An Integrated Approach to Remaining Life Prediction of Rotating Machines

Abstract

State-of-the-art decisions related to repair and replacement of rotating machines in any process or industrial environment is still based on qualitative engineering judgment which often tends to be arbitrary and conservative in nature and has potential for loss of revenue and more importantly net production. The subject acquires significant dimension as number of rotating machines, for example, induction motors are very high, say of the order of 10–1000 or more for a plant like nuclear or process plant. Thus, there is a need to have a science or rational based approach for the plant managers to take decisions related to maintenance or ageing assessment based on well defined quantitative metrics or criteria. Industry experience suggests that prediction of health of insulation in an induction motor is one of the important parameters that require attention towards characterizing the life of the machine. This paper presents R&D work being performed on predicting the remaining useful life of insulation of the induction motors. The focus of this R&D is on development of an integrated framework where data driven approach is integrated to physics-of—failure approach towards developing robust model for predicting the remaining life of insulation.