The leak occurred on the evening of 10 March 2017 in connection with a planned activity to reconnect well S-4 on the S template, which is tied back to Åsgard A. This activity was pursued from the Deepsea Bergen semi-submersible drilling rig.

At the time of the incident, the well was disconnected and a blind had been installed at the connection point on the manifold. The leak occurred when the blind was removed.

The isolation valve from the flowline on the manifold to well S-4 was open, allowing gas and condensate to flow to the sea. Wells producing to the flowline were shut and the leak continued until the pressure in the flowline had equalised with seabed pressure, which took about 20 minutes.  

Actual consequence

The consequence of the incident was that gas and condensate escaped to the sea and the atmosphere. Currents and wind direction were favourable, so that little gas flowed to Deepsea Bergen.

Based on figures from Statoil, about 31 tonnes of gas and 1.6 tonnes of condensate are estimated to have been discharged. In addition, production from the S template was halted for 28 days.

No people were injured in the incident.

Potential consequence

Gas hazard analyses by Statoil show that the discharge could have led, under different weather conditions, to ignitable gas entering Deepsea Bergen’s moonpool. These analyses show that ignition of the gas would not have threatened the integrity of the facility, but could have led to fatalities had there been personnel in the area.

Direct and underlying causes

The direct cause of the gas leak from the S template was that the isolation valve stood in the open position when the blind was removed. Several underlying factors contributed to the failure to detect that the valve was open before removing the blind, and to test and secure this valve as a barrier.

Findings

The investigation has identified five nonconformities and one improvement point.

Nonconformities:

  • Barriers – no barriers were established to prevent discharges to the sea during work on the template.
  • Risk assessments – in connection with planning and executing the operations on well S-4, important contributors to risk and changes in risk were not identified and assessed.
  • Responsibility – responsibility for testing isolation valves was not unambiguously defined and coordinated in connection with the operations being conducted.
  • Knowledge of governing documents – personnel involved had little knowledge of barrier requirements in governing documentation for work on the template.
  • Documentation on requirements – relevant requirements in governing documents concerning isolation valves were not clarified in the operational procedures.

Improvement point:

  • Follow-up – follow-up by management has not contributed to identifying technical, operational or organisational weaknesses, errors and deficiencies.

Operator Statoil has been asked to explain how these nonconformities will be dealt with, and to provide an assessment of the identified improvement point.