Halliburton defends the cement it designed for BP in 2010
8th April 2013 · 0 Comments
By Susan Buchanan
At the Gulf spill trial last week, Halliburton staffer Jesse Gagliano said BP rejected several of his recommendations on cement for the Macondo well, putting safety at risk. He was cross-examined in the landmark trial that began Feb. 25 in U.S. District Court in New Orleans, where Judge Carl Barbier is assessing negligence in the April 20, 2010 rig explosion that snuffed out 11 lives. BP was the well operator and leaseholder, Transocean owned the Deepwater Horizon rig and Halliburton was BP’s cement contractor in the disaster that caused the nation’s biggest offshore spill.
Three years ago, Gagliano was embedded as a Halliburton technical adviser at BP in Houston, where he’d reported for almost five years. He said last week “my sole purpose was to provide cement recommendations and cement support for BP.” Halliburton was contracted to seal the Macondo drilling hole with cement to keep gas and oil from flowing into the well. Gagliano is now a Halliburton senior account representative after nearly 14 years with the company.
A week before Gagliano’s April 1 testimony, Transocean capital projects vice-president Bill Ambrose said at the trial that his company’s study found that a failed cement job was the “precipitating cause” of the April 2010 explosion.
According to Gagliano, BP made decisions that affected the integrity of Halliburton’s cement work. He heard no complaints from BP, which had final say on how cement jobs were run on its wells, until Macondo-related friction developed in spring 2010. “For the production casing job, we had recommended a number of centralizers that weren’t run,” he said. “Also, on the retarder concentration I recommended, they decided to go with more concentration.”
Oil industry practice is to center a well’s casing, or metal tubes, so that channels or paths don’t form in surrounding cement. Centralizers are placed around casing sections to allow the cement to form a strong, 360-degree seal between the casing and the borehole. Without enough centralizers, a channel for drilling mud or contaminated cement can develop, providing a path for a later blowout. A tiny crack can become a gushing channel.
Based on modeling Gagliano ran on Halliburton’s OptiCem software, he recommended that BP use 21 centralizers at its Macondo well to minimize gas flow potential. BP chose to run only six, however. Gagliano said his model showed that using just six suggested “severe gas flow potential.” He tried to get six centralizers to work in his model but “no matter where I placed them and the distance between them, nothing would take care of the channeling.”
Gagliano also recommended .08 gallons of retarder concentration for every 94-pound sack of cement at the well, but BP decided to go with .09 gallons. Judge Barbier asked if using a higher concentration of retarder meant it would take longer for the cement to set, and Gagliano said “yes.”
Gagliano suggested a full bottoms-up, or flushing, of BP’s well before pumping the cement job to see if mud at the bottom of the hole came to the surface. But BP decided to pump only a couple of hundred barrels of fluid, rather do the full bottoms-up, he said.
The Macondo well had a fragile formation, and Gagliano’s task was to design a slurry that would work best. He recommended a foam cement slurry and BP agreed. Gagliano said “we had a dry blend on the rig comprised of some additives that were mixed, and then you also add some liquid additives required for the job specifications, you mix it and then you inject nitrogen into it to get the desired weight. Then you go down hole with it.”
The dry blend used on the Macondo well remained on the Deepwater Horizon rig when it was moved from the Kodiak well, drilled in 2008 by BP and partners. “It was leftover cement from Kodiak that was actually just transferred on paper to the Macondo well,”and originally purchased by BP from Halliburton, Gagliano said. He suggested BP use it for the foam slurry he was designing even though “I could have sold them additional cement, which would be additional revenue for us.”
He said the initial dry blend contained Halliburton’s D-Air 3000 defoamer, which is added to cement “to break the air to give it a lot smoother density going down hole.” Cement mixed at the surface can become thick and entrap air, changing the cement’s weight, he said.
In his testimony, Gagliano dismissed industry safety concerns about D-Air 3000 having a destabilizing effect on foamed cement slurry. He said a Halliburton liquid additive called ZoneSeal foamer was used to compensate for the defoamer. “And then as long as you test it to verify it’s stable, there will be no issues,” he said. The slurry design also contained a retarder. On April 18, 2010, he sent an email to BP engineers listing the ingredients and additives in his design and never got any questions back.
BP has since accused Halliburton of being aware that properties of D-Air 3000 and several other additives in Gagliano’s slurry design, including the retarder or dispersant SCR-100L, should not have been used with a foam cement slurry.
Gagliano said he provided BP engineers with all the test results they requested. A couple of results from the Broussard, La. lab were identified by Halliburton as “invalid” because the slurry was mixed incorrectly at the facility, he said.
He also said that before the April 2010 explosion he believed that if the Macondo cement job failed, BP could go back and pump more cement to rectify problems. He never saw the need to call a stop to the cement job, and looking back he has no regrets about the slurry he designed. Gagliano, a former U.S. Marine now in the Marine Corps Forces Reserve, was on the witness stand most of Monday.
Late Monday and Tuesday, Ph.D. mechanical engineer Glen Stevick, Halliburton’s expert witness on blowout preventers, testified. He said the drill pipe in BP’s well should have been centered before the BOP was activated two days after the blowout. Following the April 20, 2010 explosion, the rig drifted and sank and the connected drill pipe bent so that the BOP couldn’t sever the pipe and seal the well. The BOP wasn’t designed to sever the pipe at an angle, he said.
Halliburton rested its case Thursday. Before that, however, Judge Barbier chastised the company for concealing cement samples from the Kodiak well at its Broussard lab until several weeks ago. He said the samples were relevant to the Macondo trial. “This has been drip, drip, drip, drip, where evidence all of a sudden is discovered—whether it be a bucket of cement, whether it be documents that should have been produced long ago,” Barbier said. “Now all of a sudden it’s discovered.”
BP attorneys said they will call their first witnesses Monday and hope to finish sometime between April 18 and 23.
Last week, Barbier said he won’t hold closing arguments at the end of this Phase One trial assessing negligence. He’ll allow time for parties involved to submit proposed conclusions at a post-trial briefing. After that, “I may consider whether or not it would be helpful to hold any kind of oral arguments or not,” he said.
A Phase Two trial on the size of the spill is expected to begin in U.S. District Court in September, followed by a third phase, probably next January, to decide damages. If BP is found negligent under the Clean Water Act, penalties will be $1,100 per barrel. But if gross negligence is proven, the fine grows to $4,300 per barrel. And if the U.K-based company is guilty of gross negligence, it will be liable for as much as $17.6 billion under the CWA.
This article originally published in the April 8, 2013 print edition of The Louisiana Weekly newspaper.