基于符号有向图分析的炼油厂中渣油催化裂化危险与可操作性

SDG analysis of residue fluid cracking in an oil refinery for HAZOP study purposes

作为危险与可操作性(HAZOP)研究的一门辅助技术,符号有向图(SDG)分析技术迅速得到普及。仿真工厂数据是流程工业在线数据最理想的近似,能够体现过程变量之间的相互关系。本文以石油化工典型的渣油催化裂化装置为例,将SDG分析与基于机理模型的动态流程模拟仿真工厂结合起来,通过仿真工厂数据形成专家经验,进而在SDG辅助平台实施HAZOP分析,此外,分析结果也可以进一步作为HAZOP分析的知识库。同时,仿真工厂又可以验证事故传播路径,从而起到筛选SDG生成的事故序列的作用。相比于传统的专家组会议而言,将仿真工厂模拟数据运用于SDG分析中,能大幅缩短HAZOP分析时间,减少人员及投资,并且获得更为全面、准确的分析结果。

Signed digraph（SDG）analysis is quickly gaining popularity as an aiding technique for hazard and operability（HAZOP） analysis. The data from a dynamic process monitoring platform is the closest substitute to online data in process industry since it can reflect the relationship between the process variables. This paper obtained model based dynamic process simulation platform to gain expert experiences,using residue catalytic cracking unit typical in petroleum industry as an example for analysis. The model was used in performing the HAZOP study of an existing plant with SDG tools. The process monitoring platform can also be used to check and select the accident path; therefore,the sequence of accidents generated by SDG tools can be determined. The application of process monitoring platform with SDG tools in HAZOP analysis can reduce analysis time and obtain more comprehensive results.