Modelling Under Risk and Uncertainty: An Introduction to Statistical, Phenomenological and Computational Methods

1.3 Industrial Safety, System Reliability and Probabilistic Risk Assessment (PRA)

Reliability and risk associated with industrial facilities are highly affected by their complexity, as those systems comprise numerous physical components organised into a wealth of functional chains. It may be hard to identify a modern elementary object which cannot be seen as a system in itself (e.g. a temperature captor); but think about the hundreds of thousands of such elementary pieces included in industrial facilities such as power plants, nuclear waste repositories or in industrial products such as airplanes, cars, and so on. SRA covers systems to a certain extent and has developed structural systems safety extensions to account for multiple-failure events such as those occurring in multiple-component mechanical structures (Ditlevsen and Bjerager, 1986). Nonetheless, practices in systems reliability have traditionally involved other approaches, denominated alternatively as Quantitative Risk Analysis (QRA), Probabilistic Risk Analysis (PRA) or even Probabilistic Safety analysis (PSA) in the nuclear sector (Bedford and Cooke, 2001). Similar to what was suggested in linking formally natural risk to structural reliability studies, subsequent sections will develop the theoretical continuum relating QRA to SRA, as already advocated in the literature, for instance by Aven and Rettedal (1998).

1.3.1 The Context of Systems Analysis

The overall context (Figure 1.3) can be analysed in a similar framework ...