Well design for Macondo laid foundation for problems, panel told

Nick Snow
Washington Editor

Decisions made during the design phase of the deepwater Macondo well led to conditions in the hours before the well's blowout on Apr. 20 that left the crew with inadequate and incorrect information and not enough time to respond, experts told US President Barack Obama's independent commission investigating the accident and subsequent oil spill on Nov. 9.

"A well design needs to include as much detail as technically possible. It has to show what the well intends to accomplish, and then each element needs to go through a complete design cycle which is rolled back to consider implications on what's intended," explained Steve Lewis, an advanced drilling technology engineer at Seldovia Marine Services in Alaska who the commission retained to study the Macondo well's design and operating plans.

There should be a complete design reevaluation at each phase, Lewis said, adding that he found the Macondo well's initial design adequate for planning, but deficient in detail. "It would have helped, I believe, to focus both the field and the office staff on the difficult and almost marginal nature of what they were attempting to accomplish. This might have brought in a heightened level of vigilance, allowed more time to mobilize materials, and possibly allow a further level of discussion," he said.

Producers drilling deepwater wells in the Gulf of Mexico routinely have to deal with high-porosity, high-permeability formations with weaker rocks, forcing them to contend with a very narrow margin between the rocks' pore pressure and fracture gradient, according to E.C. Thomas, a consulting geophysicist and owner of Bayou Petrophysics in Spring, Tex. "As a well is drilled, signs have to be monitored to make sure the pore pressure gradient isn't exceeded," he said. "We're happy to have 2 lb/gal separating the two [gradients]. Having a 1.8 lb/gal separation, as the Macondo well had, is very close to having to stop drilling."

Thomas said conditions at BP PLC's Macondo well had been stable for 4 days before the blowout and rig explosion that killed 11 workers. "They had solved the problem of getting down, but they could never ignore the fact that they were in an environment with a very narrow pore pressure-fracture gradient window," he said. "They would have had to pay particular attention not to exceed the fracture gradient with the weight of the cement."

Charlie Williams, chief scientist for well engineering and production technology at Shell Oil Co., generally agreed that the narrow pore pressure-fracture gradient window posed a challenge at the Macondo well. The design engineers would have faced others, including the need to manage mud in the risers and the well and working with blowout preventers on the ocean floor that would have to be pulled to the surface for maintenance, he said. A typical deepwater team with operating staff involves at least 15 people and sometimes as many as 20, he indicated.

"At an exploration well, where you don't know as much about the environment as at a production site, it can take from 8 months to a year, depending on the complexity," Williams said. "It's typical for designs to change during this period. It's a defined collaborative process. We bring in all the technical groups, including the people who are actually going to drill the well, and hone the design. At different stages, we have even bigger reviews where we bring in the entire contract rig crew to look at the design. It's a maximum collaborative environment to produce a design for executing the well."

Lewis said the three changes that BP made in its well abandonment plan during the final week were excessive. Establishing procedures earlier "would have allowed people to give the matter more thought in a less time-sensitive environment," he suggested. "A well is actually a pressure vessel. It's designed to control and contain the fluids we're trying to extract. Containment is paramount at all times.

"At the end of a well operation, many things need to be done in order to move forward," Lewis said. "There's also a natural human tendency to look toward the future operation and lose focus on what has just been accomplished. At the end of the well, it's even more important to maintain vigilance and focus. [The tendency to not do this] is extremely variable. You obviously have to have the resources in terms of manpower and the commitment to maintain focus on what you're doing instead of thinking what you'll be doing in the future. It's a natural thing not to."

• Individuals should be trained to repeatedly question data, raise concerns, and double check assumptions.

• Greater attention should be paid to the magnitude of consequences of all anomalies, even if they seem to be minor.

• Individual risk factors cannot be considered in isolation but as an overall matrix. Personnel cannot ignore anomalies after believing they have addressed them.

• There should be greater focus on procedures and training in how to respond to low-frequency, high-risk events. "How do you know it's bad enough to act fast?"

• There was a failure to develop or adopt clear procedures for crucial end-of-well activities.

• Poor communication between the operator and subcontractors deprived otherwise capable personnel of information necessary to recognize and address risks.

• There were muddled lines of authority within BP, and between BP and its contractors.

• The flow path was exclusively through the shoe track and up through the casing.

• Cement (potentially contaminated or displaced by other materials) in the shoe track and in some portion of the annular space failed to isolate hydrocarbons.

• Pre-job laboratory data should have prompted a redesign of the cement slurry.

• Cement evaluation tools might have identified a cementing failure, but most operators would not have run tools at that time. They would have relied on the negative pressure test.

• The negative pressure test repeatedly showed that the primary cement job had not isolated hydrocarbons.

• Despite these results, BP and Transocean personnel treated the negative pressure test as a complete success.

• BP's temporary abandonment procedures introduced additional risk.

• The number of simultaneous activities and nature of the flow monitoring equipment made kick detection harder during riser displacement.

• Nevertheless, kick indications were clear enough that, if observed, would have allowed the rig crew to respond sooner.

• Once the rig crew recognized the influx, there were several options which might have prevented or delayed the explosions and/or shut in the well.

