Interesting Insights on Probability and Likelihood from National Grid

My attention was captured yesterday by an interview on the Radio 4 Today Programme with the CEO of National Grid.  In the course of explaining the background to the widespread power outages last Friday, he first described the almost simultaneous loss of two producers of this scale as “rare, unique” and later as “pretty unique”.  However, he also stated that it had occurred once before during his 28 year career at National Grid!

Even a naive analysis of 2 occurrences in 28 years would suggest an annual probability between 5 and 10% which would place it in the “Possible” category according to the UK Government’s guidance for likelihood scoring in Community Risk Registers.  A more sophisticated analysis would indicate that you could not rule out an underlying probability greater than 10%, which equates to a likelihood rating of “probable”.  Whichever way you look at it, it’s not “rare” (0.01-0.1%); and the UK Government guidance does not even attempt to quantify “unique”.

Whilst the above discussion may seem pedantic and theoretical, the language used to talk about probability does have real consequences in how we manage risk.  Labelling last week’s scenario as “pretty unique” appears to have resulted in a situation where, although there was a contingency plan to shed load to preserve the integrity of the grid (and this appears to have worked well); there was no accompanying communications plan to inform affected customers and the general public what was going on in a timely fashion.

Go to our Downloads section for more guidance on crisis communications.