27th February 2017

Peak Oil

Classic car at petrol station

by Enza Ferreri

A Short History

In October 1973 the powerful OPEC (Organization of Petroleum Exporting Countries) started an oil embargo that lasted five months until March 1974, during which it stopped exporting oil to the USA, Western Europe, Japan and other nations.

The reduction in oil supply caused the price of oil to quadruple globally (from 3 to 12 US dollars per barrel), and the crisis sent a shock wave through the industrialised world, exposing its dependency on oil and on the Arab Middle East which was producing much of it, a region characterised by high political and military volatility.

A similar oil crisis in 1979, this time involving Iran, more than doubled the cost of crude oil, reignited the sense of panic and confirmed the vulnerability of the developed world to the potential problems generated by energy scarcity.

During both crises long queues at petrol stations, as well as “austerity” measures designed to reduce the use of cars and electricity, became common features in Western countries.

Since then OPEC permanently lost its dominant status, and in 1981 its oil production was outstripped by that of other countries. Russia, then the Soviet Union, became for a while the world’s largest oil producer and has since remained in a strong position, being today, with the United States and Saudi Arabia, one of the top global producers.

Production of crude oil, according to official energy statistics from the U.S. Government’s Energy Information Administration (EIA), has increased more than threefold from 20.99 million barrels per day in 1960 to 73.47 million barrels per day in 2006.

Human Choice, Not Just Geology

This short historical introduction serves to show that there are few things whose cost is affected by international geopolitics as much as crude oil and, in consequence, car fuel. But other things affect it too, like the motivation to look for new sources and ways of extraction.

Peak oil theory, instead, starts from the opposite viewpoint. It simply, deterministically states that oil supply reaches a peak, or point of maximum production in time, after which it declines permanently and irreversibly. Peak oil is reached both by individual countries' production and world production.

Like one of the forebears of the environmentalist way of thinking, Malthus with his theory of population, followed by many other doctrines of human overpopulation and/or finite resources, this kind of theories has shown again and again how hard it is to make predictions that involve human behaviour, due to the difficulty of applying to human beings the type of natural laws useful, for example, in physics.

The U.S. was supposed to have reached its peak oil state in 1970 with 9.64 million barrels per day, which seemed to confirm the peak oil theory created by Shell Oil Company’s (later working for the U.S. Geological Survey) geologist M. King Hubbert, who in 1956 predicted that oil production in the contiguous 48 U.S. states would peak about 1970.

But now the States’ oil production has increased again and has almost reached the 1970 level with 9.415 million barrels per day in 2015.

Hubbert said production follows a bell-shaped curve, also called “Hubbert’s peak” or “Hubbert’s curve”.

The problem with the peak oil theory is this: if what it says is that oil is a finite resource and therefore it will end (finish), then it’s nothing more than a tautology. It doesn’t say anything that wasn’t already contained in its premise: oil is a “finite”, meaning “it will end”, resource. It just makes explicit what was implicit in the propositional premise or even in the concept “finite” itself, exactly like saying “the triangle has three angles”. We all knew that oil is not infinite, as nothing in the natural world is. The theory, then, is rather circular and not of much value.

Russian sea oil platform

If, instead, the theory of peak oil has any empirical, i.e. informational, value, then it must be conveying something new about reality, and, like all theories that aspire to scientific status, make predictions deduced from it, namely statements about observational phenomena which may turn out to be true and confirm the theory or not true and refute it.

It seems that the latter scenario is what the creator of the theory had in mind, since he made a somewhat specific prediction about the year the U.S. oil production would reach its peak: around 1970. And that was remarkably close to what happened.

After 1970 that production declined, but not, as the theory predicted, permanently. It has now risen to nearly its 1970’s level.

There is something wrong in and with the theory, then. And what’s wrong is that it didn’t take into account two important variables: 1) the behaviour and reaction of the market, 2) human ingenuity.

