The Myth of Peak Oil

Each time there is a short-term shortage of oil or the price begins to rise, there is talk of running out of affordable oil, an idea captured by the concept of Peak Oil. Peak Oil is the theoretical point when the maximum rate of oil production is reached and after that time enters into a terminal decline. There is a lot of debate surrounding the Peak Oil theory, with some observers predicting rapid decline in oil production with serious implications for our entire economy and society.

No name is more closely associated with the concept of Peak Oil than geologist Marion King Hubbert.  Hubbert was a research geologist for Shell Oil Company and later the US Geological Service.  Hubbert is credited with developing a quantitative technique (Logistic Growth Curve) now commonly referred to as the Hubbert Curve, which he suggested could be used to predict the remaining oil supplies (or any other finite resource like gas, copper, etc.) and the time of eventual depletion.

In the 1956 meeting of the American Petroleum Institute in San Antonio, Texas,  Hubbert  presented a paper titled Nuclear Energy and Fossil Fuels where he suggested that overall petroleum production would peak in the United States between the late 1960s and the early 1970s. Since US oil production did indeed appear to peak in 1970, many Peak Oil advocates acclaim Hubbert as a prophet. However, an apparent peak in production does not necessarily represent a peak in oil availability, especially in a global market—something that Peak Oil advocates tend to overlook. In fact, a “peak” may just be one of many “spikes”.

Another point of confusion in the debate over the ultimate availability of oil and gas supplies is the question of “unconventional” fossil fuel sources like tar sands, oil shales, heavy oils, and shale oil. Hubbert did not include these other energy types in his estimates and many of the proponents of Peak Oil today tend to ignore these hydro-carbon sources.  However, since there is vastly more oil (and gas) found in these “unconventional” sources compared to “conventional” crude oil and traditional gas sources, the exclusion of them from any policy debate over oil’s demise leads to serious misrepresentation of our ultimate fossil fuel availability.

As Hubbert wrote in his paper, “if we knew the quantity (of some resource) initially present, we could draw a family of possible production curves, all of which would exhibit the common property of beginning and ending at zero, and encompassing an area equal to or less than the initial quantity.” In theory, Hubbert’s basic concept is sound. As a way of thinking about and approaching the issue of declining finite resources, Hubbert was a pioneer.  But that does not mean his predictions were accurate.

The problem for anyone trying to predict future resource availability is discerning the initial starting amount of a resource such as oil when one cannot readily see or gauge accurately the resource.  This lack of transparency presents huge opportunities for error, in particular, erring on the side of under estimation of the total resource.  And time has consistently shown that under estimation of total resource is the most common error, and as we shall see this is exactly the error that Hubbert made with regards to his estimates of our remaining oil and gas reserves.  Hubbert can be forgiven because new technology can make previously unavailable resources accessible, even less expensive to exploit. In fact, he even anticipated this to a degree in his paper, another point that Hubbert’s admirers today tend to overlook.


Few that credit Hubbert with a successful prediction have apparently actually read his paper. A reading of his presentation demonstrates that Hubbert grossly underestimated total oil supplies, and thus his predicted high point of the bell curve deviates significantly from reality. Indeed, there is good evidence we haven’t even reached the top of the bell curve, much less past it in 1970. He did not anticipate things like the discovery of oil in Alaska’s Prudhoe Bay or shale oil like the North Dakota Bakken Formation,  among many other oil discovery that have significantly changed total oil supplies.

And because US oil production did peak in 1970, the same time period which Hubbert suggested oil reserves would reach their half-way point and start an inevitable decline, few bothered to ask whether the observed decline in US production might have any other explanation other than declining geological petroleum stocks as Peak Oil advocates suggest.

Predicting future oil and gas supplies is fraught with dangers. Many factors influence oil extraction other than geological limits. A rapid shift to renewable energy, a decline in global economies, new technological innovation, energy conservation, a high oil price that dampens consumer demand, political instability and wars all significantly affects energy production, thus when and how “peak” is achieved. Many believe a more realistic model rather than a bell curve is a rapid run up in production to a spike or series of spikes followed by a long drawn out plateau and production decline with ultimately more oil production occurring after the apparent peak, but less rapidly than prior to the “peak” which of course wouldn’t really be a peak in the traditional sense of the word.


