Dealing with epistemic uncertainty in risk-informed decision making for dam safety management

Abstract

[EN] In recent years, the application of risk analysis to inform dam safety governance has increased significantly. In this framework, considering explicitly and independently both natural and epistemic uncertainty in quantitative risk models allows to understand the sources of uncertainty in risk results and to estimate the effect of actions, tests, and surveys to reduce epistemic uncertainty.

In this paper, Indexes of Coincidence are proposed to analyze the effect of epistemic uncertainty in the prioritization of investments based on risk results, which is the key issue in this paper. These indexes allow consideration of the convenience of conducting additional uncertainty reduction actions.

These metrics have been applied to the prioritization of risk reduction measures for four concrete gravity dams in Spain. Results allow for a better understanding of how epistemic uncertainty of geotechnical resistance parameters influence risk-informed decision making.

The proposed indexes are also useful for probabilistic risk analyses of other civil engineering structures with high epistemic uncertainty environments, since they analyze whether existing uncertainty could have an impact on decision making, outlining the need for extra studies, surveys and tests.