高要求模式SIS异型冗余结构PFH计算模型

PFH calculation model of SIS with diverse redundant architecture in high-demand mode

为了满足石油化工生产中对高要求操作模式下采用异型设备的安全联锁回路进行SIL定级的需求,避免因误用同型PFH公式导致SIL等级评估误差。考虑各通道差异性及其失效顺序的遍历性,以MonteCarlo仿真值为多元线性回归模型观察样本,以改进共因失效部分多样性修正因子确定方法,提出异型K oo N冗余结构每小时危险失效平均频率(PFH)的计算公式;比较该模型独立失效部分与异型1oo2结构Markov模型PFH结果,并分别将该模型和传统同型PFH公式应用于海上采油平台高完整性压力保护系统(HIPPS)异型关断阀子系统的比较分析。研究结果表明:在不同检测周期内,所提出的PFH计算模型与Markov模型PFH计算结果相对误差均保持在10^(-3)数量级;但当检测周期大于3 a时,使用同型PFH公式会出现对HIPPS子系统SIL等级的误判,造成井口压力联锁保护功能过保护或欠保护。研究结果有助于生产单位准确评估联锁保护风险和设备维护投入。

In order to meet the requirements for safety integrity level(SIL)grading of safety interlock circuits adopting diverse equipment under the high-demand operating mode of petrochemical production,and avoid the SIL rating error due to the misuse of PFH formula with same type,considering the difference between the channels and the ergodicity of failure sequence,the MonteCarlo simulation values were taken as the observation samples of the multiple linear regression model,so as to improve the method for determining the partial diversity correction factor of the common cause failure,and a calculation formula for the average frequency of a dangerous failure per hour(PFH)of the diverse K oo N redundant architecture was proposed.The independent failure part of the model was compared with PFH results of the diverse 1oo2 structure Markov model,then the model and the traditional identical PFH formula were respectively applied in the comparative analysis on the diverse shut-off valve subsystem in the high integrity pressure protection system(HIPPS)of offshore oil production platform.The results showed that the relative error of PFH calculation results between the proposed PFH calculation model and the Markov model were all maintained in the order of 10^(-3) in different test periods.But when the test period was more than 3 years,the use of identical PFH formula would cause the misjudgment on the SIL level of the HIPPS subsystem,causing the wellhead pressure interlock protection function to be over-protected or under-protected.The research results are helpful for production plants to accurately assess the risk of interlock protection and equipment maintenance investment.