Moving Circuits to Subpanel - 300.3B Debate

[hitehm]

The subpanel feeder breaker IS in the main. However we didn't want to land the solar breaker in the sub since it would've required a larger diameter conduit or more than 1 conduit for both feeders and solar and branch circuits, so we didn't want to deal with that or future load issues. Plus, we really thought it to be a cleaner install to have the solar breaker land in the main. And even though we would've been ok, we didn't want NEC 705 rules dictating our subpanel and breaker selection with the solar breaker in the sub.

You could have put the subpanel feeder breaker in the main panel where you put the solar breaker, and then put the solar breaker in the subpanel opposite its supply. That would have required clearing up 2 fewer positions in the main panel.

Cheers, Wayne

Sorry Wayne - Just noticed you said "opposite it's supply" so yes, that would've entered the sub on a different conduit. However, we again wanted to keep the installation neat and also the opposite outer wall was giving us issues with where we could re-enter the garage. In the end the physical challenges dictated a lot of design decisions but that's off topic now. Just really wanted to know what everyone would do with those Ns and Gs!
Thanks, Jon

 

[wwhitney]

Sorry Wayne - Just noticed you said "opposite it's supply" so yes, that would've entered the sub on a different conduit.

My suggestion did assume, though, that 125% of the PV inverter output current is no more than 20% of the panel busbar rating, otherwise it would have required paying attention to the subpanel feeder size vs the subpanel busbar size to satisfy the 120% rule.

Just really wanted to know what everyone would do with those Ns and Gs!

EGCs--no question the individual wires can stay in the main panel. The feeder EGC in your nipple between panels can do double duty as the branch circuit EGC for the branch circuits returning through the main panel--you never (except patient care areas?) need more than one EGC in a wiring method.

On the neutrals, as we determined here you could have left them in the main panel. Since your feeder has a neutral, it also wouldn't be any extra work (just delayed work) to go back and move them to the subpanel if an issue ever came up, e.g. because someone else has a different understanding of the relevant codes, or because one of the breakers needed to be changed to a GFCI breaker or an AFCI breaker that does GF monitoring.

Cheers, Wayne

 

[jaggedben]

Still no problem.

(I am only discussing the point of inductive heating here, not other code requirements related to the subpanel branch neutrals. )

No matter what the combination of subpanel load and PV input the _net_ current in the nipple with the subpanel feed and the 'switch loop hots' of the subpanel branch breakers will remain zero.

Without the PV input you would have the situation where the branch 'hot' currents are balanced by feed 'hot' currents.

Throw the PV in and now the feed 'hot' current might drop to zero because everything is coming from the PV source. But in this case one of the other feed conductors will be carrying the necessary balance current because the supply from the PV system is balanced in terms of current flow.

This even works if the PV is connected 120V to the subpanel. If you have a situation where current is flowing 'out' on the branch hots is supplied by the PV, then the necessary balancing current would be flowing 'out' on the feed neutral. No net current and no induction heating.

Jon

For some reason I thought this wasn't entirely true, but after diagraming a few examples I see it is correct.

The only thing I will add is that if one is relocating circuits to a backed up subpanel that will be disconnected from the grid and powered by a backup system in an outage, then one must absolutely move the neutrals (but not the grounds) to the sub, or else you'll lose the 120V reference to the neutral. But that's a totally different situation than the OP's.

 

[wwhitney]

The only thing I will add is that if one is relocating circuits to a backed up subpanel that will be disconnected from the grid and powered by a backup system in an outage, then one must absolutely move the neutrals (but not the grounds) to the sub, or else you'll lose the 120V reference to the neutral.

The only MID I'm familiar with (the Tesla Powerwall Gateway) doesn't switch the neutral. So if the neutral is unswitched, I don't see why the above would be true.

Cheers, Wayne

 

[jaggedben]

I misstated. You won't lose the reference to the neutral, but you'll have neutral current flowing in an otherwise de-energized upstream panel where the circuits are relocated from. It may just be a more extreme version of the above situation, where it's hard to pinpoint the code violation but it just seems like a horrible idea.

 

[wwhitney]

You won't lose the reference to the neutral, but you'll have neutral current flowing in an otherwise de-energized upstream panel where the circuits are relocated from.

For that to be true you'd have to be extending the circuit through the MID. Agreed, that does seem a bit more extreme compared to extending it 8" between adjacent panels.

Cheers, Wayne

 

[jaggedben]

For that to be true you'd have to be extending the circuit through the MID. Agreed, that does seem a bit more extreme compared to extending it 8" between adjacent panels.

Cheers, Wayne

Well, one could extend the circuit in some raceway that bypassed the MID. And that would cause inductive heating if you dont relocate the neutrals, both when connected to the grid and not.

 

[wwhitney]

Well, one could extend the circuit in some raceway that bypassed the MID. And that would cause inductive heating if you dont relocate the neutrals, both when connected to the grid and not.

