Uppsala University

International Energy Agency accepts Peak Oil

in Peak Oil by

An analysis of Chapter 3 of the World Energy Outlook 2004

Kjell Aleklett, Professor of Physics
Uppsala Hydrocarbon Depletion Study Group, Uppsala University, Sweden

The International Energy Agency, IEA, has released a new report, World Energy Outlook 2004, WEO2004 [1], which has been reviewed by the New York Times on October 27 under the title: Oil Demands Can Be Met, but at a High Price [2].

A month ago, at a conference in Abu Dhabi, I met Fatih Birol, the Agency’s Chief Economist and principal author of the report. During a coffee break, we discussed the upcoming release, and he stated that we in ASPO would probably appreciate it. The New York Times article was, therefore, disappointment, indicating that the IEA had once again copied the grossly implausible energy outlook released by the US Energy Information Administration (EIA) in April [3]. The New York Times message was that world oil demand would grow by about 50 percent to 121 million barrels a day by 2030. Chapter 3 in the Outlook, Oil Market Outlook, covers this forecast, but careful reading delivers an entirely different message, namely that peak oil is around the corner. Indeed Fatih Birol’s remarks were accurate: the Outlook does now start to discuss crude oil supply the same way as we do in ASPO [4], referring to peak-oil, backdating, creaming curves, and the other techniques that we have been using. The problem with Chapter 3 is that the reader needs to have detailed knowledge to understand the true message.  Below, I will try to give the required code, the Uppsala code.


In summary, the IEA states: Production of conventional oil will not peak before 2030 if the necessary investments are made. But we find in Chapter 3 that a peak on this date is premised on the USGS Mean estimate of 2626 Gb (billion barrels) for remaining conventional oil (IEA include NGL, Natural Gas Liquid, in conventional oil), adding that if this estimate proved too high, the peak of production would come by 2015 or before . It is critical to note that the IEA now accepts the notion that there is a peak in oil production, even if there is uncertainty as to the date. The range is from 2015 to 2033, coming even sooner if all the assumptions are not fulfilled. It follows that Governments are now on notice that they must make energy plans for the future that accept peak oil as a reality. That will be a departure from past practice.

Historical Background

To understand the report, it is necessary to take into account the background of the IEA and see how it compares with OPEC ( The Organization of the Petroleum Exporting Countries ), which is an international organization of eleven countries that rely heavily on oil revenues as their main source of income. OPEC was formed in 1960, and its current members are Algeria, Indonesia, Iran, Iraq, Kuwait, Libya, Nigeria, Qatar, Saudi Arabia, the United Arab Emirates and Venezuela. Its stated objective is to co-ordinate and unify petroleum policies among Member Countries, to secure fair and stable prices for petroleum producers; an efficient, economic and regular supply of petroleum to consuming nations; and a fair return on capital to those investing in the industry. In the early 1970s, certain Arab OPEC countries embargoed exports for a few months in support of the Palestinians facing Israel, and the world entered an oil crisis.

The IEA was established in November 1974 in response to this oil crisis as an autonomous inter-governmental entity within the Organization for Economic Cooperation and Development (OECD) to study energy supply and security, and advise the member nations accordingly. In general, OPEC is accepted as a political organization with its statements being colored accordingly. What many do not realize is that the IEA is also a political organization and that its reports are highly colored by the perceived best interest of its members.  It, therefore, becomes necessary to try to lift the political veil from the contents of the Energy Outlook 2004.

Political agenda

The political agenda is laid out in Figure 1, Oil Demand and GDP Growth.

Oil Demand And GDP
Figure 1: Relation between Oil Demand and GDP Growth, figure 3.1 in WEU2004

It shows a strong correlation between oil demand and GDP growth, save during the 1970s and 1980s when most of the world’s nuclear power plants come into operation reducing the demand for oil to fuel electricity generation. When the increase of nuclear power was phased out in many countries, oil demand accordingly grew by about two percent to offset the decline of nuclear power and deliver the necessary growth of GDP. The current study presents of a modest rate of economic growth at 1.6 percent a year. Starting from 77 Mb/d (million barrels a day) in 2002, this rate of increase means that production would have to rise to 121 Mb/d by 2030. Politically, it is not possible for the IEA to present an outlook contemplating less.

