ODAC Guest Commentary - Response to Peter Odell
The fundamental reasons that Professor Odell can write a piece like this are four-fold:
1. Data access
Perhaps the main reason for the current ‘oil peak debate’ is the difficulty of getting access to reliable data on how much oil the world has discovered.
There are two distinct datasets:
- Reasonably reliable data on oil discovery are given by the proved plus probable (‘2P’) data held by industry, and by the datasets of the commercial data providers (such as IHS Energy, Wood Mackenzie or Energyfiles Ltd.). Virtually every detailed forecast that uses these data (CERA’s being the only exception I know) sees a peak reasonably soon for conventional oil; and most also see a not-too-distant peak for ‘all-oil’. These include forecasts from PFC Energy; Energyfiles Ltd.; Chris Skrebowski’s calculations (see Petroleum Review of the Energy Institute, London); the German consultancy LBST; the private model of Dr. Richard Miller of BP; the calculations from the University of Uppsala; and the 1995 Petroconsultants study (with subsequent updates independently by C. Campbell and J. Laherrere).
- Very poor data on oil discovery contained in the public-domain data of proved reserves (‘1P data’). As Professor Odell’s piece makes clear, if one examines only the latter it looks as if an oil peak must be many years away.
While the industry 2P data are generally expensive, there are some very useful datasets at fairly modest price. Two co-authors and I have sought to explain why the two datasets are so different (and also how to access industry data) in a recent paper: “Assessing the date of the global oil peak: The need to use 2P reserves”, by R.W. Bentley, S.A. Mannan, and S.J. Wheeler; Energy Policy, vol. 35 (2007), pp 6364-6382. (Please note that copyright of this belongs to the publisher; it must therefore not be copied and put up on a website. The paper is available easily enough via inter-library loan.)
2: Lack of discussion
Curiously for such an important topic, there has been very little useful discussion between the two camps. Such discussion needs of course to have access to the industry data, and this can be hard to achieve without compromising sources. Nevertheless, I and several colleagues have frequently carried such data to meetings with individuals, and to public debates, but it has often proved hard to get people in the other camp to examine the data in any detail.
3. Lack of a good source book.
There is still in my view no complete academic source one can go to read up on oil peaking. This is partly because of the data access problem mentioned above, but mainly because the problem was forgotten for so long (having been covered in many engineering energy textbooks in the 1970s) that few have yet had the opportunity recently to put together an authoritative primer. This will need public release of certain illustrative industry data; a full by-field examination of some countries past peak to make the mechanism of peak clear; adequate reservoir engineering to quantitatively identify past and prospective reservoir behaviour and improvements in extraction techniques; useful aggregate 2P data on the global hydrocarbon resources by type and by likely production profile; and a critical discussion of the various forecasting techniques that are used. Sooner or later such a primer will be written, but to my knowledge it does not yet exist. Perhaps the best current primer is the report by LBST (at www.lbst.de).
4. The strength of the ‘economic view’
This reflects the apparent success of the ‘resources into reserves’ and other economic arguments. Again partly based on the poor public domain data, and to some extent on a misreading of past forecasts, a cohesive ‘economic view’ of oil availability built up since the 1970s (by, for example, Adelman, Lynch, Odell, Watkins and others) that saw any oil forecast that assumed a fixed resource base as being fundamentally flawed. It will take a lot of data, and argument, to un-do the lines of reasoning built up over many years that comprise this ‘economic view’.
Dr. Roger Bentley is a Visiting Research Fellow at the University of Reading's Department of Cybernetics. His areas of research include global hydrocarbon depletion, solar energy, and broader energy issues