Egregious Citations Issued to BP

Posted to Data Sources  |  Tags:  |  Nathan Yau

BP processes about 1.5 million barrels of crude oil per day, across six refineries in the United States. In total, 150 refineries in the United States process just under 18 million barrels per day, so BP processes about 8.5 percent of it. However, as reported by the Center for Public Integrity, 97 percent of the most dangerous violations found by OSHA were on BP properties.

There are several levels of citations. The most dangerous are egregious willful and willful, in that order. Egregious citations are flagrant violations and willful citations are intentional disregard for employee safety and health.

Between June 2007 and February 2010, there were 761 egregious citations issued in total. All but one of them were issued to BP. In the same time period, 69 of the 91 willful citations also belonged to BP.

Most citations for other companies were in the serious category, meaning there was “substantial probability” of death or serious injury. There were 1,521 serious citations issued to non-BP refineries, versus BP’s thirty. However, if we take into account barrels of oil processed by the rest, we get an even worse image of BP. Really, it’s bad no matter how you look at it:

Grab the data here to take a look for yourself.


  • If our business and government leaders were required to study the fundamentals of data and information visualization + analysis, serious problems would be prevented, or at least minimized.

    The visual is easier to understand and remember. The descriptor paragraph, alone, is not enough.

    Thanks for sharing this.

  • David Shea June 6, 2010 at 10:07 am

    I agree with Lynn, it would also be interesting to see the data normalized for volume of production, citations per location.

  • Current spill notwithstanding, the cynics amongst us may just suggest that a fine upstanding British company just wasn’t prepared to pay off the inspectors like perhaps the other companies did.

    Of course that would be an absurd allegation, had it not been for the recent revelations about Massey’s behavior prior to the Big Branch tradegy.

    • Alex,

      I’d be inclined to agree with you but it turns out that BP, despite its British origins, has been very well linked in with American politicians and regulators. The headlines are all trumpeting its connections with Obama and the Democrats (i.e. Rahm Emanuel’s apartment) but the Minerals Management Service under Bush was hardly a model of effective regulation either. Even today, the government is helping to shield BP even as Obama talks about “kicking ass.” See this NYTimes article today:

      So, no, I don’t think that it is true that BP just didn’t “pay off” regulators like the others did. I think they were playing with the big boys all along.

  • Dave shea June 6, 2010 at 10:40 am

    Data Bias? Another great point, Fundamental to presenting data clearly is providing accurate data!

  • I think it’s super important to consider how citation opportunities might differ between companies, refinery types, etc. in making these comparisons. Perhaps not all violations are equally likely to result in citations being issued?

    Differences in man-hours, HR policies, etc. might bias the use of “citations per barrel” as an informative measure of company safety.

    • I think your point would be valid Paul, if the numbers were vaguely close, but I can’t think of any normalizing factors that would make up for the massive difference in the worst category of citations

  • True – I suppose you could actually check and see what the difference would have to be in order to create such a gap.

    • Probably the best thing to do is go through the raw OSHA data, but I had trouble navigating the site and data and lost patience. So I settled for the numbers aggregated by the Center for Public Integrity. Maybe someone else will have better luck.

  • was about to nitpick on “more than almost”, but see you’ve fixed it.

  • With this data it is so clear the government shares the blame with BP.

    So many citations and the government does not do anything. And still gets to blame BP???
    No one should be allowed to do that.

  • Open question: does it matter that the BP disaster was at a production site, and the data are about refineries? On the one hand, attention to safety might be presumed to be similar across different parts of the firm. But on the other, this still doesn’t speak to BP’s practices in running its production — as opposed to refining — facilities. Presumably the underlying claim is that the spill could have been prevented if BP were not reckless and negligent, but these data on refineries don’t directly support such a statement.

    Relevance of data is as importance as accuracy of the data.

  • Well, I think we all agree that this data analysis is interesting and warrants further study. The other thing on my mind is the possibility of a “small numbers” issue here.

    One comparison could be the US airline industry. The FAA produces a large number of warnings, levies some moderate number of fines, but there are relatively few catastrophic accidents. Just because an airline received a large number of warnings or even a large fine does not necessary translate to the number of catastrophic incidents (which is a very small number).

    Is the same true for the oil industry? Don’t know the answer! However, if someone wants to fund an independent agency/thinktank to look at these types of issues, I would gladly work there.

  • Is it me, or is there not a key for the gray and green bars? It took me a bit of reading to figure out what they represent.

    Also, would it be more representative to normalize the bars to reflect the number of refineries in the first plot? For example, X% of BP’s refineries were cited; y% of all other refineries were cited for violations.

  • A critical point here is comparison to production rates, rather than just dollars. The risk exposure is the same regardless of the price per barrel. The exposure is also the same regardless of whether you have people working part- or full-time.

    What isn’t noted here is that these might not all be from US-OSHA (as opposed to state-programs), and the enforcement schedules in different regions are likely not identical. So, yes, the citation opportunities might have been different with different companies. And a company is more likely to get a repeat inspection after having problems, increasing the citation possibilities.

    @ Paul: No, not all violations are equally likely to be cited. Inspections can be focused on a specific topic, limiting the likelihood of some errors being noticed. You might be inspected by construction, safety OR health branches. Any one of these inspectors can cite you for anything, but they aren’t all equally likely to notice the same details outside of their personal expertise. BUT – since this data is all for the same severity-type of citations, I think this is a normalizing factor of sorts. These are the sort of things that should leap out and scream “cite me! cite me!”.

    And, Citation severity isn’t related to any internal corporate factor. It is based upon whether the company meets the standard, whether it is a repeat failure, and how grave the immediate danger to the employee is.

    The fact the data are from refinery, rather than drilling, operations will effect which standards are applicable. If I continue to procrastinate about my thesis today, I’ll check the OSHA website. I’m really curious as to which standards were cited. I can persuade myself it’s related to my research, maybe? :)

  • Interesting… If you eyeball total number of violations, egregious and otherwise, BP probably makes up about 40%-ish of total violations. My first thought as a layperson was “that’s not so bad,” but now that I consider it, it’s pretty awful.

  • John Jane Something or Other June 10, 2010 at 3:05 pm

    Holy crap guys, look at serious violations and then look at willful & egriogous. QQ O.o

  • This graphic was included in a collection of oil spill infographics. I thought the collection was actually really cool – just interesting to see lots of infographics developed on a common topic:


Reviving the Statistical Atlas of the United States with New Data

Due to budget cuts, there is no plan for an updated atlas. So I recreated the original 1870 Atlas using today’s publicly available data.

How We Spend Our Money, a Breakdown

We know spending changes when you have more money. Here’s by how much.

10 Best Data Visualization Projects of 2015

These are my picks for the best of 2015. As usual, they could easily appear in a different order on a different day, and there are projects not on the list that were also excellent.

Years You Have Left to Live, Probably

The individual data points of life are much less predictable than the average. Here’s a simulation that shows you how much time is left on the clock.