I had a longish critique published at FT Alphaville (an online publication of the Financial Times) today, of the widely-heralded recent report by Harvard’s Leonardo Maugeri, “Oil: The Next Revolution.” It’s unavoidably technical and wonky, but hopefully still readable by non-wonks.

Read it here: Is peak oil dead?

Postscript, August 6, 2012: Rich Turcotte assembled a useful, if tongue-in-cheek summary of links to the credulous mainstream media coverage of the Maugeri report, as well as the responses from the peak oil camp. It’s a fine snapshot of the peak oil debate at this point in time.

Is peak oil dead?

One might think so, judging by a slew of optimistic new forecasts for oil production. Even George Monbiot, notable for his thoughtful previous coverage of peak oil in The Guardian, threw in the towel with his July 2 mea culpa, “We were wrong about peak oil. There’s enough to fry us all.”

Monbiot reversed his position after reading a new report by Leonardo Maugeri, an executive with the Italian oil company ENI and a senior fellow at a BP-funded center at Harvard University.

Maugeri forecasts new global oil production capacity of 49 million barrels per day (mbpd) by 2020, a number that is “unrestricted” by real-world circumstances, and “unadjusted for risk.” This constitutes a whopping 53 percent increase over the current claimed capacity of 93 mbpd in just eight years. While impressive, this headline number obscures some important details.

First, capacity is not production. The world has never produced 93 mbpd. Global oil production was 88.3 mbpd in 2011, according to the International Energy Agency (IEA), which uses a very liberal definition of “oil” that includes biofuels, non-associated natural gas liquids, and other liquids. Under a more restrictive definition used by the U.S. Energy Information Administration (EIA), which counts crude oil plus lease condensate (natural gas liquids that are produced and naturally associated with the crude), and liquids extracted from natural gas production, world oil production was 87 mbpd in 2011. Counting only crude oil and lease condensate, world oil production was 74 mbpd in 2011, a level it has maintained since the end of 2004 despite a tripling of oil prices since 2003.

Therefore, there is a 19 mbpd gap between actual crude oil production and Maugeri’s unverifiable claim of 93 mbpd of oil production capacity, depending on how one construes (or misconstrues) the meaning of “oil.” Much of the capacity and growth Maugeri foresees includes millions of barrels per day of natural gas liquids, of which only about one-quarter are useful as vehicular fuel.

Maugeri generally refers to production capacity throughout his report, not actual production. The only nod to actual production appears in a footnote on page 4, where he says “In the first quarter 2012, average world oil production consistently reached or surpassed 91 mbd.” Since he doesn’t specify the source of this data, we must assume he obtained it from his private, field-by-field database, as the IEA shows production in the first quarter of 2012 to be 90.7 mbpd.

Next, Maugeri adjusts his 49 mbpd increase by various risk factors, and finds that adjusted new production of 28.6 mbpd might be possible by 2020. He expects most of this additional production to come from 11 countries, shown in the following figure.

Maugeri’s Figure 3, of “Worldwide potential additional liquids supply out to 2020 (crude oil and NGLs, excluding biofuels)” for the 11 countries representing the majority of his projected increase.

Maugeri devotes several pages of his report to a light treatment of the risks he accounted for in this adjusted number, offering little purchase for a skeptical reader who might discount the risks differently. We are essentially left to take his word for it.

Finally, Maugeri adjusts for the depletion of currently producing fields and reserve growth, to come up with a final projected increase of 17.6 mbpd and a total world production capacity of 110.6 mbpd by 2020. This is where the really squishy assumptions come into play, which are core to his forecast.

Depletion and decline rates

Most oil analysts are careful to distinguish depletion rates from decline rates. A depletion rate is the percentage of the recoverable oil in a field that is being produced each year; therefore, if new technology were to increase the estimated recoverability of oil in a field, the depletion rate would fall. A decline rate is an annual percentage decline in the rate of production from a given field, so it does not depend on the size of the field. Maugeri mixes up the terms, defining only depletion rate as “The natural decline of an oilfield’s output after years of production. It could be partially offset by reserve growth.”

In 2008, CERA, a consultancy which has one of the few comprehensive databases of the world’s oil fields, and the IEA, which used CERA’s database, estimated decline rates for the world. The IEA found a global average production-weighted decline rate of 5.1 percent per year. CERA estimated the global average production-weighted decline rate of all fields at 4.5 percent per year. A similar 2009 study by Mikael Höök et al. found a production-weighted average decline rate of 5.5 percent per year. Other estimates have ranged as high as 8 percent. All of these studies find that decline rates increase over time, and they are higher for “unconventional” sources like deepwater and shale than for conventional fields. (Source)

Maugeri muddles these important distinctions, claiming that the aforementioned studies are in sharp variance when they are not. He goes on to allege overestimation of “depletion” (we assume he means decline) rates in the past, without any references, and finally concludes, inexplicably, that apart from Norway, the UK, Mexico, and Iran, he “did not find evidence of a global depletion rate of crude production higher than 2-3 percent when correctly adjusted for reserve growth.”

