Optimal refueling station location and supply planning for hurricane evacuation

Fuel supply shortages that have emerged in recent hurricane events underscore the weakness of existing emergency logistics planning processes. An effective modeling approach is central to refueling station location and supply planning decisions. This paper documents a research effort based on a simulation-optimization framework integrating a mixed integer program formulation and a mesoscopic simulation model. The mesoscopic simulation model was incorporated with decision rules to select refueling stations, methods to model the impact of stalled vehicles on traffic flow, and a formula to accumulate each vehicle's fuel consumption under various running speed conditions. The mixed integer program formulation is aimed at maximizing the served demand by deciding which stations to operate and how much fuel to supply given limited resources. The interplay between the simulation model and the optimization model continues until convergence. The proposed modeling approach is applied to a case study based on the I-45 corridor between Houston and Dallas, Texas, to highlight the characteristics of the proposed modeling approach.