Monday, July 19, 2010

On the eve of British PM Cameron's visit with President Obama, BP has a plan. Again.

It's deja vu all over again. On the eve of the David Cameron's state visit and meeting with President Obama, BP has good news, same as it did the last time they met a few weeks ago at the G20 meeting in Canada.

This time, it's that the "seepage" found recently (sounds so much nicer than "leak" or "spill") is unrelated to their well. And that BP is considering permanently sealing the well they just capped with a so-called "static kill." The New York Times only had to use three qualifiers to describe the latest BP plan as their spokesman described it -- "could," "could" and "might." By BP standards, that's almost a sure thing.
He said that the procedure could speed the process of sealing the well and that the digging of a relief well, which has been seen as the ultimate solution and could be completed by August, might be needed only to confirm that the technique had worked.
Others are not so sure and worry that with their capping "experiment," BP is once again flirting with disaster for short term gain. Dimitry at The Oil Drum calls BP's constant resetting of the goalposts the "normalization of error."
But hey, if you really, really want the results to be good, you will just have to assume that these things are really responsible for the low pressure and not some prosaic, unpleasant leak.

So, to sum up, we get a bad result, rewrite the agreed upon rules, come out with a number of unverifiable assumptions backed up by uncheckable analysis and declare the actual bad results to be "almost certainly good". We then proceed AS IF we got an actual GOOD result in the 8+ ksi range. We proceed as if we actually established well integrity - this is normalization of error. It is the most often responsible for really bad errors that result in disasters - the two space shuttle disasters come to mind, underscoring the possibility that organizations can repeat the same methodological mistake twice.

The really weird thing to me, is that there is a spectacular lack of realization that an analogous process led to BP's original blow out.
Why do I get the feeling that Obama and Cameron will not be talking about "normalization of error"?

No comments: