705.12(B) "Capable of supplying multiple branch circuits or feeders"

[wwhitney]

An academic question about the nuances of the wording in the first paragraph of 705.12(B) (2017; in 2020 the language moved to 705.12). Which of the following, if any, is exempt from the requirements of 705.12(B), because it is not "distribution equipment . . . fed simultaneously by a primary source(s) of electricity and one or more other power source(s) and . . . capable of supplying multiple branch circuits or feeders"? In all cases there are three breakers of equal size, one for the utility supply, one for a PV/ESS, and one for a feeder to a load panel.

1) A 2 space MB panel with 100A MB and quad 100A/100A breaker.
2) A 6 space MLO panel with (3) double pole 100A breakers.
3) An 8 space 200A MB panel with (2) double pole 4 position 200A breakers.

For (1), I don't know if such equipment exists, but suppose it does; I wanted an example that unarguably does not fit the quoted language.

For (2), someone could come along and replace the 100A double pole breaker with a quad breaker; does that mean it is "capable of supply multiple branch circuits or feeders?" If so, there's no way to qualify it under 705.12(B)(2)(3); if you want that topology, you need to reconfigure it as a feeder splice rather than use a panelboard. [E.g. (3) separate 100A disconnects with a gutter; apparently the wire-type feeder is safer than a single enclosure with a busbar? (ironic)]

Likewise for (3), some one could come along and replace the 4 position 200A breaker with (2) 2 position 125A breakers (generally). That would create an actual chance of the bus being overloaded. So I guess it is prohibited by 705.12(B)(2)(3).

Cheers, Wayne

 

[jaggedben]

Just my supposition, but I think the purpose of the language is to exempt equipment like fused disconnects individual circuit breaker enclosures. I certainly used to cite that language to push back against people who would tell me that, for example, I couldn't backfeed more than 20A on a fused service disconnect rated 100A.

In other words, none of your examples are clearly exempt, although (1) is a bit on the bubble since 100/100 quads do not exist. (Max is 50/50). I say on the bubble because if you install a 100A breaker in a 2 space panel you make it incapable of feeding other feeders or branch circuits, bit technically the panel still has that capability if you take the 100A breaker back out. I would probably exempt it if I were the AHJ, but I'm not one.

 

[wwhitney]

The language exempts distribution equipment that is not capable of at least 4 external connections (implicitly double pole), since the equipment is supposed to have at least two power source connections and also multiple branch circuits or feeders. And Kirchoff's current law tells us we can't get into trouble with only 3 connections. [OK, it does not literally say "supply multiple branch circuits or feeders while simultaneously have two power source connections," but I think we can apply a little common sense in understanding the language. Just a little.]

Most panelboards can accept a lot more than 3 connections though; I agree that (1) in my OP is the only one that can not (and is bad example since you tell me quads max out at 50/50). Is there any practical way to take a 6 space MLO panel and render it incapable of utilizing quad breakers? Some sort of quad breaker rejection feature? If so, I would say a so-modified 6 space MLO panel would also be exempt, and might occasionally be useful in practice.

[My PI for adding "excluding one of the smallest breaker present" to the sum in the "100% rule" would also allow any panel with just 3 breakers to qualify, but I doubt it will be adopted, since it got rejected 3 years ago. The substantiation is again Kirchoff's current law and simple combinatorics.]

Cheers, Wayne

 

[jaggedben]

To somewhat repeat myself, I think the intent is to exempt only equipment that makes only series connections. Any parallel connection of any kind would mean it is subject to the rules at issue.

Put another way, any distribution equipment of any kind, by definition, supplies at least one branch circuit or feeder. If that feeder or branch circuit may carry current from a power source other than the primary source, then the equipment is fed by two sources, and all you need is the capability to connect one more feeder or branch circuit at the same location, and then the rules apply. So I think for one thing the magic number is 3, not 4.

...Is there any practical way to take a 6 space MLO panel and render it incapable of utilizing quad breakers? Some sort of quad breaker rejection feature? If so, I would say a so-modified 6 space MLO panel would also be exempt, and might occasionally be useful in practice.

Some panels are listed to take quad breakers and some are not, and modifying either would be a code violation. One could perhaps 'render it incapbable of utilizing quads' by filling all spaces with 2-pole breakers 60A or greater, but that's kind of an interpretive loophole that it is reasonable to not accept code-wise, since breakers can be replaced and don't really change the panels 'capabilities'.

Finally, don't forget the main lugs. What I'm seeing is a cascade of reasons that your idea cannot work.

First, the panel would have to not be listed for quads (unless you accept that filling it with <60A breakers renders it equal to that).
Second, it would have to be a 4-space and not a 6-space, since a six space is capable of 4 connections when including the main lugs.
Third, per my argument above, it would have to be a 2 space and not a 4 space, since a 4 space can handle 3 connections with the lugs.

 

[wwhitney]

Leaving aside momentarily the question of whether the code language says the magic number is 3 or 4, physics-wise the magic number is 4, right? Because with 3 connections to a node, each can only be carrying current in or out, so the total current through the node is limited to the maximum of the 3 currents.

[And an extension of that is basically the idea behind the PI I mentioned.]

Cheers, Wayne

 

[jaggedben]

I don't think that's correct. I think you are assuming that each connection has overcurrent protection, but the code has to account for situations where that isn't true, and I've seen such in the field.

Example: I find a 4-space 125A MLO panel which someone has landed a 100A backfed breaker as a service disconnect. The main lugs connect to a remote MLO subpanel via a 100A feeder. Can I land 25A of solar (let alone 50 or 100A) without defeating the overcurrent protection on the sub feeder? No, I actually can't any amount. Hence the beginning of 705.12, which contains the rules to prevent that, needs to say that the rules apply to equipment with only 3 connections.

 

[wwhitney]

That is a very helpful example, thank you, it bears further consideration. Preliminarily, sounds like the magic number is 3, or 4 if every connection has OCPD.

So the upshot is that in your reading of the first paragraph of (207) 705.12(B), the "multiple branch circuits or feeders" should include any power source connections beyond the primary supply. Because of the example problem with just 3 total connections.

[Part of my fixation with this section is the apparent disparity in the rules on busbars vs feeders, and the feeling or hope that from the proper point of view it should be possible to unify them. But that point of view so far eludes me.]

Cheers, Wayne