The entry tells the story of the negative pressure test on the Deepwater Horizon on 20 April 2010. This piece goes behind it twice: first through the eight days in which BP’s engineers wrote, unwrote and rewrote the procedure that reached the rig as two words, and then through the three hours of the test itself, gauge reading by gauge reading, as the Chief Counsel of the President’s commission, the BOEMRE panel, BP’s own investigators and the federal court each set them down. Where their numbers differ, the differences are given. The sources are those of the entry; the words of the witnesses come at second hand from the reports that quote them, which name the hearing dates.

Eight days of procedure

A negative pressure test underbalances a well on purpose and watches whether the reservoir comes in. In April 2010 the Minerals Management Service did not require one, and neither BP nor Transocean had a standard procedure for it. BP required the test in some conditions, but one of its people admitted the tests “could be different on every single rig depending on what the [well] team agreed to.” Transocean required it too, had no set procedure, and did not describe it in its Well Control Handbook; its crews learned the test through general work experience. The crew of the Marianas, which had begun the well, had done its negative test at Macondo in a different way from the way the Deepwater Horizon’s crew did theirs. One independent expert told the Chief Counsel’s team he had had to consult an academic text to find a written description of the procedure at all.

On 12 April BP’s drilling engineer Brian Morel circulated a temporary abandonment plan with no negative pressure test in it. It set the lockdown sleeve before the surface plug and put the plug in seawater at 6,000 feet below the sea. Ronald Sepulvado, the well site leader who was ashore on 20 April, had to tell Morel to add a test. On 14 April Morel’s “Forward Ops” email included one, but would set the plug in mud before the riser was displaced. On 16 April BP sent the MMS an application to modify its permit with a procedure different from both, which said the test would be done on the kill line. The Chief Counsel found no evidence that any of these changes went through a formal risk assessment or a management-of-change process.

On 18 April, in an email headed “Negative Test”, Morel asked the wells team leader, John Guide, whether to keep a first test with base oil on the bottom plug. “Are you ok with this, or do you think we should remove the first base oil test and just use the displacement as a negative test (shut down at the end and do a flow test)?… I have got different opinions from everyone on the team. The way we currently have it set up is the standard we have been using, but this one is slightly different because the plug is so deep and base oil doesn’t achieve the full negative load the wellbore will see. Don [Vidrine] and Bob [Kaluza] don’t seem to have strong opinions either way.” Guide answered: “I would use the seawater displacement as the negative test, as you stated, shut down at the end and do a flow test.” Twenty minutes later Morel wrote: “[d]one.” The BOEMRE panel observed that neither man told the rig crew that the change raised the risk of a well-control event, and that it “converted it from the previously-approved multi-step process.” Asked afterwards why the procedure had changed from the one the MMS approved on 16 April, Kaluza said “maybe [Guide and Morel were] trying to save time.”

At 10:43 on 20 April, about to board the helicopter, Morel emailed an “Ops Note” with the day’s procedure: a positive pressure test; drill pipe run to 8,367 feet; 3,300 feet of mud displaced with seawater and lifted above the blowout preventer; a negative pressure test; the riser displaced; the surface plug set at 8,367 feet; the lockdown sleeve. It was the first the two well site leaders had seen of it. The crew heard it at the 11:00 pre-tour meeting. Kaluza’s briefing left the test out, and Transocean’s offshore installation manager, Jimmy Harrell, put it back: “[t]he first plan I seen [sic] didn’t have a negative test in it,” he testified on 27 May. “So I told him [Vidrine] it was my policy to do a negative test before displacing with seawater.”

Harrell also described how the rig did the test: “[y]ou do it by leading [bleeding] off back to Halliburton and up your drill pipe. You pump seawater to the end of your tail pipe and all the way back up to above your annular with your spacer…You hold the mud in the riser with the annular closed…You have seawater in the drill pipe and you have seawater in the kill line and either one would be seeing the same pressure…”

