>I think this came up onlist a few years ago; you should check
>the archives. Deffeyes's book Hubbert's Peak is a good
>exposition of this. He's a geologist who uses a plausible
>statistical model to predict a worldwide peak in oil
>production no later than 2010. His model quite nicely
>accounts for the rise and continuing decline of oil
>production in the U. S. As with all statistical models,
>the model only as good as the assumptions used to create
>it; however, it's not crackpot-green propaganda.
From what I've read, it may be just the opposite - petrophiles love it because it justifies subsidies and more aggressive exploration.
Doug