Varun Sivaram

Energy, Security, and Climate

CFR experts examine the science and foreign policy surrounding climate change, energy, and nuclear security.

Print Print Cite Cite
Style: MLA APA Chicago Close


How Not To Argue That We’re Running Out Of Affordable Oil

by Michael Levi
January 27, 2012


I made a New Year’s resolution to spend less time on this blog explaining why other people are wrong.

But New Year’s resolutions are meant to be broken — and some things just beg for intervention. That’s unfortunately the case with “Oil’s Tipping Point Has Passed”, an essay in the current issue of Nature by James Murray, an oceanographer at the University of Washington, and David King, a chemist who was chief scientific advisor to Tony Blair and now heads the Smith School of Enterprise and the Environment at Oxford.

The authors write that “there is a potentially more persuasive argument [than the danger of climate change] for lowering emissions: the impact of dwindling oil supplies on the economy”, and then go on at length to explain why. Their bottom line may be correct – there is certainly much to debate here – but their many arguments are almost invariably flimsy. I want to step through a fairly long list of those claims, both because the article is prominent and likely to be read as authoritative, and because the article provides a convenient and fairly exhaustive compendium of misunderstandings about oil.

The paper begins with a summary before getting into the main substance:

“The idea of ‘peak oil’ — that global production will reach a peak and then decline — has been around for decades, with academics arguing about whether this peak has already passed or is yet to come. The typical industry response is to point to increasing assessments of global reserves — the amount known to be in the ground that can be produced commercially. But this is misleading. The true volume of proven global reserves is clouded by secrecy; forecasts by state oil companies are not audited and seem to be exaggerated.”

Perhaps reserves are exaggerated. Perhaps they aren’t. What do the authors cite to support their position? A recent paper by none other than David King. In any case, the essay immediately goes on to more important things:

“More importantly, reserves often take 6–10 years to drill and develop before they become part of supply, by which time older fields have become depleted. It is far more sensible to look instead at actual production records, which are less encouraging. Even while reserves are apparently increasing, the percentage available for production is going down. In the United States, for example, production as a percentage of reserves has steadily decreased from 9% in 1980 to 6% today. Production at existing oil fields around the world is declining at rates of about 4.5% (ref. 4) to 6.7% per year. Only by adding in production from new wells is overall global production holding steady.”

Little of this makes much sense. Yes, it takes time to convert reserves to production, during which time other fields have declined. Oil producers know that and incorporate it into their planning. The authors seem to suggest that because new investments accompany declines, absolute production increases aren’t possible. That’s silly: just because I draw on my bank account while also earning money doesn’t mean that I won’t become richer. Nor is the fact that production as a percentage of reserves has declined over time a good argument in favor of the claim that we’ve hit a limit to production. Think about it: if a technological breakthrough doubled our reserves tomorrow, the reserve-to-production production-to-reserve ratio would be sliced in half; does anyone seriously think that this would be bad news for oil?

Let’s continue:

“In 2005, global production of regular crude oil reached about 72 million barrels per day. From then on, production capacity seems to have hit a ceiling at 75 million bar­rels per day. A plot of prices against produc­tion from 1998 to today shows this dramatic transition, from a time when supply could respond elastically to rising prices caused by increased demand, to when it could not (see ‘Phase shift’). As a result, prices swing wildly in response to small changes in demand. Other people have remarked on this step change in the economics of oil around the year 2005, but the point needs to be lodged more firmly in the minds of policy-makers.”

Start with something simple: the authors seem to have misread their reference, which reports nearly 82 mb/d of conventional oil production as of 2008 (and another 3.9 mb/d of unconventional production). [UPDATE: A commenter points out correctly that the authors were referring to a narrow set of sources than I did. I think that it’s misleading to segment production in that way, but if ones chooses to do so, the 72/75 numbers hold up.] More problematic, though, is the implicit claim that this is strictly about “economics”. One would not know from this discussion that a large fraction of world oil is produced in states that exercise strategic restraint in their production. It is entirely possible that politics is a much bigger factor than pure economics in explaining what has happened in recent years.

The authors then pivot to explaining why they think it will be difficult to expand production:

“We are not running out of oil, but we are running out of oil that can be produced eas­ily and cheaply. The US Energy Information Administration optimistically projects a 30% increase in oil production between now and 2030 (ref. 2). All of that increase is in the form of unidentified projects — in other words, oil yet to be discovered.”

