Application of automated hazard analysis by new multiple process-representation models to chemical plants

It has been recognized that traditional hazard-analysis methods have had dr awbacks because they are arduous, tedious, and time-consuming works, requir e multidisciplinary knowledge, and demand considerable cognitive efforts fr om the analysts. To overcome these problems, using multiple modeling concep ts, new process-knowledge representation models for hazard analysis are dev ised. The models consist of the unit function model, the unit behavior mode l, the process structure model, and the process material model. They are us ed to describe chemical processes from a safety-oriented paint of slew. Fro m these models, three hazard-analysis algorithms (deviation analysis. malfu nction analysis, and accident analysis algorithm) are proposed. In this art icle, the overall system, which is embodied using G2 language, is described and applied to olefin dimerization plants. The results show that more poss ible accidents can be identified and that the developed methodology has the ability to capture process hazards in terms of both functional failure and unexpected variable deviations, thereby improving the quality of the hazar d analysis.