120% rule and end-of-busbar

[tortuga]

Ahh then what about a more complex scenario, they have two 32A inverters one in each panel and a lug kit in the center of panel A?

 

[wwhitney]

Ahh then what about a more complex scenario, they have two 32A inverters one in each panel and a lug kit in the center of panel A?

Panel A still has 200A from the utility and 64A of continuous inverter output connected to it. So for 705.12(B)(1) it would need a bus of at least 200A + 1.25 * 64A = 280A (not likely). For 705.12(B)(2), you'd need to move the feed thru lug kit to the bottom to be next to the 32A of inverter output directly connected to panel A; then the busbar rating would only need to be 280A / 120% = 233A (still not likely).

For 705.12(B)(3), that section is still miswritten to not reference what's connected to the feed-thru lugs, so it may or may not comply just looking at the breakers in panel A. Ideally 705.12(B)(3) should be rewritten to require adding the lesser of the 200A main breaker in panel A1, or the sum of all breakers in panel A1.

Cheers, Wayne

 

[jaggedben]

That wouldn't comply with 705.12(B)(1), (2), or (3).

 

[wwhitney]

That wouldn't comply with 705.12(B)(1), (2), or (3).

It could in theory comply with 2023 NEC 705.12(B)(3), either the as-written version that just looks at breakers in Panel A, or the as-it-should-be-written version that would look at all breakers in Panel A and Panel A1.

Cheers, Wayne

 

[jaggedben]

It could in theory comply with 2023 NEC 705.12(B)(3), either the as-written version that just looks at breakers in Panel A, or the as-it-should-be-written version that would look at all breakers in Panel A and Panel A1.

Cheers, Wayne

Ok it could possibly comply with the as-written version, correct. It wouldn't comply with the as-it-should-be-written version with the 200A breaker shown in Panel A1; the sum is already too great even without knowing the other breaker ratings.

 

[ggunn]

Ok it could possibly comply with the as-written version, correct. It wouldn't comply with the as-it-should-be-written version with the 200A breaker shown in Panel A1; the sum is already too great even without knowing the other breaker ratings.

Agreed. Even if it would comply with some interpretation of the letter of the NEC it would be exploiting a loophole to get around the intent of the code. I don't do loopholes.

 

[wwhitney]

It wouldn't comply with the as-it-should-be-written version with the 200A breaker shown in Panel A1; the sum is already too great even without knowing the other breaker ratings.

My "as-it-should-be-written" version of 2023 NEC 705.12(B)(3) would allow as an option, for the case of side by side panelboards, looking beyond the 200A main breaker in Panel A1 to instead count all the breakers in Panel A1. So then it could be satisfied. Although it wouldn't be likely, as why have two side by side panelboards if the sum of all the relevant breakers in both is less than 200A?

For distant panelboards, even though the same electrical logic would apply, as an enforcement/practical matters, the labeling and verification of compliance would just be too difficult to be practical.

Cheers, Wayne

 

[ggunn]

For distant panelboards, even though the same electrical logic would apply, as an enforcement/practical matters, the labeling and verification of compliance would just be too difficult to be practical.

Difficult or not, I have had to do that.

 

[wwhitney]

Difficult or not, I have had to do that.

For what rules? I was talking about a hypothetical version of 2023 NEC 705.12(B)(3) that would require labeling two panels connected as in post 21 each something like "sum of all breakers in this panel and that panel 50' over there, excluding the breakers protecting the busses, may not exceed 200A."

Cheers, Wayne

 

[ggunn]

For what rules? I was talking about a hypothetical version of 2023 NEC 705.12(B)(3) that would require labeling two panels connected as in post 21 each something like "sum of all breakers in this panel and that panel 50' over there, excluding the breakers protecting the busses, may not exceed 200A."

Cheers, Wayne

I did it for PV system interconnections with taps on the load side of existing beakers that were feeding remote MLO subpanels. I designed them with OCPD on the load sides of the taps. Is that what you were talking about?

 

[tortuga]

Here is another one that I think meets the letter of the 2023 code:

 

[ggunn]

Here is another one that I think meets the letter of the 2023 code:
View attachment 2571353

Rube Goldberg called; he wants his machine back.

 

[wwhitney]

Here is another one that I think meets the letter of the 2023 code;

Nope, that one also has 64A of inverter output connected to Panel A. So the same considerations as to the schematic in post 21 would apply. Although as the inverter output is already at the opposite end of the bus as the primary supply to Panel A, using the 120% rule would be easier, if you could get a bus rated 233A. More likely, get a bus rated 225A and use a 175A MB in Panel A.

Cheers, Wayne

 

[wwhitney]

I did it for PV system interconnections with taps on the load side of existing beakers that were feeding remote MLO subpanels. I designed them with OCPD on the load sides of the taps. Is that what you were talking about?

No, that just uses the feeder interconnection rules and is a great option in general. I was talking about a hypothetical version of the "sum of all breakers" rule that looks across two panelboards interconnected via feed thru lugs.

