VFD Causing Bearing Failure 5

[richanton]

I have searched on this and found some discussion, but I am curious as to what anyone has found to be a good solution to the problem. In my case, I had a new 200HP motor/pump setup(pumping chilled water) that the bearings failed after about 6 months. The motor is fed by a new IGBT VFD. A single ground wire initiating at the source is run straight through to the motor and grounded at the termination box. There are currently no chassis grounds.

I have found much discussion online, particularly a white paper by Reliance Motor, but I was looking for some real solutions that have worked.

Right now we are looking at a ring assembly sold by a company called Shaft Grounding Systems. I am also considering adding some chassis grounds, and/or refeeding the motor from the VFD with continuously corrugated cable.

Any input is welcome.

 

[davidbeach]

Shaft grounding should go a long way toward solving your problem, if it is not a complete solution.

 

[DickDV]

Shaft grounding rings are usually the end of the problem and are made by ElectroStatic Technology in Mechanics Falls Maine and Parker Inc, the hydraulics people.

 

[Drivesrock]

And make sure the motor is well grounded and that this is continuous to the drive and the feeding supply neutral if you have. The bearing currents have to go somewhere.

 

[ozmosis]

You should look to using a low impedance chassis ground too, like a braided cable or metal strap. These are low cost and part of the solution.
Does your motor have insulated NDE bearings?

 

[richanton]

The motor does not have insulated bearings.

 

[ozmosis]

If the bearing has been replaced and the cause is seen to be ciculating currents pitting the bearing, it might have been/would be prudent to fit insulated bearings in the NDE.

 

[DickDV]

Insulated bearings are a poor substitute for grounding rings in my opinion. First, it is expensive to do. Second, they often continue to pass the highest frequency pulses since the insulation is very thin, and third, insulated bearings simply move the problem elsewhere as in the gearbox or pump bearings unless you insulate the coupling too (Not so much a problem if belt coupled).

Grounding rings can be installed in place without motor disassembly and have the added benefit of blocking some contaminants from getting into the bearing.

 

[starkopete]

Try this link, it has a lot of good information.

http://www.reliance.com/pdf/motors/whitepapers/raps868.pdf

 

[ozmosis]

DickDV
I agree, insulated bearings are expensive and I wouldn't generally advise unless they were taking damaged ones out and refitting new ones. It might make sense to do this at this time.
Richanton
Also have a look at this:

http://www.gke.org/beppe/files/Bearing currents on one page 1.pdf

designed and developed by one of our own top contributors in Eng-Tips. No affiliation to me but if he is looking for an agent in Australia...

 

[LionelHutz]

DickDV is right about insulated bearings alone. But, an insulated bearing on the NDE may be required to be used in combination with a shaft grounding brush on the drive end should to fix it.

Is this a Weg motor by chance?

 

[richanton]

Thanks for all the input. I have a much better idea on how to proceed.

That link to the Beppe meter is very interesting. The Swedish engineer even gave a name to it, EDM.

 

[dpc]

DickDV,

I thought insulated bearings were required on a NEMA "inverter-duty" motor?

"An 'expert' is someone who has made every possible mistake in a very narrow field of study." -- Edward Teller

 

[DickDV]

No, dpc. Sadly, there are no specs or even guidelines regulating terms like "inverter duty", "inverter rated", and "inverter ready".

Every manufacturer does their own thing on these terms. For some, like the former Reliance, they are very conservative. For others, any Insulation Class F motor qualifies for at least Inverter Ready.

For my money, stick with MG1 Part 31. A motor with that endorsement on the nameplate has been tested with high frequency pulses up to 1600V and has high temp grease in the bearings. Even MG1 P31 does not require insulated bearings.

 

[starkopete]

Hi,
The following is the section out of MG 1 that pertains to “Definite-Purpose Inverter-Fed Polyphase Motor” bearings.

