An unhappy anniversary, this: Eleven people were killed in an explosion aboard the Deepwater Horizon five years ago in the Gulf of Mexico. That blowout immolated the huge offshore drilling rig, which sank two days later on the 40th anniversary of Earth Day, leading to the worst oil spill in U.S. history.
In some ways it was a prototype disaster of the future, one involving complex technology that failed in unexpected ways. This will happen again — count on it. Maybe it’ll be another oil spill, but more likely it’ll be something else. We’re unprepared for catastrophes that break the patterns of previous disasters, that pose exotic challenges and involve technological systems that are not easily repaired.
The BP oil spill five years ago was horrifying for many reasons, starting with the loss of the 11 lives and the polluting of the Gulf. But it also was disturbing in the way it exposed our collective inability to fix a seemingly simple technological problem: a hole at the bottom of the sea. It took nearly three months to plug the Macondo well. The spillcam showed the oil gushing into the Gulf, but no one seemed able to put an end to it. That kind of powerlessness is unsettling to those of us who like to think that experts can fix anything.
One example of the next Big One would be a massive grid outage as a result of a powerful solar flare. Our technological systems, on land and in space and in the cloud, are increasingly complex, and are vulnerable to accidents, sabotage, unanticipated design flaws, network outages, corrosion, bad luck, X-factors, hiccups and the shenanigans of gremlins.
You know the people who run the government are liberal arts majors, right? Big technological disaster hits, the president looks around and says: Is there anyone here who understands this stuff? The default move is to send in the military, the repository of technical competence in the U.S. government, but guess what, even the military doesn’t know anything about deepwater drilling.
President Obama tapped Steve Chu to fix the oil spill, as if he’d go in there and plug the well with his Nobel Prize in physics. My book gives sufficient credit to Chu and others in the government for doing their best to help solve the problem, but ultimately this was a disaster created by the oil and gas industry and it would have to be solved by the oil and gas industry. Specifically, BP. The company discovered that it wasn’t prepared for a deepwater blowout. It hadn’t ever happened, anywhere. The deep is different. A mile down, all rules change. You can’t get there except robotically and strange forces of chemistry and physics are working against you.
BP ended up plugging the well after 87 days, using a piece of hardware that was sitting on a dock all along in southern Louisiana — the 3-ram capping stack. This was after multiple failed efforts (blind shear rams that didn’t work, containment domes that floated away, a “top kill” with mud that was a huge disappointment) and much consternation about what exactly was going on in the blown-out well.
Only much later did we learn a key feature of the accident: When gas surged up the well after an inadequate cement job, the violent kick bent the drillpipe that had been threaded through the blowout preventer. Thus when the blind shear rams closed, to cut the pipe, they couldn’t get a clean bite on the drillpipe and it remained open, allowing the gas to reach the rig. You don’t have to understand the mechanics of this to grasp the central concept that there was a single-point failure lurking in the drilling protocol. The initial loss of well control — when gas from the deep reservoir infiltrated the supposedly cemented well — quickly disabled the machinery designed to react to just that kind of a gas kick. The backup plan wasn’t truly a backup: It was in the line of fire. For nearly two weeks, BP’s engineers thought they could use ROVs (robotic submarines) to make the blind shear rams close in the well, but they didn’t know that those cutting blades had already tried and failed and were never going to succeed. And the oil continued to gush.
“I think another major spill is actually more likely to happen as we go farther offshore drilling deeper, higher pressure wells,” Amos tells me.
There have been some regulatory changes designed to make blowout preventers more robust, and to have capping-stack hardware in place that’s ready to be deployed in case of a deepwater blowout, Amos says. But every blowout has novel features; what would happen, for example, if a rig like the Deepwater Horizon were to sink directly on top of the well, rather than (as was the case April 22, 2010) 1,500 feet away? You couldn’t get to the wellhead to cap it.
[T]here are fast-moving catastrophes like the BP spill that are tailor-made to get our attention: sudden, spectacular and lethal. And then there are the slow-moving disasters, like the chronic “day to day” pollution that accompanies offshore drilling, that are all too easy to ignore. This chronic pollution threat to economies based on tourism and fishing should be as big a concern for coastal communities (Virginia Beach, Nags Head, Ocracoke) as the rare catastrophic spill. But it’s very difficult to get a handle on the severity of this chronic problem. We are still in the ridiculous position of relying on pollution reports that are generated by the polluters themselves, even though those reports are demonstrably incomplete and systematically inaccurate.
A few more lessons learned, directly from the last chapter of my book on the spill:
When doing something risky, remember that risk builds like plaque.
Make sure that your backup plan is really in back and won’t get blown up out front along with your plan A.
Remember that low-probability, high-consequence events become more likely given enough time and opportunity. This is why you tell your teenager that you don’t want her running around the party district at 2 in the morning. Sure, she’ll probably stay out of trouble tonight, but it’s her entire adolescence that you have to worry about.
Measure your misery. Don’t shy away from knowing precisely how badly you’re screwed.
Keep your wits about you. It is extraordinarily unlikely that the disaster you are dealing with is qualitatively worse than the many calamities that humans have survived to this point.
[From Science News, here’s a report on long-term ecological consequences of the spill (sperm whales are going nowhere near the wellhead — they’re not stupid).]
[Here’s a story I did on complex technologies failing in complex ways.]