Assessment and Importance of Oil Depletion - Part II
Misleading ReportingThe main reasons why this subject is not better understood are the ambiguous definitions and unreliable reporting practices of the industry.
Conventional and Non-Conventional Oil
Oil is oil from the standpoint of the motorist filling his tank, who does not much care from whence it comes. But the analyst of depletion needs to identify the different categories because each has its own costs, characteristics and extraction rates, and hence can contribute differently to peak production. The term Conventional is widely used to describe the traditional sources, which have contributed most oil produced to-date, and which will dominate all supply far into the future.
There are, in additional, Non-Conventional sources, which will be increasingly important when Conventional oil declines after peak, but there is no standard definition of the boundary. Here, the following categories are treated as Non-Conventional, being described in greater detail in a later section.
- Oil from coal and "shale" (actually immature source-rock)
- Extra-Heavy Oil (density <10o API)
- Heavy Oil (density 10-17.5o API)
- Polar Oil and Gas
- Deepwater Oil and Gas (>500m water depth)
Production and Supply Reporting
Measuring production is simply a matter of reading the meter, but national statistics are confused by the inconsistent treatment of natural gas liquids, war loss, which is production at least in a technical sense, and frontier changes. Supply is not the same as production but includes stock change and refinery gains. The reporting of gas production is still more confused, referring variously to raw gas or marketed gas after the removal in inert gases such as nitrogen and carbon dioxide, which are often present. The differing treatment of flared and re-injected gas adds to the uncertainty.
The practices of reserve reporting evolved early, being much influenced by the environment of the old onshore fields of the United States, which were characterised by a highly fragmented ownership. Being onshore and close to market, the wells could be placed on production as soon as they had been completed.
To prevent fraud, the Securities and Exchange Commission (SEC) introduced strict rules whereby owners could treat as proved for financial purposes only the reserves "behind pipe", meaning those to be drained by existing wells. As the fields were drilled up, the reported reserves naturally grew.
In practice, the reserves of the old fields were mainly estimated by extrapolating the decline rates of the wells. The highly fragmented ownership meant that there was little interest, or indeed possibility, of making field-wide reserve estimates. The companies did however recognise additional Probable and Possible Reserves, that did not qualify for Proved financial status.
Different conditions obtained overseas and in offshore areas, where it was normal for fields to be developed as single entities by one or more companies, acting as a group. They were more interested in what the field as a whole would produce over its full life, especially offshore, where they had to design appropriate facilities in advance of production. They were still lumbered with the SEC rules, which required the reporting of Proved Reserves, although in practice they reported better estimates of what the fields would deliver over their full lives. There is naturally a degree of latitude in estimating future production.
For a variety of commercial reasons, it was found expedient to report conservative estimates on discovery, which consequently grew over time, delivering an attractive impression of gradually appreciating assets to the stockmarket, and serving to reduce tax in countries operating a depletion allowance. Many of the large North Sea fields, for example, were initially under-reported by about one-third.
This luxury is not however available to the more recent small fields, with a short life and high economic threshold. In fact, they sometimes give disappointing results, yielding negative reserve growth.
A further confusion has arisen from the application of probability theory in which reserves are equated against differing subjective probability rankings. Under this system, Proven Reserves (1P) are commonly equated with a 95% probability, whereas Proved + Probable +Possible Reserves (3P) are held to have a 5% probability. Mean, Median and Mode values are then computed. This system appeals to the scientifically inclined, but in practice adds to the confusion.
In plain language, Proved Reserves relate to the current status of development, whereas Proved and Probable Reserves (or Mean under the probability system) are estimates of what the field as a whole is expected to produce over the rest of its life. The probability range is unnecessarily wide, as engineers with modern methods can make good estimates. Why should anybody be interested in an assessment having no more than a 5% probability of being correct, and a subjective one at that?
A simple way to escape from all these definitional confusions is to avoid the term Reserves altogether, and instead refer to estimates of what is to be produced in the future from known fields. This is particularly useful in connection with Non-Conventional heavy oils and bitumen, which are subject to extraction rate rather than resource constraints. Trying to estimate their resources and possible reserves is a pointless exercise.
If we further introduce a time cutoff of, say, 2075, we avoid worrying about irrelevant tail-end depletion that is of little significance. We may also forget about probability theory, and base our conclusions on what in plain language can be called "best estimates", which, like all estimates, are not exact but as good as we can do.
The Dating of Reserves Revisions
For financial purposes, reserve revisions are reported on the date that they are made, but this gives a misleading impression of the discovery trend. Year on year comparison of national reserves, with subtraction of the intervening production, gives the impression that more is being found than is the case, which has misled many analysts working with data in the public domain.
To determine a valid discovery trend, it is necessary first to make sure that the reported production and Proved and Probable reserves relate to the same categories of oil; and second to back-date any reserve revisions to the discovery of respective fields. In practice, this cannot be done without access to the industry database to identify the details, but its cost puts it out of range for most analysts.
Several OPEC countries announced colossal overnight reserve increases in the late 1980s when they were vying for quota based on reserves. While some upward revision was called for, as the earlier numbers were too conservative, having been inherited from the private companies before they were expropriated, the revisions have to be backdated to the discovery of the fields containing them, some of which had been found up to fifty years earlier.
Dating reserves is as important, if not more so, than estimating the amounts. The explanation for the revisions is another highly important issue to grasp. The industry, not wishing to admit to poor reporting practices, has found it expedient to attribute revisions to technological progress when they were in fact mainly a reporting phenomenon. This in turn carries the danger of unjustifiably extrapolating reserve growth into the future on the assumption of an inexorable march of technological progress. No one disputes the progress to-date, but its main impact has been to hold production higher for longer, which makes good economic sense but accelerates depletion, adding little to the reserves themselves.
Two trade journals, the Oil and Gas Journal and World Oil, have compiled information on production and reserves for many years on the basis of a questionnaire sent out to governments and others. Many of the reports remain implausibly unchanged for years on end, simply because the country concerned has failed to update its estimate, despite production. There are also substantial discrepancies between the two data-sets, despite the fact that they are compiled in a similar fashion.
A third source is the BP Statistical Review of World Energy, which is the most misleading of all, because many analysts wrongly assume that the reported oil reserves have at least the tacit blessing of a competent and knowledgeable oil company in a position to assess their validity. In fact, BP simply reproduces the Oil and Gas Journal's oil reserve data, save in one or two specific cases.
These public sources contain information very different from that in the industry's own database, which is compiled on a field by field basis directly from the companies' own records. This itself contains certain anomalies and seems to be deteriorating in quality as it faces the increasingly difficult challenge of compiling information from the proliferation of small companies and ever less reliable State sources. Particular difficulties are faced in interpreting data from the former Soviet Union that operated its own system of reserve classification that tended to exaggerate by ignoring economic constraints.
In short, although there are no particular technical difficulties in estimating the size of an oilfield, especially with the advantage of modern technology, the reporting of production and reserves remains highly unreliable. In these circumstances, it is well to confirm, wherever possible, the estimates of individual fields by extrapolating the decline, which plots as a straight line on a graph relating annual to cumulative production (see Figure 3).
Next week... Estimating Future Discovery
blog comments powered by Disqus