Unbalanced open delta loads

[kclark8799]

Is A-N utilized for anything? Is there a building at this pump station or just a control panel?

I'm wondering if xA or xB is loose inside the center tapped transformer

They replaced the original transformer with 2 additional sets of transformers. I know new doesn't mean much anymore. Not much on the 120 volt side. Control panel only

 

[kclark8799]

Are you saying that you get these unbalanced current measurements when across-the-line, as well as when feeding the motor with a VFD?

A suggestion is to put a current clamp around all of the conductors feeding the VFD (i.e., A, B, and C) to verify that you don't get a significant current reading. That's just to make sure there isn't some other current path involved. It's not likely to be a problem, but it's probably easy to check.

Done this with a Fluke data logger

 

[Joethemechanic]

Control panel only

That more or less eliminates any kind of hidden stuff. It's got to be something upstream of the utility power terminals of the transfer switch. On your end that doesn't leave much, I guess just a main breaker and some service conductors.

Is this an aerial drop, or underground from the pole?

Have you checked the transfer switch contacts?

 

[kclark8799]

Done this with a Fluke data logger

That more or less eliminates any kind of hidden stuff. It's got to be something upstream of the utility power terminals of the transfer switch. On your end that doesn't leave much, I guess just a main breaker and some service conductors.

Is this an aerial drop, or underground from the pole?

Have you checked the transfer switch contacts?

Thanks for the ideas. I took the power company s feeder from the meter can Polaris some 1 kv gen. Cable and landed it to the main in the control panel (by passing everything on customer side.) Then had the Poco throw the jacks in on the pole. Stayed the same. Beginning to think Poco has a lead lag issue. Need proof.

 

[synchro]

It is not uncommon to have unbalanced VFD input currents when fed from an open-delta (see paper from Yasakawa below).
The line-to-line output impedances of the open delta will not be symmetrical. For one thing, the center-tapped transformer is typically larger and so the impedance across A and B will be lower than that across B and C from the smaller transformer. And the impedance across the "open" part of the delta between C and B will be the highest because the secondaries of both transformers are in series.
Of course, the open circuit voltage on the transformer outputs will also have a major affect on the balance between VFD input currents.

If you don't already have a line reactor and/or DC link choke in the VFD, they could help equalize the input currents.

Everything was checked both ways. Start p reports 3 years ago showed issues. Now that we sized all gen sets using VFDs the dc buss keeps faulting the drives. Sometimes during the prime time they will not reset. Dirty power what I call it.

Line reactors might also reduce voltage spikes and other distortion that could be causing such issues.

Paper frpm Yasakawa:
https://www.yaskawa.com/delegate/ge...PS.01&cmd=documents&documentName=PR.PS.01.pdf

This article is somewhat dated but still has some relevant info:
https://www.ecmweb.com/content/article/20891755/beware-of-linecurrent-unbalance-on-vfds

 

[ptonsparky]

It is not uncommon to have unbalanced VFD input currents when fed from an open-delta (see paper from Yasakawa below).
The line-to-line output impedances of the open delta will not be symmetrical. For one thing, the center-tapped transformer is typically larger and so the impedance across A and B will be lower than that across B and C from the smaller transformer. And the impedance across the "open" part of the delta between C and B will be the highest because the secondaries of both transformers are in series.
Of course, the open circuit voltage on the transformer outputs will also have a major affect on the balance between VFD input currents.

If you don't already have a line reactor and/or DC link choke in the VFD, they could help equalize the input currents.




Line reactors might also reduce voltage spikes and other distortion that could be causing such issues.

Paper frpm Yasakawa:
https://www.yaskawa.com/delegate/ge...PS.01&cmd=documents&documentName=PR.PS.01.pdf

This article is somewhat dated but still has some relevant info:
https://www.ecmweb.com/content/article/20891755/beware-of-linecurrent-unbalance-on-vfds

Supposedly all VFDs are set the same for operation with hi leg open delta if they are supplied by it.

Why just this motor when operated with the VFD, or ATL without it, on this feeder?

 

[zbang]

Supposedly all VFDs are set the same for operation with hi leg open delta if they are supplied by it.

I'd still like to know if the generator is wired as a high-leg delta, corner-grounded delta, or wye.

