Contents:  Introduction, Conclusion, Table of Forecasts, Notes to Table, Discussion of The Limits to Growth, References




One of the main reasons that people, oil experts in particular, are disinclined to believe the situation forecast by current global oil depletion calculations is their conviction that past oil forecasts have been wrong, particularly those made in the 1970’s. This view sees the present calculations as just another example of ‘crying wolf’. 1


On examination, it turns out that most reputable oil forecasts made in the 1970’s were substantially correct.


Oil forecasts made in the 1970’s nearly all fit into one of four classes.

  -  General, non-quantitative, fears of global supply scarcity, based on the experience of shortages that occurred during the oil shocks.

  -  Predictions of global oil would run out (i.e., reach exhaustion) in 30 years or so, based on the then-proved oil reserves of about 30 years’ worth of current production.

  -  Predictions of oil global exhaustion in a much shorter timescale, based on the then proved oil reserves (or some assumed larger amount), but with growth assumed to rise at a fast exponential rate, as had been the case until fairly shortly before the shocks.

  -  Predictions that global oil would reach a production peak (very different to oil running out) around the year 2000.


It was this fourth view that characterised the forecasts from nearly all reputable organisations at the time, and which was also reflected in many textbooks and monographs on energy published at the time (see a summary of some of these in the Table of Forecasts, later in this section).


This fourth, ‘production peaking’, forecast was based on:

  -  the then well-accepted estimate for the world’s conventional oil ‘ultimate’ (i.e., original endowment of recoverable oil), of roughly 2000 Gb;

  -  the knowledge that global production peak would not occur until something like half of this, 1000 Gb, had been used;

  -  the knowledge that only ~300 Gb had been consumed at that date;

  -  the assumption that production would follow an ‘unrestricted’ logistic (‘Hubbert’) production profile.

On this basis, the global midpoint was calculated to lie around the year 2000, (a precise calculation by Hubbert giving the date as 1996).


In the event, global demand was substantially curtailed by the price rises of the oil shocks, and an unrestricted logistic profile was not followed; with the result that the estimate of conventional ultimate of around 2000 Gb (still to-day, for this purpose, the best estimate to use) simply moved the global conventional oil production peak to around 2010. This is illustrated in the following Figure.


   [FIGURE 11: ] 1970’s ‘Logistic Curve’ Forecast, and Actual Demand

This graph is for ‘narrowly-defined’ conventional oil (i.e., it excludes oil with API < 17.5, oil in waters deeper than 500m, oil in polar regions, and NGLs).

The graph shows:

          -  Global oil discovery by year (finds prior to 1930 shown in 1930);

          -  An unconstrained logistic (‘Hubbert’) curve with an area of 1800 Gb;

          -  Actual production to 1999;

          -  Estimated future production, also with an area of 1800 Gb. (The short

plateau in production ( ~2000 to 2008) is based on the assumption that price will curb demand.)

As can be seen, the high prices from the oil shocks lopped the top off the unconstrained curve, and shifted the date of peak by about 10 years.

        Source:  C.J. Campbell.




It is true that there have been calculations in the past warning of oil’s approaching scarcity that turned out to be misleading. Some, like many of those enumerated by Butler of the DoE, correctly indicated that production from a particular region or country was soon to decline, but overlooked the scope for new discoveries elsewhere. Other have been clearly erroneous, such as a CIA forecast from the 1970's that was predicated not on resource limits, but on assumed structural decline in the Soviet Union; or an early forecast by the UK’s UKOOA that was based only on areas so far licensed. Still others, as mentioned above, predicted oil exhaustion based on only proved reserves, or assumed the continuation of very high growth rates in demand.


But, as pointed out above, nearly all the forecasts made by reputable organisations in the 1970’s combined mid-point peaking arguments with realistic estimates for the world’s original endowment of conventional oil. Hence these forecasts gave, in quantitative terms, exactly the same warnings of the ‘wolf’s’ approach as given by to-day’s oil depletion calculations: that global conventional oil production will peak when roughly 1000 Gb has been produced.


