VFD Controlled Motor Fried

[ESMD]

I have come across four 7 1/2 HP Baldor motors, each one controlled by a Danfoss Graham VSD (VFD). Two of these motors supported two chiller units in one building and the other two supported two chiller units in the building adjacent. All four motors went bad on the same day around the same time. The megger shows no insulation between the motor windings and ground. How can this be possible? What could be the cause, or have all four motors failed simultaneously for no apparent reason? and if so how is it possible that all twelve fuses have blown in the vfd cabinets and all four breakers tripped at the 480 volt panels, yet did not stop the motors from burning up? Any ideas?... Keep in mind, these are two completely separate buildings.

 

[StephenSDH]

Any storms roll through your area at the same time? Any electricians roll through the area?

I just had a customer who lost every siemen's drive on his line of a certain type. During a storm the utility lines had 2 phases bolted together so he had A, B, and B. Apparently the type of drive couldn't protect itself.

 

[ptonsparky]

Danfoss drives do everything but hand you the paper, if they are told to. Programming was no doubt identical on all drives. Something missed perhaps? Major external event already mentioned?

The burning up motor and drives are most likely what caused the fuses and breakers to operate, not the other way around.

 

[StephenSDH]

ptonsparky is right that it could be programming. If you have the wrong gains the motor can cook until the insulation fails, even though it is not turning. How long were the drives running before they burned up? I haven't used that drive before, but all of the higher end models have an auto-tune feature. Before another motor burns up call the rep and have him walk you through configuring the drive and then autotuning it. ptonsparky is also right that the breakers/fuses likely a result of the insulation failure, and don't have anything to do with why it happened.

 

[nakulak]

this is one of the reasons why its usually worth the incredibly small investment in phase loss monitors.

 

[ESMD]

In Addition

In Addition

Thank you gentlemen. This helps. But I have an addition that may add to the confusion. The two chiller units also tripped around the same time at the panel as well. The max protection is factory reccommended at 250 amps, but the OCP was at 350 amps. What in the world would cause the breakers to trip if the compressor motors on the chiller units are all meggering out just fine? Has anyone ever come across anything like this before? Also these two buildings were built in 2005. The motors, vfd's, panelboards and chillers are only 5 years old.

 

[Jraef]

The chances of "programming related" component failures all happening at the same exact time in 4 different drives in 2 different sites is astronomically small. All evidence points to an event of some sort.

Something I just saw recently that I never saw before was a VFD failure like that where there was a packaged piece of equipment on rubber vibration isolators that was properly grounded when originally installed, but some time in the last year or so (when this was happening a lot), someone stole the EGC. So the packaged equipment had been "floating" for who knows how long and an overhead cable fell across it. VFD transistors are an easy place for energy to transfer and once it happens, everything blows.

Were your chillers on a packaged air handler? Did you check the integrity of the ground connection? Ground conductors on rooftop air handlers were easy targets for copper thieves because nobody could see them cutting wires up there. A passing storm could have discharged onto framework attached to them, and the lack of a ground path would make the transistors an easy target.

 

[ESMD]

I have not, but will do so. Thank you.

 

[charlie k.]

Just curious what the meggar readings are. I am thinking you had lost a phase somewhere. Could be possible that the utility dropped a phase momentarily. It does not take long.

Charlie

 

[StephenSDH]

Just curious what the meggar readings are. I am thinking you had lost a phase somewhere. Could be possible that the utility dropped a phase momentarily. It does not take long.

Charlie

I was wondering about phase loss, and I would understand that if it was a soft start instead of a vfd. If it is a VFD I would think a phase loss would only effect the diode bridge and not the motor. I am not familiar with compressors though.

 

[Cow]

Wouldn't the VFD simply have faulted out upon phase loss? Or could the programming have tried to keep it running with just two(?) phases until it all burned up?

 

[jim dungar]

Wouldn't the VFD simply have faulted out upon phase loss? Or could the programming have tried to keep it running with just two(?) phases until it all burned up?

Yes it is possible to program some VFD's to ignore input phase loss. This allows them to act as 1-phase to 3-phase converters, but only if the output load is reduced.

 

[mcclary's electrical]

I was at a storage facility a few months back that had smoked 4 heat pumps simulaneously. I found a faulty phase loss monitor that would not not open the service disco. Once I replaced that with not only a phase loss monitor, but a low voltage monitor also, it has never happened since(crossing my fingers and knocking on wood.)

 

[sparky=t]

i have seen a vfd taken out by lightning on the secondary side of the drive, most times back feeds are not protected

 

[ATSman]

Restriking Grounds on a HRG System

Restriking Grounds on a HRG System

The chances of "programming related" component failures all happening at the same exact time in 4 different drives in 2 different sites is astronomically small. All evidence points to an event of some sort.

Something I just saw recently that I never saw before was a VFD failure like that where there was a packaged piece of equipment on rubber vibration isolators that was properly grounded when originally installed, but some time in the last year or so (when this was happening a lot), someone stole the EGC. So the packaged equipment had been "floating" for who knows how long and an overhead cable fell across it. VFD transistors are an easy place for energy to transfer and once it happens, everything blows.

