Troublesome electrical issues

Merry Christmas

cpedersen

Member
Location
USA
Occupation
Facilities Mgr
So we have a piece of equipment that requires 310 amps at 400 V. It is the only piece of equipment on the circuit.

The primary side is supplied by the utility with 240 nominal high leg delta. The transformer is a 225 KVA delta / Wye transformer.
The secondary side is 400 volts WYE. No neutral has been brought to the equipment only three hots and a ground.

The equipment has 30 VFD's with low duty cycles, each rated at 30 Amps. Peak current (200ms) has been 120Amps on each leg.

I have monitored the power for the last month. During this time in operation we consistently get the ground showing intermittent 5.5V. Also current, but I cannot recollect amount (an amp or two?) Internal to equipment there is a transformer 400/24 transformer. The VFDs are fed by a power supply, measuring from the VFD to the ground bar, I show 45 ohms while running. The metal case (also supposedly grounded to the bar) only shows 2.6 Ohms.

Total harmonic distortion (current) ranges from 45-70% - which I think contribute to the voltage on the ground.

When turning on, random VFD's will trip going into error conditions. This goes on for about 45-135 minutes. After which the equipment runs all day without incident.

Random problems always brings to mind grounding issues.

I suspect that the voltage on the ground is causing the trips.

Question: What avenue would you attack this?
1). Bring a neutral, with the idea that this would allow us to cleanup the ground? (Does that even make sense?)
2). Install line reactors, with the idea that lowering the harmonic distortion by 10% would result in less voltage on the ground.
3). Same argument as #2, except active filtering would lower the distortion down to 5-8%. (much more expensive).
4). Something else?


Appreciate the thoughts
 
Last edited:

cpedersen

Member
Location
USA
Occupation
Facilities Mgr
Yep, I'm not. Voltage and amperage clamp around each phase, and one around the ground bar. I fixed the original post to reflect. Thanks.
 

cpedersen

Member
Location
USA
Occupation
Facilities Mgr
The Wye is grounded to a ground bar in the disconnect switch, which itself is tied to the ground rod outside. The ground rod shows 30 Ohms.


The cable from the disconnect switch to the ground rod is too small. Fixing that is underway. I haven't sized the grounding cable from the transformer to the disconnect switch, nor from the transformer to the equipment.
 

synchro

Senior Member
Location
Chicago, IL
Occupation
EE
I have monitored the power for the last month. During this time in operation we consistently get the ground showing intermittent 5.5V.
Between what two points are you measuring this voltage? For example, is it between the ground bar in the disconnect and the equipment housing or ground terminal, and utilizing a separate wire hooked to a test lead to make the connection?

The VFDs are fed by a power supply, measuring from the VFD to the ground bar, I show 45 ohms while running. The metal case (also supposedly grounded to the bar) only shows 2.6 Ohms.
So is this power supply providing a common DC bus to all of the VFD's?

I don't think taking a resistance measurement when the equipment is running is very meaningful or reliable. In the mode to measure resistance, a meter is applying a small current and measuring the resulting voltage that's developed to calculate the resistance that's present. Having the equipment running may produce stray currents that can interfere with the measurement.

Does the equipment have more than one terminal that's supposed to be grounded, and if so are they all connected properly?

Are there data cables going to the machine, or are all controls and user interfaces self contained in the machine?
 

cpedersen

Member
Location
USA
Occupation
Facilities Mgr
The equipment has 4 ground bars. They are all bonded to a groundwire leading from the transformer. The metal cabinet is also grounded to the same ground bars - which is why I don't understand the difference in ohms.

I measured cabinet to power outlet ground at 2.6 ohms
Groundbar to power outlet ground while running at 46 ohms, 37, ohms 36 ohms and 26 ohms (4 separate bars).

Regarding between which two points, its the legs at the equipment, and the ground bar in the equipment. Voltage and amperage clamp around the three legs, and on the ground bar.

The equipment is self contained. Data cables run from a computer, to control cards, which go from there to each of the components, predominantly VFD motors. All drawn on common power and ground.

Thank you!
 

topgone

Senior Member
So we have a piece of equipment that requires 310 amps at 400 V. It is the only piece of equipment on the circuit.

The primary side is supplied by the utility with 240 nominal high leg delta. The transformer is a 225 KVA delta / Wye transformer.
The secondary side is 400 volts WYE. No neutral has been brought to the equipment only three hots and a ground.

The equipment has 30 VFD's with low duty cycles, each rated at 30 Amps. Peak current (200ms) has been 120Amps on each leg.

I have monitored the power for the last month. During this time in operation we consistently get the ground showing intermittent 5.5V. Also current, but I cannot recollect amount (an amp or two?) Internal to equipment there is a transformer 400/24 transformer. The VFDs are fed by a power supply, measuring from the VFD to the ground bar, I show 45 ohms while running. The metal case (also supposedly grounded to the bar) only shows 2.6 Ohms.

