Vulnerability of road networks

Current evaluations of the vulnerability of a road network tend to focus on the probability of damage and the change of traffic demand. The forecasting of low-probability but high-consequence events is a major difficulty. In this paper, a new theory, using a systems-thinking approach, for examining the vulnerability of the form of the network is presented. Our purpose is not to simulate traffic flow but to identify high-consequence scenarios that may arise from vulnerable weaknesses in the form of the network. Such scenarios are independent of models of traffic demand or the source of the damage and can subsequently be combined with specific demands to assess risk. A hierarchical model with clusters of road circuits formed at various levels of granularity of a road network is developed for use in a search process. Only free uncongested flow is considered. A search algorithm for finding vulnerable failure scenarios is described. A vulnerability index is proposed as a measure of the disproportionateness of the consequences of a series of events within a failure scenario in relation to the damage causing those events. The theory is illustrated with two examples.