What went wrong with the Deepwater Horizon blowout

Bruce Thompson
The forensic report on the Deepwater Horizon's blowout preventer (BOP) was released to the public on March 23, 2011. The next step in the investigation will be to try and reconcile it with the report of the Chief Counsel of the Oil Spill Commission . There are two major public policy issues: worker safety and spill prevention; we don't want either dead workers or oil spills.

While the process goes forward and the issues are debated, a preliminary finding has emerged. The crew of the Deepwater Horizon in the drill shack left the activation of the BOP so late that they were dead before it finished closing. Figuring out why they did not act earlier is a matter of debate; the fact that they acted too late is not.

Based on the testimony of the surviving witnesses, the chief counsel states, "Sometime between 9:40 and 9:43 p.m., mud overflowed onto the rig floor, shot up to the top of the derrick, and poured down onto the main deck." (pg 181). Now the flow path at that initial moment is not clearly defined in the report. It could have been straight up out of the rotary table in a classic gusher, or it might have been diverted to the mud gas separator located high on the deck and overflowed there. I am leaning toward the notion that the crew activated the diverter at 9:41 p.m. as subsequent events demonstrate that the diverter was activated at some point. But the key piece of new evidence is that the upper variable bore ram (VBR) was found closed and there is no provision for activating it via the Remotely Operated Vehicles (ROV) on the sea floor. So it must have been activated before the connection between the drill shack and the BOP was lost in the explosions. The chief counsel continues:

Although Anderson had activated the annular preventer, that action had not fully shut in the well. Instead of reaching the expected shut-in pressure (approximately 6,000 pounds per square inch [psi]), drill pipe pressure plateaued at about 1,200 psi. In response, the drill crew either tightened the annular to create a seal or activated a variable bore ram. At 9:47 p.m., drill pipe pressure increased dramatically. At this point, the well may have been shut in.

From the forensic report, we now know that it was the variable bore ram that was activated at 9:47 p.m. The pressure on the drill pipe, which was sensing the pressure 8,367 feet below the deck, started a steep climb. By 9:49, it had reached 5,730 psi when the second explosion cut off electrical power to the rig and the data stream abruptly ends. It is useful to note that when the well was finally shut in on July 15, the initial shut-in pressure at the BOP was 6,700 psi despite nearly three months of depletion of the oil reservoir, so the chief counsel‘s estimate of a shut-in pressure of 6,000 psi is clearly too low.

The crew of the Deepwater Horizon tried to use the blowout preventer as a blowout suppressor and it was not up to the task. If the well was initially gushing out of the wellhead 200 feet into the air, the velocity would need to have been 113 feet per second (77 miles per hour). Based on a net cross-sectional area of the 18.5" diameter riser minus a 7" diameter drill pipe inside it, that would put the flow at 1,921 barrels per minute. Compare that to the estimated flow of oil to create the oil spill of a measly 62,000 barrels per day. Having lost the controlling back pressure at the rig, the gases in the riser were allowed to expand. Yet the methane detectors had not gone off. This may be the explanation for why the workers failed to realize why they were in so much trouble until it was too late. The initial burst of gas was not methane as one would normally expect, rather it was nitrogen that had been injected into the cement used to create the primary cement job, phenomenon known a nitrogen breakout. While the volume of nitrogen at the 14,000 psi pressure of the bottom of the well was measured in mere barrels, at atmospheric pressure its volume would be measured in thousands of barrels. Once those thousands of barrels were depleted however, the rig would suddenly be overwhelmed by explosive methane gas from the well. The duty officer on the bridge reported that she got a slew of simultaneous gas alarms as the rig exploded, as one would expect under this scenario. Had the crew closed the variable bore ram at 9:41 p.m., the scope of the accident would have been much smaller and the blowout preventer could have prevented the worst of this incident. The crew would likely be alive and the rig afloat.

