Development of a Plant Health Index Monitor

Since 2008, BNF Technology Inc. and Kyung Hee University have developed the 'Plant Health Index (PHI)' which is a software package to detect 'unhealthy conditions' of plant equipment in advance. While the difference between a setpoint and an operational condition is called 'process margin', the residual between an anticipated normal condition and an operational condition is called 'process uncertainty' or 'healthiness' in this study. It is obvious that the anomalies in process uncertainty can be observed earlier than those in process margin, which is the concept of 'early-warning' proposed in the recent condition-based maintenance (CBM) studies. One of the key factors for implementing the early warning capability should be how to expect the anticipated normal conditions using available information. The PHI was developed on the basis of empirical models, and we have published a few papers with regarding to the core technologies of the PHI. However, the overall architecture and features of the PHI have not been introduced to academic area so far. This paper delineates the overview of the PHI, and focuses on the recently developed module, which is the health index generator