Lost Phase from Utility / Backfeeding from PV System

[jmkeng]

Hi all - I recently posted something very similar in the PV portion of the forum, but wondering if it may be a better suited here since I'm trying to conceptually wrap my mind around the behavior of the transformers in this situation.

We have a PV array is installed on a facility, and this feeds into the utility via a net meter.
The utility recently dropped a leg of power, but the inverter failed to shut down and kept back-feeding the utility.

The utility transformer is a Delta-Wye (208/120, 3P, 4W on the secondary) which feeds the buildings main distribution gear. The PV array is 480V off the inverter, which feeds into a Wye-Delta (Primary/High-side: 480V Wye, no neutral connected, Secondary/Low-side: 208V Delta, feeding the distribution gear/net meter).

The relationship between the Wye/Delta configurations is getting me a little confused; when the utility dropped a phase, could that cause an amperage spike in the two "live" legs and cause the PV transformer failure? I also understand that it can be a little tricky for an inverter to sense a dropped phase from the utility because the wye side may not see a voltage difference - is that accurate?

Thanks in advance for your feedback!

 

[Hv&Lv]

“The utility dropped a leg of power...”

The utility lost a phase.
Which side? Delta side(Primary) or Wye side(secondary)

Edit: I see now this is the same post as earlier in another category.

 

[winnie]

Out of my depth here and mostly guessing:

If the utility transformer really is Delta-Wye (often a utility will use wye-wye), then when one phase is lost, the remaining two phases power all _three_ transformer legs. One leg is fully powered, and the other two legs are placed electrically in series. So I would expect your 208/120 service (L-N of 120V,120V,120V) to change to 120V,xV,(120-x)V. Also all three legs would be 'in phase' (same phase angle or inverted relative to each other).

When this wacky voltage 3 wire single phase feeds the delta of a transformer, then you have 3 coils 'expecting' 208V, but instead you would get 120+x, 240-x and 120V. These three wacky voltages would then get stepped up in the ratio of 277:208V.

Not sure where things would go from there...
Jon

 

[synchro]

As winnie (Jon) mentioned, utilities often use wye-wye. Are you sure the utility transformer is delta-wye?
Wye-wye transformers with a three-legged core can regenerate voltage at the output when there's a "lost" phase" on the input. From the paper at the link below:

https://www.cce.umn.edu/documents/cpe-conferences/mipsycon-papers/2013/openphaseconditionsintransformersanalysisandprotectionalgorithm.pdf

On pg. 9:
"In summary the important aspects of voltage re-generation
in a Yg-Yg transformer upon loss of a phase on the high side
are:
1) Only in three-leg core type Yg-Yg transformers is the
voltage of the lost phase re-generated. In transformers
with four or five leg cores or a bank of three single phase
transformers there is no voltage re-generation.
2) The voltage of the lost phase is re-generated by the
flux that goes through the coil on the third leg in
accordance with equation 3. Energy is transferred to the
high side coil with the lost phase by magnetic fluxes of
the two remaining phases which is then transferred to the
low side coil."

On pg. 10:
"We saw that generally a transformer’s response to an open
phase on the high side can be categorized as one of the three
cases we discussed: 1) Yg- transformers in which lost phase
is always re-generated, 2) transformers with or ungrounded
Y primaries in which the lost voltage is never re-generated,
and 3) Yg-Yg transformers that the response depends on the
core construction."

If the common terminal at the input of a Y-Y is not connected to neutral/ground and one phase is open then only two connections remain at the input. Therefore any outputs can only be single phase.

 

[jmkeng]

As winnie (Jon) mentioned, utilities often use wye-wye. Are you sure the utility transformer is delta-wye?
Wye-wye transformers with a three-legged core can regenerate voltage at the output when there's a "lost" phase" on the input. From the paper at the link below:

https://www.cce.umn.edu/documents/cpe-conferences/mipsycon-papers/2013/openphaseconditionsintransformersanalysisandprotectionalgorithm.pdf

I did double-check and you all are correct - it is, in fact, a wye-wye at the utility side. Guess that teaches me to always double-check the "facts". That's a very helpful paper. I think I've got some reading today before I can throw out an educated follow-up - thank you very much for sharing.

 

[NewtonLaw]

In my Utility (40 years) we never use a distribution transformer bank configured as a Delta on the high side, however, I'll assume you are correct in your description. If I look at just the Utility side transformer bank first, if I lose one phase supplying the Delta, the transformer bank becomes a single phase device. One transformer will deliver 120 volts on the secondary, the remaining two will have the primary voltage divided across them and will deliver 60 volts each. Consider the phasor diagram below as the Utility starting point as you said.

The round dots shown on the transformer coils represent the polarity of the coil. In the next image open "C" phase supply to the primary side. The phasor configuration chages to the following:

This in turn changes the secondary phasors such that they look like:


This means that anything attached to the secondary "A" phase will see 120 volts while "B" and "C" phases now see 60 volts 180° out of phase with "A" phase. So, if your inverter only detects loss of source on say only "A" phase, it remains in service and the PV transformer will now have to contend with significant circulating current.
Hope this helps.