Modeling and analysis of uncertainties for risk-informed decisions in infrastructures engineering

TLDR

Engineers must make decisions under uncertainty, and in infrastructure engineering design codes aim for safety and reliability despite incomplete information, distinguishing aleatory randomness from epistemic knowledge gaps. The paper proposes a framework for properly modeling and treating aleatory and epistemic uncertainties in risk‑informed engineering decisions. The framework is illustrated through applications to bridges and offshore marine structures.

Abstract

Engineers deal with uncertainties in all their activities, and must often make decisions under conditions of uncertainty and risk. Infrastructures engineering is no exception—design codes are developed to ensure a desired level of safety and performance, or to ensure a specified operational life with a prescribed level of reliability; the required decisions must often be formulated without complete information and thus contain uncertainties. In considering uncertainties, it is important to recognize two broad types; namely, the aleatory type which is associated with natural randomness and the epistemic type which is associated with imperfect knowledge. Proposed here is a framework for the proper modeling and treatment of each type of uncertainty in the formulation of risk-informed engineering decisions. The concepts are illustrated with applications to bridges and offshore marine structures.