Total harmonic distortion (current) ranges from 45-70% - which I think contribute to the voltage on the ground.

When turning on, random VFD's will trip going into error conditions. This goes on for about 45-135 minutes. After which the equipment runs all day without incident.

Random problems always brings to mind grounding issues.

I suspect that the voltage on the ground is causing the trips.

Question: What avenue would you attack this?
1). Bring a neutral, with the idea that this would allow us to cleanup the ground? (Does that even make sense?)
2). Install line reactors, with the idea that lowering the harmonic distortion by 10% would result in less voltage on the ground.
3). Same argument as #2, except active filtering would lower the distortion down to 5-8%. (much more expensive).
4). Something else?


Appreciate the thoughts
This a harmonics-laden system!
Q: What was your load amps when you took THD-i measurements? I am asking this so you'll know whether your problem leans on having too much harmonics. You have to compute for the TDD there and compare with standards.
E.g. one of our systems has 17.87% THD-i; the load was 1387 A and a max. demand of 2,405 A, and the computed TDD was 10.3% (which is above the suggested limit of 5% (for a system with a short-circuit ratio below 20kA). I am trying to recommend a detuned filter to solve this.
 

synchro

Senior Member
Location
Chicago, IL
Occupation
EE
When turning on, random VFD's will trip going into error conditions. This goes on for about 45-135 minutes. After which the equipment runs all day without incident.
Can error codes be accessed on the VFD's to help isolate the root cause of their trips?

Internal to equipment there is a transformer 400/24 transformer. The VFDs are fed by a power supply, ...

Total harmonic distortion (current) ranges from 45-70% - which I think contribute to the voltage on the ground. ...

Question: What avenue would you attack this?
1.) ...
2). Install line reactors, with the idea that lowering the harmonic distortion by 10% would result in less voltage on the ground.
Line reactors would reduce the THDi somewhat, but they may increase the THDv at the equipment line inputs because the THDi will produce corresponding harmonic voltages across their impedance. Harmonics on the supply voltage might affect other circuits inside of the equipment such as those fed by the 400/24 transformer that was mentioned. But that is an unknown.

Is there just a single set of three connections (terminals, etc.) in the equipment to supply the 400V 3-phase to everything inside of it? Or is there a separate set of terminals to supply the VFD's that allows line reactors to be inserted at that point, and another set of terminals for the other internal devices so they can be fed directly and not experience an elevated THDv from the reactors?

When you say "The VFDs are fed by a power supply" is this a rectifier or switching supply providing a common DC bus voltage to all of the VFD's, or is it something else?
 

winnie

Senior Member
Location
Springfield, MA, USA
Occupation
Electric motor research
Regarding between which two points, its the legs at the equipment, and the ground bar in the equipment. Voltage and amperage clamp around the three legs, and on the ground bar.

So your metering system is connected to the 3 phase conductors and the ground bar, and nothing else?

If so, then the voltage number on the ground is probably meaningless.

By definition voltage is measured between two points, such as phase to phase voltage or phase to ground voltage, or neutral to ground voltage.

To measure neutral to ground voltage your meter needs to be connected to neutral and ground. You've only described one of these.

The meter might be grabbing a ground reference from its power supply, or it might be reported the voltage between a neutral 'derived' from the phase connections and the neutral/ground clamp, or it might be reporting a completely meaningless number. You need to figure out what the actual reference is before you try to fix the 'ground voltage' issue (if it is an issue at all).

Next: have you tried placing an oscilloscope on any of your ground conductors? I bet you will see a huge hash of random high current spikes from the 30 drives.

The way a drive works the connected motor winding ends up being 'switched' from the internal + DC point to the - point and back at the 'PWM' frequency. This high frequency switching gets coupled to the motor frame by capacitive connection, and shows up as high frequency current on all the grounded metal in the system.

This capacitive coupled switching current is flowing back to the transformer neutral. This noise current can wreak havoc with lots of systems.

I'd suggest looking for such noise current and then seeing if it changes between the initial starting problems and the fine running state.

One wild ass guess: at initial power up the VFD clocks start at the same time, so the switching events are synchronized and the noise from different drives adds up. Then the clocks drift apart and while the total noise stays the same the peaks get lower.

I don't think my WAG is really likely, but I think you will find the source of your problem if you set up to look at the switching noise.

Jon
 

cpedersen

Member
Location
USA
Occupation
Facilities Mgr
This a harmonics-laden system!
Q: What was your load amps when you took THD-i measurements? I am asking this so you'll know whether your problem leans on having too much harmonics. You have to compute for the TDD there and compare with standards.
E.g. one of our systems has 17.87% THD-i; the load was 1387 A and a max. demand of 2,405 A, and the computed TDD was 10.3% (which is above the suggested limit of 5% (for a system with a short-circuit ratio below 20kA). I am trying to recommend a detuned filter to solve this.
Completely agree that this is harmonics laden. The fluke meter has a harmonics display which I will post up also.

