Human error assessment of oil tanker grounding

Abstract

Human error plays an important role triggering ship tragedies. Human error assessment of marine accidents is troublesome due to the lack or incompleteness of data. Some studies generate results in a form of priority weights rather than a probabilistic nature. Others produce Human Error Probabilities (HEPs) without the consideration of quantitative effects caused by the individual, technological, contextual and organizational factors or using the human performance data from nuclear power plants. This paper proposes a risk study evaluating the human error contribution to oil tanker grounding. A group of 6 experts with substantial seagoing experience is invited to provide professional judgement for oil tanker navigation. This research establishes a logical safety structure of oil tanker grounding based on Fault Tree Analysis (FTA) under which a novel fuzzy Cognitive Reliability Error Analysis Method (CREAM) model is constructed. By virtue of Bayesian Reasoning, the parameter synthesis process fully deliberates all useful information from each rule and derives reasonable outcomes. This study reinforces the logicality between Common Performance Conditions (CPC) observations and Contextual Control Mode (COCOM). The probability of ship powered grounding is acquired accordingly. Furthermore, fatigue and Collision Regulation (COLREG) violations are the vital elements triggering ship groundings according to the sensitivity analysis.