Reserve growth

As a global average, current technology and prices only permit about 30 to 35 percent of the oil in an oil field to be economically recovered, up from about 20 percent thirty years ago. Over time, new technology and techniques make it possible to economically recover more oil, and that additional oil may then be reclassified from resources (the oil in place in a field) to reserves (oil that may be legally claimed as recoverable). This process is called reserve growth.

Maugeri discusses reserve growth at length, emphasizing the vast quantity of remaining resources and asserting that new technology will soon make more of it accessible, particularly from unconventional oil resources. “In fact, the current decade could herald the advent of ‘unconventional oil’ as ‘the oil of the future,’” he claims, “changing the geopolitical landscape that has marked the oil market for most of the 20th Century.”

As an example, Maugeri cites the Kern River field in California, one of the longest-producing oil fields in America. New recovery methods have substantially increased the recoverability of oil from this field over time. What he does not mention is that waterflooding and other enhanced oil recovery methods that enabled reserves growth in Kern River are now routinely used early in the exploitation of oil fields, belying his suggestion that similar reserves growth will be achieved in the future. Nor does he mention that despite the intensive application of enhanced recovery methods, Kern River production has been declining since its last peak in 1985, or that it currently produces about 10 barrels of water for every barrel of oil, at a very significant energy cost. I have visited the Kern River field and studied its production, and found that its energy return on investment ratio is now probably on the order of four, which hardly makes it a shining example of new abundance.

Kern River production history. Source: Chevron

Therefore, while it is true that reserves do grow over time with the application of new technology, it is disingenuous to imply that it will lead to the enormous increases in production, or the far lower decline rates that Maugeri claims. Again, Maugeri only presents the summary results from his private database and does not disclose the recovery factors he is using, so there is no way to judge how realistic his model is.

However, we do know from more than 60 years of history with enhanced oil recovery techniques that they tend to lengthen and thicken the tail of a field’s production, not achieve new production highs. This is even more true today than it was decades ago, when the Kern River reserve growth Maugeri highlights occurred.

Even in the U.S., much of the apparent reserve growth over the past three decades had more to do with the technical reclassification of oil as “proved reserves” under SEC rules than technology, as petroleum geologist Jean Laherrère has detailed at length.

Reserve growth and price

Maugeri’s discussion of reserve growth elides the well-known exaggerations of proven reserves among the world’s major oil producers. Producers in the Persian Gulf, North Africa OPEC, Russia, Venezuela and Canada report “reserves” estimates that can only be economically produced if oil prices are at least double the $70 per barrel assumption in his analysis. He does not provide any further details about the economics of production in his analysis, except to say that “More than 80 percent of the additional production under development globally appears to be profitable with a price of oil higher than $70 per barrel.”

This claim seems highly dubious given recent estimates of production costs. Research by petroleum economist Chris Skrebowski, along with analysts Steven Kopits and Robert Hirsch, finds a new barrel of production capacity in deepwater, some OPEC countries, the Canadian tar sands, and Venezuela’s Orinoco belt will cost up to $80 or $90 a barrel. Canada’s Globe and Mail reported in June that $80 a barrel was low enough to cause several tar sands operators to slash their expansion plans. And arecent report from Bernstein Research found that the real floor of new production in 2011 was around $92 a barrel, and will be closer to $100 a barrel this year.

We also know that the cost of new oil production has been climbing sharply in recent years, along with the cost of all commodities, as shown in the following chart.

Source: EIA

Maugeri acknowledges this fact, noting, “Over this decade, another problem affecting the production of all shale/tight oil plays in the United States will be the inevitable rising costs of services, rigs, labor, and pipelines, caused by the inflationary pressure from the frenetic activity throughout the shale/tight oil and gas sector.” This does not square with this subsequent assertion that “the advancing knowledge of shale oil development and the gradual expansion of the infrastructure necessary to each shale play should balance the rising costs, and eventually drive them down,” and he offers no empirical basis for it.

Likewise, his acknowledgement that “the oil market will remain highly volatile until 2015 and prone to extreme movements in opposite directions, thus representing a major challenge for investors, in spite of its short and long term opportunities,” doesn’t square with his assumption of a minimum $70 per barrel holding firm through 2020 and beyond.

Bakken ballyhoo

Maugeri devotes the longest section of his report to the tight oil and shale gas “revolution” in the U.S., saying it “could be a paradigm-shifter for the oil world.” He extrapolates most of his forecast from the Bakken formation, a tight oil reservoir which underlies parts of North Dakota, Montana and Saskatchewan.