The first problem with Hubbert’s prediction is that his estimates of total oil and gas reserves are far too low. If the starting amount of reserves are low, than the top of the bell curve is reached much sooner than if there are greater amounts of oil–assuming that a bell curve actually represents what is  occurring–which many people dispute. Some suggest Hubbert just drew the curve to fit his assumptions.

In his paper, Hubbert estimated that the “ultimate potential reserve of 150 billion barrels of crude oil for both the land and offshore areas of the United States.”  Hubbert’s estimate was based on the crude oil “initially present which are producible by methods now in use.”  Using the 150 billion barrel estimate he predicted US Peak Oil occurring in 1965. But to be cautious, he also used a slightly higher figure of 200 billion barrels which produced a peak in oil production around 1970—the figure that Hubbert advocates like to use to demonstrate that Hubbert was prophetic in his predictions.  However, by 2006 the Department of Energy estimated that domestic oil resources still in the ground (in-place) total 1,124 billion barrels.  Of this large in-place resource, 400 billon barrels is estimated to be technically recoverable with current technology.

This estimate was produced before horizontal drilling and hydraulic fracturing or fracking techniques were widely adopted which most authorities believe will yield considerably more oil than was thought to be recoverable in 2006.

Going back to Hubbert’s paper we find that he predicted that by 1970 the US should have consumed half or about 100 million barrels of oil of the original endowment of 150-200 billion barrels of recoverable oil. And by his own chart on page 32 of his paper if we use the assumption of 200 billion barrels as the total potential oil reserves of the US we should be completely out of oil by now. According to his curve and graph, by year 2000 we should have had only around 27 billion or so barrels of oil left in the US and fallen to zero sometime in the mid-2000s.

Yet the US government estimates as of 2007 that our remaining technically recoverable reserves are 198 billion barrels, and this excludes oil that may be found in area that are off limits to drilling (i.e. like most of the Continental Shelf).

And there are another 400 billion barrels that some suggest could be recovered with new methods (which itself is a subset of total in place oil which future technology may make available at an affordable price).

Obviously if Hubbert were correct, and we had reached Peak Oil in 1970 (point where we had consumed half of our oil) and we started out with only 200 billion, we could not have nearly 200-400 billion still left to extract—and total resources are likely even higher than this figure.

It’s also important to keep in mind that “technologically recoverable” resources are not the “total” amount of oil thought to exist in the US, so the total in-place reserves are much, much larger. It does not take a lot of imagination to predict that many of these oil resources will eventually be unlocked with new technological innovation thus added to the total “proven reserves.”

Another example of his under-estimation of oil is US off-shore oil. In his 1956 paper, Hubbert suggests we had 15 billion total barrels, but the US government now estimates there is closer to 90 billion barrels of oil left off-shore–and we have already extracted quite a bit. (I’m not sure if that figure is just for off -shore currently open to exploration or all off shore–since oil exploration is banned on 83% of the US coastline. If this figure refers only to those areas currently available to drill–then the number may be quite a bit higher if all off shore areas were opened to oil extraction).

Hubbert was even farther off in his estimate for global oil reserves, which is not surprising since in 1956 very few parts of the world had been adequately studied.  In his 1956 paper Hubbert  wrote that there was “about 1250 billion barrels for the ultimate potential reserves of crude oil of the whole world.” In his paper he estimated that the entire Middle East including Egypt had no more than 375 billion barrels of oil. Yet by 2010, the Central Intelligence Agency (CIA) estimated that just the “proven reserves” in Saudi Arabia alone totaled 262.6 billion barrels.  Similarly in his paper Hubbert uses an estimate of 80 billion barrels for all of South America, yet Venezuela has 296 billion barrels of proven reserves.