"31.4.4.3 Shaft Voltages and Bearing Insulation
Shaft voltages can result in the flow of destructive currents through motor bearings, manifesting themselves
through pitting of the bearings, scoring of the shaft, and eventual bearing failure. In larger frame size motors, usually 500 frame and larger, these voltages may be present under sinusoidal operation and are caused by magnetic dissymmetries in the construction of these motors. This results in the generation of a shaft end-to-end voltage. The current path in this case is from the motor frame through a bearing to the motor shaft, down the shaft, and through the other bearing back to the motor frame. This type of current can be interrupted by insulating one of the bearings. If the shaft voltage is larger than 300 millivolts peak when tested per IEEE 112, bearing insulation should be utilized.
More recently, for some inverter types and application methods, potentially destructive bearing currents
have occasionally occurred in much smaller motors. However, the root cause of the current is different. These drives can be generators of a common mode voltage which shifts the three phase winding neutral potentials significantly from ground. This common mode voltage oscillates at high frequency and is capacitively coupled to the rotor. This results in peak pulses as high as 10-40 volts from shaft to ground. The current path could be through either or both bearings to ground. Interruption of this current therefore
requires insulating both bearings. Alternately, shaft grounding brushes may be used to divert the current
around the bearing. It should be noted that insulating the motor bearings will not prevent the damage of other shaft connected equipment.
At this time, there has been no conclusive study that has served to quantify the relationship of peak voltage from inverter operation to bearing life or failure. There is also no standard method for measuring this voltage. Because of this, the potential for problems cannot consistently be determined in advance of motor installation."

 

[richanton]

LionelHutz asked above if the motor in question was a WEG motor. I just went and checked. In fact, it is a WEG NEMA Premium, and is not an inverter duty motor. But regardless, I wouldn't think that would affect the bearings used, or would it.

 

[DickDV]

You shouldn't blame the motor manufacturer or the bearings. You simply need to ground the shaft. As the MG1 article describes, this bearing problem comes up once in a while and fortunately is easy to fix as an afterthought.

 

[Skogsgurra]

I am sorry that I cannot partake in this dicussion. I could probably add substance to it. But I will not.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[mccla5214]

Do you have a line reactor on the OUTPUT side of the VFD going to the motor...

One poster was correct the current has to go some where, that is while the chassis of the motor, vfd and the gounding conductor coming in on the supply all have to be bonded, especially between the motor and VFD. If you check the case of the motor to the vfd ground you will find AC current on the ground when they are gounded together. I have seen motors that were being driven by a VFD draw over 100 amps on the grounding conductor. The voltage would be low but the current would be high. I have also been shocked hard from motors that were not grounded properly, this also resulted in grounding currents on the shaft of the motor. Anytime you have gounding current on the motor shaft you will have arcing across the bearings. Shaft grounding kits only work when the currents are coming into the motor from the shaft end. The kit grounds out the currents on the shaft before it can reach the bearings. Same reason you never ground a welder to a structure and then weld somewhere else so current winds up flowing across a bearing surface. The kit can bypass the bearing but only if it is grounded to the motor itself.

A pocket voltage tester like the Fluke tester can help in determining if the motor leakage is high enough to be discharging to the case and grounding is not good enough.

Something else to check is to megger the motor at 1000V (because you are running a VFD) and see what the leakage is. You could be running in the low megohms which will allow the case of the motor to be energized without tripping the motor. I found one like that and it ran for months before it failed. We haven't asked how far the motor run is from the drive or how it is run (PVC, EMT, RIGID, ALUMINUM). The longer the run the higher the ringing voltages are across the motor and this results in higher voltage spikes on the motor insulation. That is one of the things that an inverter duty motor takes care of, it uses a much higher class of insulation (I believe Class H) which reduces the leakage to the motor frame.

I wish you luck and would recommend installing a 3-5% line reactor on the output of the drive if the run to the motor very long.

Mike

 

[itsmoked]

Skogs; I don't see why you can't participate. So you sell devices that prevent this.. You can still participate! Your specialist view would be welcome.

Keith Cress
kcress -

http://www.flaminsystems.com