Using different 705.12(B)(2)(3) methods for multiple panels in a PV system

[dissonant]

What is everyone's interpretation of the wording from the 2017 NEC 705.12(B)(2)(3) when it states "One of the methods that follow shall be used to determine the ratings of busbars in panelboards"

Does this mean that you must select only 1 method to use for every panelboard the PV system is connected to? Or can it be interpreted it to mean you can select one method for one panelboard in the system and another method for another panelboard in the PV system?

For my particular case, I am looking at using 705.12(B)(2)(3)(c) for the panel that the inverters are connected to (only inverter breakers on this panel) and then 705.12(B)(2)(3)(b) for the panel that feeds the inverter panel.

 

[Designer101]

Any of the three rules can be used to qualify any of the panel board, if you have multiple inverter systems that needs to be connected to single breaker at the end of main panel then 705.12(b)(2)(3)a is usually used for Load centers down stream of Main panel and then you can use 705.12(B)(2)(3) (b) or 120 percent rule or 705.12(B)(2)(3((c) sum of breakers rule for connection at the main panel which ever is easier.

 

[jaggedben]

I agree with Designer101. There is no logical reason for all the panelboards on a site to follow the same rule.

 

[Carultch]

Take each panelboard individually, select the rule that works best for you, for protecting its busbar.
When you look at the next panelboard, justifying its protection is independent of the rule you used on the previous panelboard.

If you interconnect to the general distribution system at a subpanel, assume that all panelboards upstream of it also must qualify, up until the service point. For instance, if you tie in on a garage subpanel, that is fed from the house's main panel. Consider worst case scenario that all local load in the garage diminishes to zero, while the system performs at maximum power. That power will propagate in full to the main panelboard from which it is fed. There is an advantage to interconnect at a panelboard with a rank as high as practical.

 

[ohmti787]

So, if I'm reading this right, under section 705.12(B)(2)(3)(c), I can pull enough breakers out of the MSP, re-route them to a subpanel, install my PV breaker inside the MSP (even if it goes over the 120% rule) and still be compliant as long as the sum of the OCPD's in the MSP does not exceed the busbar rating, correct?

 

[Carultch]

So, if I'm reading this right, under section 705.12(B)(2)(3)(c), I can pull enough breakers out of the MSP, re-route them to a subpanel, install my PV breaker inside the MSP (even if it goes over the 120% rule) and still be compliant as long as the sum of the OCPD's in the MSP does not exceed the busbar rating, correct?

That's correct. To help understand why, consider the following worked example. Suppose you start with a main panel that has a 200A bus and a 200A MCB, containing the following branch breakers:

1x 50A/2P
1x 30A/2P
20x 20A/1P, balanced across both busbars.

These initially add up to 280A on each of the two live busbars, which gives you no headroom to add a new source through the rule about adding up branch breakers. At best, you could use the 120% rule, but that limits you to 40A to interconnect. The interconnection rule about adding up breakers assumes all of the 20A breakers draw their full load simultaneously, which is a conservative assumption.

It is common that panels with no source other than the main supply are loaded well in excess of the busbar rating, with the underlying assumption being that (A) they won't all draw full load at once, and (B) that the main breaker will trip if the simultaneous load ever does exceed it.

You then relocate all of 20A breakers to a subpanel, behind a common 70A/2P breaker, that is still fed from this main. You can do this, as long as their collective calculated load is 70A or less, on each of the two lines. These circuits can each individually draw 20A, but at any given instant, they may not simultaneously draw a total greater than 70A on either of the two ungrounded conductors. You are depending on models of probable end-user behavior so that it is unlikely that they all draw their full load at once. If they do, the 70A breaker will trip.

By doing this, it now gives you headroom to interconnect up to 50A of inverter OCPD. Because the remaining original branch breakers and 70A subfeed breaker add up to 150A. Busbar protection is justified, because if the 20A/1P breakers do collectively draw more than 70A at any given instant, the 70A breaker will trip before the main panel busbars are overloaded.

 

[jaggedben]

So, if I'm reading this right, under section 705.12(B)(2)(3)(c), I can pull enough breakers out of the MSP, re-route them to a subpanel, install my PV breaker inside the MSP (even if it goes over the 120% rule) and still be compliant as long as the sum of the OCPD's in the MSP does not exceed the busbar rating, correct?

