How temporary repairs contributed to offshore deaths
By Bill Campbell - February 2015
Introduction
The evidence referred to in this article relates to temporary repairs on pipes, 370 in all, of which 196 were not approved. 80 of these repairs were on pipes containing hydrocarbons. All the 196 repairs were not approved in advance of installation by a technical authority and were potentially materially defective. After the accident on Brent Bravo 10 repairs were found to be materially defective after examination including the repair that initiated the gas leak. The evidence is held by the police and Shell accepts that this data is authentic. Shell has raised no objections to this publication on legal or other grounds
_______________________________________________________________________
In 1999 a combination of surface corrosion and internal erosion was eating away like a cancer at offshore pipe work. It was a complete shambles. An Audit carried out at that time reported on the ageing assets offshore there was increasing use of temporary clamps, due to pipe-work reaching minimum acceptable wall thickness causing loss of containment. The audit found that no person at any level in the Aberdeen organisation appeared to have a concise overview of the technical integrity status of offshore installations e.g. the collective picture of loss of containment risks due to fitting temporary clamps, and the amount of thin wall pipes at any moment in time.
Directors accepted these findings and plans were put in place to recover the situation. Despite this the situation deteriorated drastically. A pipe repair was being fitted every three days somewhere in the oilfield. The number of these repairs was unknown because they were not being recorded; there was no register of these repairs. In panic and confusion the Production Director asked his operators to walk every line, the more they looked, the more they found.
In total by the time of the fatalities in 2003 some 370 temporary repairs had been carried out, one on average every day, with 196 such repairs carried out, one every 2 days or so, that were not approved, of which 80 were in hydrocarbon service, and of which 10 were found to be materially defective including the repair that initiated the accident in September 2003.
These numbers related to repairs carried out on 17 installations throughout the Shell oilfield.
The 196 non approved repairs equates to 196 breaches of Shell commitments in its Safety Cases where no changes to plant or equipment were to be allowed without the prior consent and approval of a competent person. This instruction formed part of a code of practice referred to in the Safety Case as Change Control, this code was mandatory.
The public Inquiry in 2006 established that a significant factor in the growth of unapproved repairs was the lack of resources, inadequate competent persons to deal with the backlog, such that Asset Managers simply ignored their own mandatory policy to get prior approval.
The repairs were being carried in a band aid fashion to keep up with the large level of repairs. The leaking pipes in hydrocarbon service caused a loss of containment, and such a loss from of oil or gas is reported under Government guidelines as a dangerous occurrence because of the potential of creating a flammable atmosphere.
If we consider the 80 repairs to hydrocarbon carrying pipes, every 5 days or so, a dangerous occurrence would develop when pipes with hydrocarbons and with paper thin walls leaked, it appears that many of these events, which by Law require to be reported, it is after all the principal key performance indicator reported in HSE statistics, were not reported. These leaks were then repaired, the repair was not registered, no one knew exactly how many there were offshore, and where they were, and no prior approval of the repair was sought from a technical authority.
Immediately after the accident operators were instructed to walk every line on the 17 installations hand checking for temporary repairs, the more they looked, the more they found.
From the viewpoint of the risk analyst, the mean time between dangerous occurrences offshore was circa 5 days, and yet Shell told the world in 2006 that it absolutely refutes the allegation that it operated installations with dangerously high risk levels and yet something dangerous was happening somewhere in the field every week.
The commitments accepted by the Managing Director of Shell Expro in 1999 to get to grips with the situation, register these repairs, have in place a sufficient number of competent persons to approve the repairs, or otherwise, just simply failed to get done. This failure described in the words of the Lord Advocate Elish Angolini as the haphazard management of pipe repairs contributed to the deaths in 2003.
The end result was the double fatality in September 2003, which the public Inquiry determined, could have been avoided if the repair to the pipe on Brent Bravo had not been materially defective. So here again Directors actions, or more appropriately inactions, contributed to the deaths.
This statement was endorsed by the then Lord Advocate Angolini in a reply to a member of parliament that the haphazard management of such repairs, over a prolonged period of time, had contributed to the deaths.
Bill Campbell
Note
This evidence was placed in a Report by the House of Commons select committee related to workplace safety and health in 2007 – 8, with reference number HC 246-11.