate reserves. It's a lot more fun to go into the boss's office next year and announce that there is actually a little more oil than last year's estimate. Engineers who have to downsize their previous reserve estimates are the first to leave in the next corporate downsizing.

The abrupt increase in announced OPEC reserves in the late 1980s was probably a mixture of updating old underestimates and some wishful thinking. A Hubbert prediction requires inserting some hard, cold reserve numbers into the calculation. The warm fuzzy numbers from OPEC probably give an overly optimistic veiw of future oil production. So who is supposed to know?

A firm in Geneva, Switzerland, called Petroconsultants, maintained a huge private database. One long-standing rumor said that the US Central Intelligence Agency was Petroconsultants' largest client. I would hope that between them, the CIA and Petroconsultants had inside information on the real OPEC reserves. This much is known: the loudest warnings about the predicted peak of world oil production came from Petroconsultants. <8> My guess is that they were using data not available to the rest of us.

A permanent and irreversible decline in world oil production would have both economic and psychological effects. So who is paying attention? The news media tell us that the recent increases in energy prices are caused by an assortment of regulations, taxes, and distribution problems. During the election campaign of 2000, none of the presidential candidates told us that the sky was about to fall. The public attention to the predicted oil shortfall is essentially zero.

In private, the OPEC oil ministers probably know about the articles in Science, Nature, and Scientific American. Detailed articles, with contrasting opinions, have been published frequently in the Oil and Gas Journal <9> Crude oil prices have doubled in the past year. I suspect that OPEC knows that a global oil shortage may be only a few years away. The OPEC countries can trickle out just enough oil to keep the world economies functioning until that glorious day when they can market their remaining oil at mind-boggling prices.

It is not clear whether the major oil companies are facing up to the problem. Most of them display a business-as-usual facade. My limited attempts at spying turned up nothing useful. A company taking the 2004-8 hypothesis seriously would be willing to pay top dollar for
existing oil fields. There does not seem to be an orgy of reserve acquisitions in progress.

Internally, the oil industry has an unusual psychology. Exploring for oil is an inherently discouraging activity. Nine out of ten exploration wells are dry holes. Only one in a hundred exploration wells discovers an important oil field. Darwinian selection is involved: only the incurable optimists stay. They tell each other stories bout a Texas county that started with thirty dry holes yet the next well was a major discovery. "Never is heard a discouraging word". A permanent drop in world oil production beginning in this decade is definitely a discouraging word.

Is there any way out? Is there some way the crisis could be averted?

NEW TECHNOLOGY. One of the responses in the 1980s was to ask for a double helping of new technology. Here is the problem: before 1995 (when the era began), the oil industry earned a higher rate of return on invested capital than any other industry. When oil companies tried to use some of their earnings to diversify, they discovered that everything else was less profitable than oil. Their only investment option was doing research to make their own exploration and production operations even more profitable. Billions of dollars went into petroleum technology development, and much of the work was successful. That makes it difficult to ask today for new technology. Most of those wheels have already been invented.

DRILL DEEPER. The next chapter of this book explains that there is an "oil window" that depends on subsurface temperatures. The rule of thumb says that temperatures 7,500 feet down are hot enough to "crack" organic-rich sediments into oil molecules. However, beyond 15,000
feet the rocks are so hot that the oil molecules are further cracked into natural gas. The range from 7,000 to 15,000 feet is called the "oil window". If you drill deeper than 15,000 feet, you can find natural gas but little oil. Drilling rigs capable of penetrating to 15,000 feet became available in 1938.

DRILL SOMEPLACE ELSE. Geologists have gone to the ends of the Earth in their search for oil. The only rock outcrops in the jungle are in the banks of rivers and streams; geologists waded up the streams picking leeches off their legs. A typical field geologist's comment about jungle, desert, or tundra was: "She's medium-tough country". As an example, at the very northernmost tip of Alaska, at Point Barrow, the United States set up Naval Petroleum Reserve #4 in 1923. <10> As early as 1923, somebody knew that the Arctic Slope of Alaska would be a major oil producer.

Today, about the only promising petroleum province that remains unexplored is part of the South China Sea, where exploration has been delayed by a political problem. International law divides oil ownership at sea along lines halfway between the adjacent coastlines. A valid claim to an island in the ocean pushes the boundary out to halfway between the island and the farther coast. It apparently does not matter whether the island is just a protruding rock with every third wave washing over the rock. Ownership of that rock can confer title
to billions of barrels of oil. You guessed it: several islands stick up in the middle of the South China Sea, and the drilling rights are claimed by six different countries. Although the South China Sea is an attractive prospect, there is little likelihood that it is another Middle East.

