Hidden hazard identification and causation analysis of urban gas accidents based on Bayesian network

doi:10.16265/j.cnki.issn1003-3033.2026.06.1674

Abstract

Abstract:

To address the unclear mechanisms of multi-hazard coupling and the insufficient identification of critical causal nodes in urban gas accidents, this study constructed a gas accident risk evolution model integrating BN and triangular fuzzy numbers. Based on accident cases and expert knowledge, a hidden hazard index system was established across the three dimensions of "human-machine-environment". Triangular fuzzy numbers were employed to characterize the prior and conditional probabilities of each node. Through probabilistic reasoning, sensitivity analysis, and path strength analysis, the model systematically identified key causal nodes and high-risk coupling paths within the accident evolution process. The research indicates that gas leakage is the source node of the accident chain, exerting a decisive influence on the occurrence probability of consequential events such as explosions and fires, and represents a critical link with global influence and controllability. Structural damage is the dominant factor leading to leakage (posterior probability 0.80). Heterogeneity exists in the interaction modes of hazards, with an incomplete correspondence observed between high-frequency hazards and high-intensity disaster-causing paths. Therefore, urban gas accident risk management should strengthen leakage monitoring and early warning, attach importance to the regular prevention and control of high-frequency hidden hazards, and conduct early intervention on high-risk coupling paths. It is necessary to realize the transformation from single-node control to full-chain systemic prevention and control.

Key words: gas accident, Bayesian network (BN), accident hidden hazard, causal analysis, critical node

CLC Number:

X928.03事故预防与预测

Zhu Wei, Sheng Yanzhen, Song Jiayun, Zhao Yifan. Hidden hazard identification and causation analysis of urban gas accidents based on Bayesian network[J]. China Safety Science Journal, 2026, 36(6): 178-185.

