As federal panel probes oil spill, picture emerges of a series of iffy decisions

By Joel Achenbach and David Hilzenrath
Washington Post Staff Writer
Sunday, July 25, 2010; A01

KENNER, LA. -- If there is no smoking gun in the Deepwater Horizon disaster, it is because there is smoke coming from so many places.

After months of oil-spill misery and endless recriminations about what happened and why, it is increasingly clear that the complex operation of drilling an exploratory well in the deep water of the Gulf of Mexico failed in a complex way. No single decision or misstep in isolation could have caused the blowout, but any number of decisions might have prevented it had they gone the other way.

The calamity, the evidence now suggests, was not an accident in the sense of a single unlucky or freak event, but rather an engineered catastrophe -- one that followed naturally from decisions of BP managers and other oil company workers on the now-sunken rig.

Such was the theme that began to emerge from hearings this past week in Kenner, La., where a federal investigatory panel, meeting in a nondescript Radisson hotel near the New Orleans airport, questioned survivors of the April 20 explosion that killed 11 of their co-workers. Government investigators describe a situation in which BP repeatedly had to make "risk-based decisions," and in every instance chose the least expensive option even though it potentially elevated the risk. That steadily whittled away at the margin of error until there was no margin left and gas found a spark on the Deepwater Horizon.

For example, on the day of the blowout, BP managers decided to skip a typically routine, and time-consuming, "cement bond log" test that could have detected fissures in the cementing of the well. They did not use the recommended 21 "centralizers" to position the well prior to the cement job, deploying just six instead. They used the cheaper of two well designs, one with fewer barriers to rising gas but costing $7 to $10 million less.

They were also conscious of the value of mud. On the day of the explosion, the expensive drilling fluid was taken out of the well and offloaded to an adjacent boat precisely when, as it turned out, the mud was the only thing suppressing the "well from hell," as the widow of one rig worker called it.

Then there was Transocean, the company that owned the rig. Transocean didn't fix a leak earlier this year in the control pod of the blowout preventer, and the company permitted a backlog of 390 maintenance jobs, according to a September audit.

The companies involved have yet to make their cases in full. Witnesses repeatedly testified that safety was their paramount priority. Officials with each company have suggested in recent months that the culpability for the disaster lies elsewhere. BP, for example, has argued that Halliburton failed to do an adequate cement job. BP also blames Transocean, the owner of the rig, for failure to maintain its blowout preventer -- a device designed to automatically close down a well that has become dangerously unstable .

Halliburton and Transocean, for their part, point out that BP designed the well and had final say on all decisions at every stage of the process. BP's investment partner in the well, Anadarko, has declared BP's actions "reckless" and a breach of contract -- a legal argument designed to remove Anadarko's liability for the oil spill.

One event received extended treatment in the Kenner hearings. Late in the day of April 20, rig workers conducted what is known as a "negative test" on the well to see if any potentially harmful gas had invaded the well bore. It showed 1,400 pounds per square inch of pressure in the well -- a sign of gas. But for reasons that remain hotly disputed, BP managers on the rig accepted a benign interpretation of the test, and proceeded to pump the heavy drilling mud out of the well and replace it with much lighter seawater. The well soon erupted.

A BP lawyer at one point asked a line of questions hinting that the company believes that a Transocean tool pusher offered an incorrect interpretation of the negative test that downplayed the unexpected pressure in the well.

The tool pusher could not testify -- he was among the 11 who were killed in the explosion.

* * *

The Kenner hearings, a joint investigation by the Coast Guard and the Bureau of Ocean Energy Management, Regulation and Enforcement (formerly the Minerals Management Service), have been something of an exercise in investigatory frustration. Company witnesses largely deflected blame or professed ignorance of key decisions that could have led to the disaster.

The two senior BP people on the rig at the time of the blowout, well site leaders Donald Vidrine and Bob Kaluza -- who held the position traditionally known as "company man" -- didn't show up after being listed as witnesses for Tuesday's hearing. Vidrine cited medical reasons; Kaluza's lawyer said he was exercising his Fifth Amendment right against self-incrimination.

