The cause was the same as before – a manufacturing error which permitted high local pressure to cause a collapse. See the article on “1995 – first injection flowline failure”.
On this second occasion, however, the incident was assessed to have posed the risk of serious consequences.
After the rupture had occurred, a remotely operated vehicle (ROV) was sent down to document the extent of the damage. Images showed that the flowline to the NWIT had fractured inside a J-tube leading out from the Draugen platform.
As the name implies, such tubes have a sharp bend at the base which penetrates the platform wall. A number of them are available to conduct piping to the seabed – known as pipe-in-pipe.
After further investigations using X-rays, it was determined that the break had occurred close to the end connection. About 180 metres of internal piping had been driven out of the J-tube, while the end and 50-70 metres remained inside.
The flowline had crumpled in a big heap only about 30 metres from the outlet. A lot of gravel and sandbags were displaced, and the flowline lay partly between other piping and control cables. None of these showed visible signs of damage.
Clearing up after the big rupture was a hazardous business, not least because the gas injection riser ran through a nearby J-tube.
The latter would lose its barrier against the platform if the tube carrying the flowline to the NWIT was opened – risky since the gas riser tube was not designed to cope with full internal explosive pressure.
It was therefore thought safer to cover the damaged flowline with rocks and switch to one of the many other J-tubes on the platform. The drawbacks of this approach were regarded as minimal.
The Norwegian Petroleum Directorate (NPD) feared that the cessation of water injection after the incident could have a big impact on reservoir pressure and oil production from Draugen.
Just two weeks after the rupture, wells A4 and A5 were showing a steady decline in pressure. The drop was expected to be 16-18 bar over an estimated injection downtime of two months.
Reducing reservoir pressure posed the threat of more sand in the wellstream and earlier-than-expected water breakthrough in the wells. In the worst case, Draugen could risk coming off plateau production and seeing its output fall sooner than forecast.[REMOVE]Fotnote: Draugen technical committee, section for marine technology, division for safety and the working environment, Norwegian Petroleum Directorate, 1 February 2000.
Despite the failures, flexible flowline manufacturer Coflexip has faith in its product. The pipeline type it had delivered in 1991 was still being produced in 2000.
But the company would not exclude the possibility that fabrication faults might be the cause, and reported that manufacturing methods had been greatly improved in recent years. Shell had not used the flowlines “incorrectly”, it was stressed.
The flowline to the NWIT had been laid in five 1 200-metre lengths, joined with connectors. Only the section closest to the platform was replaced and connected to the rest of the line.
It was therefore convenient that Shell had a 1 200-metre flowline section stockpiled at Vestbase in Kristiansund which could be used. The repair cost was put at NOK 120-170 million.
To prevent a repetition of the incident, Shell was asked by the NPD to present a programme for replacing the flowlines in the longer term.[REMOVE]Fotnote: Draugen technical committee, section for marine technology, division for safety and the working environment, Norwegian Petroleum Directorate, 1 February 2000.
The failure of the water injection line to the NWIT occurred at a time when Shell was giving high priority to the Draugen gas export development.
Involving a tie-in with the Åsgard Transport gas pipeline, this project was pursued in parallel with the repair job, and cross-utilisation of vessels and equipment was explored.
The NWIT flowline repair had top priority in the spring of 2000, without compromising the schedule for the gas export pipeline. Shell succeeded in accomplishing both tasks.