That morning the M-I SWACO mud engineer, Leo Lindner, was asked twice how the rig did it. “He [Kaluza] wanted to go over the method by which the rig had been doing its negative test and displacing. I explained – I explained it to him. He seemed satisfied with it. Shortly after that, I was called by Mr. Brian Morel. We had basically the same conversation. He seemed satisfied with it. He informed me they were going to be displacing further down the hole than usual. Usually it’s 300 feet below the mud line, but this was going to be at 8,367 feet [3,367 below the mud line]. I left the office and I make my calculations. I type up a displacement procedure.” Lindner’s sheet was the most detailed written procedure anyone on the rig had. It gave the volumes of spacer and seawater to pump. Its second step was to “[d]isplace choke, kill, and boost lines and close lower valves after each,” and it said nothing about reopening them, which a test on either line would need. Its instruction for the test was “[c]onduct negative test.” It said nothing of expected pressures, expected bleed volumes, how to read the result or what to do about an odd one. It ended: “[g]ood communication will be necessary to accomplish a successful displacement. If you are not sure, stop and ask.” The Chief Counsel’s report called it inexcusable that the most detailed procedure for the test had been written by a mud engineer in the course of specifying fluid volumes.

Asked afterwards what a successful test was, Guide said “[a] successful test needs to be run for 30 minutes with no flow from the well”; Morel said “[n]o flow for 30 minutes”; Vidrine said “[c]heck[ing] for flow or no-flow.” Neither Kaluza nor Vidrine calculated expected pressures or volumes beforehand, which other BP well site leaders routinely did. Kaluza said, “[w]e didn’t talk about what if the negative test fails.” Mark Hafle in Houston, who spoke to Kaluza that afternoon, “had [the] impression that Kaluza wasn’t really clear on neg[ative pressure] test procedure,” and the call ended without resolving it. BP’s engineering managers said after the blowout that the Transocean crew knew how to run a negative test and that the limited instructions should have been enough. Transocean argued that its crew were tradesmen, not engineers, and could not have been expected to interpret the results; the Chief Counsel answered that a negative test is an invitation to a kick under controlled conditions, and that a crew trained to see a kick while drilling ought to see the same signs while testing.

Three hours by the gauge

The readings below follow the Chief Counsel’s report, which the court adopted in substance, with the other accounts’ figures where they differ. All times are the rig’s, on the evening of 20 April.

About 3:00: the pre-job meeting, led by Lindner. Kaluza attends and leaves for his office. Shortly after, the driller, Dewey Revette, displaces the boost, choke and kill lines with seawater.

Just before 4:00: 454 barrels of spacer go down the drill pipe, taking about thirty minutes, then about 352 barrels of seawater, taking about twenty-five. BP’s investigators calculated afterwards that this would have left the tail of the spacer twelve feet above the preventer; Transocean and an independent expert calculated that it did not reach the preventer at all; the Chief Counsel thought it likely that some of the sixteen-pound spacer had fallen back through the eight-and-a-half-pound seawater on the way up.

4:53: the upper annular preventer closed around the drill pipe. Drill pipe pressure 2,325 pounds per square inch (the BOEMRE panel: 2,339). The panel’s calculation of what it should have been: 1,610. The excess, about 700 pounds, was consistent with spacer below the preventer. Nobody had calculated the expectation and nobody investigated the excess.

4:55: the crew bleed the drill pipe to 1,250 to match the kill line, which read 1,250, then open the kill line valve at the stack. The two should have stayed equal. The drill pipe pressure jumped and the kill line pressure dropped. (The panel’s reading of the data: the drill pipe bled from 2,324 to 1,427; with the kill line opened, the kill line fell to zero and the drill pipe stood at 458.)

4:58 to 5:00: the drill pipe opened at the rig and bled towards zero. It would not go below 260 (the Commission’s report: 266). Three witnesses: twenty-three to twenty-five barrels returned; the Chief Counsel’s team could not say how much, and no one had worked out beforehand how much to expect. The pipe shut in; pressure rose to 1,262. The panel: shut in at about 5:05, the drill pipe pressure rising at once, “evidence of an unsuccessful negative test,” and the choke line reading below zero, which it took as an instrument fault or an inaccurate test. One independent expert told the Chief Counsel that this alone was a failed test.

About 5:10: the fluid level in the riser falling; spacer leaking down past the preventer. Men looking down the rotary with a torch. The drill shack filling for the 6:00 shift change and the executives’ tour. Harrell has the preventer’s closing pressure raised; Wyman Wheeler tops the riser up with twenty to twenty-five barrels of mud; the level holds. The panel put the end of the first test at 5:25, and the first “long” discussion of the drill pipe pressure between 5:17 and 5:27, with Harrell, Kaluza, Vidrine and the drill crew.

