by Leanan » Fri 17 Mar 2006, 13:57:41
$this->bbcode_second_pass_quote('', 'I') am not familiar with what data Ken uses, but it seems Campbell and others don't agree with him.
It's not so much the data he uses as what he does with it. Deffeyes uses "Hubbert linearization." That is, he assumes that the after finding the first few fields, you can calculate how much more there will be. Because the largest fields are found first. The strength of this method is that once you have enough data, more finds really don't change the peak. Just as Prudhoe didn't change the U.S. peak.
The weakness of the method is if you try it too early - when not enough early fields have been found - it may be misleading. So if this new field means there's a lot more out there to be found, it might mean it's too early to do a Hubbert linearization for Mexico.
I think what some others are doing is trying to add up all the current production, subtract declines in production, and add in new production. The problems are obvious with this method. You have to estimate decline rates, which isn't easy, and you have to assume that the scheduled new project will come online as when predicted, and produce the amount predicted, which rarely happens.
There's also the question of what they are counting as "oil." Tar sands, oil shale, NGLs, etc. Some count 'em, some don't.
$this->bbcode_second_pass_quote('', 'W')hen looking at known projects, it doesn't seem Ken's prediction is going to hold, but we will know by year end.
I doubt it. Even the cornucopians are predicting this year is likely to be a bad one for oil production. Next year they hope for an increase. If 2007 shows a decline, too, that will indicate to me that Deffeyes is right.