Lately I've been getting annoyed by bituminous tar sands. This, I'll grant, makes me "not normal", but it seems like every time someone mentions peak oil, some Doubting Thomas trots out the vast, relatively untapped deposits of oil sands that will save us from our bleak future. I think this is bollocks, and that optimism is wholly misplaced.

There's a whole host of reasons why this is true. There's essentially two mega-fantastico deposits of oil sands in the world, in Alberta, Canada (the Athabasca sands) and Venezuela (the Orinoco belt). These are huge deposits: the figures quoted are on the order of trillions of barrels for each. That's enough oil to last us for centuries. But I don't think it will ever be produced.

A few years ago Canada decided to include oil sands in their reserve estimates, and ever since then there's been growing excitement about the coming boom, with Canada being compared favorably to Saudi Arabia. Despite that, relatively little oil is actually forthcoming so far: only about a million barrels per day. The Kingdom, meanwhile, produces 9 Mbd, and claims to be able to produce 11.5 Mbd.

There's two ways of getting oil from oil sands: The "easy" method, the one mostly employed now, is essentially a mining operation. You mine the oil sands out of the ground, just like you would mine coal or bauxite or whatever. Then you cart it away in trucks to your processing facility, wash the oil out from the sand, do some post-processing, and send it off to market. The harder method is "in-situ" production, which works basically like a lot of modern late-stage conventional crude operations: you pump hot steam into the ground, which displaces and loosens up bitumen, which you suck up out of the ground and process, etc.

This latter method will be more critical in the future, as the vast majority of the oil sands in Canada are not accessible via mining methods: they're too deep underground. Mining operations require a lot of water, but it can be recycled (mostly - the untreatable water is put into a holding lake. Syncrude's is currently 4.5 miles in diameter). In-situ methods mean losing the water, at about a one-to-one ratio with recovered oil. In other words, if you want to produce 9 Mbd of oil, you have to pump 9 Mbd of water into the ground. By comparison, the flow rate of the Athabasca river varies between 2.5 Bbd and 40 Mbd. That water is permanently removed from the hydrological cycle.

Let's pretend we don't care about that, or the other horrendous environmental effects associated with this operation, as with any other mining operation: let's pretend we're willing to dessicate the aquifer and it won't end up being prohibitive to production. We're still stuck with the problem of natural gas. In order to get bitumen up to snuff, it's necessary to add hydrogen, to get the higher grade of fuel needed for jets, automobiles, etc. This requires gas inputs (as do other parts of the process e.g. heating water for steam injection). Right now, gas inputs for hydrogen upgrading alone amount to 400 cubic feet; some day, in order to produce high quality fuel this may reach as high as 1700 cubic feet. 5487 cubic feet of gas is considered to be one barrel of oil equivalent. That's a significant energy input, merely to upgrade bitumen to the standard necessary for high quality fuels.

Let me quote Alan Greenspan from his testimony before the House in 2003:

Because gas is particularly challenging to transport in its cryogenic form as a liquid, imports of LNG have been negligible. Environmental and safety concerns and cost have limited the number of LNG terminals and imports of LNG. In 2001, LNG imports accounted for only 1 percent of U.S. gas supply. Canada, which has recently supplied a sixth of our consumption, has little capacity to significantly expand its exports, in part because of the role that Canadian gas plays in supporting growing oil production from tar sands.

I.e., Canada (like everyone else) is strapped for gas. It's going to be difficult for them to match the demands of their burgeoning oil sands industry. Especially when you consider that Canadian gas production is due to peak in only a few short years - after that, oil sands will very quickly (in on the order of ten years from now) become completely unsustainable if they continue to rely on natural gas.

So - gas requirements, water requirements, general environmental devastation. What else can we add to the mix? How about - CO₂? Since producing tar sands is so fuel-intensive, there's a much higher burden on the production of greenhouse gases. This is the last thing the world needs right now, and this is certainly going to become difficult for Canada as things like Kyoto actually start to get some teeth.

The most optimistic scenarios for Canadian oil sands project them being able to produce 5 Mbd by 2030. The National Energy Board of Canada has a more conservative 3 Mbd. I think this is reasonable, and underscores what Colin Campbell has to say about the subject: "The key point about tar sands is that the resource is huge but the extraction rate is very low."

So don't hold your breath.