A two-stage stochastic post-disaster humanitarian supply chain network design problem

Abstract

We consider the planning problem of designing and operating humanitarian supply chain networks (HSCN) after natural disasters. Specifically, we focus on the design of a three-layer network under demand and capacity uncertainty to support short-term recovery, i.e., to distribute critical supplies to the affected population. We aim to analyze the effect of unmet demand accumulating over the planning horizon in order to better understand and respond to natural disasters. To this end, we explicitly consider the impact of unmet demand through time under uncertain conditions by introducing a spread factor. We develop a two-stage stochastic model that retains the uncertainty pertaining to the demand along with the transportation and storage capacities of the HSCN. Then, we apply our model to a case study using real-world data from the 2018 earthquake in Indonesia. Various aspects of the problem are studied over a set of experiments, including the importance of modeling uncertainty, the effect of the budget on the solution performance, and the role of the spread factor in the accurate understanding of the crisis. According to the results obtained, considering lower values for the spread factor parameter can irreparably misguide the decision-makers by an inaccurate presentation of the crisis’ depth and consequently increase the damage caused to people’s health