This simply isn’t true. Unidentified projects don’t involve oil that is “yet to be discovered” – they involve oil that has not been specifically targeted for extraction. In fact, as best I understand, all of the production increase that EIA anticipates comes from oil that has already been discovered. Moreover, not all of it comes from unidentified projects. Indeed the authors seem to acknowledge that just a bit:

“Non-conventional oil won’t make up the difference. Production of oil derived from Canada’s tar sands — sometimes called the ‘oil junkie’s last fix’ — is expected to reach just 4.7 million barrels per day by 2035 (ref. 6). Production from Venezuela’s tar sands is currently less than 2 million barrels per day7, with little prospect of a dramatic increase.”

That’s right, but it’s beside the point. Among the authors’ apparent blind spots are offshore drilling and tight oil in the Americas and beyond; the potential for increased production in Iraq as security improves; and the possibility of higher Saudi output. None of this gets mentioned in the essay. None of it, of course, shows that oil production will be sufficient going forward, but the argument in the paper falls woefully short of demonstrating the opposite.

What about resources other than oil? Here’s how the authors deal with that:

“Many believe that coal will be the solution to our energy needs, and will stay cheap for decades. But several recent studies suggest that available coal is less abundant than has been assumed. US coal production peaked in 2002, and world coal-energy production is projected to peak as early as 2025 (ref. 8)…. The US National Research Council’s Committee on Coal Research, Technology, and Resource Assess­ments to Inform Energy Policy noted in 2007 that ‘present estimates of coal reserves are based upon methods that have not been reviewed or revised since their inception in 1974 … updated methods indicate that only a small fraction of previously estimated reserves are actually mineable reserves.’”

Let’s start by getting the fact straight again. Yes, U.S. coal production peaked around 2002 (actually in 2001); then, in 2003, it reversed course again and started heading back up. Coal production in 2005, 2006, 2007, and 2008 exceeded 2001 and 2002 levels. More important, production is a function not only of supply but demand. Coal production in the United States has been restrained not because it’s scarce but because it’s increasingly undesireable.

What about the observation that “only a small fraction of previously estimated reserves are actually mineable reserves”? It’s important to put that in context. Previous estimates of coal reserves are astronomical. The mere fact that they have been revised downward tells us essentially nothing about whether the world has enough coal.

And then there’s natural gas:

“Natural gas is still abundant and large discoveries have been made recently, notably in Israel and Mozambique last year. Power plants using natural gas provide 25%, and rising, of electricity generation in the United States. Production of conventional natural gas in North America peaked in 2001 (ref. 2), but energy companies have worked hard to promote the idea that hydraulic fracturing of shale rock will lead to ‘the age of natural gas’. There is no doubt that US shale-gas resources are immense, but recent reports suggest that both reserves and future production rates have been substantially overstated….”

It’s not clear what the authors’ point is here. On the one hand, they cite some of the biggest pessimists in the natural gas world; on the other, they talk about “immense” resources. In any case, the bigger question mark for the purposes of their essay would seem to be whether natural gas can penetrate the transport sector in a big enough way to dent demand for oil. (I happen to be slightly skeptical.) The essay, though, is silent on that count.

Having tried to make their case for an inevitable oil shortage, the authors turn to consequences:

“What does this mean for the global economy, which is so closely tied to physical resources? Of the 11 recessions in the United States since the Second World War, 10, including the most recent, were preceded by a spike in oil prices. It seems clear that it wasn’t just the ‘credit crunch’ that triggered the 2008 recession, but the rarely-talked-about ‘oil-price crunch’ as well. High energy prices erode family budgets and act as a head wind against economic recovery.”

Any scientist knows that correlation and causation are different things. The authors cite a serious paper that claims that expensive oil helped bring the 2008 recession forward by a few months, but don’t have any backup (aside from the fact that it “seems clear” to them) for their claim that pricy crude actually “triggered” it. Instead, they turn to anecdotes:

“Another powerful example of the effect of increasing oil prices can be seen in Italy. In 1999, when Italy adopted the euro, the coun­try’s annual trade surplus was $22 billion. Since then, Italy’s trade balance has altered dramatically and the country now has a deficit of $36 billion. Although this shift has many causes, including the rise of imports from China, the increase in oil price was the most important. Despite a decrease in imports of 388,000 barrels per day compared with 1999, Italy now spends about $55 billion a year on imported oil, up from $12 billion in 1999. That difference is close to the current annual trade deficit. The price of oil is likely to have been a large contributor to the euro crisis in southern Europe, where countries are completely dependent on foreign oil.”