Let us evaluate how plausible such a forecast might be. In 2002, the World produced 77 Mb/d or 28 Gb/a (billion barrels per year), which has to rise under the IEA forecast to 44 Gb/a in 2030.  It implies that total production from 2003 until 2030 will amount to 1020 Gb. That is more than was consumed during the 20th Century and is in the same order of magnitude as the existing reserves.  It does not sound plausible.


In Table 3.1, the Outlook presents oil demand by region and for some large countries. It is surprising to see that Russia will have an increase of only 1.6 percent, matching the world average. Russia is the only major country in the world that is energy independent and can stay independent for many years to come. The population has adequate education and the potential for economic growth is great, being no doubt at least comparable with that in China and India, namely around 3.0 percent. The IEA cannot admit to such an internal growth rate because it would limit exports, undermining the possibility of meeting the indicated needs of the rest of the world. A 3.0 percent increase for Russia from today’s 2.7 Mb/d would give a consumption of 6.2 Mb/d in 2030 compared with the 4.2 Mb/d proposed by the IEA. It is evident that the IEA has had to massage its input assumptions to deliver a pre-desired outcome.


Shell’s spectacular downward revision of reserves during 2004 has raised awareness of the general issue global oil reserve reporting. We know that oil reserves are far from evenly distributed, and we know that oil fields have widely different sizes. It is not surprising that it has been easier to find a giant field with more than 500 million barrels than smaller fields.  The trend of the discovery of giant fields is revealing, being illustrated in figure 2.

Giant Oil Field
Figure 2.Discovery in giant oilfields per decade and the number of such fields [5].
The bulk of the Worlds oil lies in these giant fields.  Since the world has now been very thoroughly explored, it is inconceivable that a dramatic change in the trend will occur in the 21st Century. Adding the smaller fields and back-dating the revisions gives the plot as published by ExxonMobil [6]. In Figure 3 we present backdated discovery together with consumption.

Past Discovery
Figure 3. Backdated world discovery of crude oil in comparison with production.

The decline in discovery has been relentless for decades, being also depicted by the IEA in the following Figure 4

New Discovery
Figure 4. Additions to world proven oil reserves from the discovery of new fields and production (figure 3.14 in WEO2004).

Plotting the discovery trend per decade on a logarithmic scale, Figure 5, allows extrapolation into the future, clearly demonstrating that 300 Gb will not be found during the next 30 years.  A more plausible estimate is between 150 and 200 Gb.

Future Discovery
Figure 5. Extrapolation of the discovery of new fields for the next 30 tears.

Another way to evaluate this declining discovery rate is to look at the so-called creaming curve, which matches cumulative discovery against cumulative new field wildcats, the boreholes that either do or do not, find new fields. The IEA Outlook presents such a study for the period 1963 to 2002 for oil and gas, clearly showing that it becomes ever harder to add reserves by discovery.

Creaming Curve
Figure 6. Cumulative oil and gas discovery as a function of wildcat drilling (figure 3.16 in WEO2004).

It is evidently difficult for the IEA to accept the obvious truth that most of the Worlds oil has been found. It tries instead to blame the oil companies and producing countries with large reserves to for not trying hard enough.

The Outlook reviews as follows the validity of the reserves as reported in different journals and the BP Statistical Review, confirming the unpredictable nature of the reports as has long been emphasized by ASPO. The Outlook comments:

According to BP, reserves increased dramatically in the 1980s and 1990s, from 670 billions barrels at the end of 1960 to 1147 billion barrels at the end of 2003 (Figure 7). But most of the increase occurred in OPEC countries, mainly in the Middle East, in the second half of the 1980s. Saudi Arabia and Kuwait revised their reserves upward by 50%, while Venezuelan reserves were boosted 57% by the inclusion of heavy oil in 1988. The United Arab Emirates and Iraq also recorded large upward revisions in that period. Total OPEC reserves jumped from 538 billion barrels in 1985 to 766 billion barrels in 1990. As a result, world oil reserves increased by more that 30%. This hike in OPEC countries estimates of their reserves was driven by negotiations at that time over production quotas and had little to do with the actual discovery of new reserves. In fact, only a few exploration activity was carried out in those countries at that time. Total reserves have hardly changed since the end of the 1980s.

