Survey of Energy Resources 2007
A Contribution to the Peak Oil Discussion
The Industrial Revolution was born in the 18th Century when countries tapped into their coal resources as a new and convenient source of energy to fuel industry and transport, along with the subsequent development of the railway system.
Oil seepages on the earth's surface had been known from antiquity, being locally exploited in shallow hand-dug wells. Then in 1859, drilling technology, already in place for the extraction of salt-brine, was adapted to drill for oil in Pennsylvania, and a shallow deposit, at a depth of 67 feet, was found. This small step led to the growth of one of the world's largest industries, which began to deliver increasing amounts of this cheap and convenient source of energy, leading to the rapid expansion of industry in general, together with transport, trade and agriculture, which allowed the world's population to expand six-fold over the next 150 years. This chapter of history also saw the rapid expansion of financial capital, as banks lent more than they had on deposit, confident that Tomorrow's Expansion was collateral for Today's Debt, without necessarily recognising that it was the abundant supply of largely oil-based energy that made the expansion possible. In short, the world changed greatly over what has been described as the First Half of the Age of Oil.
Oil and gas were formed in the geological past, meaning that they are natural resources subject to depletion. Therefore it is time to take stock of the situation and try to determine the status of depletion. Knowledge of the physical conditions for oil generation has improved greatly, meaning that the search for oil can now be driven by scientific principles, although speculative projects are still sometimes undertaken. In practice, the oil industry has searched the world, always looking for the biggest and best prospects, and it has generally enjoyed a favourable economic climate, insofar as even small discoveries are highly profitable and much of the cost of exploration can be written off against taxable income. Given that there is a finite limit, past success means that there is less and less left to find in the future. The industry has made remarkable technological progress, such that it has become routine to drill 5 000 m wells in the stormy waters of the North Sea. But there is a certain irony in that improvements in technology have tended to increase extraction rates, thereby accelerating depletion (unless counterbalanced by enhanced recovery).
Lastly it is axiomatic to state that oil has to be found before it can be produced, which means that the discovery profile has to be mirrored in the production profile after a time lapse. If the peak in world discovery occurred in the 1960s, as the data appear to suggest, it follows that a corresponding peak in production may be imminent. The word appear is used advisedly, because great difficulties are experienced in interpreting the data as a result of differing reporting practices and ambiguity in defining the different categories to measure. These two subjects need to be addressed carefully before coming to an assessment of the status of depletion.
As will be explained more fully below, it is important to recognise, first, that there is an Oil Age, which in fact promises to be a relatively brief chapter of human history; and second that inevitably the Oil Age is divisible into a First Half, when discovery and production rise; and a Second Half, when they decline. The world will not finally run out of oil for very many years, if ever, but the onset of decline may prove to be a discontinuity of historic proportions, given the key role oil plays in modern economies. The transition to decline threatens indeed to be an age of great economic and geopolitical tension.
Reporting Production and Reserves
At first sight, it might seem a straightforward task to assess the size of an oilfield and report its production, but in fact reporting practices range widely and are subject to much misunderstanding and confusion.
The reporting of oil production is relatively straightforward in many countries, although it is effectively a state secret in some places, even relying on no more than the reports of shipping agents, counting tankers leaving the terminals. Domestic consumption in such countries may also be inaccurately reported. There are in addition a number of factors affecting the reports, including those listed below:
" war loss: battles have been fought over oilfields, in some cases leading to the escape of oil. Such loss is to be considered production in the sense that it reduces the reserves by a like amount. Some 2 billion barrels of Kuwaiti crude was probably lost in this way during the Gulf War;
metering: in some cases, production from several fields is metered together at a central facility, such that the production of the individual component fields may not be readily identifiable. This is the case, for example, with respect to several of the Lake Maracaibo fields in Venezuela;
gas liquids: liquids commonly condense from natural gas at surface conditions of temperature and pressure, being termed condensate. This may be either metered separately or fed back into the gas or oil streams.
operating fuels: in some cases gas, condensate and even oil are used as an operating fuel, and not metered. Spillage likewise cannot be counted.
It should be noted in addition that production is not synonymous with supply, namely the amount available for consumption. Refinery gains (running at 2-3%) have to be added and losses in storage and transportation taken into account.
These distortions are however relatively minor compared with those associated with the reporting of reserves, which deserve to be explained more fully.
The traditional industry classification of reserves, arising from the early days in the United States, was to recognise 'Proved',' Probable' and 'Possible' categories, with the meanings the words normally convey. The mineral rights in the United States mainly belong to the landowner, which means that individual fields were physically divided, even to the extent in some cases of separate reservoirs having separate owners. Oil in the ground was a financial asset against which money could be borrowed, leading the Securities and Exchange Commission (SEC) to introduce strict rules, which were designed to prevent fraudulent exaggeration, while smiling on under-reporting as commercial prudence. It recognised two prime categories, namely 'proved developed' for the anticipated future production of current wells and 'proved undeveloped' for the expected production from infill wells between the existing ones, before they have actually been drilled.