By 2000, the point when Hubbert estimated that we would reach global Peak Oil we would have only around 625 billion barrels of oil left. Just the 558 billion barrels of proven reserves known to exist in Saudi Arabia and Venezuela alone (and a lot more in-place resources) is nearly equal the total global oil supplies that Hubbert estimated would remain in global reserves.  Obviously once again Hubbert’s global estimates were way too low.

The world has already burned through more than a trillion barrels of oil, clearly demonstrating how far off his prediction of oil supplies were. The estimated “proven reserves” left globally are today more than 1.3 trillion for the top 17 oil producing countries alone.


Part of the confusion in the Peak Oil debate is that people, agencies and organizations use different definitions and accounting methods that are often not explicitly acknowledged. For instance, most Peak Oil advocates rely upon “proven reserve” numbers to argue we have limited oil supplies remaining. However, it is important to note the term “proven reserves” has a very precise meaning that only includes oil that has a 90% certainty that the oil can be extracted using current technology at current price. It does not represent total oil that may over time be produced. The total estimated amount of oil in an oil reservoir, including both producible and non-producible oil, is called various terms including oil in place. Due to technological, political and other limitations, only a small percentage of the total “in place” oil can be extracted at the present time.  However, proven reserves are the bare minimum amount of oil that reasonably can be expected to be extracted over time.

One of the wild cards in predicting oil reserves is the recovery factor. Recovery factors vary greatly among oil fields. Most oil fields to this point have only given up a fraction of their potential oil holdings—between 20-40%.     By 2009 the average Texas oil field had only about a third of its oil extracted, leaving two-thirds still in the ground.     Using Enhanced Oil Recovery (EOR) techniques, many of them not even available when Hubbert wrote his paper, recovery can often be boosted to 40-60%. In essence if EOR were applied to many of the larger US oil fields, we could effectively double the oil extracted, hence “proven reserves.”

Even Hubbert recognized that we may eventually extract more oil from existing fields, though he still underestimated the effect of new discoveries and new technology.   Hubbert wrote ”… only about a third of the oil underground is being recovered. The reserve figures cited are for oil capable of being extracted by present techniques. However, secondary recovery techniques are gradually being improved so that ultimately a somewhat larger but still unknown fraction of the oil underground should be extracted than is now the case. Because of the slowness of the secondary recovery process, however, it appears unlikely that any improvement that can be made within the next 10 or 15 years can have any significant effect upon the date of culmination. Amore probable effect of improved recovery will be to reduce the rate of decline after culmination…..”

While no one realistically believes it’s possible to get every last drop of oil from an oil reservoir, new technologies are often able to get significantly more oil from existing fields than was possible in the past. The important fact is that the recovery factor often changes over time due to changes in technology and economics. Since the bulk of global oil still remains in the ground, and any shift upward in price and improvement in technology suddenly makes it profitable to exploit reserves that were previously not included in the “proven reserves” estimate. Thus proven reserve estimates are a minimum, not the maximum amount of oil available.

To demonstrate how technology and price can affect “proven reserves” estimates, just a few years ago Canada’s “proven reserves” of oil were only 5 billion barrels. Today, due to higher prices and improved technology that makes tar sands production economically feasible; Canada now has “proven” reserves of 175 billion barrels of oil. Nothing changed other than the price of oil and the technology used to extract it. Oil companies knew there was a lot of oil in the tar sands, but it took a change in technology and price to move it into the “proven reserves” category.  Even more telling is that the total minimum estimate of in place oil for the tar sands exceeds 1.3 trillion barrels of oil. Keep in mind that 1.3 trillion barrels is more oil than Hubbert thought existed in the entire world when he presented his 1956 paper.

People knew all along there were tremendous amounts of oil locked in Alberta’s tar sands.  But it took a change in price, along with some technological innovation to make it profitable for extraction. So proven reserves are not a static figure based on geology, rather it reflects economics and technology. Unfortunately too many writing about the presumed Peak of oil in the United States appear to ignore the distinction, and regularly use the “proven reserves” figure as if it were the ultimate geological limit on oil and/or gas supplies.