BP reserves
Figure 7: Proven oil reserves in different regions according to BP Statistical Review (Figure 3.10 in WEO2004)

Aleklett and Campbell describe the same thing in reference [4].

The main OPEC countries expropriated the holdings of the foreign companies during the 1970s, following the precedent of Iran’s action against BP in 1951. State enterprises were formed to produce the oil, inheriting the technical data and reserve estimates from the private companies. In 1985, Kuwait reported a 50% increase in its reserves overnight although nothing particular changed in the oilfields. It did so to increase its OPEC production quota, which was based on reserves. Then three years later, Venezuela doubled its reported reserves by the inclusion of large amounts of long-known heavy oil that had not previously been reported. It led Iraq, Iran, Dubai, Abu Dhabi and later Saudi Arabia to retaliate with massive increases to protect their quotas. Some revision was called for, as the earlier estimates were too low, having been inherited from the private companies before they were expropriated. But the revisions, whatever the right number might be, have to be backdated to the discovery of the fields concerned, which had been made up to fifty years before. In total about 300 billion barrels were added in this way during the late 1980s, significantly distorting the apparent discovery record. It is noteworthy too that in several cases the reported reserves remain implausibly unchanged for years on end despite production. It is staggering that such obviously flawed information is recorded in the public database, substantially without comment or qualification.

Table OPEC reserves

The following statement from the Outlook is explicit: Reserves data from the main public sources are not backdated. In other words, the historical time series of reserves for a given country is not revised in light of new discoveries or increases in recovery rates. The need for proper back-dating has long been stressed by ASPO and figure 8 shows the backdated data from reference [4].

Back Dated Cum Disc
Figure 8: the Current trend of the discovery of crude oil according to BP compared with backdated revisions.

Many analysts assume the estimates reported in the BP Statistical Review have at least the tacit blessing of a knowledgeable and prestigious oil company. They do not, as footnotes in the publication make plain.  A derivative of the trend reported by BP in figure 7 is shown in Figure 9. This curve reaches a maximum around 1990 when the real peak, as confirmed by Exxon Mobil, was in the 1960s, amply demonstrating the flawed nature of the BP material. That is hardly surprising since it is directly compiled from other published sources, not reflecting the company’s considerable knowledge.

Disc Per Year
Figure 9: Discovery per year according to BP Statistical Review.

The Big Provider

At this stage of the analysis, it must have been difficult for IEA to find a way to make a forecast that would fulfill the political agenda of having production at 120 Mb/d in 2030. The Reserve data, as discussed above, does not support such a proposition. But it was fortunate to be able to find a life-saver in the form of the United Stated Geological Survey. The IEA lends credibility to this institution by describing it as the most authoritative source of estimates of global ultimately recoverable resources of conventional oil.  The emphasis is on the words most reliable source, which is far from claiming it to be the most accurate source. The USGS commenced its study of world oil following the shocks of the 1970s, and for many years issued sound evaluations at consecutive meetings of the World Petroleum Congress.  But a departure came with the study of 2000, under the project’s new director, claiming a Mean estimate of the total discovery to 2025 of 3.3 trillion barrels. The following Figure illustrates the record of some 65 prior estimates by major oil companies, institutions and the USGS itself, which average 1.93 trillion barrels, indicating that the latest USGS estimate is far from the consensus.

Ultimate Recovery
Figure 10: Summary of reported estimations of ultimate recovery of crude oil

Jean Laherrere has extrapolated the discovery trend in the United States itself as illustrated in Figure 11, showing that the USGS estimates are far above the trend [6].

Figure 11: US Cumulative Production and Ultimate Recovery from Energy Information Administration with USGS 2000 forecasts for the USA.

