摘要
由于我国页岩气藏具有高压特性,地质条件复杂,勘探开发面临着巨大风险,针对页岩气开发事故的风险分析有积极意义。建立了溢流到井喷事故处置情景,提取31个基本事件和13个中间事件构建事故树,然后转化为贝叶斯网络模型,分析了基本事件对整个系统的重要度(风险促进值重要度、风险降低值重要度和变化衡量重要度),根据各基本事件的历史概率数据确定贝叶斯网络节点先验概率。结果表明,页岩气井喷失控事故是人、机、管理、地质等多因素耦合作用的结果。运用贝叶斯网络得出页岩气井井喷失控的概率约为1.04×10-9。通过贝叶斯网络重要度分析,得出喷出油气浓度不符合标准、喷出油气遇明火等14个关键因子事件,可作为实际生产重点防治对象的参考。
The paper is to quantitatively assess the risk of the blowout accident at shale gas exploitation and production sites,to explore the probability of the accidents and the impacts of the various factors,so as to solve the problems concerning the indefinite causes of the blown-out accidents during the shale gas exploitation and the indefinite foci of the accident prevention and control. For the above mentioned purposes,we have,first of all,established the accident scenarios in the paper from the overflow to blowout on the shale gas exploitation sites and in turn identified and determined 5 principal factors,including the geological design,the engineering design,the well control facilities,the well-production control technological means and the working staff spiritual and qualification psychology, all of which involved should be taken into comprehensive consideration for their impact on any of the blowout accidents. At the same time,a total of 31 elementary accidental events and 13 intermediate events have been extracted to construct an event tree,which has then been converted into a Bayesian network. Thus,in the said Bayesian network,the prior probabilities of the elementary events have been determined in accordance with the historical data,whereas the analysis of the structural importance has been done for the elementary events in the event tree. Besides,the Bayesian network-based importance analysis has also been chosen to assess the elementary events(the importance of the risk promotion value,the importance of the risk reduction value and the importance of the variation measurement). And,gradually,the key factors have been gained through comparison with the respective accidents. And,so,it can be concluded that the shale gas blowout accident should be taken as a comprehensive multi-factor involved results of equipment,personnel,management and geological unexpected factors. The probability of the out-of-control blowout events(T) that have been worked out with the Bayesian network can only be about 1. 04 × 10-9,yet the data found can still provide valuable reference for further subsequent probes.Henceforth,the analysis results of the importance analysis can still suggest a total of 14 key progammatical events,including the standard incompliance of the erupted oil and gas concentrations,encounter of the oil/gas eruption with unexpected naked flames,etc. Apart from those above likelinesses,there may come about the equipment factor and personnel factor,which may bring about significant impacts on the unhappy events. Thus,the conclusion of the paper can be used as valuable references for major quantitative accidents prevention and control foci in the actual production of shale gas.
作者
陈坤
陈序
魏鑫
米衍泽
CHEN Kun;CHEN Xu;WEI Xin;MI Yan-ze(School of Safety Engineering,ChongqingUniversity of Science and Technology,Chongqing 401331,China)
出处
《安全与环境学报》
CAS
CSCD
北大核心
2019年第6期1876-1883,共8页
Journal of Safety and Environment
基金
应急管理部安全生产重特大事故防治关键技术科技项目(Chongqing-0006-2018AQ)
重庆市教委重点项目(KJZD-K201901501)
重庆科技学院研究生科技创新计划项目(YKJCX1820705)
重庆科技学院安全工程学院科技创新计划项目
关键词
安全系统工程
动态风险分析
贝叶斯网络
页岩气
safety systems engineering
dynamic risk analysis
Bayesian network
shale gas
