The main point to counter the 2000 USGS ultimate is to accept the estimate from Ged Davis of Shell International on the BBC, The Money Programme, Nov 8, 2000, where he stated that the amount of undiscovered is 250-260 Gb and the amount for improved recovery is the same. An additional potential of 500 Gb is roughly three times less than the 1420 Gb (oil only) and 1670 Gb (oil plus natural gas liquids) as stated by USGS.
I met John Wood and his assistant Gary Long in Denver in 1998 when I came to present my approach on the assessment of the potential of a petroleum system to the USGS (Reston & Denver). They came especially from Houston to listen to my presentation and John Wood did present a display of all the US oilfields which was the most beautiful example of a parabolic fractal distribution. In his 2000 long term powerpoint presentation John Wood presents our 1998 Scientific American graph (11) and our 1998 Petroleum Economist (the world's non-conventional oil and gas) graph (12) but he was obliged to present the official USGS assessment with the ultimate of 3003 Gb (ridiculous accuracy which should have been rounded to 3000 Gb).
His graph 10 assumes that the oil in place is 6 Tb but it is pure wishful thinking in order to justify an ultimate 50% recovery factor; he left out the non-conventional. Anyway oil in place is a poor approach as there is no way to check its accuracy, in contrast to reserves at abandon time. Recovery factor is used only as a way to present development, but it is rarely used in modern assessment by simulation to forecast future production and therefore reserves (cumulative production at the end). In the last French Petroleum Institute manual on reservoir engineering, recovery rate is not used in any approach.
His graph 14 with a growth of 2%/a and a decline of 2%/a is a fair display of the 3000 Gb ultimate, giving a peak of 35 Gb/a (96 Mb/d) in 2016. But the graph 15 assumes a decline of R/P =10, when in fact it shows a decline of 10%/a. For the last 50 years the US R/P has oscillated around 10 years as it is a rule of thumb of most independent oilmen for estimating reserves. Even the USGS in some reports when they have no data on reserves assume the reserves at 10 times the present annual production. This R/P ratio stood around 10 (12 to 8) during the growth of production up to 1970 and afterwards, during the decline.
The US practice of Proven reserves and R/P of 10 is a bad practice to please the banker but it is misleading to state that more than 90% of the present additions of reserves come from revisions. This means simply that the past estimates were wrong. Happily this practice does not apply for the rest of the world where proven + probable reserves are considered before decisions are made.
Anyway a decline of 10%/a after 2037 where a peak of 53 Gb/a (145 Mb/d) is an insult to the future generation (Saint Exupery wrote: "We do not herit from our parents, we borrow from our children") and this scenario should be deleted as such.
It is obvious that John Wood and Les Magoon do not agree with the last USGS estimate. This is why they show our estimate (2000 Gb ultimate for conventional liquids and 2750 Gb for all liquids) to counterbalance, but they cannot say that the 3,000 Gb (assumed to be only conventional) is too high.