This investigation report refers to a hydrocarbon leak which occurred from a well flexible riser during a planned shutdown on Visund on 9 April 2011. Oil film and gas bubbles were observed in the sea near the production risers on the east side of the facility.
The use of flexible pipes in Norwegian oil production started in the late 1980s. The first applications were static underwater connections between pipes and subsea facilities and dynamic applications between fixed and floating facilities. Early in the 1990s, development started on fields in deeper water with floating facilities where there was a need for dynamic, flexible risers from the seabed and up to the facility.
The dynamic risers that pose challenges are so-called ”un-bonded” risers, which means that there is a layered structure where the various materials have functions for withstanding external and internal pressure, withstanding tensile forces, preventing hydrocarbon leaks, protecting against seawater, etc. There are various types of designs as regards the number of steel and plastic layers, types of steel and plastics, etc. The dynamic risers that experienced carcass collapse on Visund, Njord A and Snorre B are constructed with three layers of Coflon (plastic material). Previous experience shows that dynamic risers with a pressure barrier in multiple layers are more exposed to collapse than if there is only one layer.
The challenges related to collapse of the carcass (inner steel structure) in dynamic risers with three layers of Coflon pressure barriers are an issue that has been known since the early 2000s. Normally, pressure builds up outside the carcass, between the plastic layers, by gas diffusing through these over time. In the event of too quick depressurisation of the riser, it has been observed that overpressure can occur between the plastic layers that serve as the pressure barrier and carcass, which can lead to this expanding and collapsing. The cause of the external leaks on Njord A and Visund in 2011 is still not known and the extent of the damage to the leaking risers is also unknown.
During the summer of 2010, Statoil had two hydrate incidents on Njord A and identified collapse of carcass. On 26 November 2010, carcass fragments were discovered in the sand trap on Njord A. Based on these incidents, a decision was made in December 2010 to inspect all three-layer Coflon dynamic risers on Njord A.
In January 2011, a decision was made to inspect the three-layer Coflon dynamic risers on Visund and Snorre B and a task force was to be formed in Statoil to assess the operation of dynamic risers in Statoil. During the inspection on Njord A in March, the carcass was found to have collapsed and the production was shut down. The task force started its work on 6 April.
On 8 April a decision was made to shut down Visund following risk assessments based on experience from Njord A and Snorre B, as well as previous experience on Visund. In retrospect, Snorre B’s dynamic risers in three-layer Coflon have also been shut down as their configuration was similar to those on Njord A and Visund and, following safety-related evaluations, it was considered to be necessary to review the operational history of the risers.
During shutdown of risers on Visund, a leak occurred on well A21’s three-layer Coflon dynamic riser. Evacuation of personnel was started and the PSA’s emergency preparedness centre was established. Both Statoil and the PSA decided to investigate the incident.
On 15 April, risers on Njord A were restarted following an internal inspection with a camera and on 24 April, a leak occurred in the three-layer Coflon dynamic riser with subsequent shutdown. Following an inspection of the risers on Visund on 28 April, torn carcass was discovered at the top of the riser. This corresponded to the type of damage that was previously discovered in the riser with an external leak on Njord A, and which was a new and unknown damage mechanism.
The pictures below show a typical construction of a dynamic riser (left picture) and an example of carcass collapse (middle picture), as well as the configuration of risers on Visund (right picture).
The next figures show a sketch of risers with a carcass and the three pressure layers in the plastic material Coflon, and pictures that show an example of carcass collapse on Visund in 2005.
The leak occurred during shutdown of dynamic risers on Visund after Statoil had performed risk assessments for Visund with experience from, for example, Njord A. Oil film and gas bubbles were observed in the sea near the production risers on the east side of the facility. The production was shut down and all risers were immediately isolated. A systematic search for the leak point located the hydrocarbon leak in the dynamic riser connected to well A21. 63 of 123 people were evacuated to nearby facilities. The volume of discharged hydrocarbons is uncertain as the leak site and size were unknown and immediate depressurisation was carried out. During the incident, the maximum shut-in volume in the riser was estimated to around 4000 kg of gas.
If the leak had been larger, hydrocarbon gas could have blown into the facility and created a possibility of ignition and explosion. Due to the fact that this riser was shut-in, a limited gas volume would have been discharged, even in the event of a possible full breach of the riser. The volume of gas that would have been able to flow in over the facility in the event of a leak from a riser will depend on many factors, including where the hole is located and its size.
The PSA investigation identified three non-conformities within the following areas.
1. Establishment and follow-up of preconditions for safe operation of dynamic risers.
The PSA investigation discovered through conversations with operations personnel that the risers’ operational limits were known, but that they were difficult to satisfy and that there were no clear written procedures for the operation. Operations personnel were not significantly aware of the consequences of operation beyond the operational limits.
The PSA investigation also revealed that the risers on Visund were challenging to operate due to the facility’s design with dynamic risers connected directly to wells.
Dynamic risers in three-layer Coflon have been operated beyond operational limits on Visund. This was documented through monitoring and registration of data and summaries in bi-annual and annual status reports.
Furthermore, an operations history could not be presented to document the integrity of the dynamic risers on Visund throughout their lifetime. During the investigation, Statoil was in the process of carrying out an activity to review documentation for operation of dynamic risers on Visund. This was also true for other facilities with the type of riser and riser configuration used on Njord A and Snorre B.
Based on previous experience with this type of riser on Visund, Njord A and Snorre B, the party responsible for the discipline in Statoil was aware of the consequences of operating this type of riser beyond operational limits. However, it could not be demonstrated how risk assessments related to operation of three-layer Coflon dynamic risers on Visund had been carried out, and that these had lead to the establishment of clear requirements and follow-up of how they were complied with.
2. Governing documentation.
Before the investigation activity at Statoil’s offices in Bergen on 18-20 May, the PSA asked to be provided with procedures/routines/operational limits for dynamic risers with three-layer Coflon in Statoil and on Visund. Before the investigation, we received governing documentation for safe operation of three-layer Coflon dynamic risers on Visund. Through conversations and presentations, we were informed that this
Documents regarding procedures/routines/operational limits for dynamic risers with three-layer Coflon in Statoil and on Visund had not been made known, nor was it part of the governing documentation on Visund until after the incident on Njord A and Visund at the end of March and early April 2011. The status of this document was also unclear during the PSA investigation. The document was prepared in 2006, but was revoked as it was not considered relevant for Visund. The incident with dynamic risers on Snorre B in 2008 showed that this was relevant for Visund, but through conversations it was confirmed, however, that this was not reassessed before the recent incidents with dynamic risers on Njord A and Visund.
Parts of the content of the mentioned procedure could be found in several procedures that did not clearly state anything regarding operation of three-layer Coflon dynamic risers. None of the documents made any reference relating to what to do following operations exceeding operational limits.
3. Training and expertise.
Offshore personnel who were responsible for daily operation of the dynamic risers on Visund were not fully informed of or trained to understand the risk associated with operating the dynamic risers beyond the operational limits. This was confirmed in conversations and the review of documentation, such as bi-annual and annual status reports for dynamic risers on Visund, which showed several operational nonconformities. In connection with these operational nonconformities, there was no formal reporting from offshore to land, and following review of the reports by onshore personnel, no actions were taken to include this information until after the incident on Visund on 9 April 2011.
Source: Norwegian PSA
Officer of the Watch 
Leave a Comment