Dr. Jeffery Johnson from ExxonMobil was invited to the ASPO conference in Berlin in May 2004. He is the Operations Manager for Production Geosciences responsible for Europe, North America, Southeast Asia and Australia. He declined to answer a question of why his company was not vigorously drilling for oil in the United States, given that the USGS predicts that there is more than 80 billion barrels are there to be found before the year 2025. The USGS also exaggerates the world potential as evidenced in the following figure.

USGS and World
Figure 12: The world’s cumulative production and mean discovery for conventional oil and condensate. The average estimate of 3021 Gb of just crude oil from USGS year 2025 is at the limit of the figure. The oil and condensate estimate of 3345 Gb cannot be fitted in the figure.

But if we read the IEA Outlook carefully, we find a strong hint that they realise that the USGS Mean estimate is implausibly high, as they add the important qualification: If ultimately recoverable resources at the low end of the current range of estimates and reserve additions are slower than expected, conventional oil production will peak within the next two decades.

Table 3.3 gives the range of the USGS predictions

Table USGS

It introduces the concept of the statistical range of probability. Evidently, the 95% case has the highest probability of being confirmed, while the 5% case is near meaningless. However valid regarding statistical theory, it poses the question of the validity of the Mean value in the real world.  It is important to understand this factor. The USGS study itself may not be as flawed as it seems, remembering especially that the USGS does not itself forecast what will be found and placed on production. It is rather that the Mean value of what is assessed to be available for discovery under a system of subjective probability ranking is widely misunderstood.

The trends of new field discovery have been discussed above. In its reference scenario giving 120 Mb/d by 2030, the Outlook implies that as much as 939 Gb will be found in new fields. It contrasts with the reality of experience, as the record since 1996 has given less than 150 Gb, suggesting by extrapolation that no more that between 150 and 200 Gb will be found over the next 30 years.

The other questionable issue is so-called reserve growth. The growth of past officially reported reserves mainly reflects the reporting practices of the entity responsible. Oil companies are subject to strict stock exchange rules leading to understatement, whereas OPEC countries have had motives to exaggerate as they compete for production quota based on reported reserves. In any event, the growth of reserves has little impact of peak production, as illustrated by the example of the Prudhoe Bay field in Alaska, illustrated below, Figure 13. The growth in reported reserves typically coincides with the onset of decline, prolonging the fields life by extending the tail end of production.

PB reserves production
Figure 13: Reserves growth in Prudhoe Bay compared with production per year. The reserves growth cannot prevent the decline in the oil field.

Future oil production

In conclusion, the Outlook illustrates its forecast of future production in Figure 14.

Future Production
Figure 14: World oil production by source (figure 3,20 in WEO2004)

The figure is not explained in detail, but the IEA does discuss the decline in existing fields, concluding that it ranges from 5%  to 11% per year. Starting at 75 Mb/d in 2003 and ending with 15 Mb/d in 2030 is equivalent to a decline rate of close to 6%, which we can certainly accept.

The study adds the telling comment by 2030 most oil production worldwide will come from capacity that is yet to be built.  We may examine the possibilities of that occurring in the real world. New discoveries are indicated to yield 25 Mb/d. The North Sea had a peak production of 6 Mb/d, meaning that a production level of 25 Mb/d implies the discovery of four new provinces equivalent to the North Sea, the largest new area found since the Second World War.  The Outlooks claim that this can be achieved over the next 25 years certainly stretches credulity.

Table 3.3 gives remaining reserves and reserve growth. Around 650 Gb will be consumed from 1996 to 2030, which sounds plausible, leaving the balance for the period following 2030. It implies that Development of existing reserves and Enhanced oil recoveries will have to come from the reserve growth listed in Table 3.3. But this means that Reserve Growth will have to rise to levels similar to-days production, which seems singularly implausible.

Production by region

The last difficulty for the IEA was to find where on the Planet they could reach the oil supply of 120 Mb/d. Table 3.5 in the Outlook gives the answer.

World Oil Production

To secure world production of 77 Mb/d 2002 and 121 Mb/d 2030, it is necessary to combine Non-OPEC, OPEC, Non-conventional oil and Processing gains. In the following discussion Non-OPEC will be divided into Transition Economies (typically called FSU, Former Soviet Union) and the rest of the world. In Figure 15 we present discovery and production for FSU.