Although the major point of his paper was the potential depletion of traditional oil and gas reservoirs, he did mention “unconventional oil.” Unconventional oil reserves are oil or hydrocarbons found in geological formations other than a traditional oil reservoir. Examples of unconventional oil include  Alberta’s tar sands, oil shales of the Green River Basin of Colorado, Utah, and Wyoming, the heavy oils of Venezuela,  and other non-traditional hydrocarbons. There are far more of hydro-carbons in these formations than traditional oil reservoirs—a fact that many Peak Oil advocates frequently ignore. Or if they acknowledge their existence, they dismiss them as uneconomical or technologically impossible to exploit and therefore will never make a significant contribution to global energy supplies.

Hubbert failed to appreciate the potential contribution of these unconventional sources of synthetic oil. For instance, he put the total for US oil shales at around a trillion barrels of oil equivalent. Recently the USGS estimated that the Green River drainage area of Colorado, Wyoming and Utah may contain as much as   4.2 trillion barrels of in place oil equivalent in oil shale deposits. To put this into context, the US currently consumes around 24 billion barrels of oil in 2010, so even if a fraction of these oil shales are exploited it will significantly increase available energy to the US.

With unconventional oils like tar sands, oil shales, heavy oils, etc. included, it seems we have huge amounts of potential energy–even acknowledging that much of that oil may not be extracted until some future date due to cost and/or lack of technology.


As he did with his estimates of oil, Hubbert also appears to have underestimated natural gas supplies as well. He put total natural gas supplies to be around 850 trillion cubic feet (TCF) and maximum US production would be 14 TCF annually.  The Energy Information Agency (EIA) estimates that shale gas reserve alone total  750 TCF and shale gas is only one source of natural gas.Total natural gas reserves are increasing. Estimates vary about total gas reserves, but they run between 1400 to 2000 TFC.  I see no reason to doubt these estimates.

If correct, then his estimate of natural gas was also a vast underestimate.  This link shows that gas supplies are increasing well into the future.   And new estimates for gas hydrates (methane locked in frozen ice) suggests there may be twice as much energy locked in these resources than all the coal, oil, and traditional natural gas supplies combined.  One estimate suggests there may be a 3000 plus year supply of natural gas in gas hydrates. Whatever the ultimate number may be, the important point is that we are not in any danger of running out of fossil fuels in the near future.


Was it just coincidence and luck that Hubbert picked 1970 as one of the possible peaks in US oil production even though his starting numbers were way too low?

This raises the question whether declining US production since 1970 is due to depletion of oil fields as asserted by Peak Oil advocates or whether economics explains it better. (This is not to deny that at some point we will see declining production due to real limits–the question of importance however is when that will occur).

Another explanation requires looking beyond the US. Keep in mind that oil is a commodity. Just because we may see a decline in production of some commodity does not mean we are running out of that substance or resource. The Northeast US was once the major producer of timber in the US. Today if you buy lumber in New England, there’s a good chance it was cut and shipped from the Pacific Northwest, not because there are no trees to cut in New England. Rather due to climate, vegetation, and infrastructure factors, it’s less expensive to cut trees in Oregon or British Columbia than to log New England forests.  It would be wrong to conclude that because New England imports most of its lumber that there are not enough trees left to provide wood locally.

Similarly attributing declining US oil production to geological depletion ignores the effect of global oil production. Immediately after WWii the US was easily the global leader in oil production. This dominance of global oil markets by US production and companies continued throughout the 1950s and 1960s. Then in the late 1960s and early 1970s oil production in other parts of the world began to increase substantially. In particular, Middle East oil production improved dramatically due to foreign investment and technology. For a variety of factors, once the oil infrastructure (pipelines, tanker ports, oil fields,) was built in these places, it became less expensive to import oil from Saudi Arabia, for example, than to build a new oil field in Wyoming or Texas.  Indeed in some cases producing oil wells in the US were capped and retired even though they were perfectly capable of producing more oil. Not only was oil production increasing in Saudi Arabia, but all over the world at this time including Venezuela, Mexico, and the Soviet Union. All of these new fields were producing lower cost oil than one could get from most US oil fields at the time. So could it be that US producers just decided it was a better business plan to invest in and/or buy oil from other oil producing countries? Did this low cost oil cause oil companies to import oil rather than invest in US oil production?

