联立GO-贝叶斯的气田站场系统失效可能性研究  

Research on Failure Possibility of Gas Field Station System by Simultaneous GO-Bayesian Method

作　　者：田帝 朱丽霞 李丽锋 罗金恒 李霄[1] 郝晓东[3] TIAN Di;ZHU Li-xia;LI Li-feng;LUO Jin-heng;LI Xiao;HAO Xiao-dong(Xi’an Shiyou University,Xi’an 710064,China;Tubular Goods Institute of CNPC,Xi’an 710077,China;Planning Institute of China Petroleum and Natural Gas,Beijing 100083,China)

机构地区：[1]西安石油大学,陕西西安710064 [2]中国石油集团工程材料研究院有限公司,陕西西安710077 [3]中国石油天然气股份有限公司规划总院,北京100083

出　　处：《全面腐蚀控制》2023年第12期14-20,38,共8页Total Corrosion Control

摘　　要：气田站场是一个动态复杂系统,其中某一设备的失效可能会产生连锁反应导致整个站场失效。通过对安全敏感输出变量的限制发现传统静态风险评价方法忽视了站场的风险动态关联特征。在分析站场关键设备失效模式和失效概率的基础上总结出站场失效概率计算的准确性与事故发生概率偏差较大。为了有效提高站场的可靠性,基于站场工艺流程建立GO图,将GO图建模的便捷性与贝叶斯网络强大的推理能力结合,最后对典型站场进行了实例应用,在风险减缓措施方面可为站场提供决策依据和参考。The gas field station is a dynamic and complex system,in which the failure of a certain equipment may have a chain reaction leading to the failure of the whole station.By limiting the safetysensitive output variables,it is found that the traditional static risk assessment method ignores the dynamic risk correlation characteristics of the station.Based on the analysis of the failure mode and failure probability of key equipment in the station,it is concluded that the accuracy of the calculation of the failure probability of the outbound station is quite different from the accident probability.In order to effectively improve the reliability of the station,the GO diagram is established based on the process flow of the station,which combines the convenience of modeling with the powerful reasoning ability of Bayesian network.Finally,a typical station is applied as an example,which can provide decision-making basis and reference for the station in terms of risk mitigation measures.

关 键 词：气田站场 失效概率 风险评价 贝叶斯网络 GO图 

分 类 号：TE88[石油与天然气工程—油气储运工程]