Risk based, multi objective vehicle routing problem for hazardous materials: A test case in downstream fuel logistics

The paper analyses a practical case of study related to the distribution of fuels for the Total Erg Oil Company to the service stations located in the Province of Rome (Italy). The problem is formulated as a capacitated vehicle routing problem with time windows, where several heuristic procedures have been tested, considering both static and dynamic travel times. With respect to the standard operational costs used typically, a multivariable objective function has been proposed which takes into account also a new risk index. The risk index proposed is function of the population density of the zones covered by each path and of the estimated number of road accidents on each road link. In such a way, we take into account the population's exposure to the risk associated with an incidental event involving a fuel tank. The obtained output is the set of planned routes with minimum service cost and minimum risk. Results demonstrate how an accurate planning of the service saves up to 3 hours and 30 km on a daily basis compared to a benchmark. Moreover, the distribution company can parameterize the configuration of the service, by varying the weight adopted in order to include the risk index. Including the risk index may bring to a higher safety route planning, with an increase of the operating costs of only 2%.