Figures/Tables 11

Table 1

Table 2

Table 3

Table 4

Table 5

Table 6

Fig.1

Fig.2

Fig.3

Fig.4

Table 7

References 14

[1]	中华人民共和国应急管理部. 宁夏银川富洋烧烤店“6·21”特别重大燃气爆炸事故调查报告[EB/OL]. [2024-01-27]. https://www.mem.gov.cn/gk/zfxxgkpt/fdzdgknr/202401/t20240127_476487.shtml.
[2]	河北省应急管理厅. “3·13”较大燃气爆燃事故调查报告[EB/OL]. [2025-10-31]. https://yjgl.hebei.gov.cn/portal/index/getPortalNewsDetails?id=114bdba7-dba6-4987-a65a-6fd4599cce04&categoryid=3a9d0375-6937-4730-bf52-febb997d8b48.
[3]	Sheng Kai, Lai Xiaolin, Chen Yong, et al. Risk assessment of urban gas pipeline based on different unknown measure functions[J]. Tehnicki Vjesnik-Technical Gazette, 2021, 28(5): 1605-1614.
[4]	Tian Jiangxue, Lyu Shuran. A risk assessment model of gas pipeline leakage based on a fuzzy hybrid analytic hierarchy process[J]. Sustainability, 2024, 16(20): 1-23. doi: 10.3390/su16010001
[5]	Williams D A, Glasmeier A K. Evaluation of error across natural gas pipeline incidents[J]. Risk Analysis, 2023, 43(5): 1079-1091. doi: 10.1111/risa.v43.5
[6]	Saito S, Kawamura H. Impact assessment of mandatory safety device installation for gas stoves[J]. Risk Analysis, 2024, 44(10): 2403-2414. doi: 10.1111/risa.14313 pmid: 38616486
[7]	Pal I, Kumar A, Mukhopadhyay A. Risks to coastal critical infrastructure from climate change[J]. Annual Review of Environment and Resources, 2023, 48(1): 681-712. doi: 10.1146/energy.2023.48.issue-1
[8]	Alibek K, Faisal K, Yang M, et al. Gas leakage detection using spatial and temporal neural network model[J]. Process Safety and Environmental Protection, 2022, 160: 968-975. doi: 10.1016/j.psep.2022.03.002
[9]	郑彬彬, 冯婷婷, 王佳贺, 等. 基于文本挖掘的城镇燃气事故致因及关联分析[J]. 中国安全科学学报, 2023, 33(7): 190-195. doi: 10.16265/j.cnki.issn1003-3033.2023.07.1644
	Zheng Binbin, Feng Tingting, Wang Jiahe, et al. Causes and correlation analysis of urban gas accidents based on text mining[J]. China Safety Science Journal, 2023, 33(7): 190-195. doi: 10.16265/j.cnki.issn1003-3033.2023.07.1644
[10]	孙逸林, 郑小强, 刘险峰, 等. 基于AcciMap模型的燃气管道泄漏爆炸事故分析[J]. 中国安全科学学报, 2023, 33(7): 140-146. doi: 10.16265/j.cnki.issn1003-3033.2023.07.1011
	Sun Yilin, Zheng Xiaoqiang, Liu Xianfeng, et al. Analysis of gas pipeline leakage and explosion accident based on AcciMap model[J]. China Safety Science Journal, 2023, 33(7): 140-146. doi: 10.16265/j.cnki.issn1003-3033.2023.07.1011
[11]	向鹏成, 张子薇, 李卉, 等. 城市地下燃气管道泄漏事故致因因素系统分析[J]. 中国安全科学学报, 2023, 33(12): 140-147. doi: 10.16265/j.cnki.issn1003-3033.2023.12.2162
	Xiang Pengcheng, Zhang Ziwei, Li Hui, et al. Systematic analysis on causative factors of gas pipelines leakage accidents in urban underground[J]. China Safety Science Journal, 2023, 33(12): 140-147. doi: 10.16265/j.cnki.issn1003-3033.2023.12.2162
[12]	王轶宏, 翟越, 李艳, 等. 城市燃气管网泄漏蒸气云爆炸事故风险评估[J]. 中国安全科学学报, 2023, 33(2): 194-201. doi: 10.16265/j.cnki.issn1003-3033.2023.02.0285
	Wang Yihong, Zhai Yue, Li Yan, et al. Risk assessment of vapor cloud explosion accident of urban gas pipe network under leakage condition[J]. China Safety Science Journal, 2023, 33(2): 194-201. doi: 10.16265/j.cnki.issn1003-3033.2023.02.0285
[13]	Soomro A A, Mokhtar A A, Kurnia J C, et al. A review on Bayesian modeling approach to quantify failure risk assessment of oil and gas pipelines due to corrosion[J]. International Journal of Pressure Vessels and Piping, 2022, 200: DOI:10.1016/j.ijpvp.2022.104841.
[14]	姜子建, 周荣义, 石云霄, 等. 基于贝叶斯网络的危化品爆炸事故隐患关联与溯源分析[J]. 中国安全科学学报, 2024, 34(6): 173-180. doi: 10.16265/j.cnki.issn1003-3033.2024.06.1848
	Jiang Zijian, Zhou Rongyi, Shi Yunxiao, et al. Correlation and traceability analysis of hazardous chemical explosion accidents based on Bayesian network[J]. China Safety Science Journal, 2024, 34(6): 173-180. doi: 10.16265/j.cnki.issn1003-3033.2024.06.1848