These are warnings it would be wise to heed.


Table of Forecasts of World Oil Supply


Date of Forecast


Forecast Date of Conventional Peak

Assumed Ultimate*




“Oil to become increasingly scarce from about the year 2000.”

   2100 Gb



Report for the UN

  Confr. on Human


  likely that peak production will have been reached by the year 2000.”

   2500 Gb



SPRU, Sussex

University, UK


   1800 – 2480 Gb



UK Dept. of


Peak: “about .. 2000.”





Peak: 1996.

   2000 Gb (Nehring)



Ehrlich et al.

Peak: 2000.

   1900 Gb




“.. plateau within the

       next 25 years.”




BP (Oil Crisis …


Peak (non-communist world):  1985.




World Bank

“.. plateau around the

       turn of the century.”

   1900 Gb




Peak: 2005.

   1800 Gb, (excl. NGLs)




Peak: 2010.

~ 2000 Gb




Peak: 2020.

   2836 Gb



IEA: WEO 1998

Peak: 2014.

   2300 Gb refnce. case



USGS (Magoon)

Peak: ~2010.

~ 2000 Gb




Peak: ~2010.

   2000 Gb (incl. polar,





Peak: 2004, or 2019.

   2000, or 3000 Gb



IEA: WEO 2000

Peak: “Beyond 2020.”

   3345 Gb (from USGS)




Peak: 2016 - 2037.

   3003 Gb (from USGS)




Peak: 2003 - 2008.

~ 2000 Gb




Peak: 2011 - 2016

   2180 Gb




Peak: 2004 - 2011

   1950 - 2300 Gb equiv.


* Gb = billion barrels.


Notes to Table

[1]. The Ecologist. ‘A Blueprint for Survival’, Penguin, London, 1972; see pp 18 and 130. This report looked at the impact of continued exponential demand growth on oil’s lifetime, but also presented the ESSO forecast given here. (As mentioned above, the calculations of the 1970’s did not foresee the global demand reduction from the oil shocks, so assumed production would rise to peak at about 100 million barrels per day. This put the conventional oil peak earlier than if based on actual demand.)

[2]. B. Ward & R. Dubos, ‘Only One Earth: the Care and Maintenance of a Small Planet, Penguin Books, UK, 1972, p 184. This was a landmark report. Its status was ‘an unofficial report commissioned by the Secretary-General of the United Nations Conference on the Human Environment’, Stockholm, 1972. A committee of 158 extraordinarily eminent ‘scientific and intellectual leaders from fifty-eight countries served as consultants’ in the report’s preparation. The full extract (p184) is: “One of the most quoted estimates for usable reserves [of oil] is some 2,500 billion barrels. This sounds very large, but the increase in demand foreseen over the next three decades makes it likely that peak production will have been reached by the year 2000. Thereafter it will decline.”

[3]. H.S.D. Cole et al., Eds. Thinking about the Future: A critique of 'The Limits to Growth', Science Policy Research Unit, Sussex University, Chatto & Windus, 1974. This quotes a variety of estimates of ultimately recoverable oil reserves made mostly in the 1960's, including those of Hubbert in 1969 and Warman in 1971.

[4]. W. Marshall. Energy research and development in the United Kingdom, Energy paper No. 11, UK Department of Energy, 1976; p 12.

[5]. M.K. Hubbert. Project Interdependence: U.S. and World Energy Outlook Through 1990. Congressional Research Service, Washington, 1977, p 624; quoted in: ‘The Global 2000 Report to the President’, Penguin’, 1982, p 353. Hubbert took Nehring's world ultimate oil reserves estimate of 2,000 billion barrels, and assumed that oil production would be limited only by resource availability. On this basis, he calculated that world production would reach a peak at about 100 Mb/d, around the year 1996.

