Y-Y step down transformer for photovoltaic system

[hmspe]

A colleague has several new PV sites where the PV inverters are at 480V but the service is 208V. The step-down transformers are 30KVA and the utility has required that they be Y-Y. The PV system uses the neutral on the 480V side for monitoring only, but a neutral connection is required.

The neutral current on the 208V side is unusually high and the phase currents are badly imbalanced if the neutral on the 208V (utility) side is connected. The currents are balanced if the 208V side neutral is lifted, but the utility requires the neutral be connected. The 208V side has neutral and ground bonded by the MBJ in the electrical service. On the 480V side neutral and ground are bonded in the transformer. The representatives from the transformer manufacture who have been to the site say that it's a third harmonic problem that the utility needs to fix. The utility representatives say there's a problem with the transformers.

One test that was run had all building loads disconnected in the electrical service and had the PV system disconnect on the 480V side of the Y-Y transformer open, so the Y-Y transformer wan the only load being served by the electrical service and the transformer had no load connected. Phase loads measured at the transformer were balanced at about 21 amps and the neutral was about 63 amps. That seems really high for magnetizing current for an unloaded transformer.

Not finding much on the internet other than that Y-Y at this voltage level is not generally considered a good idea. Any help or suggestions on what may be causing the current imbalances and/or how to set up the grounding an bonding to make this work would be appreciated.

 

[GoldDigger]

A colleague has several new PV sites where the PV inverters are at 480V but the service is 208V. The step-down transformers are 30KVA and the utility has required that they be Y-Y. The PV system uses the neutral on the 480V side for monitoring only, but a neutral connection is required.

The neutral current on the 208V side is unusually high and the phase currents are badly imbalanced if the neutral on the 208V (utility) side is connected. The currents are balanced if the 208V side neutral is lifted, but the utility requires the neutral be connected. The 208V side has neutral and ground bonded by the MBJ in the electrical service. On the 480V side neutral and ground are bonded in the transformer. The representatives from the transformer manufacture who have been to the site say that it's a third harmonic problem that the utility needs to fix. The utility representatives say there's a problem with the transformers.

One test that was run had all building loads disconnected in the electrical service and had the PV system disconnect on the 480V side of the Y-Y transformer open, so the Y-Y transformer wan the only load being served by the electrical service and the transformer had no load connected. Phase loads measured at the transformer were balanced at about 21 amps and the neutral was about 63 amps. That seems really high for magnetizing current for an unloaded transformer.

Not finding much on the internet other than that Y-Y at this voltage level is not generally considered a good idea. Any help or suggestions on what may be causing the current imbalances and/or how to set up the grounding an bonding to make this work would be appreciated.

My guess is that the inverter may be internally built to source power from line to line rather than line to neutral. Under some circumstances (like the inverter trying to source equal power on all three phases) this can cause the inverter to appear to the transformer as a delta winding.
The usual reason for not connecting the wye star point on the POCO side in a wye-delta transformer then applies.

Another way to look at it is that the inverter is apparently doing something that tries to force all three phase to neutral voltages to be identical. When the actual line to neutral voltages from POCO are not equal the transformer will draw unbalanced current in the phases and neutral (sometimes even larger than the nominal full load current of the transformer) so that the voltage drops in the supply conductors(including the neutral) make the line to neutral voltages seen at the transformer terminals be equal. Or at least match the line to line voltages on the inverter side to add up properly in a delta vector triangle.

Possibly your best bet would be to use a zig-zag transformer to develop the bonded neutral on the inverter side or the wye side instead of using X0 or H0 for that purpose.

If POCO requires their neutral to be connected to the wye transformer primary, and they will not take any steps to balance the open circuit line to neutral voltages on their side, you may have to put in a delta-wye transformer before the wye-wye transformer that POCO wants to see.

The best of all possible worlds would be for the POCO to accept a delta-wye transformer in the first place (with the delta on their side.)

 

[ggunn]

My guess is that the inverter may be internally built to source power from line to line rather than line to neutral. Under some circumstances (like the inverter trying to source equal power on all three phases) this can cause the inverter to appear to the transformer as a delta winding.
The usual reason for not connecting the wye star point on the POCO side in a wye-delta transformer then applies.

Another way to look at it is that the inverter is apparently doing something that tries to force all three phase to neutral voltages to be identical. When the actual line to neutral voltages from POCO are not equal the transformer will draw unbalanced current in the phases and neutral (sometimes even larger than the nominal full load current of the transformer) so that the voltage drops in the supply conductors(including the neutral) make the line to neutral voltages seen at the transformer terminals be equal. Or at least match the line to line voltages on the inverter side to add up properly in a delta vector triangle.

Possibly your best bet would be to use a zig-zag transformer to develop the bonded neutral on the inverter side or the wye side instead of using X0 or H0 for that purpose.

If POCO requires their neutral to be connected to the wye transformer primary, and they will not take any steps to balance the open circuit line to neutral voltages on their side, you may have to put in a delta-wye transformer before the wye-wye transformer that POCO wants to see.

The best of all possible worlds would be for the POCO to accept a delta-wye transformer in the first place (with the delta on their side.)

I have never tried to do this with a wye-wye transformer; in the past I always used a 208V delta to 480/277V wye with a grounded neutral on the 480V side. Lately, however, I have started using autotransformers that pass the neutral through from the 208V to the 480V side. They are cheaper and you don't have that pesky separately derived stuff to deal with.

 

[hmspe]

Thanks for the comments. It makes no sense to me that the utility requires that their side of the transformer be wye. This is a 15KW inverter PV system. The utility's concerns seem to be centered on a delta connection on the utility side not being able to supply zero sequence loads, according to a PowerPoint slide set they provided.

 

[ggunn]

Thanks for the comments. It makes no sense to me that the utility requires that their side of the transformer be wye. This is a 15KW inverter PV system. The utility's concerns seem to be centered on a delta connection on the utility side not being able to supply zero sequence loads, according to a PowerPoint slide set they provided.

Utilities who do not understand PV are getting more scarce as time goes by, thank goodness.