Our load amps runs between 20-40 A per leg usually while measuring THD current. Usually the loads are fairly well balanced. The distortion always runs 45-70%. I will upload pictures this afternoon.
 

cpedersen

Member
Location
USA
Occupation
Facilities Mgr
So your metering system is connected to the 3 phase conductors and the ground bar, and nothing else?

If so, then the voltage number on the ground is probably meaningless.

By definition voltage is measured between two points, such as phase to phase voltage or phase to ground voltage, or neutral to ground voltage.

To measure neutral to ground voltage your meter needs to be connected to neutral and ground. You've only described one of these.

The meter might be grabbing a ground reference from its power supply, or it might be reported the voltage between a neutral 'derived' from the phase connections and the neutral/ground clamp, or it might be reporting a completely meaningless number. You need to figure out what the actual reference is before you try to fix the 'ground voltage' issue (if it is an issue at all).

Next: have you tried placing an oscilloscope on any of your ground conductors? I bet you will see a huge hash of random high current spikes from the 30 drives.

The way a drive works the connected motor winding ends up being 'switched' from the internal + DC point to the - point and back at the 'PWM' frequency. This high frequency switching gets coupled to the motor frame by capacitive connection, and shows up as high frequency current on all the grounded metal in the system.

This capacitive coupled switching current is flowing back to the transformer neutral. This noise current can wreak havoc with lots of systems.

I'd suggest looking for such noise current and then seeing if it changes between the initial starting problems and the fine running state.

One wild ass guess: at initial power up the VFD clocks start at the same time, so the switching events are synchronized and the noise from different drives adds up. Then the clocks drift apart and while the total noise stays the same the peaks get lower.

I don't think my WAG is really likely, but I think you will find the source of your problem if you set up to look at the switching noise.

Jon

Hi Jon


I agree with almost everything you said.

a). Yes the metering is connected to the 3 phase, and the ground bar and nothing else. I don't have anything else to connect to. There is no neutral, and no other grounds.

I agree with your prognosis that the meter might be grabbing a ground reference from its power supply, from a derived neutral (more likely).

But I don't know how to get an actual reference considering the elements described. This is why I was asking should a neutral be extended to the equipment. How involved is this likely to be?

b). Yes,the fluke meter (435-ii) has pretty good oscilloscope like functions - and I have mapped a great deal of noise on the ground bar. I also have oscilloscope functions built into the drives.

c). "The way a drive works the connected motor winding ends up being 'switched' from the internal + DC point to the - point and back at the 'PWM' frequency. This high frequency switching gets coupled to the motor frame by capacitive connection, and shows up as high frequency current on all the grounded metal in the system."

Don't know enough to comment, except to say the drives are switching at 6khz, and that generally I agree with what you say.
The drive / motor cable is grounded at both ends (theoretically) to the motor frame and then again at the drive.

d). There is noise and current on the ground. I don't know how narrow it down to high frequency.

e). Running state and starting state is exactly my suspicion. I can demonstrate very large inrush currents for the motors; I have not noticed a synchronization to the ground, except that every time the machine is turned on there is a very large spike on the ground.

f). What would you suggest to look for switching noise?
 

cpedersen

Member
Location
USA
Occupation
Facilities Mgr
Can error codes be accessed on the VFD's to help isolate the root cause of their trips?


Line reactors would reduce the THDi somewhat, but they may increase the THDv at the equipment line inputs because the THDi will produce corresponding harmonic voltages across their impedance. Harmonics on the supply voltage might affect other circuits inside of the equipment such as those fed by the 400/24 transformer that was mentioned. But that is an unknown.

Is there just a single set of three connections (terminals, etc.) in the equipment to supply the 400V 3-phase to everything inside of it? Or is there a separate set of terminals to supply the VFD's that allows line reactors to be inserted at that point, and another set of terminals for the other internal devices so they can be fed directly and not experience an elevated THDv from the reactors?

When you say "The VFDs are fed by a power supply" is this a rectifier or switching supply providing a common DC bus voltage to all of the VFD's, or is it something else?
Can error codes be accessed on the VFD's to help isolate the root cause of their trips?


Line reactors would reduce the THDi somewhat, but they may increase the THDv at the equipment line inputs because the THDi will produce corresponding harmonic voltages across their impedance. Harmonics on the supply voltage might affect other circuits inside of the equipment such as those fed by the 400/24 transformer that was mentioned. But that is an unknown.

