Carbon Sequestration Validation

So much for keeping a low profile at Fall AGU.

Oh well, as my putative scientific career finally evaporates altogether, I suppose it's good to have a mission. I'm going back to my self-defeating old habit of always being most interested in the things I know least about. This is pretty self-defeating scientifically, but maybe now that Revkin is semi-retired I can find a niche for myself as a science reporter. So consider me your man on the scene.

My favorite of the talks I attended today was by Sally Benson of Stanford, an invited talk on verifying carbon sequestration.

As another carbon sequestration speaker said, "it's all about the credits": energy interests aren't especially interested in carbon sequestration unless and until we put a price on carbon, but once we do so, they'll be very interested in it. Of course, this is a famously scrupulous and fastidious industrial sector, so we can presume they won't cut corners and just pump carbon randomly into pipes in the ground whether it will stay there or not. But just in case, perhaps it would be best if we could somehow discriminate actual sequestration from random pipes into the depths.

Sarcasm aside, there's little doubt that a significant amount of CO2 can be hidden underground. There's every evidence of serious people doing serious experiments to demonstrate all the pieces. And in the end, some form of sequestration is necessary to reverse carbon concentrations; whether it's what is now being called CCS or not, some way of taking carbon out of circulation in massive quantities is already the legacy of our generation, and likely of a generation or two to come.

It's the case that vast ancient deposits of methane remianed underground for millions of years. So gas can stay underground. It's also the case that we can pipe gas underground; the oil industry already does this to increase extraction. It makes sense, when you pull a resource fluid out, to pump its waste product into the same hole. (It's not really a hole. It's gaps in a granular medium. There never was an ocean of oil under that gusher.) But just because it's possible doesn't mean it's easy.

A leakage rate of 1% a year is as good as worthless. Even 0.1 % leaves only 37% of the gas underground after a millenium; you really have to get to 0.01% before the project starts doing substantial good (90% after a millenium) and a good goal is a leakage rate of 0.001% (99% after a millenium, 90% after 10,000 years).

And it's easy to get it wrong. Put too much gas underground (thereby avoiding building an expensive new facility) and you start inducing cracking in the cap rock (before I got to Texas, "cap rock" was not even in my vocabulary!) and your leakage rate starts to skyrocket.

Now in field tests, you can drill holes to check your theories, but in an actual deployment, I figure you can't be drilling too many holes to get samples. (I'm not actually sure about this, but my intuition rebels against making a swiss cheese out of your cap rock to make sure it is intact.) Anyway, Dr Benson's talk was about how to determine if you were getting the sequestration to work. The idea is that a regulator could refund you carbon points if you were successfully burying your carbon.

OK, so now there are two ways to proceed. You could try to measure the inventory, or you could try to measure the leakage. Both are problematic. You measure the inventory using seismic inversion; essentially sonar. The trouble with the sonar methods is twofold. First of all, it's messy: a lot of people make a career of having enough intuition to look at these things and find good places to explore or drill for this or that (usually that). Also, some CO2 dissolves or reacts. Even if your sonar were absolutely perfected you could easily lose 10 to 20% of your inventory. That is far too much uncertainty to be of use here.

The other approach is to look for leakage at the surface with arrays of sensitive CO2 detectors or isotopic measurements, to find sources of fossil carbon-based CO2 which is C-14 depleted. A source as "small" as 100 tons/year can be detected. Compare this to the rates of sequestration at a commercial site, on the order of 50 megatons of carbon inventory.

So that's sensitive enough. Are we done?

No, because the test is expensive and tightly focused in space, while the artificial reservoir is as large as 100 square kilometers. You have to kn ow eher to put the sensors. Here, though, is where seismic imaging does work. It can tell you something about the structure of the subsurface and give you good candidate locations to look for leaks.

Are we happy yet? Benson thinks so.

I have some doubts, myself. (The following is not part of Benson's talk but speculation of my own.)

The time scales are at issue. How long between springing a leak and detecting it? It seems like it could easily be decades before the leak erupts to the surface: very fast by geological standards but very slow by commercial ones.

Commercially, as a carbon combustion business, I want to be rewarded for my good deeds (really, non-penalized for my non-bad deeds) immediately. But as a government, I don;t want to credit you with the sequestered carbon until it doesn't leak. Most businesses will not take a thousand year lead time on their investments. So the problem, ultimately, isn't technical. It's the same old social problem. We don't know how to think on long enough time scales to make incentives for effective carbon sequestration work. Or so it looks to me. Did I miss something?