Energy Efficiency (2010) 3:115–122 DOI 10.1007/s12053-009-9063-9

The expected dates of resource-limited maxima in the global production of oil and gas R. W. Bentley

Received: 15 December 2008 / Accepted: 17 November 2009 / Published online: 18 December 2009 # Springer Science+Business Media B.V. 2009

Abstract Combining geological knowledge with proven plus probable (“2P”) oil discovery data indicates that over 60 countries are now past their resource-limited maxima of conventional oil production. The data show that the maximum of global conventional oil production is imminent. Many analysts who rely on only proven (“1P”) oil reserves data draw a very different conclusion. But proven oil reserves data contain no information about the true size of discoveries, being variously underreported, overreported, and not reported. In addition to conventional oil, the world contains large quantities of nonconventional oil. However, many current detailed models show that past the conventional oil peak the nonconventional oils will not come onstream fast enough to offset the decline in conventional oil production. The paper concludes that the likely dates of hydrocarbon production maxima are: Non-Organization of Petroleum Exporting Countries Global

Conventional oil

Now–2015

Conventional oil

2010–2020

Global

All oil

2015–2025

Global

Oil+gas

2015–2030

Global

Gas

2020–2040

R. W. Bentley (*) Visiting Research Fellow, Department of Cybernetics, The University of Reading, Reading RG6 6AY, UK e-mail: [email protected]

Keywords OPEC . Oil . Gas . Production . Peak

Introduction This paper sets out the author's current expectations of the likely dates when the global production of oil and gas will reach maxima as determined by the resource limits. The paper starts with a number of definitions, and then looks at the mechanism that drives “resource-limited” production peaking. Past and current forecasts from a number of sources for oil production are then presented and compared, and conclusions are drawn on likely future supply. The definition of conventional oil is not settled. It is frequently is taken, however, as oil which is obtained by primary or secondary extraction methods, where these include own pressure, physical lift, and water flood; as well as pressure maintenance from water or natural gas injection. On a volume-weighted basis, these methods currently extract globally about 35% of the oil in place. By contrast, nonconventional oil is usually taken to include very heavy oils and oils extracted from tar sands and shale deposits. Natural gas liquids (NGLs) are sometimes treated as nonconventional, but here, we include them in the conventional oil data. Oil can also be derived synthetically, from gas (as GTLs) or coal (CTLs), or produced from biomass. It is usual to class these last three sources as “liquids” to separate them from naturally occurring geological oil. When analysing future oil production, an important distinction is between proven (often called “1P”) and proven plus probable (“2P”) reserves.

116

Energy Efficiency (2010) 3:115–122

Data on proven oil reserves are widely available (for example, from annual editions of the BP Statistical Review) but these are extremely unreliable. This is explained in Bentley et al. (2007). The main problems with proven reserves are: –

–

–

They are conservative data for many countries, reporting only oil that is “close to the market” in terms of government sanctions, and extraction technology in place or committed. For these countries, the proven reserves can be significantly smaller than the more likely “proven plus probable” reserves data. The data are probably overstated in a number of large Middle East producer countries, such that the published proven reserves for these countries are larger than (sometimes up to twice the size) the corresponding “proven plus probable” reserves data held in industry databases. For many countries, the proven reserves data are rarely updated and remain unchanged for many years in succession.

The data that should be used to forecast oil production, therefore, are the more likely “proven plus probable” reserves. These data are available from a number of sources, but in collected form usually only from industry datasets that can be expensive to access.

The mechanism of a “resource-limited” peak in oil production A “resource-limited” oil production peak occurs when production climbs to a maximum, and then declines from this peak driven by a combination of the limited quantity of the recoverable resource, the size distribution of the reservoirs, the individual reservoir behaviour, and the degree of investment. The details of this mechanism are fairly complex, and to an extent are counterintuitive. This is because the peak occurs when a region still contains large quantities of reserves, new fields are being discovered, and technology is still raising recovery factors. Resource-limited production peaks have now been exhibited by over 60 oilproducing countries around the world (for a listing of these, see e.g., Campbell and Heapes 2008; or for a map, Energyfiles 2007).

