Online technique for insulation assessment of induction motor stator windings under different load conditions

This paper presents a new condition monitoring system which is able to evaluate and diagnose the stator winding insulation deterioration (SWID) at its initial stages. During the last few decades, considerable efforts have been made in condition monitoring of SWID in induction motors. Nevertheless, an effective diagnosis technique that can lead to a fast and accurate detection of SWID shortly after its occurrence remains a challenge. A new modelling technique is introduced using finite element method which enables detailed simulation (two turn-to-turn resolution) of insulation degradation in a stator slot. Experimental setup is then used to validate the proposed modelling technique. The detail and approximation coefficients of the discrete wavelet transform are shown to provide effective detection of SWID within a few seconds after its occurrence. The proposed system continuously monitors the status of the induction motor and is able to detect the insulation deterioration of only 1.2% turns of phase winding when it starts to take place. The proposed method is only applicable to turn insulation failure detection. Its viability under different load conditions is validated for a small-sized induction motor with simulated turn insulation fault.