Study On Improved HAZOP Analysis In Petrochemical Unit

In order to more effectively improve the HAZOP(Hazard and Operability) analysis method in the petrochemical unit, the traditional HAZOP was improved. Firstly, we established a node criticality evaluation model based on fuzzy comprehensive evaluation. In the model, the set including 1st level and 2nd level of evaluation factor set, review set and the membership functions of each factor was built. Then, AHP was used to determine their relative weigh of all evaluation factors. Finally, based on fuzzy comprehensive evaluation result of two grading levels, the node criticality grade was obtained. An application in a hydrocracking plant was performed as a case study. The result showed that the quantitative evaluation grades of the node criticality for all nodes in the plant were achieved with the use of the proposed approach. Therefore, the approach provided theoretical basis for following simplified HAZOP.Based on the node criticality, the principle of HAZOP deviation analysis was established. For nodes of the very important grade, all the process parameter deviation of each device was analyzed. The process parameter deviations were considered in full range combined guide words with process parameter. For nodes of the important grade, the deviations of analysis focused on important devices and others were simplified. For nodes of the general grade, the deviation of devices with small fault effect was ignored. As a result, comparing with traditional HAZOP, the amount of deviations was obviously reduced. The efficiency of HAZOP was improved.In order to achieve quantitative HAZOP analysis, the dynamic fault tree method was introduced in HAZOP. According to HAZOP analysis, the dynamic fault tree was built. The method of BDD(Binary Decision Diagram) and Monte Carol simulation were introduced to calculate the probability of top event and bottom event importance. An application of hydrocracking reactor overtemperature as top event for dynamic fault tree was performed as the case study. The result showed that dynamic fault tree could more truly reflect the dynamic characteristic of petrochemical unit comparing with traditional fault tree analysis method. So the results could provide more reliable basis of decision for maker. The result of HAZOP analysis was improved.