Safety assessment in oil drilling work system based on empirical study and Analytic Network Process

• Published in: Safety science 2018-06, Vol.105, p.86-97
• Main Authors: Sun, Zhi-Yu, Zhou, Jian-Lan, Gan, Lin-Fei
• Format: Article
• Language: English
• Subjects: Analytic network process; Cognitive ability; Drilling; Empirical analysis; Empirical study; Human factors; Mathematical models; Occupational safety; Oil; Oil drilling; Priorities; Safety; Safety assessment; Safety management
• ISSN: 0925-7535; 1879-1042
• DOI: 10.1016/j.ssci.2018.02.004

•Safety questionnaires are conducted in oil drilling industry.•A SEM-ANP method to evaluate the importance of human factors is proposed.•The frequency of factors is calculated with accident cases.•The causal chain and safety recommendations are discussed by the results. Safety assessment is an essential work to guarantee the safety of oil drilling. There are relations and dependencies between human factors in oil drilling work system. Therefore, the safety of oil drilling work system should be analyzed in a comprehensive way. The Human Factors Analysis and Classification System (HFACS) is applied to establish rational and applicable index system for investigating human errors. The Analytic Network Process (ANP) method is used to obtain the priorities of human factors considering the interdependences, however, the deficiency of ANP is that the obtained results are subject to experts’ cognitive limitations and psychological biases. The Structural Equation Modelling (SEM) is used to form the ANP model auxiliary, which may be expected to overcome subjective opinions from experts and provide a more pertinent and practical safety strategies. A survey is conducted to explore the importance of human factors through questionnaires of which 283 pieces made up the original data. Afterwards, the human factors’ weights are calculated by the ANP method. As a comparison, a frequency-based method is also used to obtain the frequencies of factors and observations causing accidents using accident reports. The causal chain and the priorities of the importance of human factors are explored by this hybrid method; the results are consistent with the experience and knowledge of safety management. We discuss the interdependencies between the human factors and the priorities in general, whilst, the specific safety requirements and recommendations in the hoisting and lifting system are also provided as an example.