ROUTING AND SCHEDULING OF HAZARDOUS MATERIALS SHIPMENTS: ALGORITHMIC APPROACHES TO MANAGING SPENT NUCLEAR FUEL TRANSPORT (MULTIOBJECTIVE, SHORTEST PATH, CURFEWS)

Much controversy surrounds government regulation of routing and scheduling of Hazardous Materials Transportation (HMT). Increases in operating costs must be balanced against expected benefits from local HMT bans and curfews when promulgating or preempting HMT regulations. This thesis describes algorithmic approaches for evaluating HMT routing and scheduling regulatory policy.;These algorithms are applied to the case study of U.S. highway transport of spent nuclear fuel from reactors to permanent repositories. Two data sets were used. One data set included the U.S. Interstate Highway System (IHS) network with reactor locations, possible repository sites, and 150 "heavily populated areas" (HPAs). The other data set contained estimates of the population residing within .5 miles of the IHS for the Eastern U.S. Major results were: (1) For Eastern reactors and repositories, alternate efficient routings exist which substantially reduce the resident population exposed to truck transport of spent fuel with modest mileage increases. (2) A small number of HPAs are exposed to much greater spent fuel flow than most others. Restrictions on spent fuel flow through these HPAs can be imposed with little systemwide ton-mileage increase. (3) If all HPAs imposed "reasonable" peak-hour curfews on spent fuel shipments, travel times increase less than 15%. Curfew delay is dramatically reduced by optimally scheduling departure times unless inter-HPA travel times are highly uncertain. Rerouting shipments to avoid HPAs is a less efficient approach to reducing delay.;A review of current U.S. HMT regulatory policy is presented to provide a context for the analysis. Next, a multiobjective shortest path algorithm to find the set of efficient routes under conflicting objectives is presented. This algorithm generates all efficient routes under any partial ordering in a single pass through the network. Also, scheduling algorithms are presented to estimate the travel time delay due to HMT curfews along a route. Algorithms are presented assuming either deterministic or stochastic travel times between curfew cities and also possible rerouting to avoid such cities.