Is there just a single set of three connections (terminals, etc.) in the equipment to supply the 400V 3-phase to everything inside of it? Or is there a separate set of terminals to supply the VFD's that allows line reactors to be inserted at that point, and another set of terminals for the other internal devices so they can be fed directly and not experience an elevated THDv from the reactors?

When you say "The VFDs are fed by a power supply" is this a rectifier or switching supply providing a common DC bus voltage to all of the VFD's, or is it something else?
Yes. I mentioned the errors previously, but I will upload the help files for the error codes momentarily. The errors are the drives go to inhibit (lack of 0 on braking resistor), phasing error, and OHT inverter error.

There are multiple switched power supplies in the equipment. Each of the power supplies drives a number of VFD's.

Q: Is there just a single set of three connections (terminals, etc.) in the equipment to supply the 400V 3-phase to everything inside of it?

There is a bus that provides 400 V, which feeds devices attached.
There are 400 to 24 V AC power supplies that feed the drives.

Q: Or is there a separate set of terminals to supply the VFD's that allows line reactors to be inserted at that point, and another set of terminals for the other internal devices so they can be fed directly and not experience an elevated THDv from the reactors?

A: Yes, I think we could insert line reactors between the power supply and the drive.
 

__dan

Senior Member
Anytime you measure 5.5 Volts from one grounded point to another on an industrial system. I would automatically think that the path between them has a high resistance somewhere. So it's a reading under current flow that drops voltage at or due to the resistance. Two grounded points require a path is effective at carrying fault current(s), so dropping 5 volts under load with no faults present is not "effective".

The isolating distrubution transformer sounds correct to give you grounded Y secondary to feed the drives. The secondary Y is seperately derived and has a GEC to the ground rod (#6 max to the ground rod may meet code idk). The transformer secondary has a system bonding jumper that connects the neutral, the GEC, all other Earth ground systems nearby, the building steel, and the equipment grounding system.

The drives do not use the neutral. The drive front end has a filtering system, MOV's and caps, that are Y connected and grounded. These devices connected line to equipment ground can pass substantial current, are intended to and connected to an intended rated effective ground fault path. I believe there can also be some substantial running noise due to this, the caps may try to balance some reactive imbalance, connected line to EGC.

So if I expect some running current flow on the EGC path and then I measure some at a GEC or EGC conductor coming from the drives and landed on one of the common grounding busbars, I may (or may not) think the cause is normal to the runtime. The EGC is used for an intended or coincident current flow by the drive front end Y connected filter caps.

Measuring any Voltage drop on the grounding points is a problem, indicating a resistance in the path that should not be there. Some connection to the rest of everything grounded is missing.

I did not check any loads number. Transformer heating could be an issue with the all high harmonic drive load.
 

__dan

Senior Member
OK I see the Emerson motor driver board in the picture. I am assuming that's typical for all 30 and the motor load Voltage is around the supply Voltage, 400 V.

Compared to how beefy a standalone premium regular drive unit could be, this is just the board and could be built to be cheaply disposable. The only thing you could do for the board is heat dissapation, make sure the heatsinks are bolted to what they're supposed to be bolted to and the fans are running. I would not expect that board mounted input module to have much in the way of line side filtering, the filtering would be before the bridge rectifier. Same for the manufacturer's input harmonic filtering, probably not a lot in that input module, would be another input filter on the board or before it. Looks like nothing for the DC link bus from the rectifier to the inverter.

So you would have to live with some harmonics and noisy grounds issues, or the equipment is built for those conditions. For the thing that is causing the outages and suspected Voltage related, there should be no Voltage raises anywhere on the grounding system, even as the drive uses the grounding system for some noisy currents. Some drive noisy currents are normal.

The drive board grounding should be extensively done, from the motor to the drive and from the drive to the supply. Any resistance in the drive grounding path between itself, the motor, and the drive, certainly anything giving a Voltage raise over 1 Volt could trip a drive.
 

cpedersen

Member
Location
USA
Occupation
Facilities Mgr
Yes. All the drives are Emerson M700s; most are 28 are 30 amp and there are 1x32, 3x27, 4x5A and I think a 7.5 amp. I have more info on the drives. Pics attached.

I have not checked voltages between the drive and the motor. But the ground shield seems to be well done and grounded at both ends.

The units include a small module about half the size of a dollar bill that is billed as an EMC filter. I see nothing on heat syncs, however, the units are in a climate controlled cabinet that is kept around 27C.

1637786570479.png 1637786555987.png 1637786594134.png
 

__dan

Senior Member
I would take it to a certain point and then I would be quick to call Emerson. I would be looking to tap their experience with the fault codes and get their recommendations. They will have a good idea of which fault codes point to bad connections on the grounding or where else to look.

At second look that drive is likely a pretty beefy standalone unit. I usually like everything Emerson builds. I would try to work on the fault codes with the factory.

Anything with a Voltage raise on the grounding should not have a Voltage raise.
 
Top