Petrol Depletion and Economics: Part I

Disclaimer: I have no special expertise in the areas of economics, geology or petroleum engineering. I happen to find the intersection of non-renewable resources and economics an interesting area. What follows is my attempt to cover the salient issues without getting bogged down in too much detail. Given my lack of expertise, it’s possible that I’ve made errors.

The problem comes down to one question:

Should oil supply be treated as exogenous or endogenous in economic models?

Here’s my translation of what the depletion community is saying:

Within a decade, the oil market will reach a point where supply will be dominated by an exogenous component. Furthermore this component will decay slowly, a few percent a year. Although the endogenous component may mask this decline in the short term, it will be unable to do so indefinitely. The demand for oil is relatively inelastic and oil is a significant component of international trade for many countries. This combination could result in enormous disruptions to markets and society.

In order to minimize these disruptions, steps need to be taken to adjust the demand side as soon as possible. This will be complicated by the time and investment required to develop alternatives. Another complication is that a relatively large amount of oil-dependent capital stock faces accelerated depreciation.

The rest of the discussion is divided into three parts, supply, demand and ‘other considerations’ - mainly investment and international trade.

Supply

Endogenous forces dominate the supply of traditional renewable resources. High prices spur additional development, which yields additional supply, which drives prices down. Although exogenous forces come into play, (e.g. weather) over the long term their effect on the market can be ‘averaged out’. This is not to say these markets are static. They can still exhibit a boom/bust cycle and be subject to shocks.

In the depletion view of a non-renewable resource, the relative magnitude of exogenous and endogenous components changes over time. When the resource is first exploited, supply is dominated by market considerations. Over time as the largest/easiest/highest quality deposits are exhausted, more effort is required to extract the same quantity. This exogenous effect will grow until the resource is exhausted, at which point the supply is zero - regardless of price.

A plot of supply vs. time would start at zero, rise to a peak and then decline. The area underneath the peak would equal the overall endowment of the resource. The shape of the peak would be dictated by both market and depletion forces. The latter would be a function of the distribution of the deposits and the ease of their extraction. A resource whose deposits are concentrated in a few, large, easy to extract deposits would have a different depletion characteristic than one with many deposits which varied widely in size and ease of extraction.

The key determinants of the depletion effect are the overall endowment of the resource, the distribution of the deposits and the ease (costs) of extracting them.

In order to address these questions, it helps to split supply into two stages, discovery and production. Discovery is the location of deposits. Production is the exploitation of these deposits to yield supply. The only ‘hard’ relationship between the two is that one can’t produce more than has been discovered. A look at discovery will answer the first two questions.

The rate of discovery peaked a few decades ago (mid 1960s) and has been declining ever since. At present, we actually produce around three times what we discover in a year. Although it has fluctuated with market conditions, the downward trend is unmistakable. It is clear that we have been in the ‘tail’ of the discovery curve for some time. Because of this, the overall endowment can be estimated with some confidence. Over the years the estimate of ultimate recovery has remained fairly consistent we’ve used up about half of our endowment of ‘regular oil’.

The second question involves the distribution of deposits. Ignoring other factors (e.g. quality and ease of extraction), the size and frequency of deposits follow a power-law relationship, similar to a Pareto distribution of wealth. A significant portion of the oil wealth resides in relatively few giant/super-giant deposits. What is disturbing is that most of these giant/super-giant deposits were found a long time ago. The last discovery of a giant deposit was made almost two decades ago. This is true despite an impressive record of technical progress. One can’t rule out discovering another giant deposit. At the same time, given the power-law relationship, technological improvements and the length of the ‘dry spell’, one can start to put a harsh upper bound on its likelihood.

There are four key points to take from the discovery portion of supply:
1) The estimate of ultimate recovery has remained relatively unchanged about twice what has already been consumed.
2) The size vs. frequency distribution follows a power law relationship.
3) Discovery peaked several decades ago. Although it has fluctuated with market conditions, it has fallen considerably over time. This is true despite improvements in technology.
4) In terms of finding giant/super-giant deposits, there has been a long ‘dry spell’. The likelihood of finding more such deposits is becoming smaller.