基于故障树分析法的液氨泄漏事故分析

Renovated fault analysis method for ammonia leakage based on the fault tree analysis

采用故障树分析法对液氨储罐泄漏事故进行分析。通过定性分析找到导致顶事件发生的重要因素;在定量分析过程中,处理含有模糊不确定因素的底事件时,采用模糊集理论与专家综合评判相结合的方法。该方法将专家自然语言通过运用模糊数学的相关知识转化为模糊数,再利用左右模糊排序法转换为模糊失效概率。结合液氨储罐泄漏事故树的分析,给出了相应的防范措施。

This paper is inclined to explore and examine ammonia leakage by using fault tree analysis method from the perspective of qualitative and quantitative analysis, in combination of the fuzzy set theory via the expertise comprehensive evaluation method to obtain the failure probability of the basic events in a fuzzy and uncertain manner. For our research purpose, this paper has first of all analyzed the mechanism of the ammonia leakage and set up a fault tree of am monia leakage in hoping to get the structural importance degree coeffi cient of the basic events through quantitative and qualitative analysis to work out the top event probability. We have paid particular atten tion to the fuzzy set theory and the expertise in making comprehensive evaluation to deal with the basic events with fuzzy factors in qualita tive analysis. First of all, the method we have resorted to change the natural language of experts to fuzzy numbers via fuzzy math means. And, then, we have managed to convert the natural value into fuzzy numbers with fuzzy ranking method. And, finally, we have changed fuzzy probability data into fuzzy failure probability ones. Thus, inte grating the results of the qualitative and quantitative analysis of am monia leakage fault tree results, it has made us clear that the reckless actions of human beings and ammonia tank corrosion have been found being the main causes leading to ammonia leakage, and, therefore, the human operators of the chemical installations and facilities should be warned to take necessary precautions to avoid the said leakage based on the smallest cut sets. On the other hand, careful study of the structural importance of the degree coefficient, the probability im portance of the safety coefficients may help to provide the reference needed for the routine maintenance and safety prevention. The results of the trial application of our renovated method show that it is feasible to deal with the probability of fuzzy and unpredicted events by joining the fuzzy set theory and the expertise comprehensive evaluation. Thus, it can be seen that it is possible to analyze the ammonia leakage by the fault tree analysis method from the point of view of qualitative and quantitative analysis.