Transformer question

[ggunn]

I have three 20kW 480V 3 phase 4 wire inverters that need to interconnect with a 208V service through a step down transformer. I have been advised not to use a wye to wye transformer, so the minimum 60kVA transformer will have to be 480V/277V wye on the primary and 208V delta on the secondary. My question is this: do I ground the neutral on the primary side?

My guess is no.

I'll hang up and listen...

 

[jim dungar]

If your service is 208V then you need a step-up transformer. They are commercially available, form every major manufacturer, with a 208V delta primary and a 480Y/277 secondary.

Unless you are a utility, you never never never never ground a primary wye side connection.

 

[chris kennedy]

My question is this: do I ground the neutral on the primary side?

No, 250.6 and 250.24(A)(5) come to mind.

 

[GoldDigger]

If your service is 208V then you need a step-up transformer. They are commercially available, form every major manufacturer, with a 208V delta primary and a 480Y/277 secondary.

Unless you are a utility, you never never never never ground a primary wye side connection.

Jim,

Interesting confusion here. GGunn is calling this a step-down transformer because the power is going from the 480 volt side to the 277 side. But the 277 side is POCO.

The 277 (POCO) side will be a delta. But on the inverter side (his primary, your secondary ) it needs to be a wye because of the design of the inverters (phase to neutral)

There will have to be a neutral from the inverter connected to the X0 of the 480 side of the transformer, and unless either the inverters or the X0 provides a bond from neutral to ground the inverter AC output will be totally ungrounded. I do not think this is what is wanted or needed.

I imagine that this would be considered a separately derived system on the 480 side and should be treated that way per code, with its neutral bonded to ground. It is just that the inverters themselves are a separately derived source rather than a load on the separately derived system originating at the transformer. :blink:

 

[jim dungar]

Jim,

Interesting confusion here. GGunn is calling this a step-down transformer because the power is going from the 480 volt side to the 277 side. But the 277 side is POCO.

The 277 (POCO) side will be a delta. But on the inverter side (his primary, your secondary ) it needs to be a wye because of the design of the inverters (phase to neutral)

There will have to be a neutral from the inverter connected to the X0 of the 480 side of the transformer, and unless either the inverters or the X0 provides a bond from neutral to ground the inverter AC output will be totally ungrounded. I do not think this is what is wanted or needed.

I imagine that this would be considered a separately derived system on the 480 side and should be treated that way per code, with its neutral bonded to ground. It is just that the inverters themselves are a separately derived source rather than a load on the separately derived system originating at the transformer. :blink:

The OP listed one side as 480Y/277 and the other side as 208V. In your explanation did you really mean 277V?
In this situation the terms HV and LV maybe more appropriate than primary and secondary.

As I picture your description I see a 208V delta connection to the POCO and a 480Y/277 connection to the inverters. The 208V side should not have a neutral connection at the same time that the 480V side has one.

 

[GoldDigger]

The OP listed one side as 480Y/277 and the other side as 208V. In your explanation did you really mean 277V?
In this situation the terms HV and LV maybe more appropriate than primary and secondary.

As I picture your description I see a 208V delta connection to the POCO and a 480Y/277 connection to the inverters. The 208V side should not have a neutral connection at the same time that the 480V side has one.

Exactly how I interpret it, and thank you for catching my error.

The GTI inverters have a 277 volt output and are wye connected to the HV side of the transformer. I think the OP was actually asking whether to bond the center wye terminal on the HV side. (H0???)
The POCO side is 208 volt delta I guess. I suspect that POCO derives it from a wye secondary and the neutral just does not appear at the service point, although it certainly could be corner grounded. In any case that does not matter.

 

[jim dungar]

Exactly how I interpret it, and thank you for catching my error.

I used the phrase step-up, really just to indicate the type of commercially available transformer.
I am sure someone would be willing to custom build a Y-D step down, but why go through the hassle.

 

[ggunn]

The GTI inverters have a 277 volt output and are wye connected to the HV side of the transformer.

Actually, the inverters have 480Y output (AE/REFUsol 020's). The neutral is only a reference line; it carries no current. It should only care about the center point of the wye, not absolute ground. I will call their tech support to confirm that.

 

[GoldDigger]

Actually, the inverters have 480Y output (AE/REFUsol 020's). The neutral is only a reference line; it carries no current. It should only care about the center point of the wye, not absolute ground. I will call their tech support to confirm that.

The question that leaves is whether or not the inverters will work properly if there is absolutely no ground reference on the four wire output network. That is definitely a matter for their tech support. If the neutral carries any current at all (even minuscule for sending?) NEC requires it to be grounded.

BTW, I would not call the output 480Y, since 480 is the phase to phase voltage. I would call it 480 Delta with a reference wye point.
Am I correct that the neutral is there only because UL requires a test for phase voltage balance before operating? Logically the same protection could be achieved be comparing the magnitude (and phase?) of the three line-to-line voltages. The check for line to neutral balance and phase to phase balance are sort of the same thing in a split phase system.

