Data-driven Bayesian analysis of port-call bottlenecks and turnaround time at the ship–shore interface in crude oil tanker operations

Ship consumption during port periods, main and auxiliary engine usage, cargo handling operations, and waiting times contribute significantly to ship-related emissions. Despite increasing regulatory pressure regarding carbon emissions, the assessment of ships' energy efficiency and performance during port calls remains limited. The study aims to identify the main bottlenecks and drivers of turnaround time at the ship–shore interface as a proxy indicator with implications for ship energy efficiency by examining the entire process from ship arrival until the commencement of the sea passage. Real-time data on crude oil carriers’ statements of facts during port calls has been processed to identify inefficiencies across the entire operation. Tree Augmented Naïve Bayes has been utilised the process the data. The findings indicate that the NOR Tender-All Fast (time from Notice of Readiness tender to the vessel being all fast at berth) interval is the most operationally significant contributor to non-productive time in crude oil tanker operations, based on mutual information and the variance-of-beliefs ratio. An integrated approach is required among stakeholders such as ports, ships, and charterers, while utilising digital communication tools that enable data sharing and optimise ship operations before and during port calls. The findings contribute to knowledge of ship-shore interface turnaround time and its bottlenecks by outlining micro-sequences of port calls, with implications for decarbonization strategies and energy-efficiency practices.