Were your chillers on a packaged air handler? Did you check the integrity of the ground connection? Ground conductors on rooftop air handlers were easy targets for copper thieves because nobody could see them cutting wires up there. A passing storm could have discharged onto framework attached to them, and the lack of a ground path would make the transistors an easy target.

I tend to agree with Jraef. Any simultaneous failures I have seen are almost always related to an overvoltage condition. One thing to consider here is the phenomenon of the restriking ground on an ungrounded or HRG system.
Even though the VFDs may be fed from a Y transformer with a grounded neutral, if they are at a great distance from the supply transformer then it could act like a HRG system. A restricking (intermittent) ground can cause the phase voltages on this system to rise sharply and puncture winding insulation in transformers, circuit boards & motors. This is exactly what happened several years back on the shuttle train project I was on at SFO airport (see attached root cause analysis report.)
A loose traction motor brush was touching ground on a HRG system while the train ran on the track and caused the ground detection control transformers to fail and trip the smoke alarms in all 5 substations at the same time.

I would go thru the grounding system with a fine tooth comb and look for
cut wires or loose connections. Good luck.

 

[Jraef]

Yes it is possible to program some VFD's to ignore input phase loss. This allows them to act as 1-phase to 3-phase converters, but only if the output load is reduced.

Some VFDs detect and act on input phase loss, some do not. It really is somewhat irrelevant because a VFD doesn't know or care if the supply is 3 phase or single phase. All it is doing with the incoming power is converting it to DC, incoming AC is just a "raw material source". Thatis why VFDs can be used as "phase converter. It affects the output rating of the drive, as jim said, but not necessarily the operation. If the VFDs had no input phase loss detection (or it was disabled), all that would happen is that the VFD would attempt to continue running the motors with the 3 phase output, but the input current would increase by the sq. root of 3, so it would jump by 1.732X. MAYBE if the SCPD did not open at that increased current the input diode bridge would fail and/or the caps would fry. But that doesn't tend to cause a cascading failure of everything else down stream. VFDs have a lot of protection systems and all they need to do to protect the motor is disable the transistor firing circuit (called a base block), one of the first things they do at any sign of trouble.

But if the trouble is coming from the other direction, i.e. from the motor side back into the transistors, a sufficient voltage spike would make the base block irrelevant.

 

[wirenut1980]

I tend to agree with Jraef. Any simultaneous failures I have seen are almost always related to an overvoltage condition. One thing to consider here is the phenomenon of the restriking ground on an ungrounded or HRG system.
Even though the VFDs may be fed from a Y transformer with a grounded neutral, if they are at a great distance from the supply transformer then it could act like a HRG system. A restricking (intermittent) ground can cause the phase voltages on this system to rise sharply and puncture winding insulation in transformers, circuit boards & motors. This is exactly what happened several years back on the shuttle train project I was on at SFO airport (see attached root cause analysis report.)
A loose traction motor brush was touching ground on a HRG system while the train ran on the track and caused the ground detection control transformers to fail and trip the smoke alarms in all 5 substations at the same time.

I would go thru the grounding system with a fine tooth comb and look for
cut wires or loose connections. Good luck.

Can you elaborate more on how you can get high phase-phase voltage on a solidly grounded wye service during a restriking fault? I had thought that on an ungrounded system you could get high phase to ground voltages, but a resistance grounded wye worked to limit these voltages. So even if on long runs the solidly grounded wye "looked" like a resistance grounded wye, I don't see how that could result in high voltage.

Thanks!:grin:

 

[Besoeker]

I have come across four 7 1/2 HP Baldor motors, each one controlled by a Danfoss Graham VSD (VFD). Two of these motors supported two chiller units in one building and the other two supported two chiller units in the building adjacent. All four motors went bad on the same day around the same time. The megger shows no insulation between the motor windings and ground.

I'm inclined to agree with the others that it sounds like an overvoltage transient. But did the VFDs survive the incident.

 

[ATSman]

Can you elaborate more on how you can get high phase-phase voltage on a solidly grounded wye service during a restriking fault? I had thought that on an ungrounded system you could get high phase to ground voltages, but a resistance grounded wye worked to limit these voltages. So even if on long runs the solidly grounded wye "looked" like a resistance grounded wye, I don't see how that could result in high voltage.

Thanks!:grin:

Let's back up a bit. The OP did not indicate whether the source transformer
is a grounded wye or an ungrounded delta. If it is an ungrounded delta then
a striking ground condition caused by one of the motor windings intermittently going to ground could have damaged all 4 motors. If it is a grounded wye and somehow the neutral to ground connection got cut or loose, corroded,etc then it could act as an ungrounded or HRG system causing the same results. (This is why I said to check all ground connections.)All the books say that, yes, with a HRG system the
phase to phase voltages are limited and cannot increase to damaging levels. I only mentioned the train incident to point out that I saw it with my own eyes that the insulation damaging overvoltage can occur on a HRG system with a solid state train traction motor system, which is basically a VSD.
Maybe someone else can add to this thread since this topic is not talked about very often?