A survey of research in hazardous materials routing

This tutorial session will survey research done in the area of routing vehicles that carry hazardous materials. The mode of transportation can be by rail, truck or ship. The code of Federal Regulations [1] has a special chapter (49 CFR Parts 100-177) for HAZMAT listings, HAZMAT transportation, and oil transportation and pipeline safety. Approximately 1.5 billion tons of hazmat (of which 65% are carried by truck and rail) is being transported yearly in the United States. Worldwide generation of HAZMAT is estimated as 3-4 billion tons per year. Routing is a critical factor to consider in HAZMAT logistics. Although the fatalities due to hazmat-related traffic accidents are almost negligible as compared to the deaths in ordinary traffic accidents, HAZMAT risks are considered unacceptable. The main objective of HAZMAT routing is to determine the optimal path(s) for routing the HAZMAT on a network subject to certain criteria. The objective, which can be either a single or multiple criteria, is typically based on risk, equity and cost considerations. The choice of objective criteria will highly influence the selection of the "best" route. These decisions are plagued with the NTMBY (Not Through My Backyard) syndrome. We consider routing of the HAZMAT truck with the objective to minimize the "risk" involved in the transport. Risk is represented in different ways in the literature. It has been modeled as expected consequence [5], or the population exposed to consequences due to impact [6], incident probability [7], or conditioned on the probability of the first incident [8]. Many models for HAZMAT routing have been proposed. The objective of this tutorial is to introduce and discuss this new application area and to provide a synopsis of the major research over the past two decades. Several special issues of major journals and special sessions/tracks at conferences have focused on this topic, and a large number of researchers from around the world are actively involved in research in this field. The area is particularly interesting because it requires precise modeling of the inherent probabilistic events that lead up to a HAZMAT accident as well as optimization methods for route selection.