I'd say more often you end up putting the PV in the subpanel. Because if you're moving loads to the sub, you'll need a large subpanel breaker in the main. And because you can size the new subpanel to meet the 120%rule.

A loophole has been to use a subfeed lug instead of a breaker to feed the subpanel, thus eliminating the large subpanel OCPD. I believe this defeats the intent of the rule, and will be prohibited in the 2023 NEC.

 

[Carultch]

I'd say more often you end up putting the PV in the subpanel. Because if you're moving loads to the sub, you'll need a large subpanel breaker in the main. And because you can size the new subpanel to meet the 120%rule.

A loophole has been to use a subfeed lug instead of a breaker to feed the subpanel, thus eliminating the large subpanel OCPD. I believe this defeats the intent of the rule, and will be prohibited in the 2023 NEC.

Interconnection rules propagate to upstream panelboards/busbars as well. Adding the subpanel in your example might seem like it helps you, because you can control its busbar rating, but the underlying issue still persists.

If the local load in the subpanel diminishes to zero, all the inverter current moves on to feed the main panel. The main panel therefore also needs to comply with a busbar protection rule, as if the total inverter current were feeding it directly.

 

[jaggedben]

Interconnection rules propagate to upstream panelboards/busbars as well. Adding the subpanel in your example might seem like it helps you, because you can control its busbar rating, but the underlying issue still persists.

If the local load in the subpanel diminishes to zero, all the inverter current moves on to feed the main panel. The main panel therefore also needs to comply with a busbar protection rule, as if the total inverter current were feeding it directly.

I wasn't trying to suggest otherwise. I was just saying that more commonly the details will make it advantageous to put the PV in the sub.

 

[ohmti787]

Interconnection rules propagate to upstream panelboards/busbars as well. Adding the subpanel in your example might seem like it helps you, because you can control its busbar rating, but the underlying issue still persists.

If the local load in the subpanel diminishes to zero, all the inverter current moves on to feed the main panel. The main panel therefore also needs to comply with a busbar protection rule, as if the total inverter current were feeding it directly.

I'd say there's always the potential that the local load in the subpanel will drop to zero. So even in a case where I don't have to pull any breakers out and want to go straight with my PV breaker into the MSP, I'd still have to follow the 120% rule no matter what.

It's a little confusing, but it seems to me that there's no way around the 120% rule and you can use 705.12(B)(2)(3)(c) ONLY for subpanels and ONLY if all the other upstream panelboards follow the 120% rule.

 

[jaggedben]

I'd say there's always the potential that the local load in the subpanel will drop to zero. So even in a case where I don't have to pull any breakers out and want to go straight with my PV breaker into the MSP, I'd still have to follow the 120% rule no matter what.

The 120% rule is concerned with the opposite case, where the load is more than the rating of the busbar but the main doesn't trip because solar supplies part of the load.

It's a little confusing, but it seems to me that there's no way around the 120% rule and you can use 705.12(B)(2)(3)(c) ONLY for subpanels and ONLY if all the other upstream panelboards follow the 120% rule.

Not quite. Each panel has comply with (at least) one of the rules, but they are independent of each other. So if you have main and sub, they could comply with (c) and (b) respectively, or vice versa, or both with the same rule, or with one of the other rules. They don't affect each other code wise. What I was trying to say above is that if you are dealing with an existing panel and moving around loads, the existing circuits are going to limit you somewhat.

The (c) rule which has been under discussion here was put in to allow solar AC combiner panels without having to comply with the 120% rule. But people are also finding ways to make upstream panels comply with it.

 

[Carultch]

The 120% rule is concerned with the opposite case, where the load is more than the rating of the busbar but the main doesn't trip because solar supplies part of the load.

Both cases are scenarios to consider, as to whether or not a busbar will overload. The 120% rule considers the panelboard you are currently examining, for the possibility branch circuit overload as a blindspot to the main supply breaker. It seemingly could be a 200% rule, by reasoning through Kirchhoff's laws alone, but collective heating among the breakers is why the NEC uses 120% as an industry consensus.

