Transformer Issues

[aelectricalman]

Some time ago I posted a discussion about an ungrounded delta system that has 2.4 volts phase to ground. I know that when evaluating an ungrounded delta there is no reference to ground however, can referring to ground be used to evaluate potential issues?

What happened was that an anomaly was detected on the secondary side of a 12.47kV Delta /480V ungrounded Delta transformer. Phase A on this XFMR was 2.4 volts to ground while the other two phases were 480V. This is indicative of a corner grounded Delta system. The problem was that this is not a corner grounded transformer.

A few of you said that maybe the corner ground comes from a VFD, and sometimes a reference ground is pulled off of one of the phases, to get 277V. I was a little puzzled by that so I dug a bit deeper because I didn't think the situation applied.

What I found is that there is a 480V ungrounded Delta/480V Wye Iso transformer in line between the drive and the primary transformer mentioned earlier. There are three undgrounded conductors coming into the XFMR and three undgrounded conductors coming out of the Iso XFMR. Basically no neutral or grounding (which I understand the grounds will not be derived at the XFMR). In the top of the VFD, there are three ungrounded feeders coming in. There are no rigid or metal conduits attached as the cables ride in a cable tray to the VFD then are transformed to rigid, 1 foot away ( no neutral).

So, with this being said, I do not see how the VFD could be causing this corner ground. With the Iso XFMR as a buffer, the corner ground has to be originating in between the 480v ungrounded Delta secondary and the 480 v high side of the iso transformer, since this anomaly cannot pass through the induction of the Iso transformer.

It can't be from the power company because it is taking place on the secondary side of their XFMR.

That would be a dead short to ground on one of the phases or a defect in the XFMR's (since they are designed and installed at the same time. The interesting thing is that this is not happening only once. The issue is replicated next to it in another 12.47UD/480 UD XFMR. Interesting at minimal.

 

[GoldDigger]

Some time ago I posted a discussion about an ungrounded delta system that have 2.4 volts phase to ground. I know that when evaluating an ungrounded delta there is no reference to ground however can referring to ground be used to evaluate potential issues.

What happened was that an anomaly was detected on the secondary side of a 12.47kV Delta /480V ungrounded Delta transformer. Phase A on this XFMR was 2.4 volts to ground while the other two phases were 480V. This is indicative of a corner grounded Delta system. The problem was that this is not a corner grounded transformer.

A few of you said that maybe the corner ground comes from a VFD, and sometimes a reference ground is pulled off of one of the phases, to get 277V. I was a little puzzled by that so I dug a bit deeper because I didn't think the situation applied.

What I found is that there is a 480 ungrounded Delta/480 Wye Iso transformer in line between the drive and the primary transformer mentioned earlier. There are three undgrounded conductors coming into the XFMR and three undgrounded conductors coming out of the Iso XFMR. Basically no Neutral or grounding. In the top of the VFD, there are three ungrounded feeders coming in. There is no rigid or metal conduits attached as the cable ride in a cable tray to the VFD then are transformed to rigid 1 foot away.

So, with this being said, I do not see how the VFD could be causing this corner ground. With the Iso XFMR as a buffer, the corner ground has to be originating in between the 480v ungrounded Delta secondary and the 480 v high side of the iso transformer, since this anomaly cannot pass through the induction of the Iso transformer.

It can't be from the power company because it is taking place on the secondary side of their XFMR.

That would be a dead short to ground on one of the phases or a defect in the XFMR's (since they are designed and installed at the same time. The interesting thing is that this is not happening only once. The issue is replicated next to it in another 12.47UD/480 UD XFMR. Interesting at minimal.

You still have not told us how "hard" a ground this is. One possibility is simply a (severely) unbalanced capacitance to ground in the transformer windings.
If it is a solid ground (able to carry current of the same order of magnitude as the phase loads), then, yes, it could only be happening between or in the two transformers of each similar circuit.
Among other things the VFDs might have caused severe ringing in the ungrounded input wires, leading to insulation breakdown in the nearest isolation transformers. Which phase failed first would determine which phase the "ground" appeared in, explaining the difference in grounded phase between the two circuits.

I would do a resistance measurement (megger) on the two isolation transformers closest to the VFDs. Chances are the transformers you were initially worried about are just fine, unless you still see the "ground" when the leads to the isolation transformer are disconnected (instead of just opening the disconnect to the VFDs, which removes the load but does not take the second transformer out of the circuit).

 

[aelectricalman]

You still have not told us how "hard" a ground this is. One possibility is simply a (severely) unbalanced capacitance to ground in the transformer windings.
If it is a solid ground (able to carry current of the same order of magnitude as the phase loads), then, yes, it could only be happening between or in the two transformers of each similar circuit.
Among other things the VFDs might have caused severe ringing in the ungrounded input wires, leading to insulation breakdown in the nearest isolation transformers. Which phase failed first would determine which phase the "ground" appeared in, explaining the difference in grounded phase between the two circuits.

