Aleatory and Epistemic Uncertainties in PSA. G. E. Apostolakis, MIT, Cambridge, MA
This paper shows that the distinction between aleatory and epistemic uncertainties is, in fact, important and can make a difference in PSA (Probabilistic Safety Assessment). It is often assumed that this distinction is made for communication purposes only; while it is true that communication is facilitated, there are more important reasons for maintaining this distinction. This paper presents the general formulation of the problem in terms of the model of the world (which can be deterministic or probabilistic, i.e., aleatory) and the epistemic model, which quantifies our uncertainties regarding the structure of the model and the numerical values of its parameters. An example is presented that involves an aging structure (e.g., a section of pipe) in which failure occurs when the load (pressure) exceeds the capacity (the design pressure). It is shown that, for a deterministic model of the world, familiar quantities such as the failure rate can not be defined, even though there are epistemic uncertainties.
Go to . . .
1999 SRA Table of Contents
1999 SRA Author Index
Main Abstracts Menu Page
RiskWorld Home Page