Identifying critical node set in supply chain network considering hybrid operational risk management

Identifying and protecting critical node set (CNS) is an important measure to enhance the resilience of supply chain network (SCN), especially in the face of increasingly frequent and uncertain disruptions. To respond to the urgent and real-world issue, this study proposes a two-stage stochastic programming (TSSP) framework for the problem of identifying CNS, considering hybrid operational risk management strategies. The first stage decides to pre-protect (proactive strategy) certain nodes against uncertain future risks, while the second stage determines which nodes to post-protect (reactive strategy) in response to realized risks, thereby mitigating potential risk propagation. The CNS identification problem is proven NP-hard. However, the second stage problem demonstrates monotonicity, which facilitates more efficient algorithm design. This paper introduces a load capacity model to simulate risk propagation process, and utilizes a neighborhood search and simulation-based hybrid optimization approach to identify the CNS as well as obtaining the optimal protection policy. Numerical experiments conducted on a practical SCN show that: (1) The CNS identified by our model outperforms traditional measures in enhancing supply chain resilience (SCR). (2) The optimal CNS configuration evolves in response to variations in budget constraints, the number of pre- and post-protection nodes. (3) The combined pre- and post-protection strategy demonstrates superior performance compared to individual protection approaches. The findings offer actionable insights for managers to mitigate risk propagation and enhance SCR through targeted protection of CNS.