Hazard and operability (HAZOP) analysis is a widely recognized process hazards analysis (PHA) technique used for hazard identification in the chemical process industry in the world.To overcome the shortage of the existing HAZOP expert systems with regard to “non-routine” analysis
a case based reasoning framework is proposed in this paper.The case base and the structure of cases are briefly described and the case search and matching strategies are presented.To facilitate the management of the knowledge base
a case constructor is developed.The industrial case study demonstrates that the CBR-based HAZOP expert system has overcome the technical bottleneck of the existing HAZOP expert systems in machine learning and “non-routine” HAZOP analysis.