This is a simple guide to help you determine the world's endowment and depletion of oil. Why would you need to know? For as long as you can remember, you were able to drive to the nearest filling station and fill your tank. Briefly in the 1970s, there were shortages when certain Arab countries imposed an embargo on exports as a weapon in the Yom Kippur War with Israel, but it was a passing incident, now long forgotten. Well, not quite forgotten, for lingering in the back of everyone's mind is the vague knowledge that oil is after all a finite resource that one day must run out. At the time of the oil shocks there was all sorts of Doomsday talk, but it proved misplaced as we have had plenty of oil for the past twenty years, and there are still many reports that this happy state will continue. Will it?

Think for a moment: what if? Oil may not run out for a long time, but can we rely on cheap oil such as we have known? Is it cheap? I hear you say. Well, the answer is that although oil prices are well above average producing cost, yielding profit to the companies and huge revenues to the producing and consuming states, it is very cheap in relation to its replacement cost. In fact it is almost infinitely cheap. We are now finding less than six billion barrels a year but using 23: if we lived like that in our personal lives we would be facing bankruptcy. Discovery peaked in the 1960s and it should surprise no one that thirty years on we face the corresponding peak in production. The sad truth is that we have to find it before we can produce it. So a great discontinuity is approaching.

It is hard to think of any aspect of our modern life that is not vitally dependent on cheap oil-based energy: driving to work; the tractor in the field; the ships on the sea; the airliners overhead; not to mention the military with its thirst for oil. It is the same whether we depend on a rural bus in central India or a commuter train into Manhattan. We all use oil and have come to depend on it. If it were suddenly to cost double, we would certainly notice; and the new economy that would inevitably result would surely affect our lives in many important ways. It is not impossible that the new ways would be better, but different to be sure. It certainly would be a discontinuity.

Should we seriously expect such an upheaval? Well, yes. If we look back we see that history is full of them both local and regional. For centuries in western Norway, they lived by catching herrings, and in due course learnt to preserve them by canning. King Oscar sardines were famous the world over. Nobody much noticed the arrival of the domestic refrigerator at the end of the war, but within a few years it had almost ruined the fish canning business. Frozen fish was easier, and for the canners it certainly was a terminal discontinuity. Who would have imagined the once all-powerful Soviet Union would so suddenly implode from internal tensions; and what a discontinuity that proved to be for the world's arms manufacturers, who figuratively had to turn to making plough shares that were much less profitable. It appears that economists dedicate themselves to studying the trends and patterns of trade in the periods of calm between the discontinuities but don't often successfully predict the arrival of a discontinuity. It is unfortunate because it is the discontinuities rather than the intervening calms that shape our destiny. In any event, we can say that those who fail to react to a discontinuity will suffer; those who react survive; but those who anticipate prosper.

So why should I, as a humble individual, care about such things, you may wonder? Surely there are better qualified institutions, world authorities and governments who are aware of the situation and getting prepared ? Unfortunately, there are not. Oil and oil price are hot potatoes. The politicians don't get votes for preparing for a crisis; only for solving one when it comes as a perceived Act of God. Furthermore, there are many vested interests with motives to conceal, confuse or be economical with the truth. Besides oil is a very slippery substance that is not easy to understand or measure.

So if you want to be counted amongst those who prosper from the coming discontinuity, you have no alternative but to figure it out for yourself. If it were easy you would have no advantage. No one can fully evaluate the situation and know exactly what to do, but if you are marginally better prepared than the uninformed you may still enjoy a great advantage.

So this is a step-by-step guide to help you through the maze.

In the summer of 1987 the tourist industry on the Adriatic coast of Italy faced disaster. The beaches were clogged with an evil smelling glutinous material; and the fishermen reported that the same stuff was clogging their nets and killing the fish. What had happened was that exceptional weather conditions in the near tide-less and current-free Adriatic had led to a huge proliferation of algae. These soft bodied microorganisms were extracting the oxygen from the sea and poisoning all life. The crisis lasted a few weeks before the weather changed. The glutinous masses sank to the seabed and things returned to normal.

This material is the origin of oil in a long and complex process. Simply stated, the process involves the following steps. First there have to be the climatic and physical conditions for prolific algal growth to take place. Second, the organic debris that falls to the sea (or lake) bed has to be preserved from oxidation by bottom-dwelling organisms or current action. Third, it has to be concentrated to provide the quantities needed to charge an oilfield. Lastly, it has to buried by younger sediments over geological time so that it becomes heated sufficiently to be converted into petroleum. Algal material is the main source of oil, but sometimes there are admixtures of other organic debris, including vegetal remains that are gas prone.