5:26: the drill pipe bled again, to zero. Fifteen barrels returned; three to five expected (BP’s investigators: about 3.5). Shut in; pressure rose to 773 and was still rising when the crew began bleeding it off.

5:53: on Kaluza’s instruction, by one account (“Let’s open the kill line and see what happens”), the well bled down through the kill line. Drill pipe pressure fell almost to zero; the kill line went on flowing and spurting water until the crew closed its upper valve on the rig. Witnesses: three to fifteen barrels. Kaluza leaves to find Vidrine. (The panel: around 5:30 the crew set up again, bleeding the drill pipe and closing an internal valve in it for twenty minutes, during which the drill pipe pressure could not be read; around 6:00 the drill pipe bled to the cementing unit and shut in there.)

6:00 to 6:35: with everything shut, drill pipe pressure rises to 1,400 (BP: from about 50 pounds; the court: “steadily rose to 1,400 psi., where it remained until the negative pressure test concluded around 7:55 p.m.”). The panel: the kill line pressure rose at the same time, but only to 140, “another indicator that the negative test was not successful.” The Chief Counsel: “This was the clearest indication yet that the well lacked integrity. The 1,400 psi pressure buildup can only have been caused by hydrocarbons leaking into the well from the reservoir formation.” One expert: a “conclusive failure.” Later analysis: 1,400 pounds is about what the reservoir’s pressure would show at the surface, through the seawater column, with no cement at the bottom of the well.

About 6:00 to 7:00: the discussion in the drill shack, and the bladder effect; the entry gives it.

About 6:40 to 6:45: the crew pump a small volume into the kill line to be sure it is full, and plumb it to the mini trip tank by the shack.

7:07 (the court) or about 7:15 (the Chief Counsel and the panel): the kill line opened at the surface and bled to zero. Less than a barrel came out (BP: 0.2 barrels). Then watched: thirty minutes by the Chief Counsel’s, BP’s and the court’s accounts, about forty by the panel’s. No flow, no pressure. Drill pipe: 1,400 throughout.

About 7:55: Vidrine declares the test a success; nobody disagrees. The crew make ready to displace the riser.

Why the kill line read zero was never settled. The Chief Counsel offered four candidates: spacer that had leaked below the preventer and migrated into the three-inch line and clogged it; mud never displaced from the line; a hydrate, the ice-like solid that gas and cold water form at the seabed; or a valve left shut by mistake. BP’s investigators named the first and the last. The court found the zero “obviously a false reading”: “Something had blocked the kill line such that the pressure that existed in the well and registered on the drill pipe did not affect any fluid in the kill line nor did it register on the kill line’s pressure sensor.” If there had been a clear path, the Chief Counsel wrote, the crew would have seen 1,400 on the kill line too, and might have recognised the second test for what it was. The court went further: had the test stayed on the drill pipe, “the test would not have been misinterpreted.”

The Chief Counsel’s summary of the three hours stands for all four accounts: “There were three instances in which pressure built up after being bled off, including the buildup experts have deemed a ‘conclusive failure’ wherein pressure inside the drill pipe rose from 0 to 1,400 psi. On at least one occasion, bleed-off procedures produced more flow than should have been expected. And while the rig crew observed no flow from the kill line during the second negative pressure test, the drill pipe pressure remained at 1,400 psi.”

Sources

• National Commission, Chief Counsel’s Report, chapter 4.6, and its notes, which cite the hearing testimony of Harrell (27 May 2010), Ezell (28 May), Pleasant (28 May), Lambert (20 July), Winslow (24 August) and the independent expert John Smith (23 July and 9 November). govinfo.gov
• BOEMRE, Report Regarding the Causes of the April 20, 2010 Macondo Well Blowout, section VII, with the Morel and Guide emails of 18 April, Harrell’s and Lindner’s testimony and Lindner’s procedure. bsee.gov
• BP, Deepwater Horizon Accident Investigation Report, key finding 3 and its chronology. bp.com
• Findings of Fact and Conclusions of Law, Phase One, paragraphs 241 to 250. uscourts.gov