Avoiding Motor Insulation Breakdown via PWM VSI 5

[HenryOhm]

All,

I have read some whitepapers that indicate that IEC 60034-18-41 & -42 are superior to American standards NEMA MG 1, Section 31, IEEE 522, etc when it comes to helping avoid premature insulation breakdown of motors downstream of PWM VSI's:

"New IEC standards for qualifying stator insulation systems for PWM converter drives" (Stone & Stranges)

"IEC 60034-18-41: A New Draft Technical Specification for Qualification and Acceptance Tests of Inverter Duty Motor Insulation" (Stranges, Stone, & Bogh)

"Stator Insulation Problems Caused by Variable Speed Drives" (Culbert, Lloyd, Stone)

In my application, I am specifying motors likely to be 690VAC, 1-3MW in size, with form wound coils. Can anyone in their experience speak to the advantage of the IEC over NEMA/IEEE and whether IEC 60034-18-41 or IEC 60034-18-42 would apply to my case?

Thanks for any and all help!

 

[Drivesrock]

Here is another paper for you to read.

http://www.gambica.org.uk/web_image...tion_Voltage_Stresses_Report1_3rd Edition.pdf

It does have some useful information.

For these higher rating motors you mention and their likely application, it would also be advisable for the inverter manufacturer to inform the PWM characteristics like the pulse rise times so that this can be sent to the motor manufacturer.

 

[LionelHutz]

MG-1 for the most part just specifies a faster rising waveform so it wouldn't be hard for another standard to be more comprehensive.

The only standard that would be useful is one that has tests the motor has to pass, such as a PD level test, to determine the suitability of the motor. Otherwise, the manufacturer can claim anything they want about the insulation yet that won't stop the motor from failing.

I would expect form wound motors to be good on a VFD unless the manufacturer really screwed-up when building them. It's the small sized mush wound motors where the coils are thrown in and allowed to be touching such that phase-phase voltage or near phase-phase voltage is applied directly across the enamel of two wires without other insulation which typically have issues.

 

[vfdmotor]

A few years ago I was advised that certain types of Medium and High Voltage drives do not have the same dv/dt spiking issues as low voltage drives. (Article attached)
Of course Robicon (at that time, now Siemens) wanted to promote thier design benefits. In doing so they compared five types of MV drives.
1) Load-commutated inverter
2) Neutral-point clampedinverter
3) Filter-commutated current-fed inverter
4) GTO/IGCT current-fed inverter
5) Multi-level series-cell inverter

Types three four and five above had no dv/dt issues so the MV Motor insulation is not a problem as the switching spikes normally don't exceed 1000V.
Robicon, Toshiba (same as GE, Mitsubishi and Tmeic) are type five above. there may be others I dont know of.

(Note: Current fed inverters typically require an isolation transfomer as an add-on. Type fives usually have a transformer in house already.)

So I suspect that depending on the MV VFD type, insulation breakdown is not a problem. Even Random wound MV Motors should have no issue with types 3 through 5.

In MV Drive applications the more important obstacles to watch are 1)Voltage drop on long cables, 2) induced circulating currents on the shaft and frame 3) Induced capacitive voltage on the shaft.

 

[electricuwe]

From the five inverter topologies mentioned by vfdmotor, the only one you might find in the power ( 1..3 MW )and voltage range (690 V) mentioend by the OP is the Neutral-Point Clamped Inverter. But the vast majority of designs in this range is using standard two-level topology with 1700 V IGBT-modules.

If dV/dt issues need special consideration depend on the complete drive setup.

- New motor or inverter to operate an existing motor ?
- lenght of cable to motor
- type of cable
- diode frontend or active frontend
- grounded or ungrounded

In most cases a simple dV/dt is sufficient to limit the dV/dt. If possible rely on a supplier that can offer the complete drive setup.

 

[HenryOhm]

Thanks for all the great posts. My understanding was that low voltage and medium voltage motor insulation systems downstream of a PWM VSI were treated quite differently in IEC 60034-18-41 vs -42. The former (Type 1) attempts to eliminate any chance of partial discharge (PD). The latter (Type 2) tries to manage anticipated PD throughout the life of the motor. I'm glad I'm in the former low voltage territory!

But, am uncertain whether -41 was specific to only random wound machines or could also apply to form wound. Per LionelHutz, sounds like I will just have to buy the -41 IEC standard. My understanding is that testing is a part of these, thus why they're better than MG-1(?).

My project will have a new motor (same contractor as drive supplier), cable lengths will be footprint and installation dependent, and type of cable and possibility of AFE's a contractor decision.