In the same way as the Italian experience of 1987 was a very unusual event, so the conditions that led to the deposition and preservation of prolific oil source-rocks were extremely rare in the geological record both in time and place. Most source-rocks were laid down in tropical regions, some of which were later moved to higher latitudes by the plate tectonic movements of the continents. Periods of global warming may have helped. The late Jurassic, 150 million years ago, was one such period, responsible for the main oil source-rocks in the Middle East, North Sea and parts of Siberia. Another occurred in the mid Cretaceous, 90 million years ago, and was responsible for the oil in northern South America. Much of the oil in the United States comes from older sources, such as the Permian , 230 million years ago.

Conditions for the preservation of oil were equally rare. The main requirement was stagnant water to prevent the organic material being oxidized. The North Sea for example was a stagnant rift during the Jurassic, something like the present Red Sea, that was opening as the North American continent split apart from the European landmass and began to drift westward. The Middle East region was an extensive marine carbonate platform with internal stagnant sink holes and lagoons ideal for preserving the organic material. Sometimes lakes were satisfactory environments, as along the west coast of Africa before the Atlantic opened.

In global terms, the bulk of the oil occurs in a geological province known as the Tethys, a zone of rifting between the southern and northern continents, of which the Mediterranean and Gulf of Mexico are relics. Segments of what were previously in this zone were subsequently moved northward, explaining the paucity of oil in the southern hemisphere.

The world has now been so extensively explored that all the large oil provinces have been found, and the scope for finding an entirely new one of any size is now greatly reduced if not entirely removed. The question of source-rock is critical and explains the easily made observation that oil fields are clustered together in clearly defined geological trends, separated by huge areas that are entirely barren of oil.

Although the early explorers had a general idea that oil came from dark-looking shales, it is only in the last twenty years that advances in geochemistry have made it possible to clearly identify where and when it was generated. This knowledge made the world a much smaller place, emphasising the very finite nature of the resource.

References:

Once the organic material was buried to a depth of 2000-3000 m depending on the geothermal gradient, it was heated sufficiently of chemical reactions to transform it into oil and gas. The reactions involved expansion which together with the difference in density between the oil and the water preserved within the rocks led to the onset of migration. The oil and gas gradually moved upwards through the pore space in the rocks or along fractures until it encountered a porous and permeable interval, such as a sandstone layer, along which it could more easily flow.

The sediments in the oil basins have been generally folded and faulted by earth movements. The oil migrates upwards along the porous conduits until it is trapped at the top of a fold, against a fault or where the bed carrying it pinches out. Geologists and geophysicists can map these structures and determine if they are closed and sealed, such as to form prospects for drilling. Sealing is an important factor, because over long periods of geological time, oil and gas tends to dissipate and leak from the traps that once held it. Salt deposits form the most effective seals, and their widespread presence in the Middle East and for example the Permian Basin of Texas contributes to the prolific production of these areas. Oil and gas may migrate vertically to be trapped above where it was formed, or it may migrate laterally depending on the geological circumstances. In some areas, large quantities have migrated to shallow depths around the margins of the basins, where the deposits have been attacked by bacteria that leave behind bitumen and heavy sticky oil. If the oil source-rocks or traps are buried too deeply, the high temperatures crack the oil to gas. Finding more oil is not simply a case of drilling deeper.

This short description is sufficient to explain how oil and gas are formed and how once formed they migrate, much to be lost, but some to be trapped in accumulations large enough to be exploited. Probably only about one percent of the oil generated has collected in accumulations large enough to be produced. The older the oil, the greater the risks of it being lost. Each individual field has its unique characteristics reflecting the long and varied geological circumstances to which it has been exposed. The oil itself varies greatly in composition and physical property for the same reason. This diversity explains some of the confusion surrounding how much there is.

Reference

The family of hydrocarbons is a large one, and each member has its own endowment in nature, its own characteristics and above all its own depletion profile. Ideally we should identify each of these species and consider its contribution, although in practice that is not easily done on public domain data

It is common to distinguish two broad categories of oil and gas: conventional and non-conventional (also called unconventional), although there are no standard definitions. Generally speaking, conventional oil refers to those categories which have yielded most production to-date and will continue to dominate supply until well past peak. Much flows freely from giant fields, found long ago.

The production of conventional oil rises fairly rapidly to a peak and then declines exponentially, whereas the production of non-conventional oil rises only slowly to a long low plateau before eventually also declining.