It was a perfectly sound system for the circumstances for which it was designed, but it was also adopted by the international oil companies, which were quoted on the American exchanges, although the circumstances were rather different, including in some cases greater commercial uncertainty. It made good sense for the companies to under-report discovery and then revise the reports upwards over time, which gave a comforting but misleading image of steady growth. The larger fields were normally developed in phases, with each phase being reported as it was committed. The upward revisions were termed 'reserve growth', which was widely attributed to technological progress, when in fact it was in large measure simply an artefact of reporting. But as the stock of ageing large fields declined, so did the scope for under-reporting, with many of the smaller fields being developed in a single phase, in some cases even delivering disappointing results. Accordingly it is unlikely that the apparent 'reserve growth' of the past will be matched in the future.
Estimating the size of an oilfield early in its life poses no particular technical challenge, though it is subject to a degree of uncertainty. This in turn prompted some analysts to apply Probability Theory to the issue. Under this system, alternative reserve estimates are plotted against a range of subjective probability rankings. It is common to refer to P95 Reserves, such being deemed to have a 95% probability of exceeding the stated value, and P5 Reserves for those with a 5% probability of doing so. From this range, median (P50) and mean values are computed.
The 'best estimates' so-to-speak of future production are described under the alternative systems as proved + probable or as having a mean or median (P50) probability ranking. As may be imagined, there are plenty of grey areas in the application of these systems, with a general tendency for the international companies to report cautious estimates subject to upward revision, as already noted.
There are in addition what may be described as political reserves, especially amongst the OPEC countries, which found themselves competing for production quotas based in part on reported reserves. There is some evidence to suggest that some of these countries started reporting 'original' not 'remaining reserves' during the 1980s at a time of weak oil price, while others simply aimed to match or outshine the reports of their neighbours. It would explain why the reports in some countries, as for example Abu Dhabi, have since barely changed, despite production in the meantime: it being clearly implausible that new discovery would exactly match production. It might indeed have made good sense from the standpoint of quota negotiations to have a stable number unaffected by production.
Lastly, the former Soviet Union had its own system, based on an alphabetical classification with various subdivisions. The categories A + B + C1 are widely considered equivalent to the 'proved' + 'probable reserves' of the SEC classification, but decline studies of individual fields suggest that in fact they exaggerate by about 30%.
Another misleading practice is the uncritical use of Reserves-to-Production Ratios (R/P), quoted in years, whereby the indicated reserves are divided by annual production to suggest a given life-span. It is clearly absurd to postulate that production could stay flat for a given number of years and then stop dead, when all oilfields are observed to decline gradually. Still another unfortunate practice is to produce forecasts of supply and demand over relatively short spans, evading the implication that production would have to collapse immediately after the forecast period, if it were to respect the resource constraints.
The scale of confusion arising from the differing systems of classification and definition is self-evident, meaning that the various public and industry databases record widely different estimates. The principal public databases are those published annually by the Oil & Gas Journal and World Oil, which are based on questionnaires sent out to governments and industry around the world. As trade journals, they are not in a position to assess the validity of the information they receive. In addition, proprietary industry databases exist, principally those produced by IHS Energy and Wood Mackenzie. In earlier years, the former was compiled through close, albeit informal, cooperation with the major companies, but the task has become much more difficult as a result of the proliferation of small promotional companies and the growing role of State companies with a political agenda. There are also the data bases published by the oil companies BP and ENI, which are compilations from other sources, mostly not reflecting the company's own knowledge. Lastly, there is the present Survey, which brings together information provided by the WEC Member Committees, supplemented by other data obtained from governmental or industry sources.
Notwithstanding the many uncertainties regarding the validity of the data, it remains very important to make an attempt to establish the status of depletion by country, region and eventually for the world as a whole, so that governments may be in a position to adopt policies to prepare for the Second Half of the Age of Oil, when production and all that depends on it declines, owing to natural constraints that lie beyond economic or political influences. The objective should clearly be to establish a sound working model, while remaining prepared to revise and improve it, if and when greater knowledge and insight materialise. The steps to be undertaken in such a process can therefore be outlined. It is helpful to start with countries reporting more reliable data to establish the procedure, before facing the more difficult cases.
