A peak in oil production due to resource exhaustion is a familiar topic to most. However, as of now no consensus has been reached as to likelihood of whether or not we will encounter this peak in the short run, and if we will, when will it occur. Although there are many reasons why divisions of opinion would occur, among them varying levels of optimism and fields of study, Sorrell et al. review current peak oil research done by the UK Energy Research Centre (UKERC) to explain that some of the main drivers of the uncertainty surrounding peak oil estimates are that the data are sparse and often privately owned, scientists use different and changing terminology, and the assumptions underlying models of oil production and consumption are not necessarily the same. Taking this information into account, the authors then go on to predict that oil production will probably reach a bumpy plateau and then decline rather than a form a true peak. —Steven Erickson
Sorrell, S., Speirs, J., Bentley, R., Brandt, A., Miller, R., 2010. Global oil depletion: A review of the evidence. Energy Policy 38, 5290–5295
Sorrell et al. summarize the UK Energy Research Centre’s (UKERC) findings in their independent review of the evidence for a peak in oil production. They go on to analyze the various factors influencing the difficulty in both predicting peaks in oil production and getting these predictions taken seriously in the scientific community. They conclude by explaining what needs to be done to the models to make them more reliable as well as giving their opinions on current estimates of future oil production.
One of the first problems examined is the ambiguity and quality of data. First of all, publicly available data are scant and therefore not very useful. Commercial sources are better, but are also expensive and sometimes impossible to obtain. Further augmenting the problem is the division of reserves into proved (1P) and proved and probable (2P). Neither of these terms has an especially standardized meaning, and because of the lack of data availability scientists must make assumptions in order to get any sort of estimate.
The process of estimating resource size is also fraught with uncertainty. Geological models are often used for relatively unexplored regions and economic models are typically used to plot out production in areas that are well explored. According to the authors, neither of these techniques take their weaknesses fully into account. They surmise that most techniques will be pessimistic in their estimates. In order to improve the situation, Sorrell et al. (2010) suggest new models that use both historical and geologic data along with economic and political data, but they may provide very different results from previous models for estimating regional recoverable resources.
Estimating actual oil supply is just as difficult. Both curve fitting and econometric models are used to try to map out supply, but fitted lines lack theoretical basis, and econometrics may not do a better job of predicting future output. The authors state that the ideal model would be one based from the bottom up using project data, but most companies closely guard their data, making that information difficult to attain. The authors assert that although such supply models for a given level of available resources can estimate a peak to within a decade, they can not provide much precision beyond that.
Sorrell et al. conclude the paper by discussing the timing of the peak. They say that current estimates of URR made by the US Geological Survey (USGS) are around 3345 GB. This estimate puts a global resource peak at around 2030. The authors conclude that if one wants to justify that an oil peak can be put off beyond this date they must rely on some tenuous conclusions, including low demand growth, a reversal in the 40 year trend of declining oil discoveries, and a cumulative production at the date of any potential peak that exceeds 50% of the global URR.