Impact of accident precursors on risk estimates from accident databases
Dealing with major accidents implies that these events have in common the potential to affect many people. For this type of event, societal risk curves in the form of cumulative frequency distributions of multiple casualty events are often considered appropriate measures to numerically express the degree of risk. Using data from the European Commission's major industrial accidents database MARS, the conventional societal risk estimater associated with certain types of major accidents notified and related numbers of fatalities and injuries is quantified. Bayesian calculations are performed to analyse the sensitivity of such risk estimates with regard to incorporation of accident precursor events. From that, some more general conclusions on a meaningful use of industrial databases as well as on the definition of a major accident event are drawn.
Bibliographic Reference: Article: Reliability Engineering & System Safety (1996)
Record Number: 199611391 / Last updated on: 1996-12-18
Original language: en
Available languages: en