Step 1. Collect information on discovery by field, backdating any reserve revisions to the date of the original discovery, and collect information on exploration drilling. Plotting cumulative discovery against cumulative exploration drilling will produce a clear trend, which is normally hyperbolic because the larger fields are generally found first. Extrapolating the trend to an asymptote gives a good indication of the total to be produced in the country concerned (termed 'ultimate recovery'), subject to an economic cut-off for very small fields. (Fig. 2-4 )
Step 2. Collect information on past production and plot (annual divided by cumulative) against cumulative (the so-called derivative logistic plot) and extrapolate to zero, which also corresponds with the total oil endowment. In some countries, this plot delivers a firm trend that can be extrapolated confidently, but that is not always the case. Being based solely on relatively reliable production data, it avoids the uncertainties arising from unreliable reserve reporting. (Fig. 2-5 )
Step 3. Having established the total in Steps 1 and 2, subtract past production to deliver future production, which is divided into that coming from known fields (reserves) and that derived from new discovery. It is convenient to apply a percentage factor to deliver the reserve estimate such as to bear a reasonable relationship to the range of reported reserves, after deduction of any non-conventional categories. One option would be to take the average of the range, but on balance it is better to study the matter, so as to exclude any report that would otherwise distort the calculation of an average.
Step 4. Enter production, exploration drilling and discovery in a table, and calculate the depletion rate, being annual production as a percentage of what is left. This normally ranges from about 3% to 8%. If it were higher, there would be a case for re-examining the estimate of the ultimate recovery, which could be raised in order to deliver a more plausible depletion rate.
Having input the essential data into the model, it is time to forecast future production. In this regard, it is expedient to recognise three different categories of country as follows:
This group comprises those countries that have already produced more than half their indicated ultimate recovery, and are already in marked decline. Future production can be modelled on the assumption that it declines at the current depletion rate, namely in the range of 3-8% per year.
This group refers to those countries that have not yet reached their midpoint. Production has therefore to be assessed on the basis of the prevailing local circumstances, possibly being assumed to rise on the past trend to midpoint. On reaching midpoint it is assumed that production declines at the then depletion rate. Since most such countries are in fact within a few years of midpoint, the assumptions are not particularly critical to the overall model.
Middle East Gulf
This group comprises Abu Dhabi, Iran, Iraq, Kuwait, the Neutral Zone and Saudi Arabia, and presents the greatest uncertainty. They are major producers with exceptionally low depletion rates, meaning that, in resource terms, production could rise substantially. However in political terms, it seems reasonable to assume that they will prefer to hold production at current levels in order to maintain prices and to reduce the rate of depletion. They rely heavily on oil revenue and have good reason to adopt policies to make it last as long as possible. It would make sense therefore to assume that production in these countries will remain flat until the depletion rate rises to say 3% before the onset of terminal decline. Iraq is a special case, offering the possibility that production might rise to a more normal plateau should the political situation permit. There are alternative ways in which to address this group, for example treating them as swing producers, making up the difference between world demand and what the other countries can supply, but on balance the indications are that natural and investment constraints have limited their swing roles.
The various public reserve data illustrate, amongst other things, the wide range of estimates. The analyst producing a depletion profile will naturally take full note of these assessments, comparing them with such proprietary or confidential information as may be in his possession, in order to arrive at what seems to be a plausible and reasonable estimate. This is not an exact science, but calls for common sense to evaluate the trends, identify the anomalies and arrive at an acceptable answer. Some analysts may be discouraged at the lack of transparency and be reluctant to offer a conclusion without firmer foundations, but in a political sense it seems better to provide governments with a working model upon which they can begin to plan. Such a model is illustrated in Fig. 2-6.
For the reasons explained above, it is evident that the growth of oil production over the past 150 years must give way to decline as the resources are depleted. While this can hardly be denied, a debate rages as to the date of the peak. But in fact it misses the point, especially as it is not a high or isolated peak but simply the maximum value on a gentle curve. What matters, and matters greatly, is the vision of the long decline that comes into view on the other side of it.
Fig. 2-6 illustrates a plausible model, albeit one subject to revision as new, more reliable information comes to hand. It illustrates all the categories of oil, on which some comments are offered below.
This category has supplied most to date and will dominate oil supply far into the future. It is relatively easy, cheap and fast to produce, with production costs ranging from about US$ 5/bbl in the Middle East to US$ 10-15 as a world average. The price of oil has risen three-fold in the past few years, the increase representing profiting from shortage, as the production costs have not changed materially. Some 75% of it lies in giant fields (corresponding to roughly 1% of the number of oilfields worldwide), most ofwhich were found long ago. All the significant provinces have now been identified both onshore and offshore, although further small plays may yet be found in complex structural conditions, as for example in the thrust belts in front of mountain chains. The offshore too has been very thoroughly explored, having the advantage that high-quality seismic surveys may be shot relatively inexpensively. Drilling is also, somewhat surprisingly, often easier offshore than onshore, as problems of access in difficult terrain are avoided. Hopes are sometimes entertained that the former Soviet Union may have much left to find, but in reality the Soviet explorers were as efficient as their western counterparts and being free from commercial constraints could in fact plan efficient campaigns, even drilling purely to secure geological information. Indeed, the critical geochemical breakthrough that made it possible to identify the source rocks in detail owes much to Soviet scientists.
It is convenient to include in the Heavy Oil category dense and viscous oils as well as those derived from coal and immature source rocks, as they are all characterised by a high resource base but a low extraction rate and net energy yield.