Non-conventional oil, including particularly the heavy oils and tar of Venezuela and Canada will become increasingly important in the future but it will be no substitute for the oil we have known. Most of it will be available only at great cost, including environmental cost, and at comparatively low rates of production. We are not running out of oil - only cheap oil.

To summarise, we may recognise five categories of petroleum

5.1. Conventional Oil

5.2. Non-conventional Oil (also known as unconventional)

5.3. Natural Gas Liquids (NGL):

5.4. Conventional Gas

5.5. Non-Conventional Gas

Some authorities define non-conventional petroleum as that which is not yet exploitable in economic or technical terms, which means that the goal posts continually move, making it impossible to quantify the amounts. It is much better to define the terms on physical characteristics, recognizing that some non-conventional (about 2 Mb/d) is already in production.

Conventional oil is quite cheap to produce. About half of what remains lies in just five countries around the Persian Gulf where actual production costs are less than about $5/b and likely to remain so for a long time to come. Of course, it is difficult to know what the actual cost is because it depends on accounting practices, which differ from country to country and between enterprises. Even the expensive North Sea and Alaskan oil probably does not cost more than about $12/b, and there is no particular reason why it should get significantly more expensive. The difference between the cost of producing oil and buying it in the market is of course the substantial amounts of tax levied by producing and consuming governments.

Cost is not therefore a serious constraint. What will determine the price is the scarcity of the resource itself, which prompts us to ask how much there is and how available it is. The question we should NOT ask is how long it will last. That depends on how quickly the old fields decline and how demand behaves. A tank may hold 100 gallons, which if it tipped over empties in a second, but if it springs no more than a tiny leak might take several days to empty. Oil does not lie in a huge underground cavern, which could be drained quickly, but is disseminated in the minute pore space of the reservoir where it is partly held in place by capillary pressure. Production in all oilfields declines during the latter half of their lives as the physical constraints of the reservoir impose increasing restrictions to the flow of oil to the wellbores. We will return to this all important question of depletion, but first let us determine how much there is.

No one can directly measure the amount of oil in a reservoir in an oilfield far underground, which will be known absolutely on the day when the field is finally abandoned. At that point cumulative production, namely the total produced, will equal what are termed the initial (or original) reserves and also the ultimate recovery. Prior to that date, it is necessary to use geological and engineering data to estimate the size of the field. Estimates are naturally subject to change with more knowledge and in some cases new technology. Furthermore, estimates vary depending on who makes them, for what purpose, and under what definitions. It is not an exact science, and there are no hard and fast rules for how reserves should be reported. Much of the confusion over the size of the world's endowment of oil relates to differences in the reporting procedure and the definitions, stated or implied.

When the explorer identifies a prospect before drilling, he maps the size of the trap usually with the help of seismic surveys that provide something akin to an X-ray of the geological formations underground. He then uses regional knowledge to make some assumption about the likely reservoir, and calculates the pore volume occupied by oil, which will give him a value for what is known as oil-in-place. To this he applies a recovery factor, say 40%, and another factor to adjust the volumes for surface temperature and pressure to arrive at what he estimates to be the amount recoverable, namely the Reserves. These factors are influenced by the gravity of the oil: the heavier the oil, the lower the recovery. If the amount is judged to be viable after taking into account the development costs, the wildcat well will be drilled on the prospect.

If it is successful, a new estimate of its reserves will be made based on the information from the initial well, which might be above or below the first estimate. In practice, it is normally below the first estimate, which is often somewhat exaggerated by the explorers to encourage their backers to take the risk. Once the discovery is made engineers take over. They want to be quite sure that they have the reserves to justify the huge investments that follow. It is natural that they should prefer a conservative number.

The traditional practice is to recognize three categories of reserves: Proved for what is sure; and Probable and Possible for less or unsure additions as the terms imply. For financial purposes, only the Proved category counts, and in the United States, SEC regulations require that it be confined to the drainage area of a producing well, normally 40 acres. It is not surprising that Proved Reserves tend to grow over time, which is called Reserve Growth. There is thus a sort of U-curve of reserve reporting: the number starts high before drilling, falls to a low on discovery and then gradually recovers as the field is produced to come somewhere close to the original number. Furthermore, the techniques for estimating change over time. At first the estimates relate to the mapped volumes of the reservoir, but later production performance from the wells themselves can be used to plot the decline and so extrapolate to the ultimate recovery. Computer based simulations and material balance calculations are further tools. What we are after are median probability reserves (P50) namely the best estimate of what will actually be produced taking into account the impact of the application of known and foreseen technology.