Cheers, Wayne

 

[tortuga]

Nope, that one also has 64A of inverter output connected to Panel A. So the same considerations as to the schematic in post 21 would apply. Although as the inverter output is already at the opposite end of the bus as the primary supply to Panel A, using the 120% rule would be easier, if you could get a bus rated 233A. More likely, get a bus rated 225A and use a 175A MB in Panel A.

Cheers, Wayne

Yeah I am just thinking of all the cases 705.12(B) does not address in the 2023.
They could also replace the sub-feed lugs in panel A with a 200A breaker.

 

[PWDickerson]

I don't see the other rules as being ambiguous in the presence of feed-thru lugs. (1) ignores them, they are immaterial. For (2), if they are at the end of the bus opposite the primary supply, then you just can't use (2). And (3) ignores them, which is a mistake but how it's written. (3) needs an internal limitation with respect to feed-thru lugs.

So we have a few imperfect interpretations of (5):

(A) It's attempting to modify (1), (2) and (3); if you have subfeed lugs, you must satisfy (5) before you can use (1), (2), or (3). But the text at the beginning of 705.12(B) is not written in a way that allows (5) to modify (1), (2), and (3).
(B) It's providing an additional allowance not present in (1), (2), or (3), but does so only implicitly.
(C) It's useless text that does nothing.

I'm arguing for (B); I guess you're arguing for (A) or (C)?

Note that if in (2) the subfeed lugs aren't at the end of the bus, they are immaterial to compliance with (2). So the only additional allowance that (5) can offer with respect to (2) is to allow the subfeed lugs at the end of the bus and allow connecting of the secondary source at the penultimate connection.

Note further that (2) could be modified in general to permit one load (OCPD limited) connection between the secondary source connection and the end of the bus, without any risk of overloading the bus. Or maybe with most bus layouts, as there are two bottom most positions, each such position will connect the very bottom of the bus for one pole, and the second pole (and third pole) would already be not quite farthest?

Cheers, Wayne

I just read the section again, and again, and again, and maybe even a few more times... and I think it is (A). The first two sentences of (5) tell us plainly that, in the case where you have an upstream panel feeding a downstream panel with feedthrough lugs, it is ok to interconnect solar at either panel as long as the rules for sizing the feeder are followed. It seems to be essentially clarifying a practice that was called into question in previous code cycles.

The last sentence of (5) is telling us that, if there is an OCPD at either end of the feeder, the busbar can be qualified by (1), (2), or (3). This seems to imply that if said OCPD is not in place, you can't qualify the busbar with (1), (2), or (3) when you have two panels connected via feedthrough lugs.

This seems to mistakenly allow the potentially dangerous practice of installing a backfeed breaker in the upstream panel and locating load breakers between the backfeed breaker and the feedthrough lugs, thereby allowing currents in excess of the bus rating in that portion of the bus between the backfeed breaker and the feedthrough lugs. If they simply added language to prohibit load breakers between the backfeed breaker and the feedthrough lugs, that would resolve the issue.

One of these decades, they are going to get this right!

 

[wwhitney]

I just read the section again, and again, and again, and maybe even a few more times... and I think it is (A).

I agree the intent is to modify (1), (2), and (3), but the first paragraph of 705.12(B) isn't phrased in a way that makes that possible. Each subitem needs to be independent. So (1), (2) and (3) would need language precluding their use when there are feed-thru lugs, unless (5) is complied with.

Cheers, Wayne

 

[PWDickerson]

Here is another one that I think meets the letter of the 2023 code:
View attachment 2571354

I don't think you need to upsize the feeder between the panels. You have a main breaker in the downstream panel, so 705.12(A)(2)(a) doesn't apply. If current in the feeder is flowing toward A1, it is limited to 200A, and if current in the feeder is flowing toward A, it is limited to 64A.

In this example, the feeder is adequately sized, and you only need to qualify the bus in Panel A. If the sum of the breakers in Panel A is 200A or less, I think this should qualify under 705.12(B)(3) and 705.12(B)(5), and it would be a code compliant design. And if you wanted to, you could even tap three more 32A inverters to the feeder and still be compliant, though you would need to upsize the presumably 4/0 AL service conductors to 250.

 

[PWDickerson]

I agree the intent is to modify (1), (2), and (3), but the first paragraph of 705.12(B) isn't phrased in a way that makes that possible. Each subitem needs to be independent. So (1), (2) and (3) would need language precluding their use when there are feed-thru lugs, unless (5) is complied with.

Cheers, Wayne

I completely agree!

 

[jaggedben]

I agree the intent is to modify (1), (2), and (3), but the first paragraph of 705.12(B) isn't phrased in a way that makes that possible. Each subitem needs to be independent. So (1), (2) and (3) would need language precluding their use when there are feed-thru lugs, unless (5) is complied with.

(1) doesn't need a preclusion.

(2) essentially has the preclusion. The 'opposite end' part of (2) precludes using a source breaker in the panel to comply with that option if the feed through lugs are used at the opposite end. (Perhaps if one removed the lugs so that the connection could not be used, one could still otherwise use that option at the farthest breaker spot.)

(3) arguably needs the preclusion. (Although, if the source(s) are connected to the feed through lugs, it doesn't result in an unsafe installation. As you've pointed out in the past there are many situations where actually counting all overcurrent devices seems to be overly conservative.)