AN ONLINE METHOD TO DETECT INCIPIENT FAILURE OF TURN INSULATION IN RANDOM-WOUND MOTORS

Induction motors used in various mining-industry applications fail frequently because they are subjected to severe operating conditions. Periodic test procedures are inappropriate for many installations, and the cramped and hazardous mining environment makes unscheduled motor replacements difficult and time-consuming to perform. An on-line detector of insulation degradation is proposed, and demonstrated analytically and experimentally. This detector, termed the effective negative-sequence impedance, is computed from the voltage and current phasors at the motor terminals. It can be used frequently, even continuously, and does not interfere with normal motor operation. With this detector, motor replacements can be scheduled, thereby reducing production downtime associated with emergency maintenance actions.