中国安全科学学报 ›› 2023, Vol. 33 ›› Issue (7): 140-146.doi: 10.16265/j.cnki.issn1003-3033.2023.07.1011

• 安全工程技术 • 上一篇    下一篇

基于AcciMap模型的燃气管道泄漏爆炸事故分析

孙逸林1,2(), 郑小强1,2,**(), 刘险峰1, 贺艳艳1, 王冶3   

  1. 1 西南石油大学 经济管理学院,四川 成都 610500
    2 西南石油大学 能源安全与低碳发展重点实验室,四川 成都 610500
    3 中国石油西南油气田分公司 四川川港燃气有限公司,四川 成都 610051
  • 收稿日期:2023-02-25 修回日期:2023-05-19 出版日期:2023-07-28
  • 通讯作者:
    ** 郑小强(1981—)男,四川长宁人,博士,教授,博士生导师,主要从事能源经济、风险管理与可持续发展等方面的研究。E-mail:
  • 作者简介:

    孙逸林 (1996—),男,安徽合肥人,博士研究生,研究方向为复杂社会技术系统建模、城镇燃气安全管理与事故分析。E-mail:

    刘险峰,教授

    王冶,工程师

  • 基金资助:
    国家社会科学基金(21BZZ058); 国家自然科学基金青年科学基金(52004241); 西南石油大学科研攻关项目(2019CXTD12)

Analysis of gas pipeline leakage and explosion accident based on AcciMap model

SUN Yilin1,2(), ZHENG Xiaoqiang1,2,**(), LIU Xianfeng1, HE Yanyan1, WANG Ye3   

  1. 1 School of Economics and Management, Southwest Petroleum University, Chengdu Sichuan 610500, China
    2 Key Laboratory of Energy Security and Low-carbon development, Southwest Petroleum University, Chengdu Sichuan 610500, China
    3 Sichuan Chuangang Group Corporation, CNPC Southwest Oil & Gas field Company, Chengdu Sichuan 610051, China
  • Received:2023-02-25 Revised:2023-05-19 Published:2023-07-28

摘要:

为探讨燃气管道泄漏爆炸事故系统及其致因的性质,以湖北十堰“6·13”重大事故为例进行分析。采用AcciMap模型系统辨识事故致因,构建“6·13”事故AcciMap模型;引入复杂网络(CN)与PageRank算法,分析整体系统与个体因素的属性特征,据此辨识关键致因;再通过贝叶斯网络(BN)进行参数学习和诊断推理,分析事故最大致因链。结果表明:安全生产监管工作严重失察、安全生产主体责任未落实、安全生产监督指导不力是关键致因;安全生产理念和属地责任未贯彻落实→安全生产监管工作严重失察→安全管理制度存在缺陷→安全培训教育存在缺陷→燃气管道维抢人员违规处置→明火引爆蓄积燃气→燃气泄漏爆炸并造成严重伤亡是最大致因链。研究认为:应从政府和企业等2类主体出发,采取多元化的风险防控策略以遏制事故再次发生。

关键词: AcciMap模型, 燃气管道, 泄漏爆炸, 复杂网络(CN), 贝叶斯网络(BN)

Abstract:

In order to explore the attribute characteristics of causes of gas pipeline leakage and explosion accident, the "6·13" accident was taken as an illustrative example. The AcciMap model of the "6·13" accident was built by identifying causes based on the AcciMap model. The CN and PageRank algorithms were introduced into the AcciMap model to analyze the attribute characteristics of accident systems and nodes and identify key causes. The BN was applied for parameter learning and inference learning to find out the key causes chain. The analysis results prove that the serious oversight of work safety supervision, failure to implement the main responsibilities of work safety, inadequate supervision and guidance of work safety are the key causes of the "6·13" accident. The concept of work safety and territorial responsibilities are not implemented → serious oversight of work safety supervision → defects in safety management system → defects in safety training and education → illegal disposal of gas pipeline maintenance and rescue staff → fire detonates accumulated gas → gas leakage explode and causes serious casualties is the biggest causal chain. It is concluded that diversified risk prevention and control strategies should be taken from the aspects of both government and enterprises to curb the recurrence of accidents.

Key words: AcciMap model, gas pipeline, leakage and explosion, complex network(CN), Bayesian network(BN)