BP Crisis Rational Analysis: What BP Did Not Perform?

BP Crisis Rational Analysis: What BP Did Not Perform?

Jun 8th, 2010


Now it’s time to test the Risk and Crisis Evaluation Methodology we published back in 2008 on the BP Gulf of Mexico spill. We originally presented the methodology in our book (F. Oboni & C. Oboni, Improving Sustainability through Reasonable Risk & Crisis Management, 2007).

Risk and Crisis Evaluation Methodology

Let’s start with the probability of the spill. At Riskope we:

  • are not oil experts, so can only assume that this accident, that various voices in the media declared as “almost impossible”, had a probability of occurrence of  one in a million or less.
  • mean the probability of the event before it occurred, as defined by some experts (BP’s or third party).
  • will not discuss the technical details leading to that value. We just assume it as the reasonable numeric translation of the upper bound probability of an “almost impossible” event.

1. As per foreseeable public relation consequences, a large magnitude spill scenario, following our simplified methodology (refer to the link above) would have certainly fallen in the “intolerable” are of public relation consequences.

2. As per the physical consequences, even if a company “rosy scenario” attitude would have lead to ignore possible casualties on the platform, long range biological and environmental effects, damage to the fishing industry, but had just used past large spills from, say the mining industry, the scenario would have lead to easily estimating direct losses and close indirect consequences at least 1B$, may be more.

3. A loss of share values in the order of 30% could also have been foreseen, as a minimum and, based on other industrial accidents, the maximum loss possibly going, if bankruptcy is avoided, as high as 90%. Past cases have also shown that duration of the share value loss can go as long as a decade.

Risk display, not in a 4×4 or 5×5 risk matrix

If we place now the “upper estimate” probability (one in a million is expressed numerically as 10-6) and the 1B$ “minimum estimate” loss on the probability-cost diagram (See red bubble starting at 1B$ at the bottom right of the Figure)

BP Crisis Rational Analysis

Results of a ERM study with the BP “minimal” bubble added

and compare the values with a “generalized large company” tolerance curve we have been using for the 2008 study, we can see that the scenario is intolerable even with these rough and oversimplified assumptions.

Let’s put it this way: when we encounter cases like this one (and we do in our practice while performing quantitative risk assessments for our clients) we certainly ask our clients to go way more into the details of the analysis.

Was there an explicit tolerance curve to compare risks to?

Of course, one can also assume that BP’s own tolerance curve (did they develop one?) is/was so much higher than the one we used (because BP is such a powerhouse) that the risk would have turned “tolerable” in their analysis. However, we do not know if they did perform an explicit/formal one.

To that we will answer that in this paper we have let aside many physical potential consequences (see above) and purposely neglected to add to the physical losses the reputational ones to express a bare minimal estimate of losses which would be difficult to argue against. If you go back a few paragraphs you will see that the public relation costs were deemed intolerable to start with, let alone all other ethical and human implications of such a catastrophic scenario.

Reality is now showing costs of consequences that are easily hiking well above 10B$.

Reality is now showing costs of consequences that are easily hiking well above 10B$. History will tell if the company emerges alive from this crisis. We note that banks emerged because of governmental bailouts. Will BP “dirty and oily business” be bailed out because of strategic vital importance? Or because many institutional investors (pensions) count on the company’s dividends?.

We know that, of course, the large flow oil spill in the Gulf o Mexico scenario is not the only one a large company has to consider when defining their mitigative prioritizations.

For example, the figure we are using using shows the result of a large company ERM study, which resulted in defining fourteen scenarios that were lying over tolerance (see How to Rationally Prioritize Risks).
Methods exist to properly prioritize risks and bring large industries into tolerable areas without breaking their financial backs.

One last consideration. Regulators can use the same type of rational reasoning to ensure that rules and guidelines are reasonable and cover areas that really matter to the public and our environment.


In 2007/2008 we presented a book and a paper with a Generic Phased Crisis Model and a Simplified Model of Public Reactions. One can use these to evaluate crises acceptability. We stresses the fact that such a simplified approach was fine to evaluate crises at a rather crude screening level. Finally, we noted that there are still companies that do not perform Enterprise Risk Management (ERM). Furthermore, among those who perform it, some use overly crude systems that blur reality either in too dark or too rosy tones.

Three points

In this blogpost we:

  • stressed that the future lies in using Risk to weigh decisions, rather than just guiding mitigation.
  • showed a simplified form of Quantitative Risk Assessment applied to world-wide famous cases. Furthermore we explained what useful metrics to measure the “Enterprise disservice” of a company can be. For example, the Share Value drop, which, of course, compounds with direct and indirect costs of such a crisis. We also explained that other metrics are possible. In this case, for example: Loss of habitat (km2); Loss of species (biodiversity, number of species); Loss of human lives (number of casualties); One can actually derive any complex metric for example by using “multiple portfolio” analysis.
  • wrote that it is possible to derive a priori Estimates of the probability of the disservice. Of course, in terms of “framing ranges” by using simplified approaches from our book. An example has also been published (See Example).

Finally we compared the risk. In other words we compared the probability and consequences of a mishap to a “large company” risk tolerance curve. The idea was to show how significant the crises/risks were/had been for the paper’s case.

One last question

Was BP, like many other players world-wide, a victim of the syndrome we described in this post?.

Will humans ever change?

Tagged with: , , , , , , , , , , , , , , ,

Category: Consequences, Hazard, Mitigations, Probabilities, Risk analysis, Risk management, Tolerance/Acceptability

21 responses to “BP Crisis Rational Analysis: What BP Did Not Perform?”

  1. Hi Franco,
    Nice approach.
    Yesterday I was discussing with a dam consultant the risk criteria and matrix that XXX(confidential client name deleted) uses for its dams. In some cases high tailing dams located over downstream inhabitanted areas, possible with very high consequence costs on a rupture, are considered of medium risk. Potential client
    Thanks for the insight

    • foboni says:

      You made my day! Thank you for your comment.
      That’s precisely what we teach in our courses: by using “silly” qualitative methods, industries are making big, and potentially extremely costly, mistakes.
      Only wisely applied quantitative approximate methods allow to pinpoint which risks really matter.
      At the end of the day every one wins if things are done properly!

  2. Simon Cooke says:


    A very clear approach and one which, clearly, has much merit. it is effectively the same approach we offer our clients when dealing with high explosive events (low portability but potentially extremely high consequences).

    I do find that many clients still prefer their own “gut feel” and it is sometimes hard to convince them using pure logic. Good luck with taking this forward.

    Best regards


  3. Peter Whalen says:

    Hi Franco.

    I believe your analysis and discussion of the various methodologies and metrics are highly astute and insightful and, reflect very favorably on your book, papers and firm.

    I would be interested in knowing what your thoughts would be on government’s role in this incident and, related risk analysis/management approach. There is opinion, for example, quite valid in my view, that this risk exposure in the Gulf began with the current administration’s limiting new oil drilling to deep water, vs. on land or shallow water. This in my view is a throw-back to the oil shortage and long lines at gas stations in the U.S. during 1970’s. This shortage the case can be made was due to oil companies becoming so disincentivized to drill due to the related tax burdens, they in effect became importers of oil from Arabian countries, who thereby gained tremendous price leverage and control via OPEC, placing US consumers and the nation as a whole at great economic risk.

    Other gov’t related risk, getting back to the present fiasco in the Gulf, can be attributed to a laxity in the responsibility on part of administration in managing risk through the enforcement of oil rig construction permit rules. Should there be more communication and mutual risk management between private secotor entities (oil companies in this case) and government entities for such large scale ventures? Should any degree of absence of such communication and cooperation be considered a variable in a risk analysis methodology?


  4. Gareth Igloliorte says:

    I think the simple fact of the matter is that you have underestimated the risk tolerance of BP. Blowouts are rare, but certainly not uncommon. The probability of the blowout is probably higher than 1e-6, as they have occurred in the past and mitigation options for oil spills have been of historically limited effectiveness.

    Its quite possible the financial reward of the Deepwater Horizon well was worth the risk of the company going bankrupt and being bought up by its competitors.

    The real problem with low probability-high consequence events is demonstrating the costly engineering solutions are cost effective and not simply a wast of money, as there is generally very little hard data to support implementation. That is a real issue with this area of decision making.

    • foboni says:

      Dear Gareth,
      Thank you for your comment, but allow me to clarify a couple points.
      We did not even attempt to estimate the “real” risk tolerance of BP, we used instead the risk tolerance of other clients (way smaller) to build an extreme case.
      Likewise, we encoded the “almost impossible” definition given by BP as 10-6 and never suggested we had evaluated that value as the “real” probability.
      We also said from the beginning that from a reputational/image point of view the case was intolerable.
      I can hardly see how a scenario with such a heavy ethics and human loading could be considered as a good bet, or rewarding enough.

  5. Dear Franco,

    Very intresting article.

    A question: if the chance of a failure in a blow out valve is that small en “normal” quantitative and qualitative risk management precautionairy measurement were taken, hasn’t BP forgotten to minimize the enormity of the impact by creating a workable contingency plan. Such a plan set in action reduces the public relations damages by showing that your on top of things. By deploying counter measures against the oil spil right away you could have prefented a lot of the damage.

    In short when switch from precautionairy measures to damage control?

    Best regards,


    • foboni says:

      Emergency procedure, crisis management and risk management should always go hand in hand.
      At least, that’s our credo, and we teach in our courses/book.
      From what I have read in the media (of course I have no access to the original documents), emergency plans for that operation were miserable boiler plate solutions, written by people who did not even care to know that sea lions are estinct in that area since some ice age ago…so emergency plans were poor, crisis management has been poor, risk management we have already dealt with….

      Sorru for the late reply, but I have been very busy dealing with decision making support for some of our clients.

  6. oil spills can really mess up the environment, i hope we can find a very good solution to control oil spills .-“

  7. […] BP Crisis Rational Analysis: What BP Did Not Perform? June 2010 9 comments 3 […]

  8. Track Jacket says:

    ”: I am really thankful to this topic because it really gives great information ‘*;

  9. […] discussed in our BP case study adding to this vague standards the test of tolerability clearly brings value to the discussion and […]

  10. Guy Cofer says:

    I trust you and it absolutely seeing guide number of people.

  11. Hannah Sarmiento says:

    Saved your blog. Thank you for expressing. Absolutely really worth the time clear of our reports.

  12. Organic Food says:

    Thank you for this wonderful report. It was very helpful to me and to the one who read this page. You truly is a genius writer!

  13. […] We have already discussed many times how well-balanced and sustainable decisions can only be taken if risks are compared to properly defined risk tolerability/acceptability criteria. […]

  14. Alfred says:

    I’m continually searching online for suggestions that can help me. Thank you!

  15. Barclay says:

    forums who go on the exact same topics? Thank you!

  16. […] the means to embrace rational and scientific approaches to Risk and Crisis Management using clearly defined tolerability criteria, the only way to be a winner rather than a loser…We will not discuss banks and governments, […]

  17. […] and low likelihood/catastrophic-consequence scenarios (Fukushima anyone? Or BP). 2) The second step is to divide reality in discrete classes, let’s say five “steps” of […]

Leave a Reply

Your email address will not be published. Required fields are marked *

Riskope Blog

  • Integrated and convergent risk approaches need some discussion. Are they so different? And what about the good old Enterprise Risk…
  • Read More

More Info

  • Get in Touch
  • Learn more about our services by contacting us today
  • t +1 604-341-4485
  • +39 347-700-7420

Vancouver Digital Creative Agency Ballistic Arts Media Studios.