The impending Obama administration decision on the Keystone XL Pipeline, which would tap into the Athabasca Oil Sands production of Canada, has given rise to a vigorous grassroots opposition movement, leading to the arrests so far of over a thousand activists. At the very least, the protests have increased awareness of the implications of developing the oil sands deposits. Statements about the pipeline abound.
Jim Hansen has said that if the Athabasca Oil Sands are tapped, it’s “essentially game over” for any hope of achieving a stable climate. The same news article quotes Bill McKibben as saying that the pipeline represents “the fuse to biggest carbon bomb on the planet.” Others say the pipeline is no big deal, and that the brouhaha is sidetracking us from thinking about bigger climate issues. David Keith, energy and climate pundit at Calgary University, expresses that sentiment here, and Andy Revkin says “it’s a distraction from core issues and opportunities on energy and largely insignificant if your concern is averting a disruptive buildup of carbon dioxide in the atmosphere”. There’s something to be said in favor of each point of view, but on the whole, I think Bill McKibben has the better of the argument, with some important qualifications. Let’s do the arithmetic.
There is no shortage of environmental threats associated with the Keystone XL pipeline. Notably, the route goes through the environmentally sensitive Sandhills region of Nebraska, a decision opposed even by some supporters of the pipeline. One could also keep in mind the vast areas of Alberta that are churned up by the oil sands mining process itself. But here I will take up only the climate impact of the pipeline and associated oil sands exploitation. For that, it is important to first get a feel for what constitutes an “important” amount of carbon.
That part is relatively easy. The kind of climate we wind up with is largely determined by the total amount of carbon we emit into the atmosphere as CO2 in the time before we finally kick the fossil fuel habit (by choice or by virtue of simply running out). The link between cumulative carbon and climate was discussed at RealClimate here when the papers on the subject first came out in Nature. A good introduction to the work can be found in this National Research Council report on Climate Stabilization targets, of which I was a co-author. Here’s all you ever really need to know about CO2 emissions and climate:
- The peak warming is linearly proportional to the cumulative carbon emitted
- It doesn’t matter much how rapidly the carbon is emitted
- The warming you get when you stop emitting carbon is what you are stuck with for the next thousand years
- The climate recovers only slightly over the next ten thousand years
- At the mid-range of IPCC climate sensitivity, a trillion tonnes cumulative carbon gives you about 2C global mean warming above the pre-industrial temperature.
This graph gives you an idea of what the Anthropocene climate looks like as a function of how much carbon we emit before giving up the fossil fuel habit, without even taking into account the possibility of carbon cycle feedbacks leading to a release of stored terrestrial carbon The graph is from the NRC report, and is based on simulations with the U. of Victoria climate/carbon model tuned to yield the mid-range IPCC climate sensitivity. Assuming a 50-50 chance that climate sensitivity is at or below this value, we thus have a 50-50 chance of holding warming below 2C if cumulative emissions are held to a trillion tonnes. Including deforestation, we have already emitted about half that, so our whole future allowance is another 500 gigatonnes.
Proved reserves of conventional oil add up to 139 gigatonnes C (based on data here and the conversion factor in Table 6 here, assuming an average crude oil density of 850 kg per cubic meter). To be specific, that’s 1200 billion barrels times .16 cubic meters per barrel times .85 metric tonnes per cubic meter crude times .85 tonnes carbon per tonne crude. (Some other estimates, e.g. Nehring (2009), put the amount of ultimately recoverable oil in known reserves about 50% higher). To the carbon in conventional petroleum reserves you can add about 100 gigatonnes C from proved natural gas reserves, based on the same sources as I used for oil. If one assumes that these two reserves are so valuable and easily accessible that it’s inevitable they will get burned, that leaves only 261 gigatonnes from all other fossil fuel sources. How does that limit stack up against what’s in the Athabasca oil sands deposit?
The geological literature generally puts the amount of bitumen in-place at 1.7 trillion barrels (e.g. see the numbers and references quoted here). That oil in-place is heavy oil, with a density close to a metric tonne per cubic meter, so the associated carbon adds up to about 230 gigatonnes — essentially enough to close the “game over” gap. But oil-in-place is not the same as economically recoverable oil. That’s a moving target, as oil prices, production prices and technology evolve. At present, it is generally figured that only 10% of the oil-in-place is economically recoverable. However, continued development of in-situ production methods could bump up economically recoverable reserves considerably. For example this working paper (pdf) from the National Petroleum Council estimates that Steam Assisted Gravity Drainage could recover up to 70% of oil-in-place at a cost of below $20 per barrel.
