Home > Uncategorized > Deep Horizon and high-impact, low-frequency events

## Deep Horizon and high-impact, low-frequency events

from Lewis L. Smith

There is an interesting calculation which one may make with regard to the Deep Horizon platform, which incidentally is located in an “oil play”  called “Macondo” !

Following a major spill off Mexico, some 5,000 deep-water wells were drilled in the Gulf of Mexico without serious mishap.

Suppose suppose that you are the BP official in charge of Deep Horizon and are about to order the start of drilling.

You say to yourself that you should assume the worst, just to be on the safe side, and estimate that the maximum damage to be expected from a major spill would be \$50 billion.

[Actually total damages from the latest spill may well exceed this, as when one takes into account the present value of future earnings foregone from fishing, tourism et cetera.]

Now \$50 billion divided by 5,000 is \$10 million.

In other words, as hypothesized above, the mathematical expectation of the anticipated maximum disaster is the same as an absolutely certain disaster which causes \$10 million in damage.

Did BP save \$10 million or more by cutting corners in spudding Deep Horizon ?

I doubt it.

For example, note that \$500,000 per day would be on the high side for a rig rental. So if the primary saving anticipated was in terms of drilling time, BP would have had to expect a saving of at least 20 days, just to break even on the basis of the above. I doubt that they anticipated saving this many days.

So just some back-of-the-envelope calculations are sufficient to conclude that BP acted in a foolish and reckless manner.

High-impact, low-frequency events

Making judgment calls about such high-impact, low-frequency events is one of the most difficult tasks faced by managers. Historically attitudes have ranged across the spectrum  —

Private- and public-sector emergency planners discussed the possibility of an Exxon Valdez type accident a few years before it happened. But they decided that it was too improbable to plan for.

On the other hand, listen to the manager of Las Vegas’s water supply, 90% of which comes from Lake Mead. He has this to say about scientific disagreement over the future of this body of water. [It might even dry up !]  —

“… who care what the probabilities are ?  Can you say it’s impossible ? [No !] And if it’s not impossible, you’d better have a contingency plan.”
[IEEE Spectrum, June 2010, p. 35.]

By way of contrast, peaks in world oil and gas production are not improbable events. The first is highly likely by 2015 and the second, by 2025, if not sooner.

So you would expect that both Puerto Rico and the Federal government would have their contingency plans all ready. However, to this writers knowledge, there are no such plans.

If you hear of any, let him know !