A sequential barrier-based model to evaluate human reliability in maritime accident process

Human reliability is considered as the predominant causation factor in maritime accidents. The previous researches on human reliability analysis are reviewed in this paper, although these works have benefited us a lot in accident prevention, few researches focused on the human reliability in the developing stages of maritime accident process. This paper first thoroughly reviews the existing sequential accident process models, including the Swiss cheese model, accident causation model, risk-scenario based model, then the strengths and weakness of these models are compared with each other. To make a feasible and reasonable assessment on the human reliability, a sequential safety barrier-based model which includes four safety barriers are proposed by incorporating the safety barrier (SB) and risk scenario together. Moreover, the failure modes of different safety barriers, which consider the difficulty of these barriers' implementation, are also presented to further analyze the human reliability. The developing states, including initiating state (IE), mid states (MS), and end state (ES), and accident scenario of different types of accidents are also proposed under the framework of sequential safety barriers. Finally, this proposed model is introduced to investigate 97 maritime accident reports issued by Maritime Safety Administration (MSA) of China from 2000 to 2010. The result reveals that majority of the maritime accidents can be analyzed by the proposed accident process model (91.8%, N=97). It also demonstrates that different measures in terms of developing stages should be adopted to enhance the human reliability in the maritime accident process. Moreover, multiple correspondence analysis (MCA) reveals that human error is closely related with the catastrophic accident; moreover, it also demonstrates that majority of the accidents can be prevented or even avoided in the earlier stages.