Probabilistic risk assessment for structures

The existing safety culture within structural engineering has been highly successful and is likely to remain so. However, it is argued herein that it is no longer completely sufficient. It lacks the ability to interface adequately with risk assessment procedures used for other facilities in society. Such procedures require complete analysis of all sources of risk and of the various potential consequences. These are consequences to society rather than simply ‘structural failure’. Moreover, the probabilities involved are estimates for expectations rather than the nominal probabilities used to calibrate limit state design codes. The theme is illustrated with two examples: one dealing with the safety assessment of an existing road bridge and the second with potential structural fatigue failure of a large pressure vessel for liquefied petroleum gas storage. These examples show that structural safety cannot, for major structures, be divorced from more general safety analyses. Some comments are made also about the World Trade Center and related risk rationality.