On oil discovery rates

Grrr. A post at Environmental Economics dismisses the relevance of Peak Oil in the same patronizing tone I've learned to love from people who know more than enough economics, but not nearly enough geology or history.

So some history first: Global oil discovery hit it's all-time high in the late Truman administration, with another smaller peak in the 1960s. Since then, oil discovery rates - that is, the amount of new oil we're discovering every year - have declined. In the 1980s, we passed a milestone: We began using more oil every year than we discovered - global reserves were shrinking for the first time in history.

A simple, obvious, but often overlooked fact: We can only use the oil we know exists. So the fact that for 20 years we've been eating in to "our" reserves of oil without replenishing them with new discoveries is important.

So how does John Whitehead respond to this fact? The same response, it seems, that he has to every issue raised by a Peak Oil-aware commenter:

I would only be concerned if prices weren't responding. But they are. Again, prices need to rise to create the necessary incentives. Market based high oil prices are not a problem. Artificially low oil prices are.

A perfectly reasonable argument on it's own, but the implication seems to be that as prices increase discovery rates will tick up as well. This is where we run in to problems. The vast majority of the Earth's land surface has already been surveyed for conventional oil reserves, and most of this work was done by the middle of the 20th century. While I certainly won't discount the possibility of a few large discoveries, the overwhelming likelihood is that future discoveries of conventional oil fields on land will be marginal, while existing fields will eventually all go in to decline.

Which leaves us with the non-conventional sources - tar sands, shale, arctic finds, and the ultra-deepwater finds. Let's deal with the arctic and ultra-deep issue first: Despite a lot of promise, the actual discoveries in the high arctic and the gulf of Mexico have not met the expectations from the 1990s. Hell, we've been talking about gas and oil from the arctic since the 1960s, and we still haven't managed to extract meaningful amounts.

Frankly, there's no good reason to believe that there's some new Saudi-sized field out there, waiting for us to discover it. And this goes to geology - oil is an extremely rare, concentrated resource that requires a very specific number of coincidences to occur for it to form. It can only form in rocks of a certain type and age, and can only persist for millions of years in certain rock formations.

It will take some massive efforts to search the oceans for these kinds of formations, and I don't think any company - even the Exxons of the world - have that kind of cash. In any case, they keep cutting their exploration budgets - meaning prices are not yet high enough (at $60/barrel!) and need to go higher before the majors start looking for more oil.

Which leaves us the really ugly options - tar sands, shale, etc. We in Canada have truly massive reserves of a truly awful resource: one that requires massive amounts of energy to create, ruins the land it's extracted from, and consumes massive amounts of water to do so. Being perfectly rational people, we've decided that the best use of our money is to develop this resource as fast as possible.

It's true that prices will help increase the production from the tar sands, though less than is often surmised. Like any good, the tar sands are only profitable when the product (bitumen) exceeds the costs of production. And all of the costs of production for the bitumen have been going up like gangbusters - labour, infrastructure, and Canadian natural gas keep going up. Unless we build either a nuclear reactor or a liquid natural gas terminal in BC to keep feeding energy in to the tar sands, we could be facing serious production problems in less than a decade.

But, let's assume for a moment we don't. The most wildly optimistic prediction for the tar sands assumes they produce 4 million barrels a day by 2020. The Canadian government, more conservatively, assumes they'll produce only 3 million a day by then. So even by the most optimistic assumptions, Canada (the largest and most mature "non-conventional" producer of oil) will produce less than 1/20th of the world's 2005 demand. How many barrels of production will the world lose between now and then? Considering we're only talking about adding 3 million barrels a day from the tar sands (we already produce 1mbpd) I think it's almost certain we'll lose more cheap oil (from existing fields) than we gain in expensive tar sands.

The other function of high prices is to stimulate the search for alternatives, it's true. And I believe that those alternatives exist. But to blithely say that high prices are a positive sign and the market will make the transition is absurd. Plenty of people are going to be stuck with gasoline-powered cars and unable to afford the new model electric car (or whatever.) What peak oil will do is not simply increase prices: it will, if the moderately bad predictions are correct, make much of our current way of life unaffordable to most people.

To say, as Mr. Whitehead does, that a) CERA's (highly questionable) data says the peak won't occur until 2040, so relax, and b) that rapidly rising prices are a good thing and will push the peak further off, totally misses the point that the Peak Oil community keeps trying to make: oil is a much scarcer resource than is commonly understood, it is irreplaceable in many aspects of our daily life, and it will be less available in the future than it is today. This means severe economic dislocation.

This shouldn't be hard for economists, of all people, to understand. But I swear the group of people who seem most committed to an unrealistic optimism about oil all have Economics Phds.