Incorporating resilience determinants in supply chain network design model

Abstract

PurposeThe purpose of this paper is to develop a decision tool that enables supply chain (SC) architects to design resilient SC networks (SCNs). Two resilience design determinants are considered: SC density and node criticality. The effect of considering these determinants on network structures is highlighted based on the ability to resist disruptions and how SC performance is affected.Design/methodology/approachA mixed-integer non-linear programming model is proposed as a proactive strategy to develop resilient structures; design determinants are formulated and considered as constraints. An upper limit is set for each determinant, and resistance capacity and performance of the developed structures are evaluated. These upper limits are then changed until SC performance stabilizes in case of no disruption.FindingsResilient SCN structures are achieved at relatively low design determinants levels on the expense of profit and without experiencing shortage in case of no disruption. This reduction in profit can be minimized on setting counter values for the two determinants; relatively higher SC density with lower node criticality or vice versa. At very low SC density levels, the design model will reduce the number of open facilities largely leading to only one facility open at each echelon; therefore, shortage occurs and vulnerability to disruption increases. On the other hand, at high determinants levels, SC vulnerability also increases as a result of having more geographically clustered structures with higher inbound and outbound flows for each facility.Originality/valueIn this paper, a novel proactive decision tool is adopted to design resilient SCNs. Previous literature used metrics for SC density and node criticality to assess resilience; in this research, determinants are incorporated directly as constraints in the design model. Results give insight to SC architects on how to set determinant values to reach resilient structures with minimum performance loss in case of no disruption.