Combined production of crude oil, gas and condensates in the United States is on course to a hit a record this year, passing the previous peak set in 1972.
The rise in output has confounded the famous forecast made by Shell geologist M King Hubbert who predicted that U.S. oil and gas production would peak in the 1970s and then decline.
Hubbert's prediction was contained in a paper entitled "Nuclear Energy and the Fossil Fuels," published in 1956, in which he argued the coming decline of oil and gas output would make the development of nuclear energy essential.
Hubbert's prediction of peaking oil and gas production came to be known as "Hubbert's peak" and spawned the popular and influential theory known as "peak oil".
The forecast was illustrated with a series of graphs, two of which are reproduced here (Charts 1 and 2).
For almost four decades between the 1970s and 2000s, Hubbert's prediction appeared to be frighteningly accurate.
U.S. oil production peaked at 3.5 billion barrels in 1970 and then declined steadily to just 1.8 billion barrels in 2008.
Gas production peaked shortly afterwards. Output levelled off rather than falling but remained below its 1973 peak for almost 40 years.
But since 2005, Hubbert's predictions have been spectacularly falsified.
Total petroleum output (which includes oil, gas and condensates) has surged from a low of just 5.15 billion barrels in 2005 to 7.13 billion barrels in 2013.
Output is at the highest level since 1973 (7.26 billion barrels) and is on course to beat the 1972 record (7.34 billion barrels) in 2014 (Chart 3).
The combination of horizontal drilling and hydraulic fracturing, which have enabled oil and gas extraction from previously impermeable rock formations, is responsible for disproving Hubbert's thesis.
But it is worth asking why Hubbert's theory proved to be wrong because it sheds light on why oil and gas supplies will never run out in any meaningful sense.
Crude oil, natural gas and coal are all formed from the fossilised remains of plants, algae and plankton that lived millions of years ago and were buried underground or on the bottom of ancient lakes and oceans before they could decompose completely.
A small fraction of the organic remains was converted to kerogen by the action of anaerobic bacteria. As sedimentation continued, some of the kerogen was buried more deeply. With greater depth came higher temperatures, converting some of the kerogen into oil, gas and coal in a process that is similar to the cracking and coking employed in modern oil refineries.
New accumulations of fossil fuels are being laid down all the time, especially in organic rich depositional environments like the Mississippi and Niger deltas.
But the current massive fossil fuel reserves were laid down over 500 million years and the current rate of exploitation probably exceeds the rate at which new oil, gas and coal are being created.
Hubbert therefore concluded fossil fuels are a finite resource. "We can assume with complete assurance that the industrial exploitation of the fossil fuels will consist in the progressive exhaustion of an initially fixed supply to which there will be no significant additions during the period of our interest," he insisted.
Based on the historical experience of coal, oil and gas production in the United States, Hubbert noted that when a new energy source is first exploited, production tends to rise exponentially, but the doubling of output cannot continue indefinitely, and eventually production growth slows and then starts to decline.
Production would first rise exponentially, level off, then decline exponentially. Output could be described by a curve.
Hubbert drew an analogy between the production and eventual exhaustion of an individual oil field and the production and exhaustion of oil, gas and coal reserves at a state, national or world level. They would all follow the same pattern of expansion and then decline, shown as a classic production curve.
"These curves embody just about all that is essential in our knowledge of the production of energy from fossil fuels on the world, a national and a state scale," Hubbert concluded.
"No finite resource for longer than a brief period such a growth rate of production; therefore although production rates tend initially to increase exponentially, physical limits prevent their continuing to do so."
Hubbert assumed the ultimately recoverable oil reserves of the United States were around 150 billion barrels, of which 50 billion had already been produced by the mid-1950s.
Production was still increasing at that point, so peak output was still some way in the future, but given the finite nature of the resource the peak would have to occur around 1965 or 1970 before output began its inevitable decline, he argued.
Hubbert was well aware that the recoverable resources of oil and gas were dwarfed by much larger accumulations of other fossil fuels like coal, tar sands and kerogen, all of which could be converted into more usable liquid fuels if necessary.
Global oil and gas output would peak before the end of the 20th century, according to Hubbert. But "this does not necessarily imply that the United States or other parts of the industrial world will soon become destitute of liquid and gaseous fuels, because these can be produced from other fossil fuels which occur in much greater abundance."
Coal, kerogen and even biomass can be converted into oil and gas using the well-established Fischer-Tropsch process.
Moreover, Hubbert understood that improvements in technology would allow more oil and gas to be extracted in future and could boost ultimately recoverable reserves.
"By means of present production techniques, only about a third of the oil underground is being recovered," he wrote. "However, secondary recovery techniques are gradually being improved so that ultimately a somewhat larger but still unknown fraction of the oil underground be extracted than is now the case."
But Hubbert doubted improvements in technology and recovery rates would occur fast enough to have "any significant effect" on the date of production peaking.
And that points to the flaw in Hubbert's theory. He understood output rates and ultimate production were influenced by both the size of the initial resource base and technology. But Hubbert failed to appreciate just how much technology could change and how quickly.
Hubbert's defenders point out that his predictions were made with reference to the then known reserves of conventional oil and gas.
Production rates have been sustained and are now rising again, but only by exploiting new resources such as offshore deepwater fields and shales, most of which are much more difficult and expensive to produce than the conventional fields Hubbert was analysing. Output has been sustained, but only at an increasing cost and price.
British economist William Jevons made similar predictions in the 19th century about the rising cost of coal production from Britain's coal fields as seams near the surface were gradually exhausted and miners were forced to delve deeper.
In this form, the theory is true, but not meaningful. Proponents talk about the progressive exhaustion of "conventional" resources and the need to replace them with more expensive and difficult "unconventional" resources.
But improvements in technology shift the boundary between conventional and unconventional all the time. What was once conventional or even unimaginable becomes mainstream and commonplace. Even shallow offshore oil fields, once seen as very unconventional, are now considered conventional by most observers. Shale oil and gas production is well on the way to becoming mainstream.
Contrary to Jevons, Hubbert and the main proponents of peak oil theory, costs and prices do not always rise. In fact, the real inflation-adjusted price of oil, has been quite variable over the last 150 years, as shown by the price history contained in the annual BP Statistical Review of World Energy (Chart 4).
Costs and prices are as much a function of technology as output. It is simply not true to say prices (in real terms) must rise in the long term as easy resources are exhausted because that too depends on changes in technology.
"No mineral, including oil, will ever be exhausted. If and when the cost of finding and extraction goes above the price consumers are willing to pay, the industry will begin to disappear," according to MIT professor Morris Adelman.
If oil becomes too expensive it will be replaced by other forms of energy, just as coal replaced wood in the 19th century, and oil replaced coal in the middle of the 20th century.
In that sense, Hubbert's theory is technically correct but practically meaningless because it rests on an essentially static view of technology.
It is what Nobel physicist Max Planck termed a phantom problem: "soberly looked at, it is void of meaning - a phantom problem on which our labours and thoughts were wasted."
John Kemp is a Reuters market analyst. The views expressed are his own.