SPEED UP EXPLORATION. It takes a minimum of ten years to go from a cold start on a new province to delivery of the first oil. One of the legendary oil finders, Hollis Hedberg, explained it in terms of "the story". When you start out in a new area, you want to know whether the oil is trapped in folds, in reefs, in sand lenses, or along faults. You want to
know which are the good reservoir rocks and which are the good cap rocks. The answers to those questions are "the story". After you spend a few years in exploration work and drilling holes, you figure out "the story". For instance, the oil is in fossil patch reefs. Then pow, pow, pow - you bring in discovery after discovery in patch reefs. Even then,
there are development wells to drill and pipelines to install. It works, but it takes ten years. Nothing we initiate now will produce significant oil before the 2004-8 shortage begins.

To summarize: it looks as if an unprecedented crisis is just over the horizon. There will be chaos in the oil industry, in governments, and in national economies. Even if governments and industries were to recognize the problems, it is too late to reverse the trend. Oil production is going to shrink. In an earlier, politically incorrect era the scene would be described as a "Chinese fire drill".

What will happen to the rest of us? In a sense, the oil crises of the 1970s and 1980s were a laboratory test. We were the lab rats in that experiment. Gasoline was rationed both by price and by the inconvenience of long lines at the gas stations. The increased price of gasoline and diesel fuel raised the cost of transporting food to the grocery store. We were told that ninety percent of an Iowa corn farmer's costs were, directly and indirectly, fossil fuel costs. As price rises rippled through the economy, there were many unpleasant disruptions.

Everyone was affected. One might guess that professors at Ivy League universities would be highly insulated from the rough-and-tumble world. I taught at Princeton from 1967 to 1997; faculty morale was at its lowest in the years around 1980. Inflation was raising the cost of living far faster than salaries increased. Many of us lived in university-owned apartments, and the university was raising our apartment rents in step with an imaginary outside "market" price. Our real standard of living went progressively lower for several years in a
row. That was life (with tenure) inside the sheltered ivory tower; outside it was much tougher.

What should we do? Doing nothing is essentially betting against Hubbert. Ignoring the problem is equivalent to wagering that world oil production will continue to increase forever. My recommendation is for us to bet that the prediction is roughly correct. Planning for increased energy conservation and designing alternative energy sources should begin now to make good use of the few years before the crisis actually happens.

One possible stance, which I am not taking, says that we are despoiling the Earth, raping the resources, fouling the air, and that we should eat only organic food and ride bicycles. Guilt feelings will not prevent the chaos that threatens us. I ride a bicycle and walk a lot, but I confess that part of my motivation is the miserable parking situation in Princeton. Organic farming can feed only a small part of the world population; the global supply of cow dung is limited. A better civilization is not likely to arise spontaneously out of a pile of guilty consciences. We need to face the problem cheerfully and try to cope with it in a way that minimizes problems in the future.

The substance of this book is an explanation of the origin, exploration, production, and marketing of oil. This background about the industry is important because it sets geologic constraints on our future options. I describe the strengths and weaknesses of Hubbert's
prediction methods and end with some suggestions about preparing for the inevitable. My intention is to give the reader some expertise in evaluating the problems. The experts' scenario for 2004-8 reads like the opening passage of a horror movie. You have to make up your own mind about whether to accept their scary account.

My own opinion is that the peak in world oil production may even occur before 2004. What happens if I am wrong? I would be delighted to be proved wrong. It would mean that we have a few additional years to reduce our consumption of crude oil. However, it would take a lot of unexpectedly good news to postpone the peak to 2010. My message would remain much the same: crude oil is much too valuable to be burned as a fuel.

Stephen Jay Gould is fond of pointing out that we all have difficulty rising above our cultural biases. ("All" in that sentence includes Gould.) It helps if I identify the roots of my biases. Here is my confession:

I was born in the middle of the Oklahoma City oil field. I grew up in the oil patch. My father, J A "Dee" Deffeyes, was a pioneering petroleum engineer. In those days, companies moved employees around wherever they were needed. I went to nine different grade schools getting through the first eight grades. During high school and college, each summer I had a different job in the oil industry: laboratory assistant, pipeyard worker, roustabout, seismic crew.

