General approach to risk optimisation of road bridges exposed to accidental situations

摘要

Road bridges in accidental situations represent the most vulnerable components of transportation networks. Hazard situations may be caused by natural or man-made hazards such as floods, terrorist attacks or impacts of vehicles. Decisions concerning selection of protective measures can be based on the probabilistic optimisation of expected costs. In general, total structural costs consist of initial cost and consequences due to structural damage and failure. The initial cost comprises of construction cost including cost of protective measures; the failure consequences account for direct costs related to structural failure and indirect costs due to malfunction of transportation network. The paper proposes a general framework for the risk assessment based on Bayesian (causal) networks. The cost optimisation requires data on probability of occurrence and magnitude of selected accidental actions that often need to be based on expert assessments and judgements only. The present study indicates that the probabilistic cost optimisation in conjunction with Bayesian networks provide an effective tool for decisions concerning appropriate protective measures for bridges endangered by various types of hazards.