This is mostly the subject for a future post, but I’ll just point out for now that oil imports have little or no power to predict current account deficits. Both Germany and China, for example, import a lot of oil, yet they have not run current account deficits in recent years. The Italy example is nothing more than cherry picking. As for the claim about the euro crisis, one would presumably need to explain why others with big import bills haven’t been similarly affected in order to conclude as the authors do. They make no attempt to do that, but they do have a bottom line:

“Historically, there has been a tight link between oil production and global economic growth. If oil production can’t grow, the implication is that the economy can’t grow either.”

Once again, correlation and causation are different things. The U.S. economy, for example, has grown much faster than its oil consumption has. Europe has grown while its oil consumption has remained flat. Could limits to oil production restrain economic growth? Sure. But the case is far less clear cut than the essay suggests.

The rest of the Nature article gets into solutions, most of which are eminently sensible, even if one doesn’t buy the authors’ description of the problem. I won’t get into them here, since that’s not the point of this post.

But I want to finish with a broad observation. When opponents of action to deal with climate change rest their case on weak and logically unsound arguments, they are rightly assailed by serious scientists – including, quite often, the authors of the present essay. When these same climate experts cross over into talking about oil, it would behoove them to write with the same rigor that they demand of their adversaries elsewhere.

Post a Comment 14 Comments

  • Posted by David B. Benson

    John Morgan offers an Australian perspective worthy of reading (I’ve found him to write carefully):

    He mostly considers a rathr buried Australian study which was two years in the making; probably far for worthy than the obviously hurried study published in Nature.

    In any ase, the world is finite and so eventually humans will indeed run out of inexpensive petroleum and coal.

  • Posted by Nick Grealy

    The paper does sound as if they are desperately clutching at Peak Oil straws, but what is scary is the level of influence of people like David King.

    To mention Mozambique and Israel natural gas, but not (multiple) US, Canada, Argentina, Poland, UK or Australian shale discoveries tells you a lot.

    Nobody, but nobody saw shale coming out of the blue and creating an inflection point in energy: anything that happened before shale is basically besides the point. Some people saw it sooner ( Daniel Yergin, me, and Dieter Helm in that order for example), but it surprised everyone.
    Denialists like King, or Art Berman, simply don’t think anything changed. They continue to deny reality and will get swept away with it.

  • Posted by wadosy

    oh, man…

    does this mean that the PNAC/AEI/exxon people didnt have to do 9/11 in response to peak oil?

    3000 american lives, countless lives elsewhere… wasted, because nobody saw shale oil coming.

    what a tragedy.

  • Posted by jukka

    annual energy outlook, table C6

    conventional production: 79.6
    nonconventional: 4.14
    total : 83.74

    conventional in this includes apparently condensate, so just crude oil has been around 75 or a bit lower in recent years.

    [ml: this is for 2009, which saw weak demand. 2008 is higher. Fair point on condensate.]

  • Posted by mmckinl

    We are using oil at 5 times the rate of new discoveries …

    nuff said …

  • Posted by hurin

    Lets take the facts again, shall we.

    We are not running out of oil, but we have almost run out of cheap oil. Deep sea oil, shale and tar sand require an oil price of 100$ to work.

    Oil is finite, production will peak and decline. Production has been flat since 2004, and there is no reason to believe it will rise again. OPEC reserves are faked, to believe they’re real, you would have to believe OPEC nations reserves grow with the exact same amount they pump out of the ground year, after year, after year.

  • Posted by Mason Inman

    Hi Michael,

    It seems to me you’ve made a number of factual errors here, which I’ll run through. (Another version of this, with graphs, is posted on my blog: I called my post “How not to argue against the idea of peak oil.” If you’re going to try to shoot down people like Murray and King,

    First off, they weren’t arguing that we are running out of oil—at least in the sense that we would sometime soon not have any more oil to use. They’re talking about the rate of oil production, and whether we can manage to push it higher. When you equate “peak oil” with “running out of oil,” you’re pulling the same kind of switcheroo that Daniel Yergin does.

    [ML: It’s shorthand. That choice of words has zero impact on the argument in the post.]

    Also, it seems you got things upside down with the reserves-to-production ratio. You wrote: “If a technological breakthrough doubled our reserves tomorrow, the reserve-to-production ratio would be sliced in half.”

