Predicting the timing of peak oil involves estimation of future production from existing oil fields as well as future discoveries. The initial production model was Hubbert peak theory, first proposed in the 1950s. Since then, many experts have tried to forecast peak oil.
Present range of predictions As of 2024, the International Energy Agency predicts that peak oil will happen by 2030, while the US Energy Information Administration forecasts a peak in 2050 and the OPEC does not see a peak in oil demand before 2050.
Past predictions
1880s–1940s
The idea that human use of petroleum faces sustainability limits attracted practical concern at least as early as the 1880s, as did the related idea that the timing of those limits depends on the extraction technology. The concept of exhausting a natural resource to a point of diminishing returns had some antecedent examples. During the same decades when the modern petroleum industry was launching, the New England whale oil industry had just experienced a peak and was grappling with decline.
Economist and oil analyst Daniel Yergin notes that the first predictions of imminent oil peaks go back to the 1880s, when some American experts believed that exhaustion of the Pennsylvania oil fields would kill the US oil industry. Another wave of peak predictions occurred after World War I.Daniel Yergin, "There will be oil," Wall Street Journal, 17 Sept. 2011.
"... the peak of production will soon be passed, possibly within 3 years. ... There are many well-informed geologists and engineers who believe that the peak in the production of natural petroleum in this country will be reached by 1921 and who present impressive evidence that it may come even before 1920." - David White, chief geologist, United States Geological Survey (1919)David White, "The unmined supply of petroleum in the United States," Transactions of the Society of Automotive Engineers, 1919, v.14, part 1, p.227.
"The average middle-aged man of today will live to see the virtual exhaustion of the world's supply of oil from wells," - Victor C. Anderson, president of the Colorado School of Mines (1921)"The present status of the oil shale industry" 1921, Colorado School of Mines Quarterly, Apr. 1921, v.16 n.2 p.12.
A correspondent named W.D. Hornaday, quoting oil industry executive J.S. Cullinan, described the concerns in a 1918 article for Tractor and Gas Engine Review titled "Petroleum consumption enormous." The article said, "There has been considerable discussion of late as to the possible length of time that the petroleum supply of the United States and the world will hold out." The article quoted Cullinan as saying, "It is just possible, so far as the United States is concerned, that the development and the exhaustion of the supplies may occur within the course of one human life. It is certain that unless radical changes from present methods are applied promptly, all sources of supply within the range of known drilling methods will be exhausted during the life of your children and mine." It turned out that radical changes from 1910s drilling methods were, in fact, applied promptly, and thus, the predicted timeframe was premature; but the underlying concerns (that the vastness of consumption would lead to shortages soon enough to worry about, regardless of the exact decade) did not disappear.
Hubbert's model
In 1956, M. King Hubbert created and first used the models behind peak oil to predict that United States oil production would peak between 1965 and 1971.
In 1956, Hubbert calculated that the world held an ultimate cumulative of 1.25 trillion barrels, of which 124 billion had already been produced. He projected that world oil production would peak at about 12.5 billion barrels per year, sometime around the year 2000. He repeated the prediction in 1962.M. King Hubbert, 1962, Energy Resources, National Academy of Sciences - National Research Council, Publication 1000-D, p.75 World oil production surpassed his predicted peak in 1967 and kept rising; world oil production did not peak on or near the year 2000, and for the year 2012 was 26.67 billion barrels, more than twice the peak rate Hubbert had projected back in 1956.OPEC, OPEC Statistical Bulletin , 2013. OPEC world oil production is significantly lower than that calculated by the US Energy Information Administration, because OPEC does not include surface-mined oil, such as Canadian heavy oil sand production.
Hubbert's 1956 peak projection for the United States depended on geological estimates of ultimate recoverable oil resources, but starting in his 1962 publication, he concluded that ultimate oil recovery was an output of his mathematical analysis, rather than an assumption. He regarded his peak oil calculation as independent of reserve estimates.M. King Hubbert, 1962, "Energy Resources," National Academy of Sciences, Publication 1000-D, p.60.M. King Hubbert, "National Academy of Sciences Report on Energy Resources: reply," AAPG Bulletin, Oct. 1965, v.49 n.10 p.1720-1727.M. King Hubbert, "Degree of advancement of petroleum exploration in United States," AAPG Bulletin, Nov. 1967, v.51 n.11 p.2207-2227.
In 1956, Hubbert confined his peak oil prediction to that crude oil "producible by methods now in use." By 1962, however, his analyses included future improvements in exploration and production.M. King Hubbert, Energy Resources(Washington: National Academy of Sciences, 1962)Publication 1000-D, p. 73–75. All of Hubbert's analyses of peak oil specifically excluded oil manufactured from oil shale or mined from oil sands. A 2013 study predicting an early peak excluded deepwater oil, tight oil, oil with API gravity less than 17.5, and oil close to the poles, such as that on the North Slope of Alaska, all of which it defined as non-conventional.
