Resilience construction of automated sea–rail intermodal ports against extreme weather: technology integration and collaborative innovation
Yimei Chang, Zhenya Cheng, Kang JianPurpose
This study aims to explain how a short-lived extreme-weather capacity loss develops into persistent congestion in automated sea–rail intermodal terminals and evaluates how retained operating capacity affects backlog formation and operational recovery.
Design/methodology/approach
AIS-derived weekly indicators for Ningbo during 2019–2022 are used to calibrate and externally check normal-operation scale. Three seven-day weather paths observed in 2023 drive a nonstationary M/G/25 simulation for a representative container-terminal cluster within Ningbo Zhoushan Port. The weather-driven nonstationary M/G/25 queuing simulation model combines long-run mean arrival rate of 26.429 vessels/day with correlated daily demand variation, lognormal service requirements, 200 common-random-number replications and capacity-retention sensitivity over θ = 0.60–0.80.
Findings
At θ = 0.70, the event resilience indices are 0.720 for Haikui, 0.744 for Khanun and 0.707 for the convective event. Vessel-level median waiting time (WT) rises from approximately 21.2 h in baseline to 43.3, 43.1 and 44.8 h, respectively, while event peak queues reach 58.5, 56.0 and 60.0 vessels. Khanun and Haikui recover on Days 12 and 13; the convective event does not satisfy the joint WT–time in port recovery rule within 14 days. Raising θ to 0.80 reduces all event medians and restores all three scenarios within five recovery days.
Originality/value
This study advances resilience analysis for automated sea–rail intermodal container terminals by coupling threshold-derived operating states with a time-varying queuing process. Unlike static indicator-based assessments, the proposed framework explains how weather-induced capacity degradation develops into queue accumulation and continues affecting operations after meteorological conditions normalize. It distinguishes meteorological normalization from operational recovery and quantifies how retained service capacity influences both event-period congestion and post-event backlog clearance.