Modelling Peak Oil  

Mobjectivist has now completed his epic effort attempting to come up with a model describing the peak oil curve (no, its not a bell curve).

If you're willing to try and refresh all that maths that you long ago forgot, go and check it out.

I want to hammer a few points home on the Micro Peak Oil Model. The characteristic asymmetry of the consumption curves (i.e. steeper rise than fall-off) arises due to the first-order assumption that humans extract petroleum at a rate proportional to the amount left in the ground. I can't say that this follows precisely the real world, but as a fan of Occam, I normally try to use the simplest explanation for phenomenon that I can get away with. For a good example of the proportionality principle, consider the rise and decay of U.S. wildcat operations. At one time, each gusher generated a large flow, but over time the reservoir contents became depleted enough that the contents reduces to the much smaller stripper well flow. The latencies involved in collecting and delivering the oil extracted from stripper wells contribute to the extended tails we see in the post-Hubbert-peak of U.S. oil production.