One hears a lot about increased recovery as a means by which to add more oil supply.

As touched on above, the reported reserves of a large offshore field at the point of development may be as little as half those estimated prior to drilling the wildcat responsible for its discovery. It is no surprise that in such cases the reported reserves grow over time as the field approaches exhaustion when by definition its reserves will match a P50 probability ranking.

This apparent growth has confused many analysts who are in position to observe only the changing national reserve numbers and are distant from the actual business. They commonly attribute the increase to technological advances and management skills, and they are certainly encouraged so to do by many vested interests. In reality, the increase is little more than a move from a conservative to a realistic estimate. That said, it is of course true that knowledge and experience of the reservoir and its management improve over time along with continuing advances in technology, so that some dynamic increase may also be involved. Such anticipated progress is however incorporated into the P50 estimates. It is not easy to make accurate P50 estimates, but it is obvious that statistically they will show neutral revisions with as many down as up. Systematic upward revision speaks of initial under-reporting. Figure 1 shows the decline of production in the giant Prudhoe Bay Field. It has been at a constant rate since 1990, demonstrating the minimal impact of new technology. The Ultimately recovery of about 12 Gb is probably little different now from when it was first calculated by the operator twenty years ago.

A secondary related issue is the notion that technology leads to an increase in the recovery factor over time. It is said that average recoveries were once about 30% but have now risen to 50% thanks to technology. This does not bear close examination either. First, the early 30% number was little more than a rule of thumb, because few fields had then been abandoned so that no one knew what they would deliver, and second, the tools available to map the oil-in-place, on which the recovery factors are based, were less sophisticated than now. In fact, recovery factor is a consideration only at the exploration and development planning phase. Later, recovery is based on the performance of the wells themselves, and no one knows or cares what percentage that might be of a notional amount of oil-in-place, which cannot in any event be measured accurately because much of it remains forever in the ground. Recovery factor is primarily related to the quality of the oil and the reservoir, and has little to do with technology. Naturally, if a conservative early estimate of oil-in-place is compared with what the wells are actually found to deliver late in the field’s life there will be an apparent improvement.

Governments normally collect data on reserves without stating the probability ranking, which they publish for various purposes, but in recent years these reports have become increasingly unreliable in many countries. Certain countries overstate; some understate; and others simply do not update their reports to account for production which inevitably eats into reserves unless matched by new discovery. For most purposes, the validity of the reserves of a country are not a particularly vital issue, with most eyes being primarily interested in short-term production. But knowledge of the reserves is a critical part of determining the future trends, and is indispensable to a study of this sort. We need to know how much oil has been found and when, if we are to have a sound basis for projecting the future availability.

The issue of dating reserve revisions is even more important than the size of the revisions. Clearly if the revisions simply reflect a move from a cautious to a realistic estimate the revision has to be backdated to the discovery of the fields, since nothing was dynamically added. Proper backdating is essential to determining the real discovery trend.

What information is there and how can we use it? If the numbers are not reliable can we nevertheless interpret them usefully? The following are the main sources of information.

9.1. Petroconsultants in Geneva

9.2. The Oil and Gas Journal

9.3. World Oil

9.4. BP

9.5. US Geological Survey

We need to know how much conventional oil there is, and we need to know the rate at which it is being produced, found and depleted.

There are three key elements on quantities

Remembering always to identify the category of oil concerned of which Conventional oil is the most important. From these three elements, three further parameters can be derived:

Most of these elements, save Drilling and Discovery Rates, can be input from the indicated public domain information with such interpretation and adjustment as the analyst feels confident in making. For Drilling and Discovery Rates it is necessary to turn to Petroconsultants.

We will now discuss each of these important elements and give some numbers for consideration. It is useful to secure the historical trends back to, say, 1930 with the earlier statistics being lumped together as a single pre-1930 total, as will be seen later.

In principle, this is a straightforward concept : how much has been produced. However it is important to check the treatment of statistics of countries whose frontiers have changed, and to consider war-loss as in Kuwait, where about 2 billion barrels are thought to have been lost. It should in principle be treated as production, although not very useful production, because it eats into the reserves and ultimate endowment. Information on the former Communist countries is unreliable, and naturally the early production was not always properly recorded. Probably many countries are not correctly reporting their current production either. It is also important to try to determine if Natural Gas Liquids (NGL) are included or excluded in the statistics, particularly in the case of the United States.

