I was thinking about the CERA report vs the one from ODAC which came out with such different results, and I have an idea of what is the problem. CERA came out with considerably higher projected production levels than ODAC. At first people thought there must be some methodological difference, and it was speculated that CERA was ignoring depletion of existing fields. But now that more data is out, the funny thing is that there doesn't seem to be any big difference. There are some differences on a field by field basis, CERA had some that ODAC didn't, but also some differences in the other direction as well.

Here's how I see it. The problem is that it is very hard to accurately predict production levels for dozens or hundreds of oil fields, many of which are yet to begin production, for years to come! No one could realistically be expected to make an accurate prediction, given the uncertainties that lie ahead. It's no wonder that two different analyses came out with somewhat different answers. If they agree to within five or ten percent, that is all we could reasonably expect. And they do.

The further problem is that we aren't really that interested in future production levels. What matters is the difference between potential production and extrapolated demand (based on current price). Even demand levels have considerable uncertainty. The four years from 1990 had consumption growth rates (from the previous year) of 1.2%, 0.7%, 1.0%, and 1.6%. These growth rates jump around quite a bit as well.

The result is that we have two uncertain figures, and we are in effect subtracting them, to see if production levels can meet extrapolated demand. But as anyone knows who has worked with statistics, if you have two data series with considerable uncertainty levels, and you subtract them, your error bars can go through the roof. If we are already uncertain about production levels by several percent, and the differences with consumption levels are less than several percent, then we really can't extract any meaningful predictions at all. The error bars are greater than the information we are interested in.

As I have argued before, the situation is so uncertain that IMO it is premature to say that the case has been made for a near term Peak Oil scenario. When two careful and thorough studies come to such different conclusions, that is further evidence of this truth.

I recently found an amusing academic paper which further emphasizes this reality. Forecasting the limits to the availability and diversity of global conventional oil supply has this to say: "Our results predict that global production of conventional oil will almost certainly begin an irreversible decline somewhere between 2004 and 2037, at 22 to 42 billion barrels per year, depending upon how much oil is available from the earth's crust and the growth rate in its use." How's that for a prediction? Not exactly red meat for Peak Oil true believers, but I believe that any honest evaluation that admits the tremendous uncertainty in the information available will reach a similarly cautious conclusion.