    Actually, since the reserves are on the top of the ratio, if reserves doubled overnight, then the reserve-to-production ratio would also double. And that would be a good thing for delaying the peak of production.

    [ML: Thanks for catching the typo. I hope it’s clear from the context that the original argument should stand unchanged.]

    On the current level of production in recent years, you wrote “the authors seem to have misread their reference.”

    No, they’re talking about “regular crude oil,” not “all liquids” or “all conventional liquids.” Their reference on this point was the EIA’s 2011 Annual Energy Outlook; you can see that here:

    It seems that when Murray and King talked about “regular crude oil,” they meant the same thing that the EIA calls “conventional crude oil and lease condensate.” You only get up over 75 million barrels a day if you also throw in natural gas liquids, which, in my understanding, most people don’t consider “regular crude oil.”

    [ML: Fair point — updated accordingly. I think that drawing the line this way leads to confusion — the other stuff in the conventional category isn’t “expensive oil” — but, as a straightforward technical matter, the number in the paper is correct.]

    On that EIA graph, production of “conventional crude oil and lease condensate” has been flat at about 72 million barrels a day since 2005, and according to the EIA’s reference scenario projection, will remain at that level through 2025.

    On the issue of unidentified projects, I think you’re wrong again. You wrote: “Unidentified projects don’t involve oil that is ‘yet to be discovered’ – they involve oil that has not been specifically targeted for extraction. In fact, as best I understand, all of the production increase that EIA anticipates comes from oil that has already been discovered.”

    It sounds like you’re saying that you think the EIA makes a projection that is led by the supply available.

    [ML: I’m not saying that.]

    However, in my understanding, both the EIA and the IEA make “demand-led” projections. So then the question is, is there enough supply to meet the projected demand? We will undoubtedly continue discovering more oil—but how much more would we need to find to meet the demand?

    When Murray and King mention “unidentified projects,” it seems they’re referring to a 2009 presentation by Glen Sweetnam of the EIA, “Meeting the World’s Demand for Liquid Fuels.” (Note how the problem is framed not in terms of available supply, but in terms of meeting demand.)

    Sweetnam’s presentation has become sort of famous in peak oil circles for showing a huge gap between projected demand and the oil from existing and planned projects. This huge gap is labeled “unidentified projects.” Here’s the link to Sweetnam’s presentation; the graph is on page 8:

    How much of this gap could be filled from known reserves, and how much would have to come from new discoveries? For an answer to that, we can look to the IEA’s 2010 World Energy Outlook.

    There, they show production from active fields dropping off over the coming decades. Then they show production from known fields that haven’t yet been developed. Finally, they show the “fields yet to be found.” (For the graph and more details on this, see my post “The future of oil and the wedge of hope”:

    This wedge of fields yet to be developed or yet to be found is very close to the same size of the wedge of “unidentified projects” in Sweetnam’s presentation. So I gather they’re both talking about the same issue.

    [ML: This is a messy way to make an argument. You’re talking about two entirely different presentations. I wouldn’t try to match the two. In any case, the claim that I was going after — that all “unidentified projects” involve oil that has not yet been found — remains unequivocally wrong.]

    If we do fill the gap and meet this projection of rising demand, then around 2030, about half this gap would have to be filled from newly discovered oil fields, producing about 20 million barrels a day. To put that in perspective, that’s the equivalent of today’s two biggest oil producers—Russia and Saudi Arabia—combined. So I call that the “wedge of hope.”

  • Posted by cyrill landau

    How about nuclear power?

  • Posted by Mason Inman

    I just wanted to follow up about one thing. In reply to my comment, you wrote: “the other stuff in the conventional category isn’t ‘expensive oil’.”

    But the “end of cheap oil” doesn’t mean that all of the stuff that’s cheap to produce is literally gone. Saudi oil is still pretty cheap to produce—far, far below the price they sell it for now. (I’m basing that on what Paul Stevens at Chatham House has said about the Saudis’ cost of production.)

    In my understanding, the international rates now are set largely by the cost of the *marginal* supply. That is, if the only way we can boost production is by producing some expensive oil, then most oil will get sold at expensive prices.

    So some of the additional supply above the “regular crude oil” might not be particularly expensive to produce. But since we’re demanding enough that we’re tapping things like tar sands, then most all oil then becomes “expensive oil.”

