Oroville Dam risks became unmanageable?

In the case of Oroville dam, the population exposed to the potential dam failure has increased significantly. Normalization of deviance increased the probability of failure. So is it possible that Oroville Dam risks became unmanageable because of that?

Oroville Dam risks became unmanageable?

When looking at risk mitigation the old saying is that you can either lower the probability or lower the consequences. That statement requires some explanation. Indeed what mitigation alone can do is exclusively to reduce the probability of failure. In the case below this will not change the blue area.

The way to reduce consequences (i.e. the damages within the blue area) one has to change the system. In the case of Oroville Dam one can reduce the probability of failure (of any part of the system)  by reinforcing, mitigating, etc. HOWEVER, reducing the consequences may require, for example, moving population and infrastructures away from the blue zone.

Responding to a California Public Records Act request by NSPR, state officials released a 2006 map . They revised it in June 2016 showing the extent of flooding if Oroville dam failed with full reservoir.

The recurring question

Clients regularly ask us to advise if a risk is manageable or not. So we thought we could share our thoughts and analysis process to answer this question.
The question is actually more complex and should sound like this: Which risks are:

1. tolerable?
2. intolerable but manageable, thus mitigation occurs by reducing the probability of failure?
3. intolerable and unmanageable and hence require strategic shifts (altering the system)?

At inception of the dam, the state of California provided specs for the design of the dam in the form of the FoS which indirectly defines a likelihood of failure pf. But in those times, as still today, no one established the FoS-pf relationship .

The dam was therefore considered safe, without asking any further questions. Today, if an engineer evaluates a probability of failure and estimates consequences of a failure, he can compare the risk to a selected tolerance threshold. That threshold can be regional, national, etc. NB: the idea is not to allow for nefarious accidents to occur, but to evaluate if risks are, for example, higher than those normally accepted by a society.

The reply to the questions above is possible using appropriate risk analysis techniques.

The potential exposed population grew

Mitigations aiming at reducing the likelihood of failure or increasing the overall FoS are probably inconceivable, from a financial point of view. They are beyond the present repair works plan. Those, by the way,  do not influence the overall system, but bring it back to a “point in history”. Even their feasibility is uncertain, given the size of potential damages or damageable infrastructure.

Note that, in the aftermath of a major catastrophe, especially the nuclear ones ( such as Chernobyl or Fukushima), people are evacuated and the evacuation can last a few years or “at perpetuity”. Such a move, which is a “strategic shift”, is very hard to justify as a preventative measure.  Certainly any authority that would impose it would not remain popular during next elections! Small scale, well documented cases of forceful evacuations have occurred around the world. That documentation often includes legal actions against the authorities/decision-makers.
So, in the case of Oroville Dam it seems that the only other possible strategic shift could be to permanently reduce the stored volume of water.

Tools exist

Did Oroville Dam risks became unmanageable? Well, we have not performed an analysis. But it seems that the state of California did that and the number of available options are rather limited. The selection of alternatives requires an attentive comparative risk assessment as provided by ORE and cradle to grave service/maintenance risk informed cost analysis such as provided by CDA-ESM.

Tagged with: Intolerable, Manageable, Oroville Dam, Tolerable, Unmanageable

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