STATOIL CITES DESIGN ERRORS IN SLEIPNER STRUCTURE LOSS

Norway's state owned Den norske stats oljeselskap AS has pinpointed cause of the accident that resulted in the sinking of the Sleipner A platform concrete gravity structure in a Norwegian fjord in August. Statoil earlier this month announced the accident was caused by design errors leading to cracks in one of the concrete shafts (OGJ, Oct. 21, Newsletter). Loss of the structure off Stavanger stunned Norway's offshore industry and forced Statoil to scramble to implement other supply

Oct. 28, 1991

4 min read

Statoil earlier this month announced the accident was caused by design errors leading to cracks in one of the concrete shafts (OGJ, Oct. 21, Newsletter).

Loss of the structure off Stavanger stunned Norway's offshore industry and forced Statoil to scramble to implement other supply options to meet Sleipner gas field contract deadlines (OGJ, Sept. 2, p. 30).

GENERAL CONCLUSIONS

Structural failure of the tricell unit linking the D-3 drilling shaft and adjoining base units was the most likely cause of the loss of the Sleipner structure, according to an investigation by field operator Statoil. The initial failure, at 105 ft from the bottom of the structure, could have triggered one or more lesser cracks in other tricell walls of the D-3 shaft.

Statoil said the investigation team looked at 24 other possible causes of the accident and concluded the failure occurred because the tricell walls and adjacent joints were underdesigned.

The team also recommended a new concrete structure for Sleipner A could be ordered provided some design aspects are changed and monitoring of design and construction work is tightened.

HOW IT HAPPENED

The Sleipner A substructure sank during a submersion test Aug. 23.

Norwegian Contractors personnel conducting the test reported a loud bang followed by two lesser bangs and formation of fog in the D-3 shaft.

Personnel then reported a roar that sounded like a waterfall that gradually subsided. When the fog cleared, the water in the cell appeared to be boiling. The structure started to submerge quickly and sank within 18 min of the first bang.

The structure imploded as it sank, and seabed inspections showed little was left of the base cells and the four shafts.

FLAWS DETERMINED

The tricell wall and adjacent joints were underdegigned to the extent that cracking could have been expected when exposed to water pressure in the fjord, Statoil said.

Further, the wall in question in the T-23 tricell was among the most critical of tricell walls.

No other structural weaknesses that could have had corresponding consequences were established, said company officials.

Statoil found no records of deviations in quality of construction work by Norwegian Contractors that could have caused the accident. However, it is possible marginal differences in the construction work may have caused the failure to occur first in the D-3 shaft, the companies said.

The investigation team also established that the leak rate was such that leaks must have occurred in more than one tricell wall. This contention is based on the calculated strength of these walls and redistribution of forces that would have occurred after the first cracking.

DESIGN ERRORS

Statoil found three decisive errors occurred separately but taken together contributed to the structural failure:

Errors in the initial global analysis caused the area in question to be underdesigned for the forces to be expected.
Reinforcement was improperly detailed.
The tricell joint was not designed separately and only contained the same shear reinforcement as designed for the walls between the cells.

For global analysis, the concrete structure was divided into 23,000 finite elements and calculations made of forces affecting the separate elements for all relevant loads. The results were incorrect for the areas where the failure occurred. The shear forces were only 53% of the correct values.

The investigation team also determined that details of the reinforcements in the corners of the tricells were not properly designed. The way in which Threaded steel reinforcement bars were used did not enable them to handle the forces that developed in the joint.

Further, the same shear reinforcement was applied to joints as to the adjacent cell walls. Statoil said this method would not necessarily result in an amount of reinforcement sufficient to absorb forces active in the joint.

And the team concluded that even if the forces had been correctly calculated in the analysis, this would not have resulted in a correct design because the joint required a separate detailed analysis.

Design errors should have been revealed by internal control procedures by Norwegian Contractors and a subcontractor, Statoil said.

The team also noted the engineering of Sleipner A was based on the results of the global analysis with few independent evaluations of the results. And design of dimensions was conducted with a computer without sufficient control of the separate steps in the process. The result was that design errors were difficult to detect, Statoil said.