Study Of The Computer-Aided HAZOP Analysis For Batch Processes Based On Deep-Seated Knowledge Model

Aiming at system safety, process safety assessment identifies, analyses, and estimates potential hazards as well as possible losses caused by the hazards in a process or project from the system point of view according to scientific procedure and method, which provides reference for corporations to establish safety measures and management policies. Out of the many safety assessment methods in existence, SDG-HAZOP method is the most widely used one because of its systemic process and integrated result.After surveying most of the existing software and techniques of process safety assessment home and abroad, it is found that little research has been done to the HAZOP analysis for batch process. Though the traditional SDG method has good effects on the HAZOP study for continuous process, it can not be used for batch processes directly.In this paper, a new modeling method which integrates Petri Net deeply into SDG model is presented. This model aims at HAZOP study for batch process specially, with SDG as its main body and Petri Net the assistant. Firstly, the concept of dynamic SDG is brought in, i.e., a whole SDG model is established first, and the validity of branches in the model is determined by the states of operable devices in Petri net, so we can get different models in different phases of the process. Secondly, guide words related to sequence such as "early" and "late" are brought in through Petri Net in order to express wrong operation sequences which are peculiar in batch process. As Petri Net can well describe the discrete-event character of batch process, and directed graph can explicitly represent the continuous nature of it, the combination of them makes the presented model an effective tool for HAZOP study of batch process.