Resource-limited production peaks occur quite some time after the discovery of new fields has settled into decline, and takes place at the point when the production declines in the large early fields— primarily due to pressure loss—are no longer compensated by production increases from the more numerous, smaller, later fields. Large countries that are past peak of their conventional oil production include the USA, Indonesia, the UK, and Norway. Russia is past its conventional oil production midpoint, if not technically past peak. To predict the peak in a region, it is generally helpful to have access to the 2P reserves data in order to see—and extrapolate—the decline in discovery. But because regions frequently exhibit different phases of discovery (for example, for the UK, relatively small onshore discovery, followed by much greater offshore discovery), geological petroleum knowledge of current and potential future basin productivity in the region is also required. Such 2P discovery data combined with geological knowledge indicates that many more countries will soon go past resource-limited peak, including major producers such as China and Mexico. As an illustration the typical time periods between the peaking of 2P discovery and that of production in a region, we can look at data for the USA, Germany, Indonesia, the UK, and the world as a whole. The data given below for 2P discovery, and for production, come from the IHS Energy petroleum economics and policy solutions (PEPS) dataset (IHS Energy 2000); while the estimates of original conventional oil endowment come from a blend of the ultimately recoverable resource (URR) data provided by Campbell and Heapes (2008), and those given by the US Geological Survey in their year 2000 global assessment of undiscovered oil (USGS 2000). Despite quite wide differences in estimates for oil yet-to-find for different countries, Campbell and Heapes and the USGS estimate fairly similar numbers for the URR of most countries, simply because cumulative produced plus 2P reserves usually dominate the URR totals. Using the above data, we find that for “conventional” oil: For the USA: – –

“2P” discovery peaked in the 1930s, Production reached a resource-limited peak in 1971,

Energy Efficiency (2010) 3:115–122

–

The USA has now burnt about three quarters of its original endowment.

“2P” discovery peaked in the 1950s, Production reached a resource-limited peak in 1967, Germany has now burnt over three quarters of its original endowment.

–

– –

For Germany: – –

117

For the world: – – –

For Indonesia: – – – –

“2P” discovery peaked onshore in the 1940s, “2P” discovery peaked offshore in the 1970s, Production reached a resource-limited peak in 1979, Indonesia has now burnt perhaps two thirds of its original endowment. For the UK:

–

“2P” discovery peaked in the mid-1970s,

Production reached a resource-limited peak in 1999, The UK has now burnt just over half of its original endowment.

“2P” discovery peaked in the mid-1960s, Production is expected to peak around 2010, The world has burnt something approaching half its original endowment.

Forecasts of oil production Deeply embedded in the discussion of oil depletion is the notion that “all past oil forecasts have been wrong”. This view is informed in part by two facts. In the 1970s, it was widely held that the world was

Table 1 Selected forecasts of global oil production, made between 1956 and 2008, which give a date for peak Date

Author

Hydrocarbon

URR (Gb)

Date of global peak

1956 Hubbert

Conventional oil

1,250

“about the year 2000” (at 35 Mb/d])

1969 Hubbert

Conventional oil

1,350

1990 (at 65 Mb/d)

2,100

2000 (at 100 Mb/d)

Probably conventional oil

2,100

“increasingly scarce from ~2000.”

1972 Report: UN Conference Probably conventional oil

2,500

“likely peak by 2000.”

1976 UK DoE

Probably conventional oil

n/a

“about 2000”

1977 Hubbert

Conventional oil

2,000

1996 if unconstrained logistic; plateau to 2035 if production flat

1977 Ehrlich et al.

Conventional oil

1,900

2000

1979 Shell

Probably conventional oil

n/a

“plateau within the next 25 years.”

1979 BP

Probably conventional oil

n/a

Peak (noncommunist world): 1985

1981 World Bank

Probably conventional oil

1,900

“plateau~turn of the century.”

1995 Petroconsultants

Conventional oil excluding NGLs

1,800

About 2005 (Campbell & Laherrère 1995)

1996 Ivanhoe

Conventional oil

~2,000

About 2010

1998 IEA

Conventional oil

2,300 (reference case) 2014

1972 ESSO

2004 PFC Energy

Conventional and nonconventional oil

2018

2005 Deffeyes

Conventional and nonconventional oil

2005

2008 Campbell

Conventional and nonconventional oil

2008

2008 Skrebowski

Conventional and nonconventional oil

2012—“Mega-fields” study

2008 Energyfiles Ltd

Conventional and nonconventional oil

2015

2008 Miller—own model

Conventional and nonconventional oil

2018—including all fallow fields

For references and discussion of the above forecasts up to 2005, see Bentley and Boyle (2008); for subsequent forecasts, see Sorrell (2009)