[6]. P.R. Ehrlich, A.H. Ehrlich and J.P. Holden. EcoScience: Population, Resources, Environment. W.H. Freeman, San Francisco, 1977, ISBN 0-7167-0567-2, pp 400-404. A widely-quoted textbook. 2  The authors calculated a ‘Hubbert’ peak based on the ‘high-estimate’ for global conventional oil endowment of 10,900 trillion MJ (~ 1900 Gb). (Interestingly, the book also draws attention to the then-controversy which led to the USGS revising down, by a factor of 3, its estimates for US undiscovered recoverable oil and gas.) [As a side comment, it is probable that the famous Simon vs. Ehrlich, Harte and Holdren ‘commodity price bet’ failed in large measure because of the more than two-fold fall in real-terms oil price (reflected also in other energy prices) between 1980 and 1990; energy being a large component of mineral extraction costs. Since the high price of 1980 was driven, fundamentally, by the US oil peak nearly a decade earlier, the conclusions generally drawn by economists from the outcome of that bet probably need revision.]

[7]. A.F. Beijdorff. ‘Energy Efficiency’, Group Planning, Shell International Petroleum Company, London, April 1979; p 1. (Current modeling suggests the world peak may be fairly sharp, rather than the long plateau suggested in this Shell study.)

[8]. BP report Oil crisis ... again ?, published in 1979. In terms of UK views, this report is one of the more significant of the examples of 'failed' forecasts that people (e.g., J. Mitchell, P.R. Odell) choose to quote. It indicated that non-communist world oil production would peak in 1985. This forecast bears examination. The first step is to add back in communist production. Then, like other forecasts of that time, the report assumed rising production when high prices were in fact reducing demand. Adjusting for this, and for the subsequent increases in production of NGLs and non-conventional oil, makes the resulting prediction look reasonable; forecasting a fall in global conventional oil production from around the year 2000.

[9]. The World Bank. Global Energy Prospects, World Bank Staff Working Paper No. 489, 1981. See pp 37, 46. The report said: "The bulk of the world's reserves, principally in the Middle East, was built up in most part during the 1960s. Despite increased incentives to explore for oil provided by higher prices, conventional oil production is projected to reach a plateau around the turn of the century.” (Note that by the early 1980’s, the impact of the demand reduction was becoming visible, and hence the calculated global peak date, for a given assumed ‘ultimate’, falls later.) Elsewhere, p 46, this quotes ultimate recoverable oil reserves as being 1,900 billion barrels, and says: "According to some estimates, the world's ultimate recoverable gas reserves are at least equal to [those of oil]".

[10].  C.J. Campbell and J.H. Laherrère. ‘The World’s Supply of Oil, 1930 – 2050’. Report from Petroconsultants S.A., Geneva, 1995. (See also: C.J. Campbell & J.H. Laherrère, The End of Cheap Oil, Scientific American, March 1998, pp59-65.) This is one of the more detailed studies to date, and is the basis for the information provided in this website. As explained in earlier sections of this website, this study used full access to the standard industry oil resource database to carry out analyses of hydrocarbon reserves, with those in particular countries requiring adjustment. It also used a range of statistical approaches to assess the yet-to-find, and the logistic model to generate future hydrocarbon production. As critics have pointed out, this study did not explicitly include the effects of technology or price on the assessments of regional and global ‘ultimates’. But the authors maintain, with considerable supporting evidence, that price and technology have only a limited effect on the size of these ‘ultimates’, at least as they affect calculations of production peak date.

[11]. L.F. Ivanhoe. Updated Hubbert Curves Analyze World Oil Supply. World Oil, Vol. 217, No. 11, November, 1996, pp 91-94. Used USGS discovery data, and the fact that production has to largely mirror discovery. A clear warning of the problems to come.

[12].  J.D. Edwards. Crude oil and alternative energy production forecasts of the Twenty-First Century: The end of the Hydrocarbon Era. AAPG Bulletin, vol. 81 pp1292-1305, 1997. A reasonable study, but limited by lack of access to industry data, so arrives at a high global ultimate.

