I've started to believe in modeling just the crude oil component of the peak as this fits in better with estimated discoveries. For example, how do you backdate NGL to previous discovery estimates? What does this even mean? Did somebody at some prior date say that a particular reservoir would produce X amount of NGL? I think not.

http://mobjectivist.blogspot.com/2007/10/global-update-of-dispersive-dis...

Read the blog post and you will notice that I should have realized this a long time ago, and the updated oil shock model results do not attempt to overcompensate for the non-crude discoveries that we lack data for.

Why would just doing crude+condensates only be considered wrong anyway as far as peak detection goes ?

NGL's do not provide anything close to the range of products derived from crude oil so they are not directly equivalent to crude and neither is natural gas for that matter.

I even question including the very heavy high sulfur crudes and tar sands without weighting. The production and refining of these is far more expensive and energy intensive than regular crudes and overall they are not a replacement for light sweet crude.

In short the key is the production of the types of crude we have built our economy around which is centralized on light sweet crude and thus the economic impact of peak oil which is what we are really interested in will follow the availability of the lighter sweeter oils.

The interesting thing to look at is how the main depressive shock started in the 1970s and then slowed down into the 1990's. For this model to truly make sense, people should buy into the fact that significant changes in oil usage occurred spanning the two decades. Clearly, up to the 1970s, consumers used oil like it was water, applying it to all sorts of what we would consider now incredibly wasteful actions. But then the conservation movement woke up, and the USA subconsciously lead the way when they realized their own supply was petering out. So the key is to buy into the possibility of that 40% dip in extraction rate from the available reserves over two decades.

I don't think NYMEX believes the EIA:

$88.50

Arkansaw of Samuel L Clemens

Your graph doesn't provide the details. Did the adjustement change the peak date and the slope afterwards?

It pushed the peak out a few years from my original estimate that I did in 2005. The downslope is less steep as well. This is intuitively obvious in that I don't have to account for as high a peak. The estimate also includes an optimistic model for discoveries that I had developed in the interim, the dispersive discovery model linked in the blog post.