Figure 15: Production and discovery of crude oil in FSU.

The discovery of the 10 Gb field in Kazakhstan year 2000 was the largest oilfield found in the world during the last decade, but compared with the yearly world consumption of 30 Gb we can conclude that this field only can support the World for four months. This field can be compared with the Prudhoe Bay Field in Alaska.  Production can reach the market only via pipelines and rail transport, whose capacity will set the level of offtake. One pipeline with the capacity of 1 Mb/d is under construction from Azerbaijan to Turkey, and another with capacity of 0,3 Mb/d links the Caspian with the Black Sea. Prudhoe Bay has 2 Mb/d capacities, and Kazakhstan will have around 1.5 Mb/d. It is a realistic level of new field production.

Elsewhere in FSU, discovery reached a maximum in the 1960s.  In estimating future production, it is necessary to bear in mind that the fall in production following the collapse of the Soviet Government was anomalous. The increase seen today largely reflects making good the production that was lost during this sudden fall, and not discovery. Future production should, therefore, be considered as the continuation of the overall trend that peaked in the 1980s at 12.4 Mb/d.  Future production, which many like to present as production from new fields, is in reality mainly coming from declining old fields.  Furthermore, the Russian government is now moving to take closer control of its precious oil and gas reserves, seeing merit in conserving them for as long as possible

The Outlook presents a very optimistic view of future oil production in Russia as can be seen in Figure 16. To start the record in 1990 obscures the longer-term trend, which gives a very different picture reflecting a previous peak.

Russia IEA
Figure 16: Production, consumption and net exports from Russia from the Outlook.

Even with a slow 1.6% increase of consumption in Russia will have to face the possibility for export. As Russia has to support itself, it is reasonable to expect a 2% to 3% increase in the consumption. Together with a realistic decline rate in Russia, we can estimate that the volume of export can be limited. In Table 3.5 the Outlook suggest an increase in the production for FSU from 9.5 Mb/d in 2003 to 15.9 Mb/d in 2030.  A more realistic assessment is given in Figure 15.

Figure 17: Production of oil from the Rest Of the World, ROW. For countries that have passed the peak production the year for the rollover is given. The other countries are expected to reach rollover within the next decade.

Subtracting FSU from Non-OPEC and we obtain a value for what we may term the rest of the world, ROW. Werner Zittel, the member of ASPO, has studied it in detail [7].  In figure 17 illustrates how production from individual countries adds up. Around 70% of ROW production, 25 Mb/d, comes from countries that have already passed their peak of production and are accordingly set to decline at 4% to 5% per year to 2030 when they will be able to produce around 8 Mb/d. The other group with a production of 10 Mb/d can at best have a 1% decline per year to 2030, giving production of 8 Mb/d. Adding these numbers gives 16 Mb/d for the rest of the world, compared with the Outlooks 27.5 Mb/d (43.4 Mb/d minus 15.9 Mb/d).  It is evident that the Outlook forecasts are clearly unrealistically high.

The next group to consider is the OPEC countries. In the reference scenario, the estimated production should increase from 28.2 Mb/d to 64,8 Mb/d, an increase of 130%. The Outlook lists the demands on individual countries and the influence of the IEA outlook, Figure 18, is evident.

EIA and ME
Figure 18: Production Capacity year 2001 and demanded production from US Energy Information Administration in units of million barrels per day.

Two months earlier Saudi Aramco had in Washington presented what they called the Maximum Sustainable Capacity, MSC. With a production of 12 Mb/d, it claimed to have a sustainable production until 2033, see Figure 19. It is far from 22.5 Mb/d, proposed in the Outlook.

Saudi Aramco
Figure 19: Data presented by Saudi Aramco.