118

Energy Efficiency (2010) 3:115–122

likely to run out of oil in about 30 years. This has clearly not happened. Secondly, some well-known forecasts—for example, from Hubbert (1956), Campbell (1991), or Deffeyes (2005)—have given dates for the global peak that have been premature. It is important therefore to look at what these past forecasts tell us. The 1970s view of “oil running out in 30 years”, though in the consciousness of many dealing with oil today, was based simply on the size of the proven reserves. It ignored the large volume of oil already then discovered but classed as probable reserves, the oil yet to find, and that to be realised by improving recovery factors. Not surprisingly, even in the 1970s, it was recognised that a better approach was to evaluate the likely global URR, and see how long this could sustain production growth until a “Hubbert-type” peak would be reached. By the 1970s, the general consensus was for the total global URR of conventional oil to be around 2,000 billion barrels (Gb). By

World-Conventional oil

Peak Discovery 1964 Peak Production 2005 Time-lag: 41 years

Mid-point year: Ultimate 2050: To-date 1999:

2005 1800 Gb 822 Gb 40

Theoretical Unconstrained Model

Discoveries, Gb/a

120 100 80

Peak Discovery

35 30 25 20

60

High Prices Curb Demand

40

15 10

20 0 1930

Production, Gb/a

140

5 1950

1970

1990

2010

2030

0 2050

Fig. 1 Right-hand Scale Global oil production, actual, and as modelled by a Hubbert curve, in billion barrels per year (from Campbell 1999). This figure indicates the relationship between an unconstrained Hubbert curve model, based on an example global URR of 2,000 Gb, and the actual path of global oil production. Unconstrained here indicates production that is not limited by demand or other nonphysical factors. The unconstrained production curve was not followed because of demand destruction from the 1970s price shocks. Also shown (small dots) is a forecast from Campbell's model, based on a global URR of 2,000 Gb. This gives the global peak at around the year 2010, with 1,800 Gb of production by 2050. Bars and left-hand scale indicate the history of 2P discoveries of “regular” oil. The 1930 line includes all finds prior to 1930. The high-discovery years in which Burgan (1938) and Ghawar (1948) were discovered stand out, as well as the period of maximum global discovery of “regular” oil, in the mid-1960s

the same date, only about 300 Gb of oil had been produced, requiring an additional 700 Gb to be produced before the approximate midpoint of 1,000 Gb would be reached. Assuming a logistic model of supply, this implied that global production could continue to increase until reaching a production peak around the year 2000. Table 1 lists a number of the forecasts made since Hubbert's global model in 1956 that have predicted a peak in global oil production. Perhaps, one of the best examples to illustrate this oil knowledge of the 1970s was a landmark environmental study carried out for the United Nations (Ward and Dubois 1972). In this report, the authors said: One of the most quoted estimates for usable reserves [i.e., global URR of oil] is some 2500 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. In the event, the anticipated global production increase leading to a Hubbert-type peak did not take place. This was because of the demand destruction which resulted from the high prices of the 1973 and 1978 oil price shocks. Had this been factored in, these models would mostly have predicted the peak for the global production of conventional oil as occurring around 2005 or so. This is illustrated in Fig. 1. The importance of these early “peaking” forecasts has tended to be overlooked. Opposed to these “peaking” calculations were forecasts that oil production would not decline below forecast demand for the foreseeable future (generally out to 2020 or 2030). Table 2 summarises some of the more recent of these “non-peaking” forecasts. Some of these forecasts were based on a geological assessment of the resources available; others were not. In addition, a number of the nonpeaking forecasts lacked rigour in their assessment of geological data, such as mixing 1P and 2P reserves data, calculating reserve growth from 1P data alone, or accepting at face value Organization of Petroleum Exporting Countries' (OPEC) declared reserves, or the rate of discovery implicit in the USGS 2000 global assessment. An important example is that of the International Energy Agency (IEA). In recent times, the IEA first

Energy Efficiency (2010) 3:115–122

119

Table 2 Results of some more recent oil forecasts that saw no peak before 2020 or 2030 (depending on the forecast's time horizon) Date

Author

Hydrocarbon

Ultimate (Gb)

1998

WEC/IIASA-A2

Conventional oil

2000

IEA: WEO 2000

Conventional oil (+N)

3,345

2001

US DoE EIA

Conventional oil

3,303

2002

US DoE

Ditto

2002

Shell Scenario

2003

“WETO” study

Conventional oil and nonconventional oil Ditto

2004

ExxonMobil

Ditto

2005

IEA: WEO 2005

Ditto

~4,000

World production (Mb/d) 2020–2030

No peak

90

100

No peak

103

–

2016/2037

Various

No peak

109

–

Plateau: 2025–2040

100

105

No peak

102

120

No peak

114

118

Reference Sc.

