The Last Days of Deepwater Horizon

The Last Days of Deepwater Horizon

Maybe we won't get a good disaster movie out of this until the litigation is over, years from now, but it's not too soon to begin piecing together what went wrong. The Wall Street Journal is running a series whose third installment, today, begins, "On April 20, a small group of men aboard the Deepwater Horizon listened to the nearly complete well and didn't understand what it was telling them."

The brass were largely unavailable that day. The BP manager was on-shore for training, incommunicado. Two onsite Transocean managers were busy hosting visiting BP and Transocean executives who had arrived to applaud the rig's safety record. Everyone believed the troubles with this "nightmare well" were largely behind them. They prepared for what they thought would be the final test, after which the well would be temporarily plugged until BP was ready to tap and transport its contents.

The test, which is designed to tell operators whether the well's lining is gas-proof, can be performed in several ways. Often workers set a cement plug a given distance below the blowout preventer, remove part of the drilling mud above the plug, and replace it with lighter seawater. Because the heavy mud holds down any gas that has leaked into the well, typically only about 300 feet of mud is removed, leaving enough mud, below the blowout preventer but above the plug, to continuing serving as a secure cap. With the blowout preventer closed, so that the mud above it no longer is pressing down on the column below, engineers run a 30-minute test to see whether gas is escaping into the column between the preventer and the plug.

On the morning of April 20, BP engineers in Houston (one of whom took the Fifth this week rather than testify before a federal panel) issued orders to remove 3,000 feet of mud instead of 300 feet, an "unusual" procedure that perhaps was intended "to avoid damage to a key seal" -- but at this point, stories begin to diverge. The BP day-shift manager on the rig didn't know why the procedure was changed and reportedly guessed that its purpose might be to speed things up; like the BP engineer in Houston, he has now taken the Fifth. Although BP higher-ups had received authority from federal regulators for the changed procedure four days earlier, Transocean workers didn't learn about it until that morning. Transocean's most senior worker protested the change, then reluctantly bowed to the BP day-shift manager's authority before becoming distracted by the visiting executives for most of the rest of the day. By late afternoon, the 3,000-foot column of drilling mud had been removed and the pressure test had begun.

Pressure built immediately, indicating a problem. When the senior Transocean worker stopped by the drilling floor with his visiting executives, he sent them on to finish their tour and stayed behind, still not believing the problem was unduly serious. After the crew tightened down a blowout preventer valve, the senior worker and his second in command both went back to their babysitting duties, never to return to the drilling floor. The day-shift drilling crew supervisor continued to worry until the end of his shift at 6 p.m, but his night-shift replacement was not alarmed by the pressure readings that greeted him.

The senior Transocean worker, meanwhile, had temporarily shed his visiting executives and begun to confer with his night-shift replacement, who wanted to do another test. The new test's results were anomalous, showing pressure differences in areas that should have been equal, possibly suggesting a clog. At 8 p.m., however, the senior night-shift worker told his day-shift counterpart, who had remained on hand, that he was satisfied with the test results. Although anomalies persisted, including the fact that more fluid was flowing out of the well than was being pumped in, these were noted only by electronic "black box" data reviewed later by investigators. At about 9 p.m., some of the visiting executives went up to the bridge and watched the rig captain demonstrate a severe-weather training simulator.

At 9:50 p.m., the rig's off-duty second-in-command got a call from an assistant driller saying "we need your help." Alarms rang. Before the second-in-command could reach his hard hat, he felt the first explosion. The assistant driller and the night-shift toolpusher had been killed and the day-shift toolpusher badly injured. Everyone else was running for his life. The rig burned for two more days before sinking (the picture above is from April 22).

Here endeth the installment. I'm looking forward to the next one, because it's still not clear why the blowout preventer failed, what the pressure readings should have told the workers, and what role the changed test procedures played. It was widely reported yesterday that BP now claims Transocean made unauthorized changes to the blowout preventer. Earlier reports suggested that the well's casing should have had 22 of something called "centralizers" but had only six; apparently this would interfere in the bonding of the cement casing in the annulus between concentric columns, which could permit gas to intrude. A WSJ commenter asks:

What was the status of the instruments? What was the sampling process? I recall hearing stories of dead batteries on gauges and controls in the days immediately following the explosion. Ditto maintenance issues. What was the status of training / experience of those involved at all levels of the testing? To what extent did such things damage the ability to make the right decisions? . . . So, two interconnected pipes were measuring the same thing and the smaller one gave a different reading. Anyone would conclude, within 3 seconds, that either the smaller pipe was clogged or the instruments were not functioning properly. Then you deal with that. . . . The article suggests that, instead, much chin scratching and navel gazing then ensued. . . . There's something really fishy in that account - someone is hiding something.

Another commenter says:

[T]his is a bit like ordering a bowel retract moments before an appendectomy in which a critical care patient will be operated on by rotating surgical teams who will also be on hand to assist and guide visiting hospital administrators around the operating theatre and the entire surgical floor.

No comments: