705.12 Point of connection question

[Inspector1989]

I have a question about bus ratings. From my understanding, all busbars need to follow the rules of 705.12(b)(2)(3) if they have two power sources feeding into them. There are contractors and inspectors telling me that 705.12(b)(2)(3) only applies at the first point of connection. ie, subpanel or combiner box. Is this true?


Scenario in a New PV installation - 200 amp main service panel, no changes are made to existing breakers, 80 amp back-fed PV breaker on opposite end of Main breaker. Feeding into a subpanel 80amp subpanel/combiner panel that has 4-20amp back-fed PV breakers installed in them.

I am told that this is okay because the subpanel/combiner panel meets 705.12(b)(2)(3) and only the first busbar needs to meet point of connection rules. not the second busbar.

Can someone elaborate the code on this? thanks

 

[ggunn]

I have a question about bus ratings. From my understanding, all busbars need to follow the rules of 705.12(b)(2)(3) if they have two power sources feeding into them. There are contractors and inspectors telling me that 705.12(b)(2)(3) only applies at the first point of connection. ie, subpanel or combiner box. Is this true?

In a word: No. Every panel between the interconnection and the service must comply with 705.12(B)(3) (in 2020 NEC), although there are more than one way to comply and they don't have to all comply in the same way. If you think about the currents and pathways involved it should be obvious that whoever is telling you that is mistaken.

 

[jaggedben]

Ggunn is correct.

In the 2011 code there was language something along the lines of 'where panelboards are connected in series, the first breaker connected to the PV source shall be used in the calculations'. However that language was deleted in the 2014 NEC because the rules were changed from using breaker ratings to '125% of inverter output'. Evidently the code making panel thought it would be obvious that 125% of inverter output may be fed into any of the panelboards in series, although to some it is not obvious. But that old language makes it clear that all panelboards in series are covered by the rule.

 

[Carultch]

I have a question about bus ratings. From my understanding, all busbars need to follow the rules of 705.12(b)(2)(3) if they have two power sources feeding into them. There are contractors and inspectors telling me that 705.12(b)(2)(3) only applies at the first point of connection. ie, subpanel or combiner box. Is this true?

If this were the case, there would be a loophole in a situation that can't meet the 120% rule, where you could insert a dummy subpanel between the point you want to interconnect, put one load in it and call that the point of interconnection, and defeat the entire intent of this rule.

For instance, suppose you had a main panel with a 400A main breaker and a 400A busbar. Suppose you desired to connect a 100A system to it. This cannot meet the 120% rule in this panel. So you pop out the final load on this panel, and put it in a 200A subpanel behind a 125A breaker from the main panel. Your calculation in the subpanel gets to be (125A + 100A) <= 1.2*200A, which passes. But if you recognize the realistic event that this local subpanel load is temporarily zero, the 100A source will propagate to the 400A main panel, which fails the 120% rule in this example.

 

[Inspector1989]

thanks everyone for the responses.

If this were the case, there would be a loophole in a situation that can't meet the 120% rule, where you could insert a dummy subpanel between the point you want to interconnect, put one load in it and call that the point of interconnection, and defeat the entire intent of this rule.

For instance, suppose you had a main panel with a 400A main breaker and a 400A busbar. Suppose you desired to connect a 100A system to it. This cannot meet the 120% rule in this panel. So you pop out the final load on this panel, and put it in a 200A subpanel behind a 125A breaker from the main panel. Your calculation in the subpanel gets to be (125A + 100A) <= 1.2*200A, which passes. But if you recognize the realistic event that this local subpanel load is temporarily zero, the 100A source will propagate to the 400A main panel, which fails the 120% rule in this example.

So in your scenario, one way to meet code is to relocate some loads on the MSP to the subpanel to make sure the MSP meets 702.15(B)(2)(3)(c) which is (the sum of the ampere ratings of all overcurrent devices on panelboards shall not exceed the ampacity of the busbar.) That way both busbars meet the 705.12 rules?

 

[jaggedben]

thanks everyone for the responses.


So in your scenario, one way to meet code is to relocate some loads on the MSP to the subpanel to make sure the MSP meets 702.15(B)(2)(3)(c) which is (the sum of the ampere ratings of all overcurrent devices on panelboards shall not exceed the ampacity of the busbar.) That way both busbars meet the 705.12 rules?

That could be possible in some situations, yes.

