A probabilistic model for evaluating the remaining life of evaluating the remaining life of electrical insulation in rotating machines

A Markov model is introduced for estimating the remaining life of electrical insulation in large motors and generators. The model incorporates various levels of insulation deterioration and periodic (minor) maintenance. The model permits computation of the average time before insulation failure occurs. A mathematical formulation of the expected transition time from any deterioration state to the failure state (expected remaining life) is given. Data requirements are discussed, and numerical examples are presented. This model would also allow assessment through sensitivity studies of the relative effects of various other parameters in the process which are controllable, e.g. maintenance policy and stress level. Even though it is theoretically possible to apply the Markov models to any particular insulation deterioration mechanism, this approach is best applied to an overall insulation determination problem which combines the effects of all individual processes. The models are fairly flexible. They allow, for example, representation of changing stresses acting on the machine insulation as the utility operating environment changes.<
>