One other company man did appear: Ronald Sepulvado, the longest tenured of the well site leaders on the rig. Four days before the disaster, Sepulvado had taken a helicopter back to dry land to attend what he called "BOP school" -- classes where he learned about blowout preventers and well control. He said he sent his replacement, Kaluza, an e-mail discussing the job, and also talked to him for about 30 minutes during the "handoff."

While Sepulvado was gone, an iffy well turned into a killer well.

The first problem was the loss of well integrity. The well had been successfully drilled and, the day before the blowout, cemented. For some reason, the cement job didn't work, and hydrocarbons surged into places they weren't supposed to be.

The second problem was the failure to recognize the first. Wells are typically tested in multiple ways. The cement bond log, for example, sends an instrument down the well to interpret whether the cement is properly in place. Three workers for the contractor Schlumberger had flown to the rig to conduct the test. But at 11 a.m. on April 20 they were told their services were not needed -- and they flew home.

The third problem was the failure of the blowout preventer to cut the pipe when the hydrocarbons surged from the well. Had the blowout preventer worked, it might have overcome all manner of engineering sins. What went wrong may not be known until after the Macondo well is finally killed, and the huge device can then be retrieved and inspected.

The final problem was the spark that turned a gas leak into a fireball. In what may have been the most dramatic revelation of the week -- and one that Transocean disputes -- Mike Williams, the rig's chief electrician, testified that an automatic feature of its gas-detection system had been disabled in a move designed to prevent false alarms that might awaken sleeping workers.

So many potential missteps are now part of the narrative that it is difficult to know where the pivotal points lie. What's clear is that no one took the extra step that might have saved the day. Government oversight by what was then the MMS proved ineffectual. People on the rig who might have sniffed out an imminent blowout were so complacent that in some cases they went to bed just minutes before the fireball.

The challenge for BP will be persuading authorities that it didn't cut corners to save money. But the pursuit of oil in deep water is extraordinarily costly. Lee Lambert, training to be a company man, was on the rig that fateful day and knew the "spread rate," or how much the job was costing BP on a daily basis: "I understood at the time that the overall spread rate was close to one million dollars a day, yes," Lambert testified under questioning from a Transocean lawyer.

The Deepwater Horizon disaster can be attributed to "an organizational culture and incentives that encourage cost-cutting and cutting of corners -- that reward workers for doing it faster and cheaper, but not better," an academic advisory panel of scientists, chaired by University of California at Berkeley professor Robert Bea, concluded earlier this month. "We know that in a very large number of cases, the seeds for failure are sown very early in the life of a particular system -- during the concept development and design phases (e.g. the design of the Macondo well). These seeds are then allowed to flourish during the operation and maintenance phases, and, with the system in a weakened or severely challenged condition, it fails," the panel's report states.

Even as the oil companies point fingers at one another, they remain intricately bound by the disaster: BP is drilling the relief well, the main hope for permanently killing Macondo, with Transocean's Development Driller III. Among those who have been helping to drill the new well: Sepulvado, the Deepwater Horizon company man.

Because there are multiple, parallel investigations being conducted by government agencies, academics, the oil industry broadly and BP itself, not to mention a Justice Department criminal probe, the grim narrative of the Deepwater Horizon disaster will take many months or even years to put together. Any conclusions right now remain preliminary.

But the picture that has emerged so far raises questions about whether the cutting-edge technological venture of deep-water drilling, one that involves multiple companies, an elaborate decision-making network, the foibles of human judgment, and powerful geological forces, can ever be managed with zero risk of catastrophe.

And there may be lessons here for anyone involved in a complex and difficult venture. People tempted fate, hoping for the best while failing to insure against the worst. They did not take care of the little things.

And then the big thing -- the Macondo well -- didn't take care of itself.

© 2010 The Washington Post Company