254
15 Risk-Informed Decision Making
Consider:
p f = Probability failure
C = Cost of consequences of the event with p f
M = Mitigative investment, i.e., precautions taken to prevent the risk (p f * C)
The owner of the risk is considered negligent if M < P * C.
In other words, a judge may deem a company negligent only if mitigative moneys spent (per annum) are less than the annualized risks. Clearly, transparency and
rationality constitute a strong a priori defense in case something should go wrong.
Actually, going back to Fig. 15.10, it is interesting to note that the test of negligence
would induce the owner of the portfolio to invest in Dams 3 and 4 first and then in
Dam 1B, whereas Dam 1B has a fairly large part of intolerable risks. Examining
Fig. 15.9 it is evident that the probability of failure of Dam 1B is two orders of
magnitude lower than the p f of Dam 3. Despite this, the existence of intolerable risks
in Dam 1B would incite the owner to invest in Dam 1B before mitigating Dam 3.
Later we will look at the societal aspects of these failures and decide if this decision
meets the societal test.
Adding this rational and quantitative approach to risk prioritization brings value
to the discussion and adds “foreseeability and controllability” to management’s decisions. As a matter of fact, it is often impossible to act simultaneously on all required
mitigations. Thus, it might be necessary to prove that the efficacy and efficiency
were considered. Of course, the company would be the object of intense media and
regulatory scrutiny should an accident occur.
15.9 Defining Operational, Tactical and Strategic Risks
15.9.1 Operational Risks
Dam 4 can be mitigated—i.e., brought under tolerance—with minimal effort at the
operational level. A reduction of the probability by only 1.5 order of magnitude
would make it tolerable.
In the case of Dam 4 two options could be attempted:
• reducing uncertainties and thus decreasing the appalling category of the structure,
i.e., intensive monitoring, repairs, and possibly historic research and investigations
(with care taken not to trigger a failure)
and/or
• increasing the FoS by building, if space and other conditions allow, a berm. The
final choice should be carefully examined using probabilistic cost estimates (see
Sect. 15.10).
These alternatives should be carefully weighed against the negligence test, which
would require a 100 M$/year investment to pass. Clearly in the case of Dam 4,
Précédent

- 263/823

Suivant