A blended hazard identification approach to support intelligent diagnosis in process systems

In spite of the advances in hazard identification (HAZID) and fault diagnosis, process system failures and major industrial accidents continue to occur. An important aspect in addressing this situation is the necessity of improved knowledge generation through new HAZID techniques and the subsequent application of novel diagnostic methodologies to improve fault diagnosis. In this work we introduce a novel blended hazard identification (BLHAZID) method based on a conceptual model called the Functional Systems Framework (FSF). The FSF provides a model that represents the principal physical, human and informational components and interactions of a complex system that determine system functionality (Cameron et al, 2007). Based on FSF concepts, the BLHAZID technique blends both function-driven and component-driven analyses, typically represented by HAZOP and FMEA, to improve the depth and coverage of identified hazards and elicit causality information. The results of the technique are expressed in a structured language to improve knowledge exploitation and reuse. The BLHAZID results are used to support advanced intelligent diagnosis that aids operator performance. The current methodology has been illustrated in an industrial case study.