So my recommendation for any future deepwater drilling is to never use nitrified cement as a pressure barrier at these great depths. Blowout presenters do not work as blowout suppressors. There will be a formal hearing of the Joint Investigation Team on April 4th.
The forensic report on the Deepwater Horizon's blowout preventer (BOP) was released to the public on March 23, 2011. The next step in the investigation will be to try and reconcile it with the report of the Chief Counsel of the Oil Spill Commission . There are two major public policy issues: worker safety and spill prevention; we don't want either dead workers or oil spills.

While the process goes forward and the issues are debated, a preliminary finding has emerged. The crew of the Deepwater Horizon in the drill shack left the activation of the BOP so late that they were dead before it finished closing. Figuring out why they did not act earlier is a matter of debate; the fact that they acted too late is not.

Based on the testimony of the surviving witnesses, the chief counsel states, "Sometime between 9:40 and 9:43 p.m., mud overflowed onto the rig floor, shot up to the top of the derrick, and poured down onto the main deck." (pg 181). Now the flow path at that initial moment is not clearly defined in the report. It could have been straight up out of the rotary table in a classic gusher, or it might have been diverted to the mud gas separator located high on the deck and overflowed there. I am leaning toward the notion that the crew activated the diverter at 9:41 p.m. as subsequent events demonstrate that the diverter was activated at some point. But the key piece of new evidence is that the upper variable bore ram (VBR) was found closed and there is no provision for activating it via the Remotely Operated Vehicles (ROV) on the sea floor. So it must have been activated before the connection between the drill shack and the BOP was lost in the explosions. The chief counsel continues:

Although Anderson had activated the annular preventer, that action had not fully shut in the well. Instead of reaching the expected shut-in pressure (approximately 6,000 pounds per square inch [psi]), drill pipe pressure plateaued at about 1,200 psi. In response, the drill crew either tightened the annular to create a seal or activated a variable bore ram. At 9:47 p.m., drill pipe pressure increased dramatically. At this point, the well may have been shut in.

From the forensic report, we now know that it was the variable bore ram that was activated at 9:47 p.m. The pressure on the drill pipe, which was sensing the pressure 8,367 feet below the deck, started a steep climb. By 9:49, it had reached 5,730 psi when the second explosion cut off electrical power to the rig and the data stream abruptly ends. It is useful to note that when the well was finally shut in on July 15, the initial shut-in pressure at the BOP was 6,700 psi despite nearly three months of depletion of the oil reservoir, so the chief counsel‘s estimate of a shut-in pressure of 6,000 psi is clearly too low.

The crew of the Deepwater Horizon tried to use the blowout preventer as a blowout suppressor and it was not up to the task. If the well was initially gushing out of the wellhead 200 feet into the air, the velocity would need to have been 113 feet per second (77 miles per hour). Based on a net cross-sectional area of the 18.5" diameter riser minus a 7" diameter drill pipe inside it, that would put the flow at 1,921 barrels per minute. Compare that to the estimated flow of oil to create the oil spill of a measly 62,000 barrels per day. Having lost the controlling back pressure at the rig, the gases in the riser were allowed to expand. Yet the methane detectors had not gone off. This may be the explanation for why the workers failed to realize why they were in so much trouble until it was too late. The initial burst of gas was not methane as one would normally expect, rather it was nitrogen that had been injected into the cement used to create the primary cement job, phenomenon known a nitrogen breakout. While the volume of nitrogen at the 14,000 psi pressure of the bottom of the well was measured in mere barrels, at atmospheric pressure its volume would be measured in thousands of barrels. Once those thousands of barrels were depleted however, the rig would suddenly be overwhelmed by explosive methane gas from the well. The duty officer on the bridge reported that she got a slew of simultaneous gas alarms as the rig exploded, as one would expect under this scenario. Had the crew closed the variable bore ram at 9:41 p.m., the scope of the accident would have been much smaller and the blowout preventer could have prevented the worst of this incident. The crew would likely be alive and the rig afloat.

So my recommendation for any future deepwater drilling is to never use nitrified cement as a pressure barrier at these great depths. Blowout presenters do not work as blowout suppressors. There will be a formal hearing of the Joint Investigation Team on April 4th.