What is controversial about this theory is not the only-too-obvious characteristic of finiteness of oil, but the timing and consequences of it. If limitedness and shortage keep the ball of new ways of discovery and methods of extraction rolling, then the hypothetical “peak” is continuously procrastinated and pushed forward or even indefinitely delayed and never reached, and the effects of oil’s non-renewability, to use an expression very much in fashion, are not those expected but far more benign.

One suspects that Hubbert’s theory must have been more complex than that, must have said more. But no, it’s exactly as simplistic as it appears, as former director for Asian energy and security at the Center for International Studies at the Massachusetts Institute of Technology Michael Lynch well explained.

Page 19 of the Peaking of World Oil Production: Impacts, Mitigation, & Risk Management, published in February 2005 by the National Energy Technology Laboratory (NETL) - Department of Energy of the USA, reports in a table the projections made up to that year, 2005, by various experts and organisations.

It's easy to see in that document, in clear form, how the great majority of the forecasts based on Hubbert’s peak oil theory, that predicted the date of the peaking of world oil production to be from 2006 to 2025, were proven wrong by subsequent events.

From a false theory they derived wrong predictions.

The only one of those forecasters who got nearer the truth was energy economist Michael Lynch, who saw no peak in sight as he is an intelligent critic of Hubbert’s peak oil theory.

In a brilliant 2009 New York Times article entitled "'Peak Oil' Is a Waste of Energy" Lynch, remaining now, as I am, on the subject of predictions, wrote:

“When the large supply disruptions of 1973 and 1979 led to skyrocketing prices, nearly all oil experts said the underlying cause was resource scarcity and that prices would go ever higher in the future… Prices instead proceeded to slide for two decades, rather as the tide ignored King Canute.”

The popularity of Hubbert’s theory started with the March 1998 publication in Scientific American of the article The End of Cheap Oil by Colin J. Campbell and Jean H. Laherrère. The theory had not entered mainstream discussion of energy before that time, and some years later the expression “peak oil” was created.

Michael Lynch replied to this article by Campbell and Laherrère and to the general thinking behind the theory of oil depletion and irreversible decline expounded there and in other essays with a detailed analysis, which included the following observations:

“[About the pattern of Campbell’s errors in predictions.] For example, except for Italy, Tunisia and Oman, in all of the countries where Campbell's forecast is too high, the upstream sector is controlled by national oil companies. This shows the importance of policy and economics, refuting the argument that geology alone determines long-term production profiles. If Campbell is said to be projecting possible, not actual production, so that his error on these countries should be disregarded, then it leaves a total "optimistic" error of only 106 tb/d, a trivial amount. This demonstrates clearly that his method is biased in a scientific sense, always producing forecasts which will tend to be too low.

"Additionally, the enormous error in countries like Canada and the United States, Figures 5 and Figure 6 , indicates the inability of this measure to produce accurate results. He argues that these are mature producing areas, where no more [oil source] giants are to be found and cause surprise increases in oil production, and the oil field database he employs corrects for the conservative bias due to SEC reserve reporting requirements. Since the oil resources are thus so well-known, the production curve should be extremely accurate. Yet even without any giant field discoveries, both countries have outperformed his forecast far beyond levels conceivable to him. If this method doesn't work in mature regions with high-quality data, it is hardly likely to be reliable elsewhere…

If Campbell's estimation methods produce accurate, rather than conservative, resource estimates, overcoming the objections about field and resource growth made by Adelman and Lynch, and if the Petroconsultants field size estimates are accurate because they don't have to conform to SEC rules requiring conservative field size reporting, why have his production forecasts been much too low [to begin with] and why have his resource estimates increased [over time]?

Finally, Lynch provides the answer to my question of "finiteness" above when he argues that

the Hubbert method fails because it takes recoverable (not total) resources as fixed, and assumes that to be the area under the curve of total production. When the estimate of the area under the curve (resources) is increased, the entire increase must be applied to future production. This is exactly what is happening with Campbell, as Figure 15 shows. The errors in his 1991 forecast and the adjustments he has made in his latest work are thus predicted by Lynch (1996). [Emphasis added]


Enza FerreriEnza Ferreri is an Italian web author, Philosophy graduate and former journalist living in London.




Back to Top