PowerFlex 753 ground fault issue

[Ken Moore]

We have a 753, 75 HP, driving a centrifugal pump, that has been giving us ground faults for months. The drive has load, and line reactors. We have changed the motor twice, the drive once, and still have fault issues.
Megging the T leads, shows the wiring is good.

Any thoughts, anyone? Could it be the line reactor, if so, how do we test/check?

 

[baref00t25]

What type of voltage supply is powering the drive, grounded, high resistive ground, or ungrounded?

 

[PLC Pie Guy]

We have a 753, 75 HP, driving a centrifugal pump, that has been giving us ground faults for months. The drive has load, and line reactors. We have changed the motor twice, the drive once, and still have fault issues.
Megging the T leads, shows the wiring is good.

Any thoughts, anyone? Could it be the line reactor, if so, how do we test/check?

Every time I have an issue like this, it turns out to always be the motor circuit.

How long is it from the VFD to the Motor?
Is it VFD cable? Is it shielded?
Meggering sometimes doesn't prove anything I have found (@ 1000V). Plugs, cord connectors even motors and simply cables that are too long can sometimes pass a megger test, but still ground fault on the drive. I had a very long run that would trip the ground fault on a PF525 with no motor connected to it recently! It had me and another electrician stumped for while I tell you. Out of 10 drives in my cabinet, 3 would run the circuit in question. Once the lead was shortened, they would all run it. This was very handy a 300' run. Since the situation allowed, I was able to switch that VFD for a contactor.

 

[Ken Moore]

High Resistance Ground via line reactors on in feed to drive.

Distance from drive to motor is about 300 meters. VFD rated cable, inverter duty motor also.

There are three pump/drive skids, identical. Only the one is having issues.

 

[tarik1978]

can you swap the cable, just to confirm is good

 

[Ken Moore]

Replacing the cable is going to be a major ordeal. It's cable tray rated cable with a special chemical resistant coating. This is all out doors, with a cable tray height of about 10 meters, so not easy to replace. We are going to try disconnecting the motor cables and hooking up a motor locally, down stream of the load reactors. If it passes, we will know it's the leads and not the reactor.

What's strange it it will run for 30-40 hours sometimes, and 3-4 minutes sometimes. Maybe and hour or two at times, before faulting.

 

[goghie]

I had problems same problems with motors with lower insulation rating and also I had similar problems with faulty motor bearings.

 

[Ken Moore]

Follow up:
The drive has ground fault trip without motor or load reactor installed. Next step replace line reactor.

 

[Maxkling]

How can the drive detect a ground fault from the upstream supply?

 

[Ken Moore]

How can a drive without a motor or wires have a ground fault?
We have contacted local AB distributor for support.

 

[EICS]

How can a drive without a motor or wires have a ground fault?
We have contacted local AB distributor for support.

sounds like drive possibly has an intermittent ground tracking issue in its output PWM stage, is it in a dirty environment and IGBT section caked in crud?

 

[Maxkling]

How can a drive without a motor or wires have a ground fault?
We have contacted local AB distributor for support.

I was honestly interested if a drive can detect an upstream fault, .

I have never seen this in an AB drive, that I can remember. But I have in many Altivars with a SCF fault. Unhook the motor and fault is still present, every time it was a faulty drive. With the PF753 being an Architect class drive there maybe some field serviceable components that can fix the fault.

 

[James Mcquade]

what has changed in the plant or in the motor area when the issues started?
is it possible a new transformer, substation, capacitor bank, inductive bank was installed and not properly grounded?
was a transformer moved or rewired?
was new wiring added and maybe nicked the motor wire and a vibration causes it to fault now and then?
just trying to throw ideas out there.
james

 

[Bit_Bucket_07]

I once had a Baldor drive tripping on ground fault, and it turned out to be another motor in the same MCC on a FVNR across the line starter. It was a screw conveyor, and we noticed that the fault always occurred just after it started. Upon inspection, the sealtight was loose in the motor's peckerhead and vibration had worn through one of the motor leads, sparking to ground occasionally as it ran, but never enough to trip it's own MCP.

I don't know how the drive was detecting a ground fault on another device, but repairing the loose sealtight connection stopped the faults from occurring.

 

[Ken Moore]

More information. The drive trips for ground fault, with output leads disconnected, drive energized, but not running.

This morning, we replaced the incoming line reactor, and the drive has been sitting for about 2-3 hours without tripping, an improvement. We are going to reconnect pump motor and test this afternoon. However, we are underneath a tornado watch, so may have to test later this week.