No peak

105

115

Deferred Invest.

No peak

100

105

2007

IEA: WEO 2007

2008

IEA: WEO 2008

2008

Reference Sc.

4,500

Forecast date of peak (by study end-date)

Ditto

No peak

Reference Sc.

Ditto

No peak

OPEC

Ditto

No peak

116 106 103

114

Shell's ultimate of 4,000 Gb was composed of: ~2,300 Gb of conventional oil (including NGLs); plus ~600 Gb of “scope for further recovery” oil; plus 1,000 Gb of nonconventional oil. WETO's ultimate of 4,500 Gb was for conventional oil only; it started with a USGS figure of 2,800 Gb, then grew by assuming large and rapid recovery factor gains to 2030. For references and discussion of the above forecasts up to 2005, see Bentley and Boyle (2008); for subsequent forecasts, see Sorrell (2009) Mb/d Million barrels per day

recognised the possibility of global oil peaking in their 1998 World Energy Outlook (Table 1). This view changed again in World Energy Outlook 2000 (WEO; Table 2), based on the availability of the USGS year 2000 assessment of the global URR. Currently, in WEO 2008, the IEA takes an intermediate view that the investment required to meet the demand for conventional oil up to 2030 is “daunting”, and that conventional oil supply is likely to have reached a plateau by this date. As a second example, Shell scenarios (predating their current models) were based on a global URR of 4,000 Gb, which included 600 Gb of “scope for further recovery” and 1,000 Gb of nonconventional oil. These scenarios implicitly envisaged that a liquids supply peak would be avoided by a smooth transfer away from conventional oil to the nonconventionals. Finally, there have been a range of “non-resourcebased” views. This was a group of opinions and analyses which ruled out any need to examine the size of oil resource base in any detail based in part on the assumption that market forces would always ensure that supplies met market demand. These views are

still prevalent in some quarters. Figure 2 compares some of the global oil forecasts made between 2002 and 2006.

Judgement of current oil forecasts Given the disagreement between current forecasts (taken here as those made between 2005 and 2008), how is a judgement to be formed on which are the most likely? In this author's view, the better results are likely to arise from detailed “bottom-up” models. The best of these are probably the largely by-field models, such as those from PFC Energy (2005), Peak Consulting (Skrebowski 2008), Energyfiles Ltd. (in Sorrell 2009), or the model originally run by R. Miller when he worked for BP (in Sorrell 2009). Such models do not assume a URR, but simply build future production from known fields and assumptions on discovery rate and “reserves growth”. The URR of such models is an output. Global URR is also an output of models which are “bottom-up” by country, as in the case of Campbell's model.

120

Energy Efficiency (2010) 3:115–122 Global production - All oil (conv. plus non-conv.) 140 IEA 2004

Production (Mb/day)

120

Odell

100

Shell-Sc. PFC

80

Bauquis BGR

60

BP-Miller

40

Deffeyes EnergyFiles

20

U/Campbell

0 1990

2000

2010

2020

2030

2040

2050

Fig. 2 Forecasts of global all-oil production (conventional plus nonconventional). Forecasts from IEA (2004); Odell (2004); Shell (2003); PFC Energy (2005); Bauquis (2003); BGR

(2002); Miller (2004); Deffeyes (2005); Energyfiles Ltd. (2006); Campbell and Sivertson (2003)

The current models from the IEA, US Energy Information Administration (EIA), and OPEC, though detailed bottom-up by field or region in the near term, fall back on exogenous estimates of URR for the medium term (Sorrell 2009). The danger of such an approach is that it is easy to conclude that there are adequate known recoverable resources to meet anticipated demand to the forecast's time horizon; but to ignore the constraints set out above (of declining discovery, and the addition of smaller fields) that drive peaking. On balance, it would seem that the current bottomup models—which nearly all see a near or mediumterm peak in the production of all oil—are the more likely.

