Monitoring off-line and on-line PD under impulsive voltage on induction motors - Part 2: testing*

摘要

Large numbers of motors driven by variable-speed drives have failed prematurely over the past 20 years. Increased partial discharge (PD) activity in the insulation is known to be the main reason for these failures, causing faster insulation degradation and accelerating aging, and leading to loss of reliability [1]?????????[13]. Whereas Type II motors (rated voltage ?????????700 V and generally form wound) are designed to withstand PD activity during their working life [14], Type I motors (rated voltage <700 V and generally random wound) are not expected to experience PD activity. To test for PD in Type I motors fed by variable-speed drives, the International Electrotechnical Commission (IEC) released a technical specification (TS) in 2006 [15] describing impulsive or sinusoidal off-line tests to be carried out on Type I insulation systems (including complete stators). Although at first glance the TS appears appropriate for assessing PD in stator insulation resulting from impulses from power electronic drives, in practice it is ambiguous and inaccurate [16] and is currently under review. It is evident that a number of offline and on-line tests must still be carried out, because few correlations between off-line tests and on-line motor performance [13], [17] have been reported in the literature. In this article, the second of a series of three dealing with the correlation between off-line and on-line PD on Type I motors fed by pulse-width-modulation (PWM) drives, the results from the on-line PD monitoring of four random-wound motors are presented. The first article (Part 1 of the series) [16] reported the results obtained from application of the off-line test procedure to these four motors, in accordance with the TS [15]. In the third article some concerns regarding the current TS will be discussed and some critical aspects emphasized. The off-line type tests described in the first article [16] were carried out bearing in mind the characteristics of the PWM converter,-n-n e.g., impulse rise time, overshoot factor at the motor terminals, length of the cables between the converter and the motor, and the rated voltage and number of converter levels. One would expect that motors passing the type test would not experience PD when connected to the converter. However, one of the conclusions of the first article was that motors B1 and C1 failed the type test, whereas motors B2 and A1 passed it.