On a recent visit to Abu Dhabi, I had the possibility to visit onshore oilfields. Discussion with operation managers gave Sustainable Capacity another dimension. They explained that the fields needed water injection to provide pressure to move the oil to the production well. If this injection is forced the total amount of oil that is possible to produce from the field will be lower compared with a sustainable production. To be able to produce 250000 barrels per day in the Bab field they need to inject 450000 barrels of water. Part of the water will be pressed through to the production well and just now the produced crude oil contents 4% water. As water flows easier than oil, you will destroy part of the field if you force is. Al old oilfields in the Middle East have about the same ration of water injection. The demand on the oil-producing countries in the Middle East to double the production is, in reality, a demand to partly destroy the possibility for future generations to get revenues in the future. We should be thankful that field managers and other responsible persons now speak up and explain the reality and us that need to import oil have to make plans according to this.


We may conclude that the International Energy Agency has delivered a forecast designed to fulfill the Agency’s political agenda, but between the lines, it clearly sets out enough evidence to show that it is indeed no more than a political statement, far removed from what is attainable in the real world. The critically important message that emerges is that peak oil will come shortly and that a peak in 2030 is nothing more than a political posture.

The Outlook does include a Low Resource Case, which advances peak to 2015. It goes on to add what it terms a High Price Scenario, which advances the date of peak to close to the present day, confirming the ASPO assessment, reproduced in Figure 20.

UHDSG 2004
Figure 20: The 2004 oil and gas liquids as presented by Uppsala Hydrocarbon Depletion Study Group [8]. Oil from Deepwater, Polar Regions, and Natural Gas Liquids are included in the Outlook as part of different countries production of oil. We have accepted the sustainable production scenario for the Middle East. As seen we are including production from tar sand etc., but we think that the increase will be slower than IEA. The increase in the polar production around 2030 is for discoveries not yet made, but we believe at they will start to drill in Alaska and find something in Russia. (Figure with better resolution )
Figure 21 illustrates the situation graphically, showing the fall of production from known fields and the desperate need to secure new production by a discovery or reserve growth, which in reality is beyond the physical limits of the Planet. It is an oblique way of describing rank shortage, and the interesting thing is that this graph is from The Lamp, the information to shareholders in ExxonMobil [9].

Exxon Mobil
Figure 21: Information to shareholders in ExxonMobil


This article opened with a quote from the New York Times, and it is fitting to end with a quote from an article from Daily Telegraph on October 31 by Liam Halligan, The West is deluded to rely on Saudi oil [10]. In the article, you find an interview with Al-Husseini, who was head of exploration and production at Aramco until March this year. Aramco is the state-owned oil company that accounts for 97% of Saudis crude production.

“The question isn’t, can we can pump 15m or 20m barrels daily?” Al-Husseini says. “The question is, how long it can be sustained? We could only manage 22 Mbps for a very short time – maybe ten years. And that would mean an awful lot of depletion, which isn’t in the best interests of the global economy.”

What does Al-Husseini say about the estimates from EIA, the US Energy Information Administration, of future Saudi production? “These are US numbers, not ours,” he says. “The American production outlook is much too high.”

Where is EIA going wrong? The IEA focuses only on demand (in this article called political agenda). That is why they overestimate not only future Middle East supplies but non-OPEC and Russian supplies too.

It is defiantly a support for the analysis presented in this article.

Confucius, said 500 BC: Study the past to define the future. We in ASPO have studied the past as carefully as it is possible and with that knowledge as a starting point, we have now defined the future.


[1]. World Energy Outlook 2004, IEA Publications, ISBN 92-64-1081-73-2004
[2]. Oil Demands Can Be Met, but at a High Price. New York Times October 27, 2004
[3]. International Energy Outlook 2004, Guy Caruso, Thirty-First Annual International Energy Conference, Boulder, Colorado, April 19, 2004.
[4]. Kjell Aleklett and Colin Campbell,
[5]. Work in progress, Uppsala Hydrocarbon Depletion Study Group, Uppsala University.
[6]. Jean Laherrere, private communications
[7]. Werner Zittel
[8]. Colin Campbell
[9]. The Lamp
[10]. Daily Telegraph on October 31 by Liam Halligan, The West is deluded to rely on Saudi oil.

Kjell Aleklett is Professor of Physics at Uppsala University in Sweden where he leads the Uppsala Global Energy Systems Group (UGES).

Leave a Reply

Your email address will not be published.