compensate for the decline of conventional oil production once this goes over the peak is not well understood. The quantity of such liquids required will be fairly large. For example, if the decline in global conventional is about 3% per year (made up of a steeper decline of perhaps 4% to 6% per year from those fields in decline, but partly compensated for by production increases from new and yet-to-discover fields), and the business-as-usual demand growth hoped for is, say, 1% pa; then as soon as 10 years past the peak of conventional oil, these other liquids must be supplying about 30 million barrels per day. This is a tall order. Not surprisingly, as a result, detailed models from, for example, Energyfiles (2007), Campbell and Heapes (2008) and others, indicate that it will be difficult to raise the production of these other liquids sufficiently in the time period required. Impediments to rapid production growth include:

The production of nonconventional oils, synthetic fuels, and biofuels The world contains large quantities of nonconventional oil and other potential liquid fuels. For example, the IEA has recently estimated the global recoverable original endowment of conventional oil (including enhanced oil recovery) as over 4,000 Gb; with heavy oils and bitumen as adding over 1,000 Gb, and oil shales another 1,000 Gb, totalling some 6,500 Gb or so for recoverable all oil. If potential liquids recoverable via GTL and CTL are added in, the total rises to around 9,000 Gb (IEA 2008). By contrast, the rate that the current world production of the nonconventional oils, synthetic fuels, and also biofuels can be increased to

– – – – –

investment availability (these other liquids tend to be very capital expensive), skilled manpower requirements; carbon dioxide emissions (generally these are more, per unit of energy delivered, than for conventional oil); in some cases, the adequacy of the water and gas supplies required; net energy constraints (which drive real costs);

and also, net energy rate limits—that is, limits to ramp-up rates that yield a positive net energy output during the ramp-up phase.

Energy Efficiency (2010) 3:115–122

Much more work is needed on the future likely rates of availability of these other liquid fuel sources—and of oil substitution. But it is recognised that these other liquid fuel supplies, though important, may not solve the problem of the conventional oil peak.

The impact of an oil peak The impact of such a conventional oil peak may be severe. There will still be plenty of oil: by then still about 40 years' worth of reserves in the ground (but to be produced at ever diminishing rates); significant amounts of conventional oil yet to be discovered; plus the heavy oils, tar sands, and shale oils; with gas-toliquids (until gas, too, peaks a little later), coal-toliquids and biofuels coming in as substitutes. But the declining availability of conventional oil is likely to have impacts greater than the shocks of the 1970s, and this time without new oil provinces to come to the rescue. It may be useful here also to speculate why knowledge of the above oil facts is so poor. There has not been space here to cover this topic, but it has to do with a number of factors, including reporting of reserves in the USA and elsewhere for far too long on a proven basis, the existence of economic theories about resource availability that—in part—do not apply to oil, and limited quantitative data-driven forecasting efforts until very recently by most of the oil majors, by government bodies, and by nearly everyone in academia.

The future of global gas production In this paper, the main focus has been on the global supply of oil. But a similar analysis as for oil points to a global production peak for gas also in the not too distant future (see for example, Campbell and Heapes 2008). However for gas, like oil, there is a range of nonconventional sources of varying degrees of technical readiness, and here, the US experience may well be a model. It is often not recognised that US gas production peaked in 1973, about the same date as the US oil peak. But unlike oil, which has seen a significant decline in production since (now to a production about 60% of that of the peak), gas

121

production has fallen relatively little. This is because the US has been able to bring on a wide range of new and alternative sources of gas, of which gas from oil shales is only the latest. So the world too, may follow this model. Nevertheless, conventional gas will probably go over its resource-limited production peak around perhaps 2030, and as with the US, this may also signal the peak of all-gas production.

Conclusions On the basis of the foregoing, the following conclusions seem valid: &

&

Proven (1P) reserves should not be used for forecasting oil production; such reserves currently are very unreliable, being variously under-, over-, and not reported. Neither should reliance be placed on reserves to production (R/P) ratios to indicate future security of oil supply. Not only are the proved reserves data themselves unreliable, but of greater significance is the fact that reserves can be large (and hence, the R/P ratio, as expressed in years of supply, look reassuringly secure) at the point when the peaking mechanism forces production to decline. Today's 1P reserves and 2P reserves both stand at about 40 years of global supply, but the world's midpoint is close, and production of conventional oil, at least, will soon decline.

