The people reporting on it all seem to use the exact same sentence without context around it, it becomes rather unclear if this 97% might mean there were, say 100 incident at refineries - it is just oil refineries - and 97 of them just happened to be all at the same place essentially tied to the same thing, like most incidents result from people normally running way under the reg because the regs are bizarrely unrealistic and they just tighten down on the day the inspector is coming by. I know for a lot of security inspections military units will come off really craptacular if they don't get advance notice and aren't normally a group expected to run mega-tight, then get flagged for 'lock on one gun rack [inside vault with two inch steel door] not specifically of the type mandated by security change from last month' and get 12 violations because of having twelve racks with the wrong lock. I'm not sure if a similiar case applies but I'd hardly be surprised if they got all most all of those violations of one or two connected incidents, like the safety supervisor was sick that week and the manager came down to the floor and asked the foreman 'we're running in accoradance with the safety regs right?' and the foreman blithely says 'sure, every things safe' and the inspector comes buy and says 'Hmmm.... widget X is required to have 2.0-3.5 liters of lubricant, you're at 1.9 on all of them, you have 97'. So while it could be some major and vile act of negligence here, it suddenly surfacing this way does make me wonder if the smell of rotten fish might be in the air.

Anyway, I've looked up the OSHA Factsheet on BP, and it list violations, starting with this omnious sounding one:

- 411 instances (New violations)

- Standard Violated: §1910.119(d)(3)(i) and §1910.119(d)(3)(ii) grouped with §1910.119(j)(5)

Okay, I've looked these up:

1910.119(d)(3)(ii)
The employer shall document that equipment complies with recognized and generally accepted good engineering practices.

This one is the only short one that does involve a lot of cross-referencing on that same page (if you want to read the whole thing just google one of those three reg numbers, same page). There are, apparently, 411 incidents of this, of course this could - they don't say, all have been from someone inspecting 412 of 412 identical widgets and only filing documentation that they met the standard on one, because the engineer assumed he only had to file one piece of paper saying that yes, all the identical widgets have the same design which meets standards. Government regs can be obnoxious like that, of course it could be something worse. Whatever those 411 violations were, Proposed penalties: $28,770,000. LAter they get cited for 242 instance of not having 'standard for safety instrumented systems' with a 50 million dollar fine, and 28 instance of 'Failure to perform relief device studies' to the tune of 5 million. So although all of those instances count as equal for the 97% figure (I gather) of 709 instance, which if it is the source of this 97% figure, would imply 731 instances in refineries during that time. Considering how many regs could be violated (pages and pages all sounding about the same in terms of chance of it happening, to me at least) my guess is that they basically had 3 violations of the same type, categorically repeated. None of those appear linked to the accidents later cited - again, to me, I'm not an expert in this area. Considering the vast majority of those were for one violation, getting fined at 70k a pop, whereas the 28 one 180k a pop and the 242 at about 210k a pop I'm guessing it is probably not exactly accurate to say each of those is 'equally bad', yet with them being counted the same for all I know Exxon might have had three errors, accounting for less than 1%, on some major parts that seriously endangered lives and got hit for 10M a pop.

Not trying to excuse these, they really could be bad, but when someone starts tossing of 97% that way when we've got so many reifneries all run by people who presumably roughly parallel BP in behavior, and they only reference a without context statement from OSHA, I have to wonder, because it seems if - as everyone blogging on it is implying, 97% of instances implies 97% of actual real tangible negligence and screwup, someone's damn head should be rolling by now at OSHA, and all the way up the chain in the DoL to Hilda Solis, she's been Sec of Labor for well over a year now, and I don't think anyone can claim underlings would have been afraid to approach an Obama nominee about what would otherwise appear to be catastrophic levels of negligence by Big Oil. As this hasn't happened, I would be inclined to say, for the moment, that someone just shuffled through data until they found a particular zone that BP appeared especially bad in, and handed it to someone who wanted a pointy hat for their witch trial. Again, for all we know instances like this could be 'Used unapproved screw to hold a plate cover, 100 screws, 100 instances, scrws required to be phillips head, not flat' while next door at other refinery it was 'Safety Operator required to have MS in engineering, operator found to be high school dropout, 1 instance', and they rack those up and company X get's 99% of safety violations.