[13].  The International Energy Agency ‘World Energy Outlook’; published Nov. 1998; ISBN 92-64-16185-6. Used the 1994 USGS mean estimate for global conventional ‘ultimate’, of 2300 Gb, for its reference case. It also used a low case of 2000 Gb, (based on the Petroconsultants report) and a high case of 3000 Gb (based perhaps on Odell’s data). The rate of discovery that would support the high case ‘ultimate’ was not examined. The study did not specifically account for the impact of likely price and technology developments, though it did examine the scope for non-conventional oils to come on-stream.

[14].  L. Magoon. USGS Open File Report, 00-320 Version 1. The main USGS 2000 survey (Ahlbrandt et al.) examined total oil available (basin ‘oiliness’), but did not look in detail at the rate at which these resources can be discovered. Magoon of the USGS endorsed data in the Scientific American article by Campbell & Laherrère on the rate at which the resources can become available.

[15].  C.J. Cambpbell. Oil Reserves and Depletion. PESGB Newsletter, Petroleum Exploration Society of Great Britain, March 1999, pp 87-90. A partial update of the 1995 Petroconsultants report. It analysed polar & deepwater oil separately, but added these back in the full analysis.

[16].  A.A. Bartlett. An analysis of US and world oil production patterns using Hubbert-style curves.  Mathematical Geology, 32/1, pp1-17, 2000. Bartlett does not have access to the industry data, so predicted peak based on these two assumed values for the conventional ‘ultimate’.

[17].  The International Energy Agency. ‘World Energy Outlook’, 2000. Used the USGS 2000 survey mean oil-plus-NGLs ‘ultimate’, including reserves growth, of 3345 Gb. The IEA state that such data are “authoritative”, but, as mentioned above, the data pay no attention to the rate that such oil can be discovered. Note that USGS 2000 data include a large allocation for reserves growth, contrary to the decision of the previous survey. The USGS team has subsequently re-evaluated its approach of basing global reserves growth on the US’ experience.

[18]. US Energy Information Administration website, 2001. Uses the USGS 2000 mean ‘ultimate’ of for conventional oil (excluding NGLs, but including reserves growth), of 3003 Gb. If the world decline rate is taken as 2% p.a., this puts peak at 2016. If a much steeper (probably unrealistic) decline rate of 10% p.a. is assumed, this puts the peak later, at 2037.As with the IEA 1998 World Energy Outlook above, this study uses USGS 2000 survey results in an uncritical manner, both on the rate of discovery of oil, and on the scope for reserves growth outside the U.S.

[19]. K.S. Deffeyes. ‘Hubbert’s Peak’, Princeton University Press, 2001; ISBN 0-691-09086-6. Uses a range of statistical techniques, based, essentially, on the discovery trend curve indicating the likely ‘ultimate’. This study has no direct access, we believe, to the industry database.

[20]. M.R. Smith. Analysis of Global Oil Supply to 2050. Consultancy report from The Energy Network, March 2002. Production estimates are based on detailed country by country exploration analyses, and use individual depletion curves to meet calculated ‘ultimates’, rather than simple ‘mid-point peaking’. Includes data on the non-conventionals, and expected oil price forecasts. Global ultimate is 2180 Gb, making the global peak in 2011 if global demand is assumed to rise by 2%/yr.; or 2016 at 1%/yr. growth.

[21].   ‘Nemesis’, in a contribution in ASPO/ODAC Newsletter, Issue 15, March 2002. This study generates a range for the dates of peak production, based on cumulative production to-date; plus reserves and ‘net discovery’ data from Campbell and BP’s Schollnberger. This approach avoids the need to use specific estimates of ‘ultimate’, but yields the approximate ‘equivalent ultimates’ listed in the Table.


