The design of resilient food supply chain networks prone to epidemic disruptions

• Published in: International journal of production economics 2021-03, Vol.233, p.108001, Article 108001
• Main Authors: Gholami-Zanjani, Seyed Mohammad, Klibi, Walid, Jabalameli, Mohammad Saeed, Pishvaee, Mir Saman
• Format: Article
• Language: English
• Subjects: Analysis; Epidemic disruptions; Epidemics; Food supply; Food supply chain; Humanities and Social Sciences; Logistics; Resiliency; Stochastic processes; Stochastic programming; Uncertain demands; Workplace accidents
• ISSN: 0925-5273; 1873-7579
• DOI: 10.1016/j.ijpe.2020.108001

Food supply chains are nowadays perturbed by an increased supply and demand uncertainty, and more and more suffering from unexpected disruptions. In the specific context of food supply chains (FSC) for perishable products, these could be linked to natural hazards, industrial accidents or epidemics and their impact could lead to huge economic losses. The case of epidemic events has been little studied in the existing literature, although there are numerous cases reported in practice. At the strategic level, this requires a novel risk modeling approach to tackle the correlation and propagation features and advanced stochastic multi-period models to design the FSC network. Our interest in this research is to propose a comprehensive two-stage scenario-based mathematical model to design a resilient food supply chain under demand uncertainty and epidemic disruptions. In order to adequately characterize epidemic disruptions, they are modeled as a compound stochastic process and a Monte Carlo procedure is developed to generate plausible scenarios. The modeling approach covers the special characteristics of FSC, such as products perishability in time and discount prices based on product's age. In addition, a number of resiliency strategies are incorporated into the core model to enhance the resilience level of the FSC network design. The developed models are solved through an efficient solution approach relying on scenario reduction technique and Benders decomposition. Numerous problem instances are used to validate the modeling approach and to derive managerial insights.