A Location-Allocation Model for Food Distribution in Post-Disaster Environments

Abstract This research investigates the optimal location decisions in a food distribution network in post-disaster environments. We propose a location-allocation model for a food bank network that minimizes the total social costs generated by distributing food supplies to disaster survivors. The social costs include the costs of the delivery operation and the external costs of the survivor’s suffering in the form of deprivation costs. This model incorporates an empirically estimated deprivation cost function for food and water supply, which is non-linear with respect to the survivor’s deprivation time. We define the deprivation time as the time a survivor has to wait for the delivery, which includes their travel time to the point of distribution (POD) and their expected wait there to receive the critical supplies. The model proposes the optimal location of PODs (i.e., food pantries) and their allocation to demand zones for food distribution following a disaster, along with a relief distribution strategy. We used the Houston Food Bank network to test the feasibility of activating the nodes as a relief distribution network that could serve the impacted community. The instances aim to mimic a potential design that would have been implemented in response to a future disaster like Hurricane Harvey in 2017. It also addresses the impact of location on delivery frequencies and shipment sizes. The results provide practical insights that can help prioritize the needs of survivors during an emergency or disaster event.