Based on the arguments above, the following peak dates for global production therefore, to this author, look reasonable: Date of peak Non-OPEC peak

Conventional oil

Global peak

Conventional oil

Now–2015 2010–2020

All oil

2015–2025

Oil+gas

2015–2030

Gas

2020–2040

Note however that analysts and policy takers need to be aware that a “mini-glut” of oil supply was—at least until recently—expected until the peak actually arrives. This was because of rapid new production increases expected from known projects either under-

122

way or anticipated in the deepwater off the USA, Brazil, Africa, and Mexico; as well as from oil from Russia, the Caspian, and Iraq; and from tar sands and Orinoco heavy oil projects. However, current economic conditions will affect this expectation, due to a new balance forming between demand destruction caused by the economic downturn versus project delay due to low prices and lack of investment. The overall conclusion of this paper is that a wider understanding of hydrocarbon depletion is needed. This includes better reserves data; calculations in terms of the impact of higher oil price and new technology; modelling of the newer and alternative liquids supply; and ideally, collaboration on an international protocol to meet the anticipated supply constraints. Recognition of the potential problems of hydrocarbon depletion is a prerequisite for planning sensible and peaceable solutions for what may turn out to be difficult times.

References Bauquis, P.-B. (2003). What energies for transportation in the 21st century? Presentation at the ASPO 2nd International Workshop on Oil Depletion (IWOOD-2003), IFP, Paris, May 26-27. Bentley, R. W., & Boyle, G. A. (2008). Global oil production: forecasts and methodologies. Environment and Planning B: Planning and Design, 35, 609–626. Bentley, R. W., Mannan, S. A., & Wheeler, S. J. (2007). Assessing the date of the global oil peak: the need to use 2P reserves. Energy Policy, 35, 6364–6382. Elsevier. BGR (2002). Reserven, Ressourcen und Verfügbarkeit von Energierohstoffen, (Bundesanstalt für Geowissenschaften und Rohstoffe) Germany. Campbell, C. J. (1991). The Golden Century of Oil: 1950-2050. The Depletion of a Resource. Springer. Campbell, C. J., & Sivertson, A. (2003). The ‘CampbellUppsala’ model. See website: www.peakoil.net, or:

Energy Efficiency (2010) 3:115–122 Campbell, C J, The Essence of Oil & Gas Depletion. Multi-Science Publishing, Essex, UK; ISBN 0 906522 19 6, pp. 233-254. Campbell, C. J., & Heapes, S. (2008). An atlas of oil and gas depletion. West Yorkshire: Jeremy Mills Publishing. Deffeyes, K. S. (2005). Beyond Hubbert’s peak. Princeton University Press. Energyfiles Ltd. (2006). (a). Smith, M R, The World Oil Supply Report 2004-2050, 3rd Edn, Douglas-Westwood Ltd. (b). Smith, M R, The World Gas Supply Report 2004-2025, 1st Edn., Douglas-Westwood Ltd. (c). Web: www.energyfiles. com, e-mail: [email protected] Energyfiles (2007). See Energyfiles data on the map on website of David Strahan: www.lastoilshock.com/map.html. Hubbert, M. K. (1956). Nuclear energy and fossil fuels. Drilling and production practice; The American Petroleum Institute, pp. 7–25. IEA (1998; 2000; 2004; 2005; 2007; 2008). World energy outlook. Paris: International Energy Agency IHS Energy, USA (2000). ‘PEPS’ database of oil data by country. Available on a commercial basis. Miller, R. G. (2004). Miller’s unofficial BP model and its results were presented at an Institute of Energy, Northern Ireland Branch meeting, 24th February, 2004. Miller’s (2008) model is potentially available on request from the author. Odell, P. (2004). Why carbon fuels will dominate the 21st century energy economy. UK: Multi-Science Publishing. PFC Energy (2005). Presentation at the ‘Global Oil Depletion Debate’, Energy Institute, London, 2005; also personal communications. Shell (2003). Scenario modelling based on data in P-R Bauquis’ presentation at IWOOD 2003, Paris. (See www. peakoil.net). Skrebowski, C. (2008). In The Oil Crunch. Securing the UK’s energy future. First report from the UK Industry Taskforce on Peak Oil and Energy Security. www.peakoiltaskforce. net. Sorrell, S. (2009). Ed. Global Oil Depletion. Report for the United Kingdom Energy Research Cntre (UKERC), London, UK. Forthcoming 2009. See: www.ukerc.ac.uk. USGS (2000). United States Geological Survey, World Petroleum Assessment 2000;/http://pubs.usgs.gov/dds/dds-060/S. Ward, B., & Dubois, R. (1972). Only one earth: the care and maintenance of a small planet (p. 1972). UK: Penguin Books.