Wye Wye Transformer tripping GFPE on pringle switch

[coop3339]

We have a Wye Wye 350kva transformer with isolated X0 and H0, 4 wire connected in a step up arrangement 120/208 - 277/480. The H0 has been grounded.

The issue is that we have ground fault current showing on the sensor in the pringle switch used as the disconnect on the primary side of about 400-500A and it sometimes trips the switch that is set at 1000A for ground trip. We have tested the conductors and everything seem fine. The neutral on the 208 side is not bonded to ground. We have not tested to see if there is actually any current coming back on the grounds yet.

Because the transformer is wye wye and the secondary H0 in this case is grounded, can this cause problems with the GF trip in the pringle switch on the primary 208V side. Is it possible we can cure this with bonding the X0 and H0 together and eliminating the ground on the H0? It seems unlikely to me that there is actually a constant 400-500A ground fault without any other symptoms. The trans is not loud or hot and no smoke or burning smell. Also has run for months in this condition. I guess it could be a problem with the ground fault trip unit?

 

[jim dungar]

I would look at the primary side neutral point connection. For low voltage non-utility I try to use floating Wye primaries, but there are reasons not to.

 

[coop3339]

The customer doesn't want to deviate from the design the engineer gave them. Lifting the neutral would also eliminate the trans operating in open delta if it loses a primary phase. This is a step up for a solar installation. Some power companies actually have you remove a fuse from the main disconnect to simulate a phase loss. If the solar system doesn't shutdown you fail.

In this case there is also a wye wye ahead of this transformer. It is a MV to 120/208v secondary. The only thing I see a little unusual is that there is a blown fuse on a capacitor bank on the MV line coming in. Also one of the bushings on the MV trans is hotter than the rest. There is only a small voltage variation of 1 or 2 volts between phases so this is probably not the issue. I guess it could cause some uneven shifts in the power factor on the line coming in. Could the lack of a cap on one phase of the feeder cause excessive harmonics that can pass through the wye wye transformers?

I'm thinking it may be the ground fault sensor.

It seems unlikely there could be 400-500A ground fault without other systems. I guess it could be harmonic current?

 

[jim dungar]

Your transformer is a wye connection. It cannot run in open delta, if it did the voltages across the windings would be incorrect Yes, running with a phase loss is one reason to connect a primary .

The last time I had an installation with GFP tripping problems, I removed the H0-X0 bond.

 

[coop3339]

You are correct it would not be open delta, it has been a while since I thought about this. The derived leg is actually from the 3 leg core acting as a delta connection for the flux. This can be eliminated by providing a trans with a 5 leg core or removing the primary neutral connection.

Floating the neutral on the primary side should not be a problem because its already solidly grounded at the source correct? They are not open to this but I think it would be a wise move.

The H0-X0 bond was my first thought. I figured the ground on the secondary was paralleling the neutral and ground on the primary side, causing the GFP to trip. In this case there is not a H0-X0 bond, they are isolated.

 

[jim dungar]

Wye-wye connections have always been of concern. I remember publications from transformer manufacturers, like GE document #GET338B, that covered these issues in a fair amount of detail.

 

[David Castor]

In a grounded wye-grounded wye transformer, ground fault on the secondary side are seen as ground faults on the primary side. So any GF device will see the fault. That's one of main advantages of delta-wye transformers. If you can run the primary winding ungrounded, this will prevent any ground fault current on the primary for secondary ground faults.

 

[coop3339]

Yes that is my understanding also David. One thing I was thinking about in this case is if a bonded neutral on the secondary side could cause the primary GFP to trip. I don't think this would be the case because it is an isolating transformer with no H0/X0 bond. If a bond between the neutrals existed it definitely would. The other thing I was thinking is if the GFP would see zero sequence current on the neutral as a ground fault? I believe the phases also contain the harmonic current but is it canceled out like circuit current or would the harmonic current on the neutral be seen as larger than the sum of the harmonic current in the phases. I would imagine this is not the case but maybe there is something that keeps the currents from canceling out?

 

[texie]

I'm trying to understand why for this application a Y primary is being specified. Is there some reason that a delta 208 primary with a 480 Y secondary was not specified? Maybe I'm not understanding the details?

 

[coop3339]

Some utilities require wye wye for solar. It was not need in this case. The engineer probably did it out of habit.

The Wye Wye definitely complicates things. I agree a Delta Wye would have been a better choice...

 

[synchro]

Is there a chance that the neutral conductor that's connected to the primary of the wye-wye transformer is not passing through the CT for ground fault detection, or that it's going the wrong direction through the CT?

 

[GoldDigger]

Yes that is my understanding also David. One thing I was thinking about in this case is if a bonded neutral on the secondary side could cause the primary GFP to trip. I don't think this would be the case because it is an isolating transformer with no H0/X0 bond. If a bond between the neutrals existed it definitely would. The other thing I was thinking is if the GFP would see zero sequence current on the neutral as a ground fault? I believe the phases also contain the harmonic current but is it canceled out like circuit current or would the harmonic current on the neutral be seen as larger than the sum of the harmonic current in the phases. I would imagine this is not the case but maybe there is something that keeps the currents from canceling out?

One thing I would certainly look out for is any intentional or unintentional ground-neutral bond anywhere on the load side of the GFP. That is pretty much guaranteed to cause a ground fault trip if there is any unbalanced current flowing on the neutral at the inverter. And a GFP device like the GFCIs in consumer use also includes a separate mechanism that will detect a spurious ground-neutral bond regardless of whether any load current is being drawn.

 

[synchro]

Is there a chance that the neutral conductor that's connected to the primary of the wye-wye transformer is not passing through the CT for ground fault detection, or that it's going the wrong direction through the CT?

If the neutral is not properly routed through the CT, and there is a delta tertiary winding, a substantial ground fault current could be indicated even with the secondary disconnected. That's because it's acting as a grounding transformer, and relativly small imbalances from the POCO side can cause high currents when the neutral is connected to the primary X0.

 

[GoldDigger]

If the neutral is not properly routed through the CT, and there is a delta tertiary winding, a substantial ground fault current could be indicated even with the secondary disconnected. That's because it's acting as a grounding transformer, and relativly small imbalances from the POCO side can cause high currents when the neutral is connected to the primary X0.

Indeed.
The same warnings that apply to wye to delta operation (other than by POCO itself), as described in detail as part of multiple threads here, also apply to a wye-wye with a tertiary delta winding. You can trip primary protection even where there is no ground fault element or severely overheat the neutral, which has no OCPD.

 

[synchro]

Indeed.
The same warnings that apply to wye to delta operation (other than by POCO itself), as described in detail as part of multiple threads here, also apply to a wye-wye with a tertiary delta winding. You can trip primary protection even where there is no ground fault element or severely overheat the neutral, which has no OCPD.

I agree with that. The current on the neutral feeding the transformer should be measured to see if it's substantially more than any current that might be drawn on the neutral of the secondary.