The Hindu Business Line : Predicting the oil peak

Analysing Hubbert's predictions of peak oil, which happens when 50 per cent of recoverable oil has been produced, SHANMUGANATHAN. N and SATISH KUMAR argue that rather than quibble about the actual peak oil date, the sooner the inevitability of Hubbert's Peak is recognised, the better for everybody.

When we first wrote about the impending $200 a barrel scenario — to happen some time between 2007 and 2009, as we had stated in "Peaked Over Crude Oil" (Business Line, November 4, 2005) — there were few serious takers. The thought was dismissed outright with reasons ranging from the Canadian tar sands and US shale oil to technology advancements and demand destruction. Now that we have set a new historical high of $75, this is probably a good time to review what we had written earlier and explain why $200 in the near future is a certainty.

Our prediction was essentially based on peak oil occurring before 2007. Most sources of peak oil theories stem from the idea proposed by Marion King Hubbert, a geophysicist at Shell Oil.

Hubbert had predicted in 1956 that the peak oil (now referred to as Hubbert's Peak) for the US would happen in 1970. Further, in 1969, he predicted that peak oil for the world would happen by 2000.

Hubbert was spot on about peak oil for the US, as indicated by the graph. Even the subsequent introduction of Prudhoe Bay, Alaska, and newer technologies that permitted deep-water drilling have not enabled the US to produce more than what it did in 1971 and, currently it produces just about 5 mbpd.

Two assumptions

Hubbert had based his predictions on two assumptions:

Oil production from well-fields would follow a logistic curve (a bell-shaped profile) and that a combination of these bell curves would also be a bell-shaped pattern.

Probability of oil discovery is directly proportional to the fraction of "undiscovered" oil and, consequently, with the progression of time, both the oil-field sizes and total oil discovered are likely to get smaller. In some ways, one can compare this with fishing — the biggest of the fishes are likely to be discovered first and, even with the introduction of advanced technologies, only the progressively smaller ones will be discovered in the future.

The first assumption is validated by looking at oil field production data, and the one for the US, shown in the graph, also indicates the same. The second assumption is validated by the discovery data presented in the bar chart. Discovery peaked during the 1960s and has been progressively getting smaller with time.

Not entirely wrong

So, why did Hubbert get his world peak oil prediction wrong? First, he was not entirely wrong; he was only off by a few years.

And for such a prediction, given the uncertainties of oil exploration, production and economic expansion over a period of 40-plus years, being a few years off is actually astonishing precision. Some of the reasons for the delay are:

The OPEC-induced oil crises of 1973 and 1979 that temporarily reduced demand.

Addition of non-conventional sources that Hubbert had not included in his calculations.

Technological advancements that have converted the peak to a plateau at the top for some of the oil fields. All that this would mean is that the fall from the plateau would be steeper as the area under the curve essentially represents the recoverable oil, and that has not changed.

So what is the current prognosis for the world peak oil date?

Predicting peak oil

There are many ways to predict peak oil. However, by definition, we will know this phenomenon only with the benefit of hindsight (the phenomenon of US peaking was denied as late as 1975 when the US Geological Survey released circulars stating that the US domestic oil production would not peak until well into the 21st century, and possibly not until the 22nd century). Some methods that can be used to predict peak oil are explained below.

Classical Hubbert Methodology: Hubbert had predicted his calculations on the basis that peak oil happens when 50 per cent of the recoverable oil has been produced. Using that methodology, Kenneth Deffeyes, a former colleague of Hubbert at Shell and author of Beyond Oil: The View from Hubbert's Peak, estimates that world peak oil happened on December 16, 2005, by which date we had consumed 50 per cent of the total recoverable reserves of 2.013 Tb.

When making this observation in February 2006, he quipped "My career as a prophet is over. I'm now a historian."

Production lagging discovery: It has been observed worldwide that peak production lags peak discovery by about 40 years. This is certainly true of the US as well as production from scores of individual sites. Of course, it is not always exactly 40 years. For example, North Sea peaked within 30 years, while Iraq took more than 60 years. The worldwide average, however, has been about 40 years. The world peak discovery happened in 1963 and, based on that, we are past the average time-line of 40 years.

Summation of individual country peaks: Of the 45 countries that account for the bulk of world oil production, 35 are well past their peak.

It is also widely acknowledged, even by the oil companies, that non-OPEC production has already peaked. Matt Simmons, author of Twilight in the Desert, says that, with 99 per cent accuracy, we can be sure that world peak production has happened when Saudi Arabia's peak production has happened.

Dr Sadad Al-Husseini stated in November 2005 that Saudi Arabian production should not cross 12 mbpd ever and that indicates a capacity addition of another 3 mbpd from their current production levels. So after world peak (excluding Saudi Arabia's), a depletion rate of even 2 per cent/year would ensure that we would indeed arrive at world peak within two years, even if we assume that Saudi Aramco can bring all of this 3mbpd to production overnight.

Of course, it could be rightly argued that a decline rate of 2 per cent is conservative and the world average for existing fields has been in the range of 4-6 per cent. Also, Saudi Aramco would take at least five years to bring this additional capacity online. Consequently, all that Saudi Aramco can do is to make the post-peak production slope a little less steep than it would have been otherwise.

Avoiding the obvious

We are pretty sure that, even at this stage, high oil prices would be attributed to greedy oil companies, Iran, Nigeria, the weather, rig shortages, MTBE, the terror premium, and so on.

All these visible (and hopefully) short-term issues are merely exacerbating a less obvious, but permanent geological phenomenon. With planned net addition of capacity of the order of just 1mbpd for the next few years, demand would exceed supplies even if peak oil was to be a distant phenomenon.

From a long-term planning perspective, it really does not matter whether the specific peak oil date was 2005 or 2008. The sooner we recognise the inevitability of Hubbert's Peak, the better it will be for everybody.

(Shanmuganathan N is a financial advisor and can be reached at shanmuganathan.sundaram@gmail.com. Satish Kumar works at Sterling Commerce and can be contacted at satish_subbiah@stercomm.com)