摘要
为有效预防和控制重油催化工艺火灾爆炸事故,针对其火灾爆炸风险分析提出基于事故树的贝叶斯网络-保护层分析方法(BN-LOPA)。以提升管反应器为例,绘制提升管反应器火灾爆炸事故树,并将其转化为贝叶斯网络。利用贝叶斯网络所具有的逆向推理计算能力,找出最易导致事故发生的薄弱环节,分别为分馏塔冲塔和反应压力上升。对这些节点设置独立保护层后,可以有效降低火灾爆炸事故发生的风险等级。结果表明,火灾爆炸风险等级在受控范围不需要增加安全措施,可以实现重油催化工艺安全生产。
In order to effectively prevent and control the fire and explosion accident of heavy oil fluidized catalytic cracking unit, against the risks of fire and explosion method of Bayesian net- work-layer of protection analysis （BN-LOPA） based on fault tree analysis was put forward. With the riser reactor as an ex- ample, the riser reactor fire and explosion accident fault tree was drawn out and turned into Bayesian network. Reverse cal- culation of Bayesian networks was done, and found the fraction- ator washing tower and pressure rise being easy to lead to acci- dent. After the independent protective layer for the points being set, fire explosion accident risk level can effectively be reduced. Results showed that fire and explosion risk was under toler- ance, and extra safety measures were not required to achieve safety in production.
出处
《消防科学与技术》
CAS
北大核心
2017年第2期262-265,共4页
Fire Science and Technology
基金
中国劳动关系学院项目(16YY001)
2016中国消防协会科学技术年会"青年消防学者论坛"交流论文
关键词
火灾爆炸
事故树
贝叶斯网络
保护层分析
fire and explosion
fault tree
Bayesian network
layer of protection analysis