Unrestricted production from shale and tight oil could reach 6.6 mbpd by 2020 in his estimation, or as much as 4.2 mbpd after considering risk factors and depletion. The U.S. is currently producing about 0.9 mbpd from tight oil, so Maugeri’s forecast amounts to a more than four-fold increase in eight years, an extremely optimistic prospect. It is also far more than the EIA expects, having recently forecasted that U.S. tight oil would reach only 1.2 mbpd by 2035. For additional perspective, total U.S. production of crude oil and condensate today is 6.1 mbpd.

He does not mention that, on the basis of Bakken well productivity, it might take 50,000 new successful tight oil wells or more to achieve his forecast, plus many more unsuccessful ones as the productive areas of new fields are delineated. In the IEA’s recent forecast, another 500,000 new shale gas wells might be drilled by 2035, doubling the number of producing gas wells in America. Many of these new tight oil and shale gas wells would need to be drilled near where people live and work, rendering an “unrestricted” forecast for new development all but meaningless. Maugeri refers vaguely to this limitation, noting that “a revolution in environmental and curb-emissions technologies is required to sustain the development of most unconventional oils,” and that if the industry continues to fail to prevent environmental contamination from tight oil projects, “massive over-regulation” could result and new development could be delayed.

Maugeri assumes that oil will sell for at least $70 a barrel to achieve his tight oil production forecast. “Most of U.S. shale and tight oil are profitable at a price of oil (WTI) ranging from $50 to $65 per barrel,” he says, but an executive with an oil company producing oil in the Bakken, who was interviewed by Steve LeVine forForeign Policy in February, said that if prices dropped to $70 per barrel, it could “create an extreme drop in drilling and field production really quickly” in the Bakken, and that “if oil drops to $70, a lot of people will lose money in the Bakken.”

Finally, Maugeri’s assumptions for the production profiles of Bakken wells appear to be far removed from reality. He uses a “combined average depletion rate for each producing well of 15 percent over the first five years, followed by a 7 percent depletion rate” for tight oil wells, while historical evidence shows that Bakken wells typically decline by 80 percent or more over the first five years.

Production profile of a typical Bakken well. Source: North Dakota Department of Mineral Resources

While tight oil production since 2005 has indeed been impressive, there is little basis for the Maugeri’s confidence that its growth trend will continue on its present trajectory through 2020, when real-world costs, siting issues, environmental concerns, and oil industry practices are taken into account.

Conclusion

Although Maugeri does not state explicitly what decline rates he is using, researchers Stephen Sorrell and Christophe McGlade derived an annual average decline rate from the data in his report of 1.6 percent, or about one-third the global decline rates estimated by IEA, CERA and others. After analyzing the IEA data, they found an aggregate global production-weighted decline rate of 4.1 percent per year. At that rate, they found that Maugeri’s forecast for 2020 would reach just 95.1 mbpd, not 110.6 mbpd—a gain of just 2 mbpd over today, not 17.6 mbpd.

We cannot independently evaluate Maugeri’s country-by-country forecasts without seeing the assumptions in his data model, but his summary expectations are optimistic in the extreme. For example, he sees production from Iraq expanding in the next eight years at rates that have never before been achieved, despite a great deal of uncertainty about the country’s stability, its ability to maintain security in the future, and its ability to attract Western oil partners with the knowledge and technology needed to exploit its resources. The failure of Iraq’s recent oil lease auctions do little to give one confidence that Maugeri’s extraordinary forecast can be realized.

More generally, his assertion that, of the countries with more than 1 mbpd of production capacity, only four will have reduced capacity by 2020 is impossible to square with the fact that production has been declining in more 50 of those countries since 2000.

Maugeri’s forecast does not mention a price ceiling at all, an obvious deficiency given the extreme volatility of oil prices over the past four years. We know that as prices approach $120 a barrel, demand shrinks, yet triple-digit prices are precisely what is required to bring much of the new supply Maugeri anticipates online.

To his credit, Maugeri acknowledges that his analysis “is subject to a significant margin of error, depending on several circumstances that extend beyond the risks in each project or country,” and he details numerous important caveats. And to the extent that he reveals the assumptions underpinning his forecast, his transparency is laudable. In the final analysis, however, it is insufficient. He fails to provide adequate justification that his assumptions, being widely divergent from most other industry estimates, are remotely realistic.

We must conclude that the key assumptions about reserve growth and its effect on decline rates in Maugeri’s report are muddled, speculative and unverifiable. And sprinkling those assertions with repeated declamations about how peak oil is a non-issue, insisting repeatedly that the only real constraints on his scenario have to do with political decisions and geopolitical risks, suggests that his report is more about grinding a political axe on behalf of the oil industry than offering a serious or transparent analysis. Finally we must note that Maugeri is well known for his hostility to peak oil, as is BP, which funded his report. After taking real-world risks, costs, and restrictions into account, the case for peak oil—which is about production rates, not production capacity or reserves—seems far more realistic.