• Diverting overboard might have prevented or delayed the explosion. Trigger the EDS prior to the explosion mighthave shut in the well and limited the impact of any explosion and/or the blowout.

• Technical conclusions regarding the blowout preventer should await the results of the forensic BOP examination and testing.

• No evidence at this time suggests there was a conscious decision to sacrifice safety concerns to save money.

Source: Presidential Oil Spill Commission Nov. 9, 2010, hearing

Lewis, who has examined communications between Macondo rig workers and BP's senior management, said the only reasonable explanation for changing the well's abandonment plan three times immediately below the blowout was that a detail emerged that was not considered in the early versions, possibly due to engineers not being available. "Once that point was reached, there was an apparent scrambling to catch up. That was based on the documents I saw. There was no real detail of abandonment in the original plan. In the final plan, while the engineering aspects were adequate, the operational plan sent to the rig was totally inadequate," Lewis said.

It takes a stated management commitment to combat inherent pressure to wrap up work at the well and move on, Lewis added. "Putting the design and field team together can provide a check-and-balance against this tendency," he said. "The overall impetus to make progress and, in some cases of design and execution, to choose a route that's quicker and requires fewer steps comes from management."

Lewis noted that conversion of the well's Weatherford float valve from two-way to one-way operation would have required 500-700 psi of pressure to function. With the weight of mud in the hole at the time, a 600 psi pressure would have been needed to pump mud through bore holes in the side at 6 bbl/min, he said, adding that BP apparently felt that 4 bbl/min was the maximum safe pressure. "The function of this type of float equipment is to isolate the cement once it's in place and prevent it from flowing back into the well. It's my opinion that the Macondo well did not convert. The preponderance of events would indicate that," he said.

BP employees on the rig relied on flow-check tests to determine that well pressure was under control, Lewis said. "The pressure differential in this design was so small that it would have not been conclusive," he said. "Also, simply watching for three minutes is a little too low. Most pressure plans require watching for 5 min. They may have watched for longer than 3 min, but that's what they reported."

Drilling a deepwater well safely under these conditions requires a longer flow check or maintaining pressure on the column for a few hours until the cement has had time to set, he continued. "Where you've had a sequence of questionable events and you've never obtained the design specifications for that particular, it's apparent that something is wrong," he said.

Darryl Bourgoyne, director of Louisiana State University's petroleum engineering research and technology transfer laboratory, said BP placed the Macondo well's cement plug much deeper than normal and left the well underbalanced while cement was being placed. Displacing 3,300 ft of mud in the well with seawater without using mud heavier than 14 lbs/gal also wasn't prudent, added John Rogers Smith, an associate professor of petroleum engineering at LSU. This resulted in an underbalanced well before the negative pressure test started, he said. BP could have used heavier mud below the cement plug, but it would have taken another 1-2 days, he said.

Having only one barrier in place in addition to the BOP during the open displacement process was "putting all your eggs in one basket," Lewis said. "Purposely bringing the well under balance and using only one barrier which had not been tested was questionable." Burgoyne said that BP could have placed a second barrier before removing the mud, while Smith said that it also could have achieved one by increasing the mud's density or setting the plug in higher density mud to begin with. "There are complications and risks associated with each one that have to be considered," he said, adding that BP did not employ any of these options.

The three witnesses agreed that the negative pressure test was a failure. "It demonstrated the well could flow," said Burgoyne, adding that he has not seen procedures for interpreting the test's results.

Smith said, "I would have expected that there would have been a calculation of pressure at the beginning of the test. I would have expected to see some statement of what to do if a test was not successful. In my own practice, there would have been additional details of volumes to pump and steps to monitor more details."

Burgoyne said, "There was not a final procedure to send out to the rig. It would have required other steps. The primary thing would have been what a successful test would have looked like, how long to wait, and calculate expected bleed-back volumes. If you start to get back a significantly greater volume, it indicates additional fluid that is not accounted for by compressability. That volume actually would represent an in-flow into the well. I haven't seen any indications that this was done at the Macondo well."

Burgoyne said he also could not find any federal regulations or industry standards and practices for negative pressure tests because the procedure is relatively rare onshore and at shallower depths where the wellhead is at the surface, but important in deepwater wells. Once the test showed there was a problem, it would have taken at least 1 day and possibly several to correct it, he said.

Bourgoyne, who worked for Chevron Corp. before joining LSU, said that if he had not understood such results, he would have contacted someone higher in the company immediately. Lewis said that a culture encouraging employees to seek advice starts at the top, and that another approach would have been to involve drilling engineers involved throughout the planning process. "That will create relationships when the person at the rig knows the person he is calling onshore when he runs into an odd situation," he said.

Other witnesses during the hearing's final day included Michael R. Bromwich, director of the US Bureau of Ocean Energy Management, Regulation, and Enforcement, BOEMRE Deputy Director Walter D. Cruickshank, ExxonMobil Corp. Chief Executive Officer Rex W. Tillerson, and Shell Oil Co. Pres. Marvin E. Odum.

More Oil & Gas Journal Current Issue Articles
More Oil & Gas Journal Archives Issue Articles
View Oil and Gas Articles on PennEnergy.com