Xfmr Service/Maintenance Disconnect PV side of Xfmr

[SteveO NE]

So this is a bit of a best practice question about the AC disconnect on the PV terminals of the transformer.

Referencing 2017 here are the abridged requirements:

• (Art 690) Evaluate each winding as pri and sec. and apply protection as determined; if xfmr FLA exceeds PV FLA then no pri protection needed on PV side of xfmr
• (Art 450) Disconnecting means is required on transformer within line of site or remote if lockable and location is marked on transformer, (Art 240) no secondary disconnecting means is required if terminating indoors, if conductors are outdoors unlimited length to a secondary disconnecting means
• (Art 690) PV equipment must have an AC disconnecting means in line of site of PV, exception is remote operated where the device to operate is within line of site of the equipment

OK, so the question is, do you need a disconnect on the PV side of the transformer? You have one for the PV anyway but from the perspective of transformer installation itself, is 450.14 requiring a disconnect on both sides of the transformer for disconnecting means? Why do I ask, you may say? Because how I as an engineer lay out the system with labels and location of equipment will encourage or discourage proper safe behavior. For instance, you would still be code compliant with a lockable PV AC disconnect on the roof and another AC disconnect out by the utility meter, as long as the transformer is placarded with both; however, the chances of an electrician going to the roof to make sure the PV is disconnected is slim - "Ahh it's OK the PV is dead". Also, if not in line of site (even if less ridiculous an example) should the transformer be placarding with both locations? My gut says yes because both sides are primary but Art. 450.14 doesn't address primary or secondary, it seems to imply that you are removing the source which could be argued is just the utility or both.

Its never actually been an issue I've had to ponder that close simply based on circumstance of the actual install but I was just looking to labels and it clicked that we label PV disconnects as "may be energized from line and load terminals." Now, I think that's a CYA label that means little with a grid-tied inverter, there are numerous protections that say that inverter is not producing anything when there is no AC grid, the inverters aren't self commutating so with the exception of large PV systems with inductive reflections from the grid itself, the inverter isn't capable of creating its own frequency and voltage (it is a current source and without voltage produces neither current nor voltage), not to mention rapid shutdown which also removes the source of energy from the inverter as well. But I don't believe there is any labeling requirement noting that the transformer may be energized from both the primary and secondary side but would article 450.14 - should there be?

 

[jaggedben]

See 705.60(B) (2017) or 705.30(C) (2020) for overcurrent protection requirements. Note the rule shifted between code cycles. I start there because even though your question was about a disconnecting means, if you need overcurrent protection then that will amount to the same.

With that out of the way...
I don't believe the presence of an interactive system on what is normally considered the 'secondary' of a transformer changes requirements for a disconnect in any way. Treat it the same as a load. As you mentioned, you may need an AC disconnect at the inverter for reasons having nothing to do with the transformer, or vice versa.

One more thing...

...we label PV disconnects as "may be energized from line and load terminals." ...

I hope you don't do this for AC disconnects on interactive outputs, and won't ask people to. (DC discos may be different.) The code has never required it, for the record. The stupidity of AHJs asking for this has frustrated me for years, and now that I'm installing lots of AC coupled ESS where the label is true, thinking of all the times that I was required to cry wolf in spite of knowing better rankles all the more.

 

[SteveO NE]

I hope you don't do this for AC disconnects on interactive outputs, and won't ask people to. (DC discos may be different.) The code has never required it, for the record. The stupidity of AHJs asking for this has frustrated me for years

You nailed it in that last point - AHJ's require it and I've argued its dumb. The only argument I've heard that made sense, though only half way, was that some people may consider the PV as the "line" and connect it to the line side lugs instead of the load side lugs, thereby leaving fuses and the terminals not covered by a dead front energized. My response was, that's stupid and exactly why we require the PV be terminated in the load side of the device. I may have accepted a less stupid response of islanding behind a transformer but that has never been attempted with me . I would still say that isn't the case (loads would need to match output to effectively island - so may have a justification on the main breaker to the building if they wanted to push it), but at least it wouldn't be a stupid answer. Alas AHJ's are good at those, or the ever present "because I don't like it," with no further explanation.

In either case nearly all of them in my locale require it and point to 690.13(B) (2017) and have trouble with reading comprehensive and the phrase "may be..."

"For PV system disconnecting
means where the line and load terminals may be energized
in the open position, the device shall be marked with the
following words or equivalent..."

I've actually taken it to the state level out of principle and lost.

This though is the basis of why I am asking this, because that leads to what then is best practice for a transformer.

now that I'm installing lots of AC coupled ESS

You should put a label underneath it that says...."No, seriously, I mean it for real this time"

See 705.60(B) (2017)

I think you mean 705.65(B) and this exception is the typical situation or at least the situation I am referring to that creates this scenario ( I meant to type 705, not 690, above in my short-hand code reference in the original post) :

"Exception: A power transformer with a current rating on the side
connected toward the interactive inverter output that is not less than the
rated continuous output current of the inverter shall be permitted without
overcurrent protection from that source."

So no overcurrent protection but if inverter is on the roof, a secondary of unlimited length shall have a disconnecting means (because the conductors go outside). However, this is exactly what I am referring to, 705.65(b) says:

"...by considering first one side of the transformer,
then the other side of the transformer, as the primary."

In which I always say it uses the word "consider" and I consider the secondary side to have no source when the utility is removed in many situations, though not always the case depending on config and go from there. So it sounds like you do the same.

 

[jaggedben]

You should put a label underneath it that says...."No, seriously, I mean it for real this time"

Not a bad idea at all, although correct spelling would be '4 reelz' as seen here.