The possibility of local subpanel load being zero, is the justification for why one should assume busbar interconnection rules apply at every rank in the panelboard hierarchy, all the way from the POI to the service point. Unload the POI subpanel, and the full source makes its way to the main panel, as if it fed the main panel directly.

Not quite. Each panel has comply with (at least) one of the rules, but they are independent of each other. So if you have main and sub, they could comply with (c) and (b) respectively, or vice versa, or both with the same rule, or with one of the other rules. They don't affect each other code wise.

Well said. Exactly the point I would've made as well.

 

[ohmti787]

Thanks for the clarification guys. Much appreciated!

 

[ohmti787]

Sooooo, in the case where the MSP that we're installing our PV breaker has feed through lugs feeding a load center inside the house, is there a potential for overloading these feeder conductors before they reach the load center?

 

[jaggedben]

Sooooo, in the case where the MSP that we're installing our PV breaker has feed through lugs feeding a load center inside the house, is there a potential for overloading these feeder conductors before they reach the load center?

Yes. See 705.12(B)(2)(2).

The easiest way I've found around that situation is to add a main breaker to the subpanel, but it isn't possible in all situations.

 

[ohmti787]

Got it. So a "MCB" at the load center inside will protect the feeders coming from the MSP outside, correct? Some inspectors are just not having it. They rather see that main breaker at the MSP outside before the feeder conductors as opposed to after.

 

[wwhitney]

I think you need to be a little more specific about your proposed configuration. What breakers are in the MSP, and what is the wiring method and length of the feeder from the MSP feed thru lugs to the MB subpanel?

Cheers, Wayne

 

[jaggedben]

Got it. So a "MCB" at the load center inside will protect the feeders coming from the MSP outside, correct? Some inspectors are just not having it. They rather see that main breaker at the MSP outside before the feeder conductors as opposed to after.

Well, they are within their rights. 240 says the OCPD should be at the point of supply. It's an unnecessarily strict reading of the code but it's hard to argue against.

 

[ohmti787]

I think you need to be a little more specific about your proposed configuration. What breakers are in the MSP, and what is the wiring method and length of the feeder from the MSP feed thru lugs to the MB subpanel?

Cheers, Wayne

In this particular case is a 200A MSP/150AMCB scenario (outside) w/feed through lugs that end up at a 200A load center inside. Our PV system is putting in 43 amps. don't know the length of the feeder conductors between panels tho

 

[wwhitney]

In this particular case is a 200A MSP/150AMCB scenario (outside) w/feed through lugs that end up at a 200A load center inside. Our PV system is putting in 43 amps. don't know the length of the feeder conductors between panels tho

You haven't provided all the details, so it's hard to answer. I'm going to assume your outside panel has an arbitrary number of other breakers supplying loads. And I'm going to assume that the outside panel feedthru lugs are at the opposite end of the busbar as the utility supply.

If your inverter maximum output current is 43A, then 125% of that is 54A, which is the figure used for busbar qualification computations. For a 200A panel busbar with a 150A main breaker, your only option to qualify the main panel would be the 120% rule. It's not really clear to me if that is allowed with the feed thru lugs, since the PV interconnection breaker can't be at the absolute opposite end of the busbar as the utility supply.

If you figure out the outdoor panel qualification, then you have a question about the feedthru lugs. It seems clear to me that if the connected feeder has an ampacity of 204A, then there should be no issue of its requiring OCPD at the outdoor panel, but I'm a little hard pressed to say exactly which rules allow that. It also seems clear to me that if it has a lower ampacity, but complies with the tap rules in 240.21(B) and 705.12(B)(2)(2), then the OCPD at the indoor panel is sufficient. It's not at all clear to me that the allowance of 705.12(B)(2)(1)(b) applies, although I think it would be logical to allow it.

BTW, all of these question could be sidestepped if you interconnected the PV by intercepting the feeder supplied by the feed thru lugs, rather than connecting it via a breaker in the outdoor panel. Then the interconnection is firmly in the realm of 705.12(B)(2)(1), meaning that 705.12(B)(2)(1)(b) clearly applies for the feeder segment to the indoor panel. And the outdoor panel definitely qualifies under the 120% rule, as the two sources of supply to the busbar are definitely at opposite ends.

Cheers, Wayne