The Azimuth Project
Peak theory



The most famous “peak theory” is peak oil. M. King Hubbert presented a paper to the 1956 meeting of the American Petroleum Institute in San Antonio, Texas, which predicted that overall petroleum production would peak in the United States between the late 1960s and the early 1970s:

  • M.K. Hubbert, Nuclear energy and the fossil fuels, presented before the Spring Meeting of the Southern District, American Petroleum Institute, Plaza Hotel, San Antonio, Texas, 7-9 March 1956

He argued that by comparison, nuclear fuels would last essentially forever. This graph from his paper illustrates the idea:

Hubbert became famous when his prediction concerning American oil production proved correct in 1970. More recently, “peak oil” has widely been used for the theory that the peak of worldwide oil production is imminent:

Although peak theories often include estimates of ultimate depletion times, it is important to realise that, in a system accustomed to growth, effects would be expected to be felt from the time of peak production onwards. As an analogy, suppose that for some reason your monthly salary over a couple of decades were to follow a bell shaped curve: although you’d clearly have huge problems at the point it finally dropped to zero, you’d become progressively more aware of the reduction – and have to choose what to economise on – from the time it peaked.

It’s important to bear in mind that a production peak can only be compellingly confirmed observationally when it is significantly in the past and a significant proportion of the decline has occurred, especially if on a global scale the extraction process has significant semi-random variation. Since for practical purposes it is beneficial to start acting as soon after the time of the peak as possible, there is an unavoidable tension between those using models which are not yet fully validated and those who feel that the models are missing crucial elements. Unfortunately, the situation often ends up with people feeling compelled to take either the position that “the peak is roughly as predicted by this model” or “the peak is not within the medium term future”, rather than allowing more sophisticated – and arguably better from the point of view of risk-hedging – positions.

We may more generally use the term peak theory for any theory suggesting that the production of some commodity will follow a roughly bell-shaped curve, although depending upon the model chosen it may not be symmetric or close to Gaussian shaped. The goal of these models is also primarily to model and predict gross features of future extraction curves, and these often do not differ much between various models.

A common rejoinder about peaks is that, whilst a peak may occur the price of the commodity will start to rise as its availability declines. This should lead to an economic incentive to find a substitute commodity which is cheaper and the world will thus transition to the substitute without any dramatic switch-over point. (For example, in the context of peak oil arguments it is often stated that it was as whale oil became more difficult to obtain that the use of petroleum oil took off.) In some cases the substitution argument may be valid, but it does depend on both the existence of suitable substitutes and a relatively slow decline (which may be less likely at a current world population and consumption levels). These arguments have to be addressed separately for each possible “peaking commodity”.


Some examples include:

  1. Peak uranium

  2. Peak coal: the claim that coal reserves are overstated in the same way as oil reserves. Very dubious on this one since unlike oil it’s not clear there’s been enough of an interest in coal to give an incentive to overstate things, and the huge reserves are in countries like the US which are “associated with transparency”… Also it’s not clear that a limit on coal availability is a bad thing from a CO2 point of view. Indeed, the problem with potential peak oil is the sudden dramatic reduction in availability of liquid fuel for building renewable infrastructure, even though it’d help reduce CO2 emissions.

  3. Peak theories for a variety of rare substances including tantalum, neodymium, and helium.

  4. Peak potable water: aquifers being emptied faster than they refill, increasing population in areas with limited river water

  5. Peak food: multiple issues combining to constrain the amount of food that can be grown

  6. Peak economy/credit: regardless of how beneficial building something is, you generally have to somehow get “money” to pay for construction in advance. The recent financial downturn could have been enough to stop huge new investments. This example is different from the preceding ones, because money is not a “commodity” in the same sense.


One of the things about “peaks” is that there’s a basic idea (“peak X may happen soon”) and detailed models which provide specific projections but which incorporate many assumptions. For individual articles on “peak theories”, we need to figure out a writing style that includes projections but makes it clear that refuting a specific model doesn’t necessarily refute the basic idea.