In 1962, Hubbert predicted that world oil production would peak at a rate of 12.5 billion barrels per year, around the year 2000. In 1974, Hubbert predicted that peak oil would occur in 1995 "if current trends continue". Those predictions proved incorrect.
In 1974, Hubbert again predicted that world peak oil would occur near 2000, this time in 1995 "if current trends continue." However, in the late 1970s and early 1980s, global oil consumption actually dropped (due to the shift to energy-efficient cars, the shift to electricity and natural gas for heating, and other factors), then rebounded with a lower rate of growth in the mid 1980s. Thus oil production did not peak in 1995, and has climbed to more than double the rate initially projected.
2000s In 2001, Kenneth S. Deffeyes, professor emeritus of geology at Princeton University, used Hubbert's theory to predict that world oil production would peak about 2005, with a possible range of 2003 to 2006. He used the observed growth of production plus reserves to calculate ultimate world oil production of 2.12 trillion barrels, noting: "No educated guesses go in." He considered the application of new technology, but wrote: "This much is certain: no initiative put in place starting today can have a substantial effect on the peak production year." His final conclusion was: "There is nothing plausible that could postpone the peak until 2009. Get used to it."Kenneth S. Deffeyes, Hubbert's Peak: The Impending World Oil Shortage (Princeton, New Jersey, Princeton University Press, 2001). As of late 2009, Deffeyes was still convinced that 2005 had been the peak, and wrote: "I think it unlikely that oil production will ever climb back to the 2005 levels."Kenneth S. Deffeyes, "Join us as we watch the crisis unfolding", When Oil Peaked website, 13 Nov. 2009.
The term "peak oil" was popularized by Colin Campbell and Kjell Aleklett in 2002 when they helped form the Association for the Study of Peak Oil and Gas (ASPO). Hubbert used the terms "peak production rate" and "peak in the rate of discoveries" in his publications.
According to Matthew Simmons, former chairman of Simmons & Company International and author of Twilight in the Desert: The Coming Saudi Oil Shock and the World Economy, "peaking is one of these fuzzy events that you only know clearly when you see it through a rear view mirror, and by then an alternate resolution is generally too late." On October 26, 2006 Simmons said that global oil production may have peaked in December 2005, though he cautioned that further monitoring of production is required to determine if a peak has actually occurred.
Phibro statistics show that major oil companies hit peak production in 2005.Oil and Economic Growth: A Supply-Constrained View Fatih Birol, chief economist at the International Energy Agency, stated in 2011 that "crude oil production for the world has already peaked in 2006."
Several sources in 2006 and 2007 predicted that worldwide production was at or past its maximum. However, in 2013 OPEC's figures showed that world crude oil production and remaining proven reserves were at record highs.OPEC Annual Statistical Bulletin , 2014.
In a 2006 analysis of Hubbert theory, it was noted that uncertainty in real world oil production amounts and confusion in definitions increases the uncertainty in general of production predictions. By comparing the fit of various other models, it was found that Hubbert's methods yielded the closest fit overall but none of the models were very accurate. In 1956 Hubbert himself recommended using "a family of possible production curves" when predicting a production peak and decline curve. The July 2007 IEA Medium-Term Oil Market Report projected a 2% non-OPEC liquids supply growth in 2007-2009, reaching in 2008, receding thereafter as the slate of verifiable investment projects diminishes. They refer to this decline as a plateau. The report expects only a small amount of supply growth from OPEC producers, with 70% of the increase coming from Saudi Arabia, the UAE, and Angola as security and investment issues continue to impinge on oil exports from Iraq, Nigeria and Venezuela.
In October 2007, the Energy Watch Group, a German research group founded by MP Hans-Josef Fell, released a report claiming that oil production peaked in 2006 and would decline by several percent annually. The authors predicted negative economic effects and social unrest as a result. They stated that the IEA production plateau prediction uses purely economic models, which rely on an ability to raise production and discovery rates at will.
Sadad Ibrahim Al Husseini, former head of Saudi Aramco's production and exploration, stated in an October 29, 2007 interview that oil production had likely already reached its peak in 2006, and that assumptions by the IEA and EIA of production increases by OPEC to over are "quite unrealistic." Data from the United States Energy Information Administration show that world production leveled out in 2004, and an October 2007 retrospective report by the Energy Watch Group concluded that this data showed the peak of conventional oil production in the third quarter of 2006.
ASPO predicted in their January 2008 newsletter that the peak in all oil (including non-conventional sources), would occur in 2010. This is earlier than the July 2007 newsletter prediction of 2011. ASPO Ireland in its May 2008 newsletter, number 89, revised its depletion model and advanced the date of the peak of overall liquids from 2010 to 2007.
Texas alternative energy activist and oilman T. Boone Pickens stated in 2005 that worldwide conventional oil production was very close to peaking. On June 17, 2008, in testimony before the U.S. Senate Energy and Natural Resources Committee, Pickens stated that "I do believe you have peaked out at 85 million barrels a day globally."