Peak Oil and Climate Change

Dieter Helm, writing in the Oxford Review of Economic Policy, offers the counter-argument to the point made recently by Andy Dessler that the peak and inevitable decline of fossil fuels may put some sort of upper limit on our ability to to put carbon dioxide into the atmosphere:

Some argue that, however intense the dash-for-resources, China (and others) will fail because they will run out. In other words, what will decarbonize the world economy is that there will not be enough carbon resources left to deplete. A particular version of this argument is called ‘peak oil’. It is argued that the peak of conventional oil production has been already reached, and that production will decline over the coming decades, just as demand rises. Prices will, therefore, increase very sharply, setting off a significant substitution effect. From a climate-change perspective, it is argued that this would be good news, but there
are major weaknesses with the peak-oil hypothesis. There may well be lots more resources to discover, notably in an increasingly ice-free Arctic Ocean, and then in Antarctica (as well as elsewhere). Existing ‘depleted’ oil wells typically retain over half the initial reserves, and new technologies (not least using CO2 to increase pressure through oil-enhanced recovery) may be effective. Then there are many near substitutes which may be available, notably tar sands. Once these are included, Canada, for example, becomes one of the top three oil-reserve countries. These sources may be even more polluting than conventional ones. There is also lots of coal.

(Thanks to David Zetland for bringing this paper to my attention.)


  1. John,
    Thanks for tracking down yet another useful article from a publication that I have never heard of.

    One of my worries about some of the climate change arguments is not whether the error bars on the estimates are 30% instead of 5% but whether the error bars are down the street or two counties over.

    In the financial meltdown and the sub-prime mess, the error bars seemed to be wrong by factors of ten or more.

    I would love to see a thorough Bayesian analysis of ‘peak oil’ including the factors listed in your posting.

  2. The cornucopians are always arguing that 1) there’s oil everywhere we haven’t looked and 2) we can get more oil out of existing fields. I don’t think evidence supports either conjecture. You run into the same rosy thinking in the climate change debate: “technology will seamlessly solve all of our problems.” As for non-conventional fuels, the problem w/ them is *rate* or production, not total reserves. In any event, this just emphasizes the need for a better assessment of our reserves.

  3. Andy –

    I agree that this discussion cries out for a better assessment of our reserves. The problem with the somewhat value-laden dismissal implicit in your term “cornucopians” is that the cornucopians have thus far frequently been right. Obviously they can’t be right forever, but until the analysis takes into account the additional resources that will become recoverable as energy becomes permanently more expensive, we really don’t have a good answer for this question one way or the other.

  4. I clicked through to Hubbert’s paper and Rutledge’s update. In Hubbert’s paper there is a function f(Q) that seems to depend on the resource that is being tapped. To a technologist like me, the function would appear to be f(Q(T),T(t)), an interesting use of parentheses meaning

    The size of a reservoir Q depends on T, the technology used to assess it and the technology used to extract it and the value of the function f at a particular time depends on the T existing at that time both in research labs and in practice, which is often behind what is in research labs.

    Are the technology and price dependent variables and their lags in implementation included in predictions of oil and coal reserves some where?

    For instance, are all the oil reserves quantitated by using the same current technology?

    When I worked with microarray data, you could not add up numbers without correcting for a lot of underlying technology dependent variables.


  5. John, see this Oil Drum post for a discussion of the potential for additional recovery from depleted oil fields. On the face of it, Helm’s take seems optimistic (quoting the post summary):

    How can you double something and still have ten times less than you started with?

    In the case of the Handil Oil Field, a concerted campaign to revitalize the field almost doubled production from 12,500 barrels per day in 2003 to 23,000 in 2007. Yet this field had once produced nearly two hundred thousand barrels per day.

    This is a representative picture of the role of technology and enhanced oil recovery: merely extending the tail end of production in oil fields that are well past their own peak in production.

    So while the Society of Petroleum Engineers and other optimists tell us that technology and enhanced oil recovery will delay peak oil, a more objective look at the data suggests that production declines are relentless and they are stacking up much faster than incremental technological gains.

    The key question in all of this seems to be about the coal.

  6. I have two friends who think Peak Oil is certainly real, coming soon, and we’d better be working hard on efficiency and substitutions. Given that each of them was Chairman or Vice-Chariman of one of the major oil companies, I tend to listen to them…

    The real issue on recoverability is ERO(E)I, Energy Return On (Energy) Invested, for which I recommend Charlie Hall.

    When 1 barrel of oil got you 100 back, life was good (1930).
    If 1 barrel of oil gets you 1 back, life isn’t good, because it doesn’t matter what the barrel sells for… At some point above that, new oil just doesn’t make any sense compared to other things.

Comments are closed.