In 1995, it was determined that the cause of failure for electric motors in inverter (variable frequency drive) applications was partial discharge, or small lightning storms occurring in voids within the insulation system.
Before that, several articles in commercial trade magazines published the suggestion by one particular engineer that the failures occurred between the first two turns of the first coil in the winding. This was what became the ‘voltage spike’ failure hypothesis.
Partial discharge is like a lightning storm inside your motor, invisible yet relentlessly destructive.
While dV/dt (voltage rise over time) is a thing, meaning that conductor insulation must be able to survive specific limits:
- For motors rated 600 Volts or less, the insulation needs to withstand peak voltage up to 3.1 times the line-to-line motor voltage. In a new insulation system it has been shown that these values in a brand new machine can be greater than 15kV in a 460 Vac motor.
- The rise time for the voltage must be at least 0.1 microseconds (i.e., ~1200 V/0.1 microseconds).
- Original windings (Chapter 30 of the NEMA MG1) were approximately 1000V peak with a 2 microsecond rise time.
In a typical worst-case scenario, considering only rise time conditions, a typical worst-case rise time at the motor terminals falls between 0.1-0.3 micro-seconds rise time, which is why the limits were set in the standard. This assumption holds if there are no output filters or built-in dV/dt output conditioning in the drive.
The primary culprits are left as two-fold. The first is the partial discharge condition, in which there would be loose conductors, conductors out of order, or other winding conditions that provide a space for conductors to vibrate or have an air gap between them (under 1 mm).
The second relates to surface currents on the electric machine, which is a topic beyond the scope of this article. The first are little lightning storms that generate ozone, which degrades the insulation material until a small short occurs between conductors.
Pinholes from partial discharge are so small that most winding testers won’t even see them.
The inverter then identifies it as a current spike, resulting in a drive trip that appears as an overcurrent trip, or similar. One of the challenges is that these pinholes in the insulation system are so small that you could run the motor across the line, and most winding testers will not detect them.
Testing for Partial Discharge Using IEC/TS 61934
How do you determine if your motor is at risk?
That’s where IEC/TS 61934, “Electrical Insulating Materials and Systems – Electrical Measurement of Partial Discharges (PD) Under Short Rise Time and Repetitive Voltage Impulses,” comes in. This test utilizes a surge comparison tester specifically designed to detect discharges during the pulses. In this test, you are looking for conditions that will most likely indicate that the motor winding is susceptible to partial discharge.
The first you will be looking for is the repetitive partial discharge inception voltage (RPDIV), which is the minimum surge voltage at which more than five discharges occur on ten voltage impulses of the same polarity.
If the peak motor voltage (not the RMS) falls within this range, then the winding will be susceptible. The second is the repetitive partial discharge extinction voltage (RPDEV), in which the discharges fall to under five discharges per ten pulses.
What the Results Mean for Motor Reliability
In a new electric machine, you would want there to be no discharge through the testing range, which is twice the voltage plus 1000 Volts on the pulses. As a machine ages, its value generally decreases, so it is essential to conduct a basic condition analysis of the machine in an inverter application.
It is also essential to understand that a good test will not guarantee that the insulation system will not fail in an inverter application, such as due to high-frequency common-mode currents.
