Last Thursday, I attended the Hydrocarbon Release Seminar, organised by Oil and Gas UK at AECC. Details of the event are here.
It was a particularly interesting event, I found. I hadn't done any significant work in the area since the days it was my job to fill in the OIR12 reporting forms and calculate/estimate/guess the release amount. The mix of people attending was quite diverse, quite a few offshore operations personnel, OIMs, technicians, team leaders, etc., as well as a good mix of operations and maintenance people. I was a bit surprised that there were not many consultancies such as Atkins represented, actually.
What was described to us was a classic problem. The industry has targeted a 50% reduction in the total number of leaks (reportable or higher) since the 2010 figure of 187 by March 2013. For an individual installation, in any given year, the target number of releases is zero. The expected number of releases if you do the stats probably works out at less than 1 per year. And apparently 60% of releases are attributable to human error during maintenance or operations, with 40% to technical factors. As we are looking at relatively low frequency events, it is easy to lose focus, or become complacent that the release won't happen on my watch.
The presenters were in many ways frank and honest. The HSE delegate described an industry with a "history of failure and a memory of success". The incidents described were in many ways repeats of what we have seen before: failures to isolate from high pressure sources; not following procedures; inadequate checking; loss of the ability of the pressure envelope to handle pressure. A number of speakers talked about there being no silver bullet to solve the problem (as an aside, if I'm correct, silver bullets are required for killing werewolves: in this case, there was no need to kill exotic creatures, the normal monsters are enough to cause the problem!). There were some complaints that the industry did not know its own standards. A rather cool high-tech camera was presented which could see gas clouds down to some grammes of flow per hour.
Some of the language was very success-oriented: "Lesson's Learned Bulletins" were circulated, but there was not much evidence that the learning opportunities were really sinking in. A lot of people talked about bringing people to Spadeadam to see the real fires and explosions, but as we are very good at understanding consequences, and rubbish at understanding risk, this might just be scaring people for scaring's sake.
What worried me a bit was that we might be chasing the wrong monsters, or at least too many of the small ones. When I presented at All-Energy last year, one of the other presenters, Guy Boyd, had talked about how in injuries to personnel, only 21% of minor incidents could have the potential to be any worse than that (see http://www.all-energy.com/Health_and_safety_across_the_sectors.html). I don't know for sure, but I don't imagine valve seepage leads to a major accident event very often, if ever. It looked to me like the approach being taken was the classic "shrink all the triangle and we'll be fine" approach, but maybe that's why the numbers are stubbornly higher than wanted. I'll need to read a bit more into this, but it would be nice to have an array of approaches to predict and find the real big ones BEFORE they happen, and permanently drive the release frequency down.
The consensus seemed to be that the target is achievable, but not easy. I hope they're right.