 

[Carultch]

thanks everyone for the responses.


So in your scenario, one way to meet code is to relocate some loads on the MSP to the subpanel to make sure the MSP meets 702.15(B)(2)(3)(c) which is (the sum of the ampere ratings of all overcurrent devices on panelboards shall not exceed the ampacity of the busbar.) That way both busbars meet the 705.12 rules?

I was playing devil's advocate with my "suggested solution", but your further suggestion is one way to make it work. I've elaborated on an example to show what this might look like in practice.

Suppose you started with a 400A main/bus panelboard that initially contained branch breakers that add up to 500A and you desire to interconnect a 100A source. As a loads-only panel, the underlying assumption is that they don't all draw their full load at once, and if they do, the 400A main will trip before anything is overloaded. Suppose you then relocate branch breakers that add up to 200A to a subpanel fed from a 100A breaker in the main. This will enable you to qualify the main panel under the rule about summing the breakers. The load-only branch breakers now add up to 300A, and introducing a 100A breaker for your subpanel and POI, will make the branch breakers sum not exceed 400A.

Now examine the subpanel. You feed it from a 100A breaker at the top for its main supply, and you feed it from a 100A inverter source at the bottom. You make its busbar 200A, and you populate it with a total of 200A for the relocated loads. It qualifies under the 120% rule, since (100A + 100A)<=1.2*200A.

By doing this, you depend on the assumption that the load breakers will never draw more than 100A at once from the main supply. You would probably need a load calculation to justify this, showing that the subpanel branch breakers are statistically unlikely to draw enough to trip the 100A branch breaker in the main 400A panel. One way you can make this more likely, is if you select a mixture of loads to relocate that are either very unlikely to run concurrently, or cannot run concurrently by design.

 

[KDough]

I have a question about bus ratings. From my understanding, all busbars need to follow the rules of 705.12(b)(2)(3) if they have two power sources feeding into them. There are contractors and inspectors telling me that 705.12(b)(2)(3) only applies at the first point of connection. ie, subpanel or combiner box. Is this true?


Scenario in a New PV installation - 200 amp main service panel, no changes are made to existing breakers, 80 amp back-fed PV breaker on opposite end of Main breaker. Feeding into a subpanel 80amp subpanel/combiner panel that has 4-20amp back-fed PV breakers installed in them.

I am told that this is okay because the subpanel/combiner panel meets 705.12(b)(2)(3) and only the first busbar needs to meet point of connection rules. not the second busbar.

Can someone elaborate the code on this? thanks

I do not believe that would be a code compliant installation. That would be the same as putting an 80a inverter circuit in the 200a meter main. You would start the calculation with the sum of the inverters max output. Take worse case 16a per 20a breaker = 64a. That would still put you 24a over the largest solar back feed you could put into a 200a main with a 200a rated buss.
The difference between a sub panel and a combiner box is loads. Combiner boxes have no loads, with the exception of the communication circuit. So when you say “sub panel/combiner panel” that is not a correct statement. It is either one or the other. You can put power sources into a sub panel but you can not put loads into a combiner box / panel.

 

[Carultch]

I do not believe that would be a code compliant installation. That would be the same as putting an 80a inverter circuit in the 200a meter main. You would start the calculation with the sum of the inverters max output. Take worse case 16a per 20a breaker = 64a. That would still put you 24a over the largest solar back feed you could put into a 200a main with a 200a rated buss.
The difference between a sub panel and a combiner box is loads. Combiner boxes have no loads, with the exception of the communication circuit. So when you say “sub panel/combiner panel” that is not a correct statement. It is either one or the other. You can put power sources into a sub panel but you can not put loads into a combiner box / panel.

I disagree. The busbar protection rules we have in 705.12, do not specify what the panel's application has to be, in order to use any given rule. There is no distinction in these rules between combiner panels and panels with a mixed application of sources and loads. The rule about summing up breakers to the busbar ampacity, doesn't "care" whether the breakers are sources or loads. They could ALL be sources, they could ALL be loads, they could be a 50/50 mix of sources and loads. No matter what current you draw/supply among them, and no matter what physical arrangement they have in the panelboard, the total current on the busbar will not exceed the busbar rating. Likewise, the total heat generated among the breakers, will not exceed what it would be, if a the breakers were load breakers, all drawing no greater than the ampacity of the main supply.