After I graduated from the Colorado School of Mines, I worked for the exploration department of Shell Oil. Right at the end of the Korean War, everybody my age got drafted. There weren't many of us. I was one of the few born right in the pit bottom of the Great Depression. I wanted to have my revenge on the army by using up my GI Bill at the most expensive school I could find. The geology department at Princeton turned out to be fabulous.

After graduate school, I was delighted to be asked to rejoin Shell at its research lab in Houston. Scientific progress happened very rapidly at the Shell lab. Jerry Wasserburg of Cal Tech, not known for passing out compliments freely, said that the Shell research lab in that era was the best Earth science research organization in the world. As I mentioned, it was Hubbert's prediction that caused me to get out of the oil business.

I taught briefly at Minnesota and Oregon State, and in 1967 I joined the Princeton faculty. In addition to teaching, I had the pleasure of cooperating with John McPhee as he wrote his books on geology. <11> The "oil boom" of the 1970s and early 1980s gave me a chance to participate in the industry again. As a consultant, I advised programs that drilled for natural gas across western New York and northern Pennsylvania. My programs drilled 100 successful gas wells without a dry hole; one of them was the largest gas well in the history of New York State. <12> I also served as an expert witness in oil litigation.

You don't outgrow your roots. As I drive by those smelty refineries on the New Jersey Turnpike, I want to roll the windows down and inhale deeply. But in all the years that I worked in the petroleum industry, I never came to identify with the management. I'm a worker
bee, not a drone.

A couple of years ago, I was testing a new treatment on an oil well in northern Pennsylvania. I picked up a pipe wrench with a 36-inch handle and helped revise the plumbing around the wellhead. I was home again; I loved it.


1. Hubbert, M K (1956), "Nuclear Energy and the Fossil Fuels", American Petroleum Institute Drilling and Production Practice, Proceedings of Spring Meeting, San Antonio, 1956, pages 7-25; also Shell Development Company Publication 95, June 1956.

2. Hatfield, C B (1997), "Oil Back on the Global Agenda", Nature 387:121; Kerr, R A (1998), "The Next Oil Crisis Looms Large-And Perhaps Close", Science 281:1128-31; Campbell, C A, and Laherrere J H (1998), "The End of Cheap Oil", Scientific American, March: 78-83.

3. Akins, J E (1973), "The Oil Crisis: This Time the Wolf Is Here", Foreign Affairs, April 1973.

4. Campbell, C J (1997), The Coming Oil Crisis, Multi-Science Publishing Company and Petroconsultants. His world scenario is on page 201.

5. Hubbert, M K (1981), "The World's Evolving Energy System", American Journal of Physics 49:1007-29.

6. Yergin, Daniel (1991), The Prize, New York: Simon & Schuster. The founding of OPEC is described beginning on page 522.

7. In this book, the reserves and production data are those reported by the Oil and Gas Journal in its last issue of each year. Campbell's evaluation of OPEC reserve increases is on page 73 in his Coming Oil Crisis.

8. Colin Campbell and Sam Carmalt have worked for, or with, Petroconsultants. In 1998, Petroconsultants was merged into the IHS Energy Group,

9. One of the best critical rejections of Hubbert's approach is Adelman, M A, and Lynch M C (1997), "Fixed View of Resource Limits Creates Undue Pessimism", Oil and Gas Journal, April 7:56-60. Other comments are in Oil and Gas Journal, February 23 1998:77, and November 2 1998:94.

10. As explained in the next chapter, David White of the US Geological Survey was a pioneer in interpreting the origins of oil. In 1929, White published "Description of Fossil Plants Found in Some 'Mother Rocks' of Petroleum from Northern Alaska", American Association of Petroleum Geologists Bulletin 13:841-48.

11. McPhee, John (1998), Annals of the Former World, New York: Farrar, Straus, and Giroux. This is a revised printing of four earlier books.

12. McCaslin, J C (1984), "Well Completions Boost Impact in New York Area", Oil and Gas Journal, March 5:121-22.

Please also see "Charts of the Coming Global Oil Crisis" at (September 06 2004)

To post a comment, or to read comments posted by others,
please click on the word "comment" highlighted at the end of
the version of this essay posted at:

Bill Totten