(572b) A Novel Dynamic Routing Model for the Transportation of Hazardous Materials Under Uncertainties and Potential Shipping Delays

Nowadays, various kinds of chemical products are daily produced and shipped out to satisfy the expanding need of modern society and life. These chemical products include not only fuel, gasoline, nature gas and propane; but also lots of intermediates, like acid, alkali, alcohol and polymers. All these chemicals are classified as hazardous materials (Hazmat) due to their dangerous properties, such as flammability, corrodibility, toxicity and even explosiveness. Any improper operations can result in extremely serious consequences to people, properties, and environment, and even leave over long-term adverse effects on health and safety. As a key stage of their life cycle, the transportation of Hazmat is very important and critical, since it has more direct and extensive impacts on human society and our environment. Therefore, great attentions have been drawn to reduce potential risks and incidents.

In previous research, different models have been developed for the routing and scheduling of Hazmat to provide safe and efficient transportation paths. However, most of works are not comprehensive, since they either employ single-criterion objective function, or ignore the concern of real-time changes and any uncertainties.[1,2] Although some studies did consider cases with uncertainties, or utilize time-dependent variables, models were solved with heuristic algorithms.

In this paper, a novel dynamic routing model is developed to route optimal shipping paths for the transportation of Hazmat, especially to dynamically route under various uncertainties. The proposed methodology contain three solving stages, identifying the risk distribution map, routing for individual shipping pairs and optimizing for the whole shipping system. Two sub-models, the individual shipping MILP model and the system shipping MILP model, are correspondingly developed to sequentially route for individual shipping pairs and for the whole shipping system, aiming at minimizing the shipping penalty, which is constructed by three criteria factors, the transportation risk, time and direction turning. When uncertainties emerge, the risk distribution map will be updated, and all affected individual shipping pairs and the whole system will be dynamically routed. In particular, if shipping time limits are violated due to uncertainties, relations between the allowable shipping time and the shipping penalty will be explored to advice stakeholders on their ultimate decisions on shipping time relaxation and shipping path selection.

The performance of the developed model has been tested by three case studies. Case I is a static routing example as base case, Case II and III are both for dynamic routing. As demonstrated, when uncertainties take place, our model can quickly response and provide an alternative shipping path with lower shipping penalty. In addition, reliable and solid information on shipping time and shipping penalty is explored in Case III for the tradeoff and path selection. Through case studies, our model displays excellent and efficient performance on the routing for the transportation of Hazmat, especially for the dynamic routing under uncertainties.

References:

[1] Erkut, E., & Alp, O. (2007). Integrated Routing and Scheduling of Hazmat Trucks with Stops En Route. Transportation Science, 41, 107-122.

[2] Fan, T., Chiang, W.-C., & Russell, R. (2015). Modeling urban hazmat transportation with road closure consideration. Transportation Research Part D: Transport and Environment, 35, 104-115.