Reference

A practical way to proceed is to take the figure of 698.6 Gb for end 1992, as reported by Masters of the authoritative US Geological Survey, which is itself probably derived from Petroconsultants material and then subtract annual production data from the Oil and Gas Journal back to 1930, lumping the balance into a single number for pre-1930. It also needs to be updated in the same way for the years since 1992.

Reference

Campbell has already done this in:

Reference

A good number for end 1997 is 795 Gb (billion barrels), excluding Alaska being treated as non-conventional see Figure 2.

We will need Cumulative Production by country and year when we come to analyse the depletion of oil. Production inexorably eats into reserves: don't forget it!

There is nothing straightforward about the subject of reserves. It is a mess. Nothing can be accepted at face value because different definitions are used and in recent years governments have been providing unreliable information. There is nothing absolute about reserves. As discussed earlier, the term reserves means the amounts to be recovered from known fields at the reference date, which is only a fraction of the oil-in-place in such fields.

Then there is the whole issue of reserve revision, as already discussed. For most purposes it does not matter whether the revision is taken on a current bases or backdated as if known at the time of the discovery. For our purpose it is critically important to backdate so as to obtain a valid picture of the discovery pattern as a basis for predicting future discovery. Failure to backdate is the cause of many misconceptions in the public domain.

To make any real inroad into this issue, it is necessary to access the Petroconsultants database that lists reserves by field, properly backdated. In the absence of that information, we may nevertheless still make some useful progress by interpreting the public domain information, using for example the Oil and Gas Journal data.

Before going further, it is worth commenting on the most blatant distortions, as already discussed by Campbell in

References:

What seems to have happened is that the eastern division of the state oil company in Venezuela decided in 1987 to add 20 Gb of heavy oil reserves, which had been discovered long before, on the strength of a pilot project even though no large scale development was in place. We treat such reserves as non-conventional. In any event it had the, not necessarily intended, consequence of increasing Venezuela's OPEC quota, and prompted the main Middle East producers to retaliate with huge arbitrary increases, adding overnight some 200 Gb. At the same time it has to be remembered that their reserve numbers had been understated previously by the companies prior to their expropriation. The important point is that nothing happened in terms of technology or knowledge of the reservoir in 1987: such reserve revision as was technically valid should have been backdated to the fields concerned which were discovered more than 30 years before in most cases (see Figure 3). Mexico has also confessed to exaggerating by the inclusion on non-conventional oil in the Chicontepec Field. Numbers for the FSU and China are unreliable, as are those from the increasing number of countries — 63 in 1997 — that report the same number year after year, which is clearly implausible.

There is no easy way to unravel these difficulties. Campbell has used a practical short cut of applying a factor to convert the reported Oil and Gas Journal (after adjustment by subtracting the amount produced during periods of unchanged reporting) to Median Probability (P50) reserves. In principle, such reserves take into account any technically justified revision. Other factors could be applied by analysts with different understandings or knowledge. It is also possible to use arbitrary alternatives to test the sensitivities.

On this basis, world reserves are assessed for end 1997 at 823 Gb, which is about 200 Gb less than the 1020 Gb reported in the Oil and Gas Journal and other such sources. It is distributed as shown in Figure 2.

The sum of the Cumulative Production and the Reserves gives the total discovered to date, namely 1618 Gb. Again, we need to access Petroconsultants' material to determine when and where it was found. A useful clue is however provided by the published listings of giant fields, namely those with in excess of 500 Mb of initial reserves (i.e. ultimate recovery) in the following references

The definition of giant fields is also not without its difficulties. Some single fields are divided for arbitrary reasons, such for example if they cross a national or concessional boundary. In other cases, fields are combined for administrative or other reasons. It does not matter for most purposes, but for us it is important because we want to analyze the field size distribution, as discussed later.

The published data suggests that 60-70% of the world's known oil occurs in little over 300 giant fields. The discovery of such fields peaked in the 1960s. Most are found early in the exploration of a new area, simply because they are too large to miss. No particularly advanced technology is required. Few have been found in the past decade and very few remain to be discovered in the future, unless some entirely new province is unearthed, which is now very unlikely (save for the Caspian), as the world has been so thoroughly explored.

Figure 4 shows the plot of giant fields and Figure 5 shows the distribution of discovered oil. About 40% is in just five countries around the Persian Gulf.

