Under Pressure

Whoops....

As worthy engineering projects go, stabilising a lethal natural booby trap and providing power to an under-developed part of the world rank pretty highly. Solving lethal problems and providing infrastructure to people are high in the civil engineers motive drivers, well that and “oh waay cool” and “blow it up again” of course and as lethal problems go, this one is a corker.

Most people don’t consider carbon dioxide dangerous. If they do, they’ll normally react around “global warming” and give you some emotive guff about polar bears and the like. They forget that CO2 is both an asphyxiant and directly poisonous. Too much in the air and you can’t breathe it, too much of it in the blood stream and you keel over. Most people also don’t’ think deep lakes are dangerous either, well apart from “I wouldn’t swim in it” and other atavistic fears, most famously played upon in the mildly terrifying “the spirit of deep water” public information film. However combine a source of CO2 and a deep volcanic lake and you have the settings for a really lethal problem. Not lethal in the “and now invoke Cthulhu” global issue, but definitely lethal for everyone in the locale. Given a deep enough lake the pressure at the bottom is enough to make CO2 readily soluble. Given enough time, seriously enormous volumes of CO2 can dissolve and worse, given really enough time, bacterial can convert some of the CO2 into CH4 (methane for all you non chemists) and allow even more CO2 to dissolve. At lake Kivu in Rwanda(read more here at Lake Kivu wiki)  the current estimate is that there is 250 cubic kilometres of CO2 and 60 cubic kilometres of CH4 dissolved in the lake. That’s almost 70% of the volume of the entire lake as dissolved gas.

So what I hear you cry?

Well the problem is with dissolved gases at these concentrations is that anything which changes the pressure can release a lot of gas. That gas release changes the density of the water column and therefore reduces the pressure. Thus releasing more gas. Over a certain size of disturbance you can trigger a positive feedback loop and the reaction races out of control. This is known as an exploding lake and it’s very dangerous. Apart from the obvious dangers of such an energetic process and flooding etc there is the problem of the gas release itself. Cold CO2 is heavier than air and will carpet an area of nearly 5000km² and in the process suffocate and poison two million people.

But what’s the risk I hear, how often does this happen?

It’s not an easily analysable problem, the boundary conditions between stable and unstable are not clear and we don’t even know the maximum amount such a body of water can dissolve before release and we’re not that sure how much gas there is in there. We don’t know the magnitude of disturbances the system can take and can’t define what would trigger an runaway release. Given that the lake is surrounded by fault zones and active volcanoes though, it’s safe to say it might be sooner than we think. It’s happened before as well. Lake Nyos in Cameroon erupted in 1986 and killed over 1700 people in the surrounding area (read more here at Lake Nyos wiki. Lake Kivu is surrounded by millions of people and some of the largest cities in the area are within reach of the projected CO2 blanket. The area is too poor and war-torn to evacuate people before any such event and after the event people would have only minutes to react.

Basically they’d be screwed if it erupted.

So it’s very heart warming to read about a great engineering project to reduce the pressure in the lake and to use some of the methane released to generate power for the area. The BBC has an article about it here. The idea is to place risers deep into the lake, allow the gas pressure to drive water & gas upwards, separate out the methane for burning and either release or re-dissolve the CO2. It’s the sort of engineering project anyone would be proud to work on but in common with my post on Lake Vostok (and rather more realistic) the risk planning on this had better be excellent. This project has a real risk of killing people. It might not be a highly probable risk but the impact is both frighteningly large and extremely rapid. I suppose the balance between “it will explode one day” and “it might explode today” is something which sounds difficult but its neither intractable or impossible to mitigate. Just so long as the risks are tolerated or transferred the project sounds like an excellent example of what engineering can do to improve people’s lives and make the world just a bit safer.

I'm just glad I'm not on the risk committee that’s all I'm saying.