    Right? If you have a different perspective on this, I’m interested to hear.

    [ML: That’s true. Can you show me where the Nature paper makes this point?]

  • Posted by Michael Aucott


    You point out well some of the weak spots in the Murray and King paper. But you seem to miss the key point: oil economics abruptly shifted around 2004-2005 from elastic to inelastic supply. This dramatic change is clearly presented in the authors’ “Phase Shift” chart. It seems highly unlikely that this change can be attributed to anything other than production of crude oil hitting a ceiling at about the same time. The idea that instead this marked change is due to some sort of political attitude shift that suddently occurred during that short period seems far-fetched.

    As far as new supplies of fossil fuels are concerned, much better data are needed on the net energy – the energy return on energy investment (EROI) – of these supplies before there can be confidence in their ability to maintain business as usual. For instance, the recent increases in total liquid supply (as opposed to crude oil supply) are due primarily to increases in natural gas liquids and biofuels, some of which (e.g. ethanol from corn) yield relatively little net energy. It is conceivable that the net energy available in the form of liquid fuels, like its main component, crude oil, has peaked or is near peak.

  • Posted by methuselah

    Since no one else has mentioned it…

    Seems like a key point should be acknowledged: I think it was NASA’s Dr. James Hansen who stated if we burn all of the fossil fuels currently safely sequestered underground, we run an actual risk of triggering drastic, rapid climate disruption, with potentially catastrophic consequences.

    Keeping that carbon sequestered by not burning it is far cheaper than building new magic tech sequestration too late to mitigate. Does it really matter how much oil is in reserves? Burn it, and we’ll all get to find out what happens when the carbon bill comes due. Seems like we can do better.

  • Posted by jim harvey

    The “wedge of hope” argument:

    It’s not only possible but reasonable to use the geologic inventory of rocks with similar compositions to those in, say, the Bakken, to infer an undiscovered resource. Bakken estimates are steadily rising…

    Tight oil and other unconventional oil production is in its infancy. Granted, it might be hopeful to expect production efficiency improvements on the order of that experienced by the semiconductor industry over the last 50 years, but techniques will certainly be refined to extract more oil with less energy – this has already happened in Alberta’s tar sands.

    And, of course, much of the US is currently off limits to drilling – almost all US offshore is restricted. What if oil companies are able to improve their drilling techniques to the point where spills are virtually impossible? Can this be opened? Did we put a man on the moon?

  • Posted by Steve

    You haven’t really dealt with the wedge of hope. Why? You’ve tried to rebut other arguments, admittedly with magical thinking about new technology improvements, always just around the corner.

    And can you reference anywhere where Saudi Arabia has said it will increase its supply by 2020, for instance, to fill this increase in demand expected by the IEA and EIA.

    The west is built on relatively cheap oil, exactly what price do you think maintenance, let alone economic growth, can continue at, because it certainly didn’t like 150 dollars.

  • Posted by Gunnar Kaestle

    I agree completely with Michael Aucott that the main message of the Murray & King paper is the change of price elasticity. You may find a regression curve with a slope of 2,5$/bbl per mbd for the dots until 2004. The slope for the data beginning in 2005 is about 80$/bbl for an extra production of a million barrels per day.

    This is crucial, as the crude oil market changed towards a sellers market. The prices is not set by the production costs of the next available wellhead, but the ability of the consumer to find a substitute or just to without. The oilprice shocks in the 70s gave a hard landing to the global economy until the OPEC opened the pipelines again.

    It seems that Saudi Arabia – if it still has the ability as a swing production state – is not willing to change the situation, as billions are exchanged for still mbd of oil which can be produced very cheaply.

    Regarding the question of nuclear power: No, NPPs won’t help as the oil peak will mainly affect the transportation sector. It’s no energy crisis, but the use of cars and trucks and trains and jets will be significantly more expensive. On the other hand, nuclear power will probably have no renaissance in OECD countries.

    The simple truth is that NPPs are extremly expensive and a very risky investment for a utility. There are less riskier options available to make money. “Government policy remains that the private sector takes full exposure to the three main risks; Construction, Power Price and Operational. Nowhere in the world have nuclear power stations been built on this

Post a Comment

CFR seeks to foster civil and informed discussion of foreign policy issues. Opinions expressed on CFR blogs are solely those of the author or commenter, not of CFR, which takes no institutional positions. All comments must abide by CFR's guidelines and will be moderated prior to posting.

* Required