Facility location and reliable route planning in hazardous material transportation

This paper considers integrated location and routing models for minimizing the expected number of hazardous material transport accidents. Two different routing policies are considered: (1) most reliable route planning and (2) multiple routing with random selection. Path reliability measurements are used to derive the expected number of accidents over a given planning horizon, where reliability refers to the probability of completing a journey without accident. Corresponding to the two routing policies, two location models are introduced: the reliable 1-median and a location framework for considering multiple routes. Based on realistic assumptions for assigning accident probabilities to segments created when an edge location is considered, vertex optimality results for both problems are obtained. An example to illustrate reliable routing and location in Massachusetts is provided.