摘要
深水井控压井作业是有效控制溢流演化为井喷事故的二级井控工艺屏障。为提高深水井控压井作业可靠性,采用BN-CREAM方法对其风险诱因进行研究。结合深水井控压井作业特点,考虑共因失效等因素,采用贝叶斯网络方法建立深水井控压井作业风险演化模型。应用人因可靠性分析CREAM法计算深水压井人因失误先验概率,参考海洋可靠性数据手册OREDA确定深水井控设备失效先验概率。依托贝叶斯网络的逆向推理能力辨识压井作业的主要风险节点,从而实现对深水井控压井作业风险诱因的有效预测和评估。研究表明:深水井控压井作业共包含6个关键风险根节点,且压井作业人因可靠性要低于设备可靠性;3级子节点"压井方法选择不合理"对深水压井作业的成功起到至关重要的作用,需进一步开展风险分析研究。
The deep- water well killing operation is the secondary well control technological barrier,which can effectively prevent the evolution from overflow to blowout accident. In order to improve the reliability of deep- water well killing operation,the BN- CREAM method was applied to study the risk factors. Combined with the characteristics of deep- water well killing operation,a risk evolution model of deep- water well killing operation was established by using Bayesian network（BN） method considering the common cause failure and other factors. The cognitive reliability and error analysis method（CREAM） for human reliability analysis was adopted to calculate the prior probability of human error in deep- water well killing operation,and the prior probability of equipment failure in deep- water well control was determined referring to Offshore Reliability Data（ OREDA）. The main risk nodes of well killing operation were identified relying on the reverse reasoning capacity of BN,thereby the effective prediction and evaluation of risk factors for deep- water well killing operation can be achieved. The results showed that the deep- water well killing operation includes six key risk root nodes,and the human reliability is lower than that of equipment. The third grade child node ＂ unreasonable choice of well killing methods＂ plays a crucial role for the success of deep- water well killing operation,being necessary to carry out a further risk analysis.
出处
《中国安全生产科学技术》
CAS
CSCD
北大核心
2016年第10期86-91,共6页
Journal of Safety Science and Technology
基金
国家工信部第七代超深水钻井平台创新专项
关键词
深水压井工艺
风险识别
CREAM
贝叶斯网络
后验推理
deep-water well killing technology
risk identification
CREAM
Bayesian network
posterior reasoning
