National Oil Spill Commission investigators have found that the Halliburton cement used to seal the bottom of BP’s wild Gulf well in April was unstable and was used despite multiple failed tests in the weeks leading up to the massive well blowout.
What’s more, the commission investigators found Halliburton knew about the problems and used the cement mixture anyway.
The finding from commission chief counsel Fred Bartlit Jr. and his investigative team could be among the most significant to date as several investigations try to establish clear causes for the disaster, which killed 11 rig workers and fouled the Gulf with nearly 5 million barrels of oil.
Rep. Ed Markey, D-Mass., who has been seeking subpoena power for the Oil Spill Commission, saw it as a watershed finding.
“The fact that BP and Halliburton knew this cement job could fail only solidifies their liability and responsibility for this disaster,” Markey said. “This is like building a car when you know the brakes could fail, but you sell the cars anyway.”
Still, the cement is just one of several possible failure points that may not have caused the blowout in and of themselves, but appear to have worked in combination to doom the Deepwater Horizon oil rig April 20.
Other questions have emerged from other investigations of the incident, and those remain on the table: about BP’s decisions to use potentially riskier designs for lining the well and to skip important steps in sealing the well closed, as well as BP’s alleged misinterpretation of the results of a final test of pressure in the well hole. The Oil Spill Commission may present more conclusions about those issues at a presentation scheduled for Nov. 8 in Washington.
But in a letter to commissioners Thursday, Bartlit focused only on the question of the cement that was supposed to have sealed the well’s metal linigs to the drilled-out bedrock. Bartlit wrote that the foam cement’s instability “may have contributed to the blowout.”
A commission staffer, who was authorized to speak for the commission in a news briefing but was not permitted to give his name, went a little further.
“Had the cement done its job, the hydrocarbons would have been isolated and there should not have been a blowout,” he said.
It is already generally accepted that gas flowed into the well through a weak barrier and then pushed up the miles-long well, blew up through an undersea riser pipe and ignited the rig, which was floating a mile above the sea floor and 50 miles off the Louisiana coast.
Halliburton had a contract with BP to supply the Deepwater Horizon drilling rig with the nitrogen-infused cement it would use April 19-20 to seal the sides of the well. Halliburton has said publicly that the cement passed stability tests before it was used.
But the Oil Spill Commission investigators, through a review of previously undisclosed documents and interviews, have determined that the first three tests Halliburton ran on the foam cement were all failures.
Also, despite not having the subpoena powers used by other investigative bodies, Bartlit managed to convince Halliburton to turn over the exact recipe for the cement mixture, something it hadn’t been willing to share previously.
Oil Spill Commission co-chairman William Reilly said Bartlit, whom he called “John Wayne in pinstripes,” was able to get information without subpoenas by reminding the companies that anything they didn’t provide willingly would be flagged for the Justice Department’s sprawling civil and criminal investigation.
With the cement ingredients in hand, Bartlit then had Chevron perform nine independent lab tests on the mixture under various conditions. Donating their services, the Chevron scientists found the slurry was unstable in all nine tests.
Bartlit’s letter to commissioners Thursday indicates that Halliburton only provided BP with the data from one of the three tests for which it had results before the cement was actually poured in the well. The one set of results Halliburton shared with BP was from February, two months before the accident, when the final conditions in the well and even the size of the hole were not known.
The data from that test showed the cement was unstable, but when Halliburton sent the results to BP by e-mail March 8, it only sent the numbers, no analysis, and there was no indication that Halliburton mentioned it was a failed test, Bartlit wrote in his Thursday letter.
Halliburton kept testing the cement and changing different variables, and finally ran a successful test in the final days before the blowout, the commission found. But Bartlit wrote that Halliburton may not have gotten the final results of that test before the cement was actually poured into the well April 19 and definitely did not share them with BP before the actual cement job was done.
Around midday Thursday, Halliburton spokeswoman Cathy Mann said the company was reviewing Bartlit’s letter and Chevron’s findings and would “publish a response later today,” but Halliburton had issued no statement as of 10 p.m.
In its Sept. 8 internal investigation report, BP said it was unable to check Halliburton’s tests because a court order barred anyone from accessing the remaining 1.5 gallons of cement slurry actually used on the rig and Halliburton refused to provide BP with its recipe.
The BP report said the test results it finally got from Halliburton showed it used a cement with a lower foam level than what was required for the well. And like the Oil Spill Commission did with Chevron, when BP tried to replicate the Halliburton cement mixture under the pressure conditions at the bottom of the well, it was unstable.
The March 8 e-mail from Halliburton to BP was eerily similar to a separate report by the cementing contractor just two days before the accident that also included alarming information that wasn’t noticed. That report contained Halliburton models warning of a severe risk of gas flowing into the well if BP didn’t use more stabilizing equipment in the well bore. But the warning was placed deep inside the report, and key BP engineers have testified that they never bothered to read the report until it was too late.