Aside from the carbon from oil in-place, one needs to figure in the additional carbon emissions from the energy used to extract the oil. For in-situ extraction this increases the carbon footprint by 23% to 41% (as reviewed here ) . Currently, most of the energy used in production comes from natural gas (hence the push for a pipeline to pump Alaskan gas to Canada). So, we need to watch out for double-counting here, because our “game-over” estimate already assumed that the natural gas would be used for one thing or another. A knock-on effect of oil sands development is that it drives up demand for natural gas, displacing its use in electricity generation and making it more likely coal will be burned for such purposes. And if high natural gas prices cause oil sands producers to turn from natural gas to coal for energy, things get even worse, because coal releases more carbon per unit of energy produced — carbon that we have not already counted in our “game-over” estimate.
Are the oil sands really the “biggest carbon bomb on the planet”? As a point of reference, let’s compare its net carbon content with the Gillette Coalfield in the Powder river basin, one of the largest coal deposits in the world. There are 150 billion metric tons left in this deposit, according to the USGS. How much of that is economically recoverable depends on price and technology. The USGS estimates that about half can be economically mined if coal fetches $60 per ton on the market, but let’s assume that all of the Gillette coal can be eventually recovered. Powder River coal is sub-bituminous, and contains only 45% carbon by weight. (Don’t take that as good news, because it has correspondingly lower energy content so you burn more of it as compared to higher carbon coal like Anthracite; Powder River coal is mined largely because of its low sulfur content). Thus, the carbon in the Powder River coal amounts to 67.5 gigatonnes, far below the carbon content of the Athabasca Oil Sands. So yes, the Keystone XL pipeline does tap into a very big carbon bomb indeed.
But comparison of the Athabaska Oil Sands to an individual coal deposit isn’t really fair, since there are only two major oil sands deposits (the other being in Venezuela) while coal deposits are widespread. Nehring (2009) estimates that world economically recoverable coal amounts to 846 gigatonnes, based on 2005 prices and technology. Using a mean carbon ratio of .75 (again from Table 6 here), that’s 634 gigatonnes of carbon, which all by itself is more than enough to bring us well past “game-over.” The accessible carbon pool in coal is sure to rise as prices increase and extraction technology advances, but the real imponderable is how much coal remains to be discovered. But any way you slice it, coal is still the 800-gigatonne gorilla at the carbon party.
Commentators who argue that the Keystone XL pipeline is no big deal tend to focus on the rate at which the pipeline delivers oil to users (and thence as CO2 to the atmosphere). To an extent, they have a point. The pipeline would carry 500,000 barrels per day, and assuming that we’re talking about lighter crude by the time it gets in the pipeline that adds up to a piddling 2 gigatonnes carbon in a hundred years (exercise: Work this out for yourself given the numbers I stated earlier in this post). However, building Keystone XL lets the camel’s nose in the tent. It is more than a little disingenuous to say the carbon in the Athabasca Oil Sands mostly has to be left in the ground, but before we’ll do this, we’ll just use a bit of it. It’s like an alcoholic who says he’ll leave the vodka in the kitchen cupboard, but first just take “one little sip.”
So the pipeline itself is really just a skirmish in the battle to protect climate, and if the pipeline gets built despite Bill McKibben’s dedicated army of protesters, that does not mean in and of itself that it’s “game over” for holding warming to 2C. Further, if we do hit a trillion tonnes, it may be “game-over” for holding warming to 2C (apart from praying for low climate sensitivity), but it’s not “game-over” for avoiding the second trillion tonnes, which would bring the likely warming up to 4C. The fight over Keystone XL may be only a skirmish, but for those (like the fellow in this arresting photo ) who seek to limit global warming, it is an important one. It may be too late to halt existing oil sands projects, but the exploitation of this carbon pool has just barely begun. If the Keystone XL pipeline is built, it surely smooths the way for further expansions of the market for oil sands crude. Turning down XL, in contrast, draws a line in the oil sands, and affirms the principle that this carbon shall not pass into the atmosphere.