The ‘Club of Rome’ Report: Limits to Growth


Because of its importance in many people’s perception of resource limits, it is useful here to also discuss the Club of Rome report: The Limits to Growth, (D.H. Meadows, D.L. Meadows, J. Randers and W.W. Behrens III, Earth Island, 1972.)


This report was a key contributor to the 1970’s understanding that resources are finite; that man’s use of these could reach limits within comprehensible time spans; and that the complex interactions between resources, population, capital and pollution demanded system thinking if a proper understanding is to result.


Prior to the report, oil use had been growing at around 7% per year, and the calculations of the Club of Rome correctly showed that if this sort of growth rate were to continue, a resource base of almost any feasible size would be exhausted in a surprisingly short time-span. The lesson, still true today, is that unfettered exponential growth is unsustainable.


The authors gave a table (p 58) listing the then-current proved reserves of various minerals, including oil at 455 billion barrels. The authors recognised that the figure they gave for each mineral represented only the resource discovered so far, and suggested that a larger amount, up to perhaps six times as much, might represent the total useful quantity of that mineral. (In oil’s case, co-incidentally, six times 455 Gb is roughly correct for conventional oil’s original endowment, i.e.,‘ultimate’).


But the authors made no use of these current resource numbers in their modeling. Instead they assumed, in their 'standard computer run', that all non-renewable resources, lumped together, had a resource base in 1970 of 250 years' supply at 1970 rates, (p 126). The standard run then showed that society would collapse in less than a hundred years due to resource depletion, itself driven by:

  -  population growth,

  -  compounded by an increasing per capita use of non-renewable resources,

  -  and further compounded by the assumption that the material capital to extract the

resources increases as the resources themselves are depleted.

Finally a point is reached where too little capital is left for future growth, as investment cannot keep up with depreciation (p 125), and the industrial base collapses, taking food and service production with it. If the authors doubled the resource base (p 127), society still collapsed, now primarily due to pollution limits, but also to severe restraints on resource availability.


Interestingly, in the sequel: Beyond the Limits (D.H. Meadows, D.L. Meadows, J. Randers; Earthscan, 1992), estimates are given for oil's ultimately recoverable reserves (as opposed to then-current proved reserves given in the previous book), an acceptable range of 1,800 - 2,500 billion barrels (Table 3-2, p 71). But the authors appeared unaware of the Hubbert 'peaking from the mid-point' argument.


Overall, the key perceptions about the Club of Rome’s report (despite the details of its simulations) are that, since no major resource shortages have appeared, the report must be fundamentally flawed; that forecasting resource limits is a fool's game; and that man's ingenuity and skill will always overcome the outdated Malthusian nightmares of resource depletion. The report is due for re-consideration.



1.  Numerous references. Recent ones include: Lord Lawson to a British meeting of energy economists; and BP’s Professor Peter Davies in the 2002 UK House of Lords report (op. cit., p 79).

2.  Other influential books from the 1970’s, at least on this side of the pond, include:

  -  G. Foley, with C. Nassim. The Energy Question, Penguin Books, Middlesex, ~1975. This contains a fascinating discussion of the then-generally available data on oil resources; including an early understanding of apparent discrepancies in the data from Professor Odell.

  -  J.G. McMullan, R. Morgan and R.P. Murray. Energy Resources and Supply, Wiley, 1976. This has an excellent graph, Figure 1.3, showing the possible future production from a wide range of fuels, including fission and fusion. For conventional oil it shows a peak soon after the year 2000. (Professor John McMullan is now at Ulster University, and was Chairman of the DTI’s ‘Foresight Programme’ Energy Futures Task Force);

   -  G. Leach et al. A Low Energy Strategy for the United Kingdom, Science Reviews, London, 1979, ISBN: 0-905-927-20-6. Page 9 has: “Forecasts show energy needs rising implacably, with widening energy gaps appearing around the turn of the century as oil and gas production begin to decline.”  (Gerry Leach is now with the Stockholm Environment Institute, and is based in London.)


Updated: 23/April/2002