I would do a resistance measurement (megger) on the two isolation transformers closest to the VFDs. Chances are the transformers you were initially worried about are just fine, unless you still see the "ground" when the leads to the isolation transformer are disconnected (instead of just opening the disconnect to the VFDs, which removes the load but does not take the second transformer out of the circuit).

I will investigate the grounding at the transformers in question. The impedance at the VFD was .35 Ohms. The voltage is 480/277 on all phases at the VFD.

 

[GoldDigger]

I will investigate the grounding at the transformers in question. The impedance at the VFD was .35 Ohms. The voltage is 480/277 on all phases at the VFD.

I don't expect that you will find anything deliberately bonded (although it would be ideal if you did.) Instead I think you will find an insulation failure of some sort.
What I am having a hard time understanding, though, is what sort of single point of failure could cause one phase line to be close to ground at the VFD and also between the two transformers. I may have to accept that it is a multiple fault situation.
If there is no electrostatic shielding between the primary and secondary of your isolation transformers, then high surge voltages with respect to ground on one side will be coupled through capacitively to the other side. That would not cause a ground resistance as small as you see to appear, but it could explain fault propagation from one side of the transformer to the other.

BTW, which direction on the iso transformer is the wye side? Ideally the wye should be facing the VFD, and you would be able to re-engineer the setup to have a grounded centerpoint, connected or not, at the input to the VFD.

 

[beanland]

Failed Insulation or Arrester

Failed Insulation or Arrester

The only way I know to isolate a ground fault in an ungrounded system is divide and conquer. Meter the phase that the 2.4V and then start turning off circuits. When the voltage jumps back to ~277V, you have found the problem. Keep chasing.

A faulty surge arrester can be a short phase-ground that will cause what you see.

 

[aelectricalman]

What I am having a hard time understanding, though, is what sort of single point of failure could cause one phase line to be close to ground at the VFD and also between the two transformers. I may have to accept that it is a multiple fault situation.

No, the only place the corner ground exists is between the two XFMRs. The Drive is 480V Wye derived at the ISO.

The interesting thing that I wanted to point out is I don't know what the potential difference is between the ISO's case ground and the Drive case ground. Even more interestingly, the drive is fed by 3 ungrounded conductors and no neutral despite the fact it is fed off of the secondary side that is wye fed. There is a ground but it appears to be run after the fact and appears to be to earth and is only good to 35 ohms. There is not a constant rigid conduit from the ISO to the Drive as its in a cable tray most of the way, so the rigid is not a good ground.

On another note, I wonder if the "corner ground" could be a product of the two transformers case grounds not being at the same potential?

FYI, Golddigger, I am finding noise on my grounding meter so maybe you are on to something with the ringing.

I will update you soon.

Thanks again.

 

[aelectricalman]

When the voltage jumps back to ~277V, you have found the problem. Keep chasing.

A faulty surge arrester can be a short phase-ground that will cause what you see.

Thanks. I'll consider this in my search!

 

[aelectricalman]

I don't expect that you will find anything deliberately bonded (although it would be ideal if you did.) Instead I think you will find an insulation failure of some sort.
What I am having a hard time understanding, though, is what sort of single point of failure could cause one phase line to be close to ground at the VFD and also between the two transformers. I may have to accept that it is a multiple fault situation.
If there is no electrostatic shielding between the primary and secondary of your isolation transformers, then high surge voltages with respect to ground on one side will be coupled through capacitively to the other side. That would not cause a ground resistance as small as you see to appear, but it could explain fault propagation from one side of the transformer to the other.

BTW, which direction on the iso transformer is the wye side? Ideally the wye should be facing the VFD, and you would be able to re-engineer the setup to have a grounded centerpoint, connected or not, at the input to the VFD.

That is correct, the wye is facing the VFD.

The system flows as follows;

Utility XFMR 12.47 UG D > 480 UG D
ISO SFMR 480 UG D > 480 WYE
480 WYE > DRIVE

The drive does not have a wye configuration run to it. Just 3 ungrounded conductors and a ground that is pulled in hindsight "that I suspect is not at the potential of the ISO.

 

[aelectricalman]

PQ analysis

PQ analysis

Check out this PQ analysis. Ouch.

https://www.dropbox.com/s/j9xhnk8jjpkbn03/SCAN102.PDF

 

[GoldDigger]

I want to upload a PQ analysis for the forum to look at but it well exceeds the forum limit. The forum will not even accept a public link to dropbox. What do I do in this situation?

You should be able to simply make the link part of the text of a post rather than trying to make it an attachment.

 

[aelectricalman]

I finally saw the light on the attachment gig..


On another note..... This is likely not related to the issue but interesting. We had a 3 phase 1000HP motor come on line and only the A phase shot up to 5600Amps at start up.

On even another note...... I am going to go back to site and set up my meter. Then I'm going to test ground potentials. I will take along piece of wire and temporarily tie the case grounds together between XFMR's to see if the results change. If nothing happens, I will Megger the Iso XFMR and look for insulation breakdown.

 

[aelectricalman]

I think I found the problem. Now I need to find the source! Any thoughts