 

[jim dungar]

BTW, I would not call the output 480Y, since 480 is the phase to phase voltage. I would call it 480 Delta with a reference wye point.

You can have grounded wye and ungrounded wye, you cannot have a "delta with a reference wye point". Delta does not mean 3-wire, Wye does not mean 4-wire.

The connection of the individual single phase inverters is either 277V or 480V. If it is 480V, connect them into a delta. If it is 277V then they get connected into a wye. The neutral point of the wye can be grounded or ungrounded, regardless the correct term is 480Y. Some sources include a G (e.g. 480GY) to indicate grounded, especially for >600V systems.

 

[GoldDigger]

You can have grounded wye and ungrounded wye, you cannot have a "delta with a reference wye point". Delta does not mean 3-wire, Wye does not mean 4-wire.

The connection of the individual single phase inverters is either 277V or 480V. If it is 480V, connect them into a delta. If it is 277V then they get connected into a wye. The neutral point of the wye can be grounded or ungrounded, regardless the correct term is 480Y. Some sources include a G (e.g. 480GY) to indicate grounded, especially for >600V systems.

Thanks Jim! My problem was in putting together "If it is 480V, connect them into a delta." with calling it 480Y. The OP's system voltage is, in the NEC, called 480Y/277 as opposed to just 480. So 480Y is just an abbreviation of 480Y/277, also written just 480/277. The classification of sources and loads as delta or wye separate from classifying the circuit as delta or wye gives me problems from time to time.

 

[ggunn]

REFUsol tech support says bond the neutral of the 480Y at the transformer to ground.

 

[jaggedben]

Am I correct that the neutral is there only because UL requires a test for phase voltage balance before operating? Logically the same protection could be achieved be comparing the magnitude (and phase?) of the three line-to-line voltages.

Not all inverters require a neutral, so it isn't (exactly) that UL requires a neutral. Apparently some inverter manufacturers have decided to solve the measurement and listing issues with a neutral, and some have decided to solve them other ways.

 

[Smart $]

REFUsol tech support says bond the neutral of the 480Y at the transformer to ground.

If the inverters output truly is 480Y, then bonding H0 to ground would create a path for neutral current on the grounding system. See 250.6.

 

[GoldDigger]

If the inverters output truly is 480Y, then bonding H0 to ground would create a path for neutral current on the grounding system. See 250.6.

Doesn't the fact that this is a separately derived system affect that? There is no reason for neutral current to flow on the grounding system since the entire secondary side of the transformer is isolated from the primary neutral. What would you expect to cause neutral current (other than a fault to ground inside the inverter or the transformer, and in that case it would not be "normal" neutral current.

What makes this different from any other separately derived system? If instead you connect the HV neutral to the LV neutral, it would be an issue, but the LV is delta, so you cannot.

 

[Smart $]

Doesn't the fact that this is a separately derived system affect that? There is no reason for neutral current to flow on the grounding system since the entire secondary side of the transformer is isolated from the primary neutral. What would you expect to cause neutral current (other than a fault to ground inside the inverter or the transformer, and in that case it would not be "normal" neutral current.

What makes this different from any other separately derived system? If instead you connect the HV neutral to the LV neutral, it would be an issue, but the LV is delta, so you cannot.

Assuming the neutral point at the inverter is bonded to ground, a picture worth a few words...

 

[GoldDigger]

Assuming the neutral point at the inverter is bonded to ground, a picture worth a few words...
Fair enough, but from what ggunn stated initially, I did not see any indication that the neutral point of the inverter set was grounded. I still don't, because if the GTI is connected, as is most common, directly to POCO rather to a separately derived system, the inverter manufacturer would know that they were not allowed to bond neutral to ECG. The inverter is a power source which has to be wired as if it were a load.
Takes a little getting used to...

So in answer to the original question, there needs to be a bond either at H0 or at the inverter. And the inverter manufacturer suggests that it be done at H0.

 

[ggunn]

Assuming the neutral point at the inverter is bonded to ground, a picture worth a few words...

The inverter company tech support was very specific that the neutral is a voltage sensing line only and not a current path. They also said that the transformer manufacturer would probably require the X0 be bonded to a ground rod.

To your drawing, the neutral is not bonded to ground at the inverter.

 

[Smart $]

Assuming the neutral point at the inverter is bonded to ground, a picture worth a few words...
Fair enough, but from what ggunn stated initially, I did not see any indication that the neutral point of the inverter set was grounded. I still don't, because if the GTI is connected, as is most common, directly to POCO rather to a separately derived system, the inverter manufacturer would know that they were not allowed to bond neutral to ECG. The inverter is a power source which has to be wired as if it were a load.
Takes a little getting used to...

So in answer to the original question, there needs to be a bond either at H0 or at the inverter. And the inverter manufacturer suggests that it be done at H0.

The inverter company tech support was very specific that the neutral is a voltage sensing line only and not a current path. They also said that the transformer manufacturer would probably require the X0 be bonded to a ground rod.

To your drawing, the neutral is not bonded to ground at the inverter.

That's fine as long as there is only one neutral to ground bond in the 480V system.