There has been a lot of talk recently about the "spike" in oil prices. Spike is, of course, a reassuring word: it implies that there's a downward slope on the other side. Just the other day, editorial writers and business-page commentators were reassuring us that oil at $75 a barrel was "unsustainable", and that prices would fall as supply and demand even out. But opinion is moving towards the Goldman Sachs 2005 forecast of a "super-spike", in which prices could go as high as $105.
At present, we are told, demand is increasing while supply is insufficient as a result of oil companies' under-investment during the 1990s, when oil prices were low. But now (says the Guardian's business correspondent), "the majors are looking for oil (and finding some) like never before". And Goldman Sachs predicts that by 2010, the price will be back to $45 a barrel. So that's all right, then. This is the economists' view. However, this is not just, or even primarily, an economic problem. It is about geology and physics. And the geology and physics are not reassuring. It's all very well talking about increasing production, but in order to produce something, it must first be there.
In 1956, the geophysicist M King Hubbert predicted that American crude-oil production in the 48 states would rise for 13 more years, then peak in 1969, give or take a year. He was ridiculed. But in 1972, on cue, US oil production began to fall. Hubbert had been proved right.
Hubbert's forecast was based not on psi but on mathematics. In the mid-1930s, he realised that in any large region, unrestrained extraction of a finite resource rises along a bell-shaped curve that peaks when about half of the resource is gone. Production lags behind discovery by about 40 years, the length of time it takes to bring a new oilfield into full production' the US oil-discovery curve peaked in the early 1930s. Hence Dr Hubbert's forecast.
Discoveries are still being made. But they are not the size they once were. In the words of the oil-industry veteran Colin Campbell, "We now find one barrel for every four we consume."
The sums are simple. If you transpose the discovery graph forward 40 years (allowing for a small lag during the 1970s, when the oil price suddenly tripled and demand fell off), you arrive at just about where we are now. …