基于HFACS-BN模型的危险化学品事故人因路径分析*

为提高危险化学品安全管理水平，收集我国134起以人为主因的危险化学品事故，构建危险化学品事故的HFACS-BN模型，并基于贝叶斯网络对危险化学品事故中的人因路径及各因素灵敏度进行分析。研究结果表明:行为违规是导致事故的最主要不安全行为；在复杂路径中，环境因素和操作者状态具有较高的灵敏度，与其相关的风险认知与处理不当是阻碍事故预防的关键因素；在非复杂路径中，操作者状态与运行计划不当具有较高的灵敏度，其所涉及的监督违规、组织过程和组织氛围，是导致危险化学品事故的根本起点，并可依此分析事故直接原因和间接原因；操作者状态在2种路径中均表现出高灵敏度，因此化工企业在人员管理上要加强关注职工状态，减少不安全行为的出现。

Analysis on human factor path of hazardous chemical accidents based on HFACS-BN model

In order to improve the safety management level of hazardous chemicals,134 human caused hazardous chemical accidents in China were collected,and a HFACS-BN model of hazardous chemical accidents was constructed.Based on the Bayesian network (BN),the human factor path and the sensitivity of each factor in hazardous chemical accidents was analyzed.The results showed that the behavioral violations were the most important unsafe behaviors leading to the accidents.In the complex path,the environmental factors and operator status had higher sensitivity,and their related risk perception and mishandling were the key factors hindering the accident prevention.In the non-complex path,the operator status and improper operation plan had higher sensitivity,and their involved supervision violations,organizational process and organizational atmosphere were the fundamental starting point of hazardous chemical accidents,and the direct and indirect causes of accidents could be analyzed accordingly.The operator status presented high sensitivity in both paths,so the chemical enterprises should pay more attention to the status of workers in personnel management to reduce the possibility of unsafe behavior.