总因素	因素	根节点	描述
人	内部人员	意外损坏A₁	第三方施工、机械作业等非故意行为导致的燃气设施物理损伤
	外部人员	操作不当A₂	运维人员未按规程操作或用户违规用气
	外部人员	蓄意行为A₃	人为破坏、盗窃燃气设备等
物	运行前隐患	规划设计缺陷A₄	管道设计不合理,安全间距不够等
		材料设备缺陷A₅	管道、容器壁厚不足或材料强度不达标
		资料不全A₆	缺乏完整的管网图纸、设备参数或历史维修记录导致运维决策失误或施工误损
		缺少必要标识A₇	未标注高风险区域或管道走向
		施工、安装缺陷A₈	焊接不合格、管道埋深不足、防腐层破损等施工工艺问题
	运行时因素	焊口开裂A₉	焊接工艺不良(如未焊透、存在气孔)或热应力集中导致的接口失效
		内壁腐蚀A₁₀	燃气中含硫化氢、水分等杂质引发管道内壁电化学腐蚀
		外壁腐蚀A₁₁	土壤电解腐蚀、杂散电流或防腐层老化导致的外部腐蚀
		设备故障A₁₂	调压器、阀门等机械部件失效(如密封圈老化),或安全装置(如泄压阀)失灵
		控制系统故障A₁₃	控制系统误报警、程序错误或传感器漂移,导致无法及时切断或预警
环	长期因素	土壤因素A₁₄	土壤酸碱度、含水率或微生物活动加速管道腐蚀,或冻胀导致管道变形
	长期因素	地质沉降A₁₅	地面不均匀沉降(如采空区)使管道受力不均,可能引发断裂或接口脱开
	突发因素	地震A₁₆	地震波造成管道剪切、挤压破坏,或引发次生灾害(如滑坡)损毁设施
	突发因素	气象灾害A₁₇	台风、暴雨可能冲刷管道裸露,雷击破坏电气设备,极端低温导致材料脆裂
		其他事故A₁₈	相邻系统事故(如电力短路引发爆炸、交通事故撞击燃气设施)的连锁反应

子节点	根节点
结构性破损B₁	意外损坏、材料设备缺陷、缺少必要标识、施工安装缺陷、焊口开裂
环境应力B₂	地质沉降、地震、气象灾害
腐蚀穿孔B₃	内壁腐蚀、外壁腐蚀、土壤因素
功能性失效B₄	材料设备缺陷、设备故障、地质沉降
人为释放B₅	意外损坏、操作不当、蓄意行为
超压破裂B₆	气象灾害、规划设计缺陷
燃气泄漏C₁	结构性破损、环境应力、腐蚀穿孔、功能性失效、人为释放
爆炸C₂	超压破裂、其他事故
火灾C₃	燃气泄漏

序号	模糊语言	三角模糊数	序号	模糊语言	三角模糊数
1	非常高	(0.83,1,1)	5	较低	(0.17,0.33,0.5)
2	高	(0.67,0.83,1)	6	低	(0,0.17,0.33)
3	较高	(0.5,0.67,0.83)	7	非常低	(0,0,0.17)
4	中等	(0.33,0.5,0.67)	—	—	—

工作年限/a	权重	职称	权重
>20	1	正高/教授级高级工程师	1
15~20	0.9	副高/高级工程师	0.9
10~15	0.8	中级/工程师	0.8
5~10	0.7	初级/助理工程师	0.7

打分	归一化权重	隶属度	三角模糊数	模糊概率	打分	归一化权重	隶属度	三角模糊数	模糊概率
0.8	0.035	0.81	(0.67,0.83,1)	(0.60,0.75,0.90)	0.8	0.028	0. 8125	(0.48,0.60,0.72)	(0.48,0.60,0.72)
0.9	0.035	0.59	(0.67,0.83,1)	(0.60,0.75,0.90)	1.0	0.035	1.00	(0.75,0.90,0.90)	(0.75,0.90,0.90)
1.0	0.035	1.00	(0.83,1,1)	(0.75,0.90,0.90)	0.8	0.028	0. 8125	(0.48,0.60,0.72)	(0.48,0.60,0.72)
0.9	0.028	0.59	(0.67,0.83,1)	(0.48,0.60,0.72)	0.8	0.028	0.81	(0.48,0.60,0.72)	(0.48,0.60,0.72)
0.8	0.031	0.81	(0.67,0.83,1)	(0.54,0.66,0.80)	0.5	0.031	1.00	(0.27,0.41,0.54)	(0.27,0.41,0.54)
0.8	0.027	0.81	(0.67,0.83,1)	(0.47,0.58,0.70)	0.8	0.028	0.81	(0.48,0.60,0.72)	(0.48,0.60,0.72)
0.6	0.031	0.59	(0.5,0.67,0.83)	(0.41,0.54,0.67)	0.4	0.038	0.59	(0.17,0.33,0.50)	(0.17,0.33,0.50)
1.0	0.024	1.00	(0.83,1,1)	(0.52,0.63,0.63)	0.8	0.035	0.81	(0.60,0.75,0.90)	(0.60,0.75,0.90)