Worse for US producers, except for a few manufactured shortages like the 1973 oil crisis created by OPEC in response to US support for Israel or the War in Iraq, the abundance of relatively inexpensive oil kept oil prices depressed throughout the 1970s, 1980s, 1990s and into the early 2000s, discouraging new investment in US oil production.

It takes up to a decade or more to bring a new oil field on line, especially if the field is not located near other infrastructure.  For instance, Alaska’s Prudhoe Bay Oil field was discovered in 1968 and it wasn’t until 1978 before the first oil was sent to market.  Oil companies will only invest in major new production if they are certain that the prices are stable and will remain at a specific break-even point into the future. This lag time between changes in price or technology and significant production is why the oil industry cannot rapidly respond to short term price increases or politically created shortages.

Peak Oil advocates continuously point to the rise in oil prices during the latter part of the 2000s and suggest that an apparent lack of significant new oil production is due to depletion.  However, there is a time lag before higher prices result in a noteworthy increase in oil production.  Given the huge investments needed to bring on line new oil production, companies have to first wait for quite a number of years after an oil price hike before they start any new development to make sure that higher prices are going to stabilize, not rise and then fall suddenly as happened in 2008 when oil reached $145 a barrel then crashed to $30 a barrel. Such volatility does not lead to greater oil production.

Nevertheless, higher oil prices in the past few years have started to spur new development in the US and around the globe. The US, for instance, has reduced its import of foreign oil from 60% to 45% due to higher production at home as well as greater efficiency spurred by higher fuel prices.  These trends point to continued reduction in imports. However, because of the long delay between start up and full production, there is no quick relief.  This is one reason why “Drill, Baby, Drill” is a foolish response to any oil price increase.

From the oil producer’s perspective, there is no advantage in increasing spare production capacity. All this will do is flood the market (global market) with cheap energy. What company wants to reduce its profits by over production?  So far global oil production has largely been able to meet all demand, except for short term shortages as a result of political change, wars, and/or price speculation.  But none of these reflect a true geological short-fall or serious effect of depletion.

Despite Hubbert’s prediction that we would be just about out of oil by now, the US oil production (and gas) have both gone up in recent years. This is in response to higher prices and new technologies. But according to Hubbert this could not be occurring because we are long past our Peak and indeed, very near our bottom line for oil.

There is no doubt that a finite resource such as oil will continue to decline, and demand will likely grow at least into the foreseeable future, both of which should lead to higher fuel costs. But whether this leads to a long term chronic shortages that cause major economic disruption or even the collapse of civilization as some predict is subject to more uncertainty than perhaps some like to admit.  For one thing there is far more oil on the planet than most people recognize, and new technologies combined with rising price for fuels is spurring development of new oil supplies.  Rising prices also spurs shifts to other energy sources, as well as greater efficiency and conservation of energy.

Rather than running out of oil and/or gas any time soon, I think the bigger danger is that we have more than enough oil and other fossil fuel energy resources to sustain us for quite a few decades if not centuries. Any efficiency and/or conservation of energy, combined with some replacement of fossil fuel energy with renewables than these finite resources, will extend hydrocarbon resources quite a few additional decades.

The real problem for the planet and human society is not the imminent danger of running out of hydrocarbon fuels, but that an abundance of these energy sources will permit population and economic growth that will gradually diminish the planet’s biodiversity, degrade ecosystems, and disrupt global climate and other systems.

George Wuerthner is an ecologist. He is currently working on a book about energy.

George Wuerthner has published 36 books including Wildfire: A Century of Failed Forest Policy