 

[Joethemechanic]

I thought he bypassed the VFD and did ATL running of the pump and the problem persisted

 

[synchro]

I thought he bypassed the VFD and did ATL running of the pump and the problem persisted

I tried to clarify that with post #40 below, but I didn't seem to get a clear answer.

Are you saying that you get these unbalanced current measurements when across-the-line, as well as when feeding the motor with a VFD?

A suggestion is to put a current clamp around all of the conductors feeding the VFD (i.e., A, B, and C) to verify that you don't get a significant current reading. That's just to make sure there isn't some other current path involved. It's not likely to be a problem, but it's probably easy to check.

Reply:

Done this with a Fluke data logger

 

[Joethemechanic]

Measuring with load and no load. ATL will work but still have 16 amps on A, 34 on B and 34 on C.

Yeah he did ATL

 

[synchro]

ATL will work but still have 16 amps on A, 34 on B and 34 on C.

Yeah he did ATL

That was his post that I replied to in post my #40 to get further confirmation that he measred the same currents ATL as with the VFD (which would not normally be expected).

 

[kclark8799]

It is not uncommon to have unbalanced VFD input currents when fed from an open-delta (see paper from Yasakawa below).
The line-to-line output impedances of the open delta will not be symmetrical. For one thing, the center-tapped transformer is typically larger and so the impedance across A and B will be lower than that across B and C from the smaller transformer. And the impedance across the "open" part of the delta between C and B will be the highest because the secondaries of both transformers are in series.
Of course, the open circuit voltage on the transformer outputs will also have a major affect on the balance between VFD input currents.

If you don't already have a line reactor and/or DC link choke in the VFD, they could help equalize the input currents.




Line reactors might also reduce voltage spikes and other distortion that could be causing such issues.

Paper frpm Yasakawa:
https://www.yaskawa.com/delegate/ge...PS.01&cmd=documents&documentName=PR.PS.01.pdf

This article is somewhat dated but still has some relevant info:
https://www.ecmweb.com/content/article/20891755/beware-of-linecurrent-unbalance-on-vfds

Thank you. We have 5 percent line reactors. Do you think if the smaller transformer was changed to match the other one that this may help ? Thanks for the article, I ll get to reading.

 

[ptonsparky]

That was his post that I replied to in post my #40 to get further confirmation that he measred the same currents ATL as with the VFD (which would not normally be expected).

When would be expected at all?

 

[ptonsparky]

Thank you. We have 5 percent line reactors. Do you think if the smaller transformer was changed to match the other one that this may help ? Thanks for the article, I ll get to reading.

Are these the only two motors on this transformer bank?

 

[synchro]

That was his post that I replied to in post my #40 to get further confirmation that he measred the same currents ATL as with the VFD (which would not normally be expected).

When would be expected at all?

Maybe when they were both turned off.

 

[Joethemechanic]

he measred the same currents ATL as with the VFD

Yeah that is kind of strange

 

[synchro]

Thanks for the ideas. I took the power company s feeder from the meter can Polaris some 1 kv gen. Cable and landed it to the main in the control panel (by passing everything on customer side.) Then had the Poco throw the jacks in on the pole. Stayed the same.

It sounds like the control panel in question is close to the service point. If there are other VFD's on the same transformer bank that are working but they are significantly further away, perhaps there's additional impedance in the conductors that's helping balance the currents, etc. That's just a guess without further info.

 

[kclark8799]

Are these the only two motors on this transformer bank?

Yes it is.

 

[synchro]

Thank you. We have 5 percent line reactors. Do you think if the smaller transformer was changed to match the other one that this may help ? Thanks for the article, I ll get to reading.

Increasing the size of the smaller transformer might help somewhat, but there will still be two transformer secondaries in series supplying one of the L-L voltages.
Also, it's possible that there could be issues on the MV side of the transformers in terms of voltage, impedance, etc.
I think load bank measurements could be useful to quantify how the L-L voltages and phase shifts change vs. load current.

 

[kclark8799]

Increasing the size of the smaller transformer might help somewhat, but there will still be two transformer secondaries in series supplying one of the L-L voltages.
Also, it's possible that there could be issues on the MV side of the transformers in terms of voltage, impedance, etc.
I think load bank measurements could be useful to quantify how the L-L voltages and phase shifts change vs. load current.

I'll try a load bank. I believe Florida rural water has one. Thanks for all inputs and advices. Ken