Right, one of the constraints specified in the OP is that the branch circuit extension follows the same path as the feeder. If it doesn't, you better extend the neutral as well.

Cheers, Wayne

 

[194hallspark]

We had to make room in our main panel that was full, so we installed a small 6 space Eaton sub next to the main connected through a short 8" section of EMT. The feeders plus the 5 returning branch circuits all went through the same 8" of EMT. Then of course the NEC 300.3.B rule came up and we had to make a decision on also splicing over the neutrals and EGCs or keeping them landed in the main panel. In the end we decided to follow the letter of the law and spliced every Hot, N and G from all 5 circuits to the sub. After we we're done I concluded it would've been better to keep the N and Gs landed in the main since it really stressed the wiring to fully relocate these circuits. I won't bore you with the details but it was very tight and very difficult on the wiring to do this.

I have seen several opposing viewpoints on this topic including on this forum. It seems split between those who move everything because of concern for EMI heating effect and staying strictly compliant to 300.3B and those who feel it's best to keep the neutrals and grounds in the main. So my questions are:

1 - Why are the EGCs also required to be brought over as well? This really seemed unnecessary from an electrical stand point since they aren't normally current carrying and also in our case no matter how gentle we were trying to be, put stress on the cables trying to trace them down and untangle them from each other.
2 - For practical purposes, is there a minimum distance that's considered a "raceway" when deciding to be strict to 300.3B? We didn't use a nipple because of the angle but there was only 8" between main and sub.
3 - What if we used PVC between main and sub? No EMI can induce in the conduit causing heat so why isn't there an exception for all nonmetallic conduit?
4 - What is anyone's opinion about how literal you need to be with 300.3B, especially when just relocating to a very close sub panel and what do you normally do in this situation?

Aren't you supposed to totally remove the circuit from the panel. Make up at a jbox at a nearby point and bring them over to the new sub panel? No conductors are allowed in a vertical section of a panel that won't terminate in the panel, right? 408.3(A)3

 

[hmspe]

Aren't you supposed to totally remove the circuit from the panel. Make up at a jbox at a nearby point and bring them over to the new sub panel? No conductors are allowed in a vertical section of a panel that won't terminate in the panel, right? 408.3(A)3

A panelboard is not a switchboard or switchgear.

 

[kwired]

Aren't you supposed to totally remove the circuit from the panel. Make up at a jbox at a nearby point and bring them over to the new sub panel? No conductors are allowed in a vertical section of a panel that won't terminate in the panel, right? 408.3(A)3

Panelboard- no

Switchboard - yes

 

[caribconsult]

QUESTION: can't you tie all the neutrals together and ground them at one point?

 

[paulengr]

In any case the NEC requires derating conductors, so I am curious as to 5 circuits and the feeder in one conduit.
Did you upsize any wires to account for that ? I normally count it out and plan how many conduit required before
adding the sub panel.

Huh? Ground is not a CCC. The neutral may or may not be. Read the Code carefully with regard to neutrals. The rule is fairly complicated.

But here is why derating doesn’t matter. If you run 4-6 current carrying conductors you must derate to 80%. But in the conduit most common cables (THHN-2, XHHW-2, MTW, RHW-2) are 90 C rated. However with few exceptions the vast majority of terminals are rated for 75 C. Presumably in the enclosure we don’t need to derate. If you check the tables carefully the 75 C column in the table is equal to 80% of the 90 C table. So from a practical point of view no derating applies up to 6 conductors because of the temperature ratings.

 

[paulengr]

So I assume this is more your practical answer not code answer because 300.3B includes the EGCs? And what is your feeling about it being in non-metallic conduit? No induction issue so why the need to relocate the Ns and Gs at all?

A neutral Is still a phase conductor if there are phase to neutral loads or common mode currents (generated by VFDs). If you don’t have any of these the neutral is not needed.

Bonding must be done which is grounding. Look here for experimental details. Essentially most of the time electricians think copper grounds are superior to relying on metallic conduit but the truth is the metal has so much more cross sectional area it provides a much better current path.

https://steeltubeinstitute.org/wp-content/uploads/2020/11/STI_ConduitTests_March_2020.pdf

 

[Little Bill]

Huh? Ground is not a CCC. The neutral may or may not be. Read the Code carefully with regard to neutrals. The rule is fairly complicated.

But here is why derating doesn’t matter. If you run 4-6 current carrying conductors you must derate to 80%. But in the conduit most common cables (THHN-2, XHHW-2, MTW, RHW-2) are 90 C rated. However with few exceptions the vast majority of terminals are rated for 75 C. Presumably in the enclosure we don’t need to derate. If you check the tables carefully the 75 C column in the table is equal to 80% of the 90 C table. So from a practical point of view no derating applies up to 6 conductors because of the temperature ratings.

You can have up to nine (9) CCC in a raceway on a 15 or 20A circuit before derating would take you below the required ampacities.