Risk evaluation for oil tanker berthing and handling operations using the improved DEMATEL-ISM approach

Abstract

The berthing and handling operation of oil tankers is an important task in the operation of oil ports. Therefore, it is crucial to identify the risk factors in the process of the berthing and handling operation of oil tankers, to explore the relationship between the influencing factors of the risk, and to find out the key factors that incur risk. In this paper, the risk factors of oil tankers’ berthing and handling operations are identified through on-site investigation, interviews, questionnaire surveys and expert consultation. Aiming at the uncertainty of the weight of each risk index during oil tanker berthing and handling, a machine learning algorithm (i.e. Random Forest Weighting) is applied. In this sense, the reliability analysis is used to calculate the generalization error of expert scoring data, and the weight of each index of each evaluation item is obtained according to the minimum error rate, which reduces the impact of subjective weighting in the identification of cause-effect chain components. In order to examine the internal relationship among various factors, the Interpretive Structural Model (ISM) of risk factors is constructed. Furthermore, we analyse the hierarchical relationship between various factors, clarify the mechanism of risk generation, and propose strategies to improve the safety management level of oil tanker terminals. Results show that the used algorithm could identify the key risk factors in the risk oil tanker terminal operations, and provide more effective managerial recommendations.