In earlier years, the world was a large place and the possibilities for finding oil in the numerous sedimentary basins that make up the continents, especially the margins, and the continental shelves, seemed almost infinite. Several earlier studies took the rock-volume of these huge tracts as a basis for estimating how much oil there was. But in the 1980s came the geochemical breakthrough, already described, which made it possible to determine exactly where oil was generated and when it migrated. Only structures in communication with such generating belts are potential traps for oil. The world became a much smaller place with this new knowledge. It means in effect that most of what will be found lies in ever smaller accumulations in the already known oil producing basins.

While we can estimate the Yet-to-Find directly, it is better for the purposes of the depletion model to derive it indirectly by subtracting the Discovered-to-Date from the Ultimate, first because some of the techniques relate to the total distribution; and second because it is better to keep the Ultimate as a constant, until it needs to be revised significantly, instead of having to adjust it every time a small discovery is made anywhere in the world.

Published estimates of the Yet-to-Find apparently range widely, but in most cases the range reflects the use of different definitions, which have to be closely examined before the numbers can be properly understood. For example, Masters, speaks of a range of range from 292 to 1005 Gb, quoting a Mean value of 470 Gb, but close reading of the text reveals that these are based on notional geological criteria, ignoring economics, the timing or the number of wells required to find them. They evidently include what are here treated as non-conventional reserves. Only the low end of this range deserves serious consideration for these reasons and when a find rate of less than 6 Gb/a is taken into account. Campbell in his latest assessment gives 182 Gb as yet to find, which considering the present falling discovery trend of 6 Gb/y probably means that it will take something like 50 years to find.

Figure 2 shows the distribution based on this estimate. Almost half lies in the Middle East Gulf, the FSU and China, where exploration is unlikely to be a high priority for a long time to come.

No one should dismiss the technological achievements: they will be needed to the full to find what remains which occurs in ever smaller and more difficult circumstances. And no one should be misled into thinking that technology can find what is not there to be found.

The bottom line is how much oil remains to produce. It is equivalent either to the sum of the Discovered-to-Date and the Yet-to-Find or in other terms, the Ultimate less the Produced-to-Date (Cumulative Production). It stands at 1006 Gb (or a rounded one trillion barrels) As shown in Figure 2, about half lies in the five countries around the Persian Gulf. Other input estimates will give other results, but the important point is to recognize the framework and methodology, including in particular the discovery rate. It will be found that there are serious constraints to much higher numbers, although such are often quoted.

The Ultimate Recovery, or Ultimate, of a field is its Cumulative Production when production ends, which is the same as its Initial or Original Reserves before production began, provided that reserve revisions are backdated. It may be determined initially by mapping the volumes of the reservoir or later by projecting the performance of the wells when they are in decline. It is the same for a basin, country, region and eventually the world as a whole, always remembering to define the category of oil being considered, here conventional oil only. It is an important cornerstone of the study. The number is in principle a constant, whereas production, reserves and discovery change every day, although it is naturally subject to periodic revision, either up or down, depending on the evolving knowledge of the endowment.

There are several ways in which to determine an Ultimate value.

Experienced explorers with access to the worldwide files of an oil company can may reasonable estimates based on their knowledge of the underlying geology in relation to the maturity of exploration.

2. Record

Something is to be learned by studying the trend or scatter of published estimates. As shown in Figure 6. the range is narrowing. But again it is necessary to check carefully into the definitions. For example, the recent estimate of Masters of 2300 Gb could be restated as about 1800 when it is realized that his Reserves are High Case (10% probability of occurrence) and only the low end of his Undiscovered is realistic in terms of discovery and drilling rate projections.

Plotting cumulative discovery against time or against wildcat wells gives a hyperbolic pattern, because the larger fields tend to be found first, and the asymptote approximates with the Ultimate subject to a cutoff for uneconomic small fields.

Plotting size against rank of fields in a natural domain plots as a parabola on a log-log format. Once a segment of the distribution is complete or nearly so, namely when no more large fields are found after a reasonable period of time, it defines the parameters of the parabola and hence the ultimate distribution, including the fields yet to find. It works best for a single petroleum system, but can give useful results for larger groupings, such as giant fields or continents that become natural domains in their own right.

References

Much can be learnt by plotting cumulative discovery by size class and by relating discovery peaks with their corresponding subsequent production peaks. Declining discovery is reflected in declining production after a time-lag, and that can be extrapolated to exhaustion, which corresponds with the Ultimate. It is an evolution of the well known Hubbert curve.

