The incident on the Statoil-operated facility began around 19.00 on 28 November 2004, when some 216 people were present on the platform.
In connection with work in well P-31A, plans called for a length of tubing to be pulled out so that a sidetrack could be drilled from the main bore.
Signals from the well had been the cause of growing uncertainty from early that morning, and various measures were adopted to normalise the position.
During the afternoon and early evening, however, it became clear that the problems were increasing. The crew gradually lost control over conditions in and around the well.
The platform manager called a crisis meeting at 19.05. His first act was to muster the emergency response management on Snorre A with the aid of a “silent alarm”.
In less than 30 minutes, however, it had become clear that the position was worsening and on the verge of getting out of control. Warnings were now sounded.
These went to the standby ship, helicopters, Statoil’s response centre on land, the joint rescue coordination centre at Stavanger’s Sola airport and the PSA.
A general alarm sounded on Snorre A, and crew mustered to the lifeboats. Gas was detected under the drill floor, on one of the cellar decks and in two modules on the seventh and eighth stories.
The latter areas were close to the flare, but the flame could not be extinguished at this time because insufficient nitrogen was available.
Personnel began to be evacuated at 20.58, with 75 people initially remaining behind to participate in efforts to kill the well and handle the emergency.
As 21.30 approached, a number of gas alarms went off. Personnel sent out to assess conditions around the platform reported that the sea was boiling with gas.
An emergency shutdown was activated in order to eliminate ignition sources and render the platform voltage-free. That left the facility without main power.
Full evacuation by freefall lifeboat was first assessed by the emergency response leadership at 21.38. This represented a critical consideration and decision, but was rejected.
Personnel could still be taken off by helicopter at this point, because a strong wind was blowing the gas away. Its direction on the helideck was also favourable.
Around midnight, the position on Snorre A became increasingly critical. The management resolved to restart the main generators to provide power for the well killing efforts.
This decision was very difficult, but proved successful. However, the wind shifted at about the same time and died down to leave Snorre A in a calm.
Statoil secured permission from the PSA to expand the safety zone to a radius of 2 000 metres around the platform and 3 000 feet into the airspace above.
No activity could now take place close to Snorre A – nobody was able to come to its assistance, either by sea or by air.
The flare was still burning, only a few metres above the gas cloud, and could not be extinguished until 03.15.
Efforts to subdue well P-31A continued by every available means. All but 35 members of the crew had been evacuated, and those left had a single overriding mission – to regain control.
Large volumes of oil-based mud had been pumped into the well throughout the night in a bid to force back the pressure. These attempts failed, and stocks of mud were starting to run out.
The gas and the safety zone made it impossible to bring fresh supplies to the platform. And the blowout could not be halted without mud.
Various emergency solutions were considered – cementing, using seawater, or mixing an emergency mud from every available substance on board.
The decision was taken to concentrate on a mix of water-based mud containing barytes and bentonite – not really a satisfactory solution, but the only option in practice.
Between 04.00 and 09.15, the residual Snorre A crew worked feverishly to mix up the 160 cubic metres of mud remaining on the platform.
The fluid was stockpiled for a single crucial bulkheading operation down in the well. According to the emergency response leadership, this represented a final effort to halt the blowout.
At 09.00 on 29 November, Snorre A made a third request for enhanced evacuation preparedness in a particularly critical phase. Fifteen minutes remained before the fateful bulkheading.
The mud operation began as planned at 09.15. The mood was extremely tense both on Snorre A and ashore as the crew began to pump the fluid down the well. This was make or break.
Everything was over at 10.22. After a life-or-death struggle during a long and extremely stressful night for all on board, the report reached land – the well pressure registered zero bar.
Incredibly, a mere eight to 10 cubic metres of mud were left on the drill floor.
Published in March 2005, the PSA’s investigation report was blunt. The incident resulted not from chance, but from general failures in Statoil’s planning, procedures and assessments.
The inquiry also noted that this event, in slightly different circumstances, could have caused a tragedy with fatalities, pollution and a major loss of material assets. The drilling contractor was also criticised.
Serious failures and shortcomings were found at every level in Statoil’s planning and execution of the P-31A job. The report cited 29 nonconformities from the regulations and a number of
It concluded that all the nonconformities would have been identified and corrected if the barriers had functioned – and that the failure of so many barriers in various phases of an operation was very unusual.
The PSA was sharply critical of the fact that the wide-ranging failures were not identified and the faults corrected before P-31A went out of control.
Statoil needed four years to close out the extensive orders issued by the PSA on the basis ofits findings. The final letter confirming that the work had finished was sent on 25 March 2009.
The incident itself and the subsequent report rank today as a point of reference for the petroleum industry both in Norway and internationally.
Snorre A represents a scenario which shares key features with the Piper Alpha disaster – although with a completely different outcome.