An interactive model for routing shipments of dangerous goods through an urban road network is presented and demonstrated. The modelcomputes minimum-risk routes based on each shipment's origin and destination, and graphically illustrates the selected paths. Alternative approximations for estimating risk are considered. The consequences of these differences are compared against routings based on objective risk exposure. Objective risk exposure is predicted by using observed accident rates, a fault-tree analysis for estimating damage potential, and various damage propagation relationships. The resultant risk estimates in this model are responsive to various environmental conditions, material properties, and location-specific parameters. An application of the model to the routing of chlorine shipments within the metropolitan toronto road network is presented. The application illustrates the sensitivity of minimum-risk routes to a range of external contextual variables and relationships. Route patterns appear to be strongly influenced by the nature of the risk measures applied to candidate links and nodes. This paper appeared in transportation research record no. 1148, Hazardous materials transportation, work-zone traffic control, and traffic marking materials. For covering abstract see IRRD no 818401.
Abstract