摘要
氢气作为一种零排放的清洁能源,具有来源广、燃烧产物无污染、热量高等特点。氢能的开发、推广及使用受到世界各国的高度关注。加氢站氢气泄漏,可能发生火灾爆炸事故,威胁生命财产安全。针对加氢站氢气泄漏风险的严重性和不确定性等问题,提出了一种融合危险与可操作性及保护层分析的动态贝叶斯网络方法,评估加氢站一年内事故发生的动态变化,并确定影响加氢站氢气泄漏基本事件。氢气泄漏概率由初始的4.27×10^(-1)增高至1 a后的6.07×10^(-1),事故后果概率从大到小依次为安全状态、蒸气云聚集、闪火、喷射火、爆炸;采用由半定量分析提出的安全措施进行改进,氢气泄漏发生概率降低了32.8%,各事故后果发生概率降低了约24.7%。利用保护层降低加氢站的氢泄漏风险,这有助于加氢站安全稳定的运行。
This paper introduces a risk assessment methodology that integrates Dynamic Bayesian Networks(DBN)with Hazard and Operability(HAZOP)analysis and Layer of Protection Analysis(LOPA).The aim is to quantify the dynamic evolution of hydrogen leakage risks at refueling stations and identify key risk factors,including hydrogen overfilling,personnel fatigue,hose failures,valve replacement issues,welding defects,loose valves and fittings,diaphragm ruptures,pressurized flange failures,gasket damage,cooling water system malfunctions,and safety valve failures.In this paper,we investigate the hydrogen leakage probability at a refueling station over one year,using 360-hour intervals divided into 24 time slices.The initial probability of hydrogen leakage is calculated to be 4.27×10^(-1).Over the course of the year,this probability gradually increases with time,exhibiting a nearly linear growth trend,ultimately rising to 6.07×10^(-1).This high probability of hydrogen leakage underscores the urgent need to implement measures aimed at mitigating leakage risks.By analyzing the probability of accident consequences,five potential accident scenarios were identified and ranked in descending order of their likelihood of occurrence:safe condition,vapor cloud accumulation,flash fire,jet fire,and explosion.This ranking offers a scientific basis for prioritizing risk prevention and control measures at hydrogen refueling stations.Additionally,the safety improvement measures proposed based on HAZOP and LOPA analyses—including optimization of equipment design,enhancement of the monitoring system,and installation of additional protective layers—can significantly reduce both the probability of hydrogen leakage and the potential consequences of such accidents.Specifically,the probability of hydrogen leakage was reduced by an average of 32.8%,while the probability of each accident consequence was decreased by an average of 24.7%.These measures have effectively improved the risk management of hydrogen leakage and mitigated the potential consequences of accidents at hydrogen refueling stations.
作者
武锦涛
谷尚硕
陈祖志
WU Jintao;GU Shangshuo;CHEN Zuzhi(School of Chemical Engineering,Dalian University of Technology,Dalian 116081,Liaoning,China;China Research Institute of Special Equipment Inspection,Beijing 100029,China)
出处
《安全与环境学报》
北大核心
2025年第10期3890-3903,共14页
Journal of Safety and Environment
关键词
安全工程
加氢站
风险评估
动态贝叶斯网络
安全措施
safety engineering
hydrogen refueling stations
risk assessment
dynamic Bayesian networks
safety measures