The three latter methods rely on data which is available only in the Petroconsultants database, and those working in the public domain are forced to rely on considering the record of published material or of trying arbitrary numbers to achieve a plot of sensitivities to alternative assumptions. The lowest estimate in recent years is Campbell at 1650 Gb, since increased to 1800 Gb, which is about the same as Masters' US Geological Survey estimate, once its definitions have been decoded. Estimates of more than 2000 Gb almost certainly include much non-conventional oil. Some oil companies, and for example Townes, the former President of the American Association of Petroleum Geologists, take the scatter rather than the trend of the same published estimates to propose numbers as high as 2500-3000 Gb, but they may have motives other than objectivity.

Reference

The reason for the fall in the published estimates since the high of 3550 Gb in 1969 is the radical fall in giant discovery and the disappointments from many offshore areas, whose potential was over-rated when the offshore was first opened. In reality the range of opinion is not that great.

In addition to these six elements that relate to quantity, it is necessary to consider three more that take into account timing. It is not enough to know how much oil remains to produce but we need to know when it can become available given the physical constraints of the reservoir.

Depletion Rate is defined as annual production as a percentage of the amount Yet-to-Produce at the end of the preceding year. It tends to increase in a country until the midpoint of depletion, when it stabilises and equates to the Decline Rate, the percentage change in production from one year to the next. The reason is that in practice it is mainly controlled by the large early fields, which are already in decline, giving a composite Depletion Rate, which effectively masks the impact of any late stage small discoveries.

Figure 5 shows the trend of discovery, with reserves properly backdated. The inflection to falling discovery in the mid 1960s, despite the high levels of subsequent drilling and the technological advances, is highly significant. Extrapolating this curve is a cogent argument in support of the proposed Ultimate.

Discovery Rate is the amount found each year in genuine new field wildcats, that is to say, exploration wells drilled on prospects that if successful will yield entirely new fields. This information is really only available from Petroconsultants on a global basis, see Figure 7. You may be tempted to try to derive it by comparing successive year end reserve reports after adding the intervening production. This does not work because the reserve reports commonly include reserve revisions, that are not discoveries, as is especially the case in the United States.

It is important to consider the number of new field wildcats drilled each year, to plot the trend of such drilling and to determine the discovery rate per well. Extrapolation of these factors imposes a serious constraint on the estimate of the Yet-to-Find. It is effectively available only from Petroconsultants on a global basis.

This concludes the list of the elements we need to have in building a depletion model.

12. DEPLETION MODEL.

Now is the time to try to use the data we have collected to build a depletion model recognizing that production from all oilfields declines during the latter half of their lives. The reason is that as each wellbore drains its catchment area, it has to draw oil from farther and farther away. The oil has to make its way through a network of minute pore spaces and constrictions and overcome the capillary pressures. Furthermore the drive mechanism in the form of expanding gas or encroaching water under a hydrostatic head may lose some of its effect as depletion proceeds. Normally an individual well's production profile is asymmetrical, rising rapidly to a peak and then declining exponentially. But the profile of a field depends on the rate at which new wells are added. There is a law of statistics that explains why a composite profile of asymmetric profiles tends to become symmetrical. The profile of a country is much the same. It depends on the rate at which new fields are added: it being normal for the larger ones to be found first.

Both theoretically and empirically there seems to be a good case for building a model in which the production peaks at the midpoint of depletion, namely when half the Ultimate has been produced.

So, why don't you try to construct one for the country in which you live as a test case. You will have to research the input data described above, paying particular attention to trying to understand exactly how the official data are defined, as already discussed. You want to aim at securing Median Probability reserves, which may be termed Proved & Probable. You probably wont find much published on the undiscovered potential, so you can either use the estimates in Figure 2 or use your own judgment. Perhaps you can find a knowledgeable explorer who can give the benefit of his experience, perhaps in some local professional society.

You will need a personal computer and a spreadsheet. We may look at the profile of a hypothetical mature country, which is well past its assessed midpoint of depletion, see Figure 8. The practical steps in filling out the spreadsheet formulae are as follows:

This is an example of a country past its midpoint, but you may face a country that is not yet at midpoint. In this case you either input future production if known from forecast developments until midpoint is reached, or make an arbitrary assumption such as a 5% annual increase to midpoint. As shown in Figure 9, most countries, except the five Middle East producers are past or close to midpoint, and an arbitrary assumption of a 5% increase to midpoint is not material to the model.