The reason we have the 120% rule, and its less-often used 100% rule counterpart, is that load panelboards in general are routinely populated with breakers that add up to far greater than the busbar rating, and far greater than the main supply. The panelboard depends on its main breaker, to stop the load on the busbar from exceeding the the main supply. And it depends on load diversity for it to be an unlikely event that you trip the main breaker. When you backfeed it with a second source, there is a chance that the two sources will add up in excess of the busbar rating, and the overload becomes a blind-spot of the main breaker. That is what could happen, when the panelboard is filled with numerous breakers that total a lot greater than the main supply. The 100% rule safeguards against this, when you interconnect on the same side as the main supply. The 120% rule enables you to take some credit for Kirchhoff's current law (KCL) so that you can have some headroom to interconnect on a panel with a main that matches the bus, and is an industry compromise for why it is 120% instead of 200% as KCL alone would seemingly allow.

If the breakers don't total to be greater than the busbar rating, it is a moot point to anticipate a condition of overload that could exceed the busbar rating. The branch breakers themselves, stop that overload condition from happening, and do not depend on the main supply to do it for them.

By using the scenario of breaking out enough loads from the main panel into a subpanel, to take credit for the rule of branch breaker summing in the main panel, you are taking the risk of tripping the main breaker out of the main panel, and moving it to the subpanel and its supply breaker instead.

 

[Inspector1989]

This scenario here would require derating the MCB, dropping down to a 150 would satisfy code?

Another option would be to install a 400 amp load center and relocate loads from MSP to load center.

Powerwall+ inverter continuous output is 32 amps. We have 2 powerwall+ so (32x2x1.25) = 80amps backfed.

 

[wwhitney]

This scenario here would require derating the MCB, dropping down to a 150 would satisfy code?

Looks like (2) PW+ and (1) PW2? That would add up to (including 125% factor) 40A + 40 A + 30A = 110A of inverters. So for the 120% rule in the 200A main service panel, you'd need to drop the main breaker to 125A.

Probably better would be to reduce the "(E) Loads" in it to 75A per pole sum of breakers, so the service panel can qualify under the 100% rule. Loads could go into the (N) 225A Load Center, as long as the 3 inverter breakers are at the opposite end of the bus as the utility supply, since it would then qualify under the 120% rule: 125A + 110A < 225A * 120%.

If I'm wrong about there being a PW2 (but what else is "Energy Storage System" on a 30A breaker?), then the inverters (with 125% factor) add up to 80A, so the service panel main breaker would only need to be reduced to 150A as you say. Or if loads are moved into the (N) 225A Load Center, the breaker order in that panel no longer matters, as 125A + 80A < 225A.

Cheers, Wayne

 

[jaggedben]

Note that Powerwall is a listed Power Control System, so they may be able to simply set the system to not output more than 32A to the main panel. Although in this case I'm a little doubtful that meets their goals.

 

[AEStech]

In a word: No. Every panel between the interconnection and the service must comply with 705.12(B)(3) (in 2020 NEC), although there are more than one way to comply and they don't have to all comply in the same way. If you think about the currents and pathways involved it should be obvious that whoever is telling you that is mistaken.

Can you point to where this is spelled out in the 2020 code?

Of course the logic makes sense, but I can't find any reference in the codebook that talks about anything other than the point of interconnection.

Thanks,

 

[pv_n00b]

Can you point to where this is spelled out in the 2020 code?

Of course the logic makes sense, but I can't find any reference in the codebook that talks about anything other than the point of interconnection.

Thanks,

It's not spelled out. The lack of it being limited to only the first panel in the system means that it applies to all. Nowhere in the code does it say it's limited to the point of interconnection, 705.12 refers to the point of connection, not interconnection. That's a very common misunderstanding.

 

[wwhitney]

The second sentence of 2020 NEC 705.12 says "Where distribution equipment or feeders are fed simultaneously by a primary source of electricity and one or more other power source . . ."

So when equipment has more than one connection that is a source of electricity, 705.12 potentially applies. That would include every piece of equipment between the service and a grid-tied inverter.

Cheers, Wayne

 

[ggunn]

The NEC's attempts to make its language so flexible and general that it covers every possible eventuality often only succeeds in my reading a section over again and saying "Whaaa...?" every time.

 

[pv_n00b]

The NEC's attempts to make its language so flexible and general that it covers every possible eventuality often only succeeds in my reading a section over again and saying "Whaaa...?" every time.