The investigation into BP’s disastrous Gulf of Mexico oil spill has hit the buffers after a drilling technology firm refused to hand over access to fundamental software used on the rig.

National Oilwell Varco rejected US government demands to provide access to its proprietary HiTech application, which was relied upon by BP engineers to determine the presence of dangerous hydrocarbons in the well during the drilling. It said that handing over the code would mislead the investigation.

The senior oil spill investigator, Fred Bartlit, yesterday wrote an angry letter to the Oil Spill Commission, highlighting the problems and pleading for help from the authorities to force NOV to give him access to the system.

The investigation could not proceed without access to the HiTech system, he said. “We write to inform you of a roadblock in our investigation. ...The roadblock is NOV [National Oilwell Varco].”

The OSC is attempting to publish a report next month on the causes of the accident, but has had serious problems accessing some of the data it is seeking.
“For over a month, we have attempted to elicit NOV’s assistance on this matter,” Bartlit wrote. “They have been generally uncooperative, either in the form of refusal or delay.”

Bartlit said that investigators needed to determine “what indications of hydrocarbon flow the Deepwater Horizon’s drilling crew could have seen in the two hours prior to the explosion”. While the precise data “sank with the rig”, a “subset” was available in a different format that could also be accessed on the HiTech software.

“In order to recreate the driller’s displays,” he wrote, “we need access to NOV’s proprietary HiTech software, conversion of the data into a format compatible with that software, and simulations of the real-time displays.”

BP had provided underlying data, Bartlit noted, and its cementing contractor Halliburton was in talks to provide key algorithms for data conversion. The rig’s owner, Transocean, was also voluntarily providing key witnesses. But NOV had not helped enough, he wrote.

Bartlit conceded that it was “impossible” to recreate exactly the screen that the drillers saw. But he said that “even a basic understanding” of the information available would “significantly” advance investigations.

The only people who knew what happened – the driller and two assistants – died in the explosion, he wrote in the letter to the commission.

“All we have is the data. We request your assistance in obtaining NOV’s cooperation.”

NOV fired back a curt statement to the media following the letter’s publication on the web, and insisted it supported investigations. It added in the statement that "manufacturing guesses as to what was displayed on the rig's computers runs a serious risk of producing a misleading picture of what actually happened."

The company said it had rejected “requests to synthesise hypothetical computer displays utilising limited mud data provided by the commission, because these would not be accurate or fair".

"It is impossible to say what channels or parameters, time scales or pages the driller had displayed at the time of the accident. Frankly, we were surprised by the commission's letter."

The systems in use on the rig have come under increasing scrutiny. Investigators are still attempting to establish what happened to the automated and manual systems in the blowout preventer, a device supposed to prevent well explosions.

Last month, a crucial presentation slide prepared by investigators briefly emerged early, in error, on the website of the OSC. It indicated that BP had ignored the advice of cement modelling software from Halliburton in an attempt to save time. BP declined to comment.