If you have come this far CONGRATULATIONS! You have assessed your first country. You may now like to make a Line-Bar graph plotting production over time, and inserting the giant fields as bars. If you look at the following reference you will see all countries covered in a similar way

Reference:

and if you want an updated much more sophisticated version with all discoveries, not only the giant fields and also critical well data and projections, you could consult

Reference

Having done one country, you can do all the rest in the same way, but it is convenient to divide up the world into a number of regions such as Latin America, Africa etc. and then consolidate the data from each country into regional totals. Two groups of countries have to be treated differently:

13.1 "Other and Unforeseen". It is convenient to consider in the above fashion only countries in production and having an Ultimate of greater than 500 Mb as described above. Other less important countries and those not yet in production can be lumped together as a group. It is helpful here to add a certain assumed Ultimate for them and the "unforeseen"so as to bring the total of the other regions up to a rounded world Ultimate. In this way, say, about 50 Gb can be held back for unforeseen discoveries or reserve revisions anywhere. By doing so, the world Ultimate remains constant, not having to be adjusted for every minor change in individual countries. Campbell in the most recent assessment proposes a world Ultimate of 1800 Gb.

13.2 The other issue is the identification of what we may call swing countries. The most uneven distribution of the amount of Yet-to-Produce oil means that some countries are at a much earlier stage of depletion than others. So they are able to maintain or even increase their production when the others are already in decline. Thus the swing countries can make up the difference between world demand under alternative scenarios and what the other countries can provide because of their resource limitations. There are several options for how to identify the swing countries. Probably the simplest and best is to limit it to the five countries around the Persian Gulf which together own half of what remains and are geographically in the same area. They are Abu Dhabi, Iran, Iraq, Kuwait and Saudi Arabia (also including the small Neutral Zone because it is owned by Kuwait and Saudi Arabia). Although there are many serious conflicts between these countries they do have much in common, not least their religion, and apart from Iran, a shared heritage in having been part of the Ottoman Empire prior to the First World War. Other options would be to consider OPEC as a swing producer, which would make sense if it did represent the larger exporters and if it did act as the swing producer it seeks to be. In resource terms its members differ from each other widely. A third option would be to expand the Persian Gulf group to include, say, Venezuela, Mexico, Libya and Nigeria.

Once the swing group has been identified, steps can be made to calculate its production and share under alternative world demand scenarios. Many different scenarios could be investigated such as:

14.1 Low Case : world production stays flat until the world midpoint is reached and then declines at the then depletion rate

14.2 High Case : world production increases at 3% a year until the Swing Share reaches a certain percentage, say 40%, when higher prices may be assumed to lead to a plateau of production until the share has reached 50% by which time these country too will be close their midpoint

14.3 Base Case : as the High Case, save that world demand grows at a slower rate, say 2% per year, and the plateau sets in sooner, say when swing share passes 30%.

Once the scenario has been selected the total swing production and share can be readily calculated and then reallocated back to each swing country individually, always recognizing that each country will be forced into decline at its midpoint, with the balance then being taken up by the other swing countries not yet at midpoint.

Figure 10 illustrates such a model. It will be noted that the several production scenarios converge over time because more today means less tomorrow given a finite total. Figure 11 shows the relationship between swing share and oil price.

The scenarios in Section 14 are driven mainly by resource constraints, although the swing share is expected to a critical factor around peak.

Many short term political scenarios affecting the issue can be considered: some advancing the onset of decline, others delaying it. Such factors apply however only over the next 10-15 years after which decline will be imposed by immutable resource constraints.

It is obvious that the Middle East swing countries are in a parlous political condition, and the future developments there are hard to predict. Oil price is critical to these countries as oil contributes the bulk of their revenue. This is not the place to review the issue in detail but some general comments can be made.

The world is not running out of oil: only cheap oil. The resource constraints are however serious and not widely understood. A doubling of oil price by about 2000 is not only on the cards, but highly probable given the growing control of world supply by a few countries. The next oil price shock will be due to circumstances very different from those that gave the shocks of the 1970s when prolific new production from alternative sources was in sight. This time it is not.

Production costs will remain similar to those of today because most of the oil comes from old giant fields which are cheap to produce. So, the price of oil is mainly determined by tax, levied either by the producing or consuming governments.

Given the pending resource driven decline in production, the world has every incentive to curb further increases in demand and take active steps to prepare for the day not more than a few years off when cheap oil, such as has driven the 20th Century's economy, is no longer available.

One great contribution would be to try to improve knowledge of how much oil has been found already which is critical to assessing how much remains to be found. So, if you can find out about the reserves of oilfields in your country or neighbourhood and feed in the information, we can build up a file of compelling evidence.

In your own lives you have many opportunities to start thinking of how you would personally react to a doubling of oil prices. There are many beneficial steps you could begin to take. The most beneficial would probably be a new mind-set to consider these issues.