So, there’s been another explosion on an oil rig in the Gulf of Mexico. This got me thinking about risk and risk management.
In the engineering sense, “risk” refers to the chance that a particular system will fail and how heavily we weight the consequences of such failures. Risk is present in any design, any system, any process. There’s no way anyone can drive risk to zero, because nobody has perfect knowledge of any system and nobody can predict the future with 100% accuracy. The question is how unlikely and how inconsequential a failure must be to represent an acceptable risk. A complimentary question is how well we plan to deal with those failures when they happen.
BP undoubtedly performed some sort of risk analysis on the Deepwater Horizon platform before it began operations. Engineers must have, at some level, looked at the drilling hardware and procedures and decided that the chance of a catastrophic failure was such-and-such percent. They must also have looked at the cost of dealing with those failures, and come up with so many billion dollars. But all this gets weighed against the potential benefits: if the Deepwater Horizon platform brought in revenue of only a thousand bucks a year, but had an chance of failure of 50%, and the cost to the company of that failure is $20 billion, then BP probably would not have set up the platform the way they did. But if the calculation came out with a one-in-a-million chance of failure, a $20 cost of failure, but revenue of $50 billion per year, then of course they’d go ahead with the project.
The failure of the actual Deepwater Horizon system could mean any one of several things. It could mean that all BP’s risk estimates were correct, and they just got supremely unlucky with that one-in-a-million chance: unlikely, but possible. It could also mean that their analysts made some error: they may have put the chance of failure too low, or the consequence of failure too low, or the potential benefit of success too high. The real trouble with this sort of thinking is that we can’t know for sure where the analysis went wrong, if it did.
However, when we look at BP’s horrendous safety record, the facts that came out about how blasé other oil companies were about drilling, safety, and cleanup in the Gulf, and this second explosion on a platform owned by another company with a dubious safety record (at least, so I heard on NPR), I tend to think there was a problem with the risk analysis. These companies are engaging in higher-risk behaviors in order to get higher payouts. In short: they are getting too greedy. This might not be a problem in some industries, but here, the cost of failure isn’t just borne by the risk-taking companies, but also by the residents of Gulf Coast states (along with the rest of us taxpayers). I hope that these incidents cause the companies in question to revise their risk analyses to be more conservative, especially now that there is wider recognition in our society of the costs of such risky behavior to the wider economy, environment, and climate.
Now, there are good reasons to pursue more high-risk activities, if the potential benefit is high. For instance, there’s my favorite kind of engineering: spacecraft engineering! I would love for NASA to take much greater risks than it currently does!
Current NASA policy, for instance, dictates that any mission should present zero risk to the safety of astronauts on board the Space Station. This policy, which appeared after the Shuttle Columbia broke up on reentry, makes little sense. Remember what I said before about zero risk? It does not and cannot exist. Yet, that’s NASA policy – and the policy has caused NASA to nix some pretty exciting missions for posing, for example, a one in 108 chance of collision with ISS. The chance of Station astronauts getting fried by solar flare radiation or baking when ISS refrigeration units fail or losing their air from a micrometeoroid are likely to be much higher than 1 in 100 million – so what’s the problem? These missions don’t add any danger compared to the dangers that already exist.
Besides, we’re taking about spaceflight. It’s not safe. I mean, we’ve made it pretty safe, but still – it involves strapping people on top of tons of high explosives, pushing them through the atmosphere at hypersonic speeds, jolting them around repeatedly as rocket stages separate and fire, and then keeping them alive in a vacuum for days, weeks, or months at a time. Honestly, it’s astonishing that we managed to pull off six Moon landings with only a single failed attempt – and a nonfatal one at that!
I would argue that those tremendous successes in the early space program came from high-risk activities. For the first American manned flight into orbit, NASA put John Glenn on top of a rocket that exploded on three out of its five previous launches. The Gemini Program pioneered the technologies and techniques necessary for a lunar landing (and that we now take for granted in Space Shuttle activities) by trying them out in space to see what happened – that program nearly cost Neil Armstrong and David Scott their lives on Gemini 8. The Apollo 8 mission, which was supposed to orbit the Earth, was upgraded to a lunar swingby – the first time humans visited another planetary body – mere months before launch. But these days, to hear NASA brass and Congressional committee members tell it, no such risks are acceptable. NASA must use “proven technologies.” NASA must accept no more than bruises on its astronauts when they return from missions. NASA must not chance any money, material, or manpower on a mission that might not succeed, even if such success could give us the next great leap forward. And so we end up with manned “exploration” of only low Earth orbit for thirty years, an Apollo reimagining to succeed the Space Shuttle, and, if the House has its way with President Obama’s proposed NASA budget, a space program dedicated to building The Same Big Dumb Rockets That It Already Built for the forseeable future.
Fortunately, we still get to see some envelope-pushing on the robotic exploration side of things. Missions to Mars have only recently broken through to a cumulative success rate greater than 50%, thanks to a string of high-profile successes, and that’s partly because of the ambition involved in landing something on another planet. It’s wonderful to see the progression from the Sojourner to Spirit and Opportunity to Curiosity rovers – but remember that the Beagle rover, Mars Polar Lander, and Mars Climate Orbiter all crashed into the Red Planet. These failures cost money and effort, and perhaps a direction of research in a few academic careers, but not lives, which makes them more acceptable to bear back on Earth. Even if the risk is high, the cost of failures is acceptable compared to the benefits.
Still, there could be more room for audacity (is audacity = 1/risk?) in robotic space exploration. Take the MER mission, for example: a pair of vehicles designed to last for 90 days have been operating for over six years – and counting. In one sense, this is a great success. But in another, it shows that spacecraft engineers are far, far too conservative in their designs. Imagine if they had actually designed the MER rovers to run for 90 days: everyone would have been happy with the mission, and the rovers would have cost less and taken less development time to the tune of something like the ratio between ~2200 and 90 sols. Or, conversely, consider if NASA had been ambitious enough to design a five-year rover mission from the start. That might have seemed laughable when the MERs were launched, but now we know that duration to be well within our capabilities. Because, in fact, we design space missions that rarely stretch those capabilities, since we do not tolerate risk.
This risk aversion in spacecraft engineering is one reason why I (and so many other people) are excited to see companies like SpaceX and Scaled Composites – which aim to turn a profit, something NASA doesn’t have to do! – doing the things they are doing. SpaceX, especially, which had to launch its Falcon 1 rocket several times before it succeeded, but used that experience to pull off a big Falcon 9 launch and secure the largest commercial launch contract ever. It’s also one of the reasons I was so excited about President Obama’s plan for NASA: it looked like NASA would be sticking its neck out for unproven technologies again.
How is it that we as a society tolerate tremendous risk when it comes to activities that affect thousands or even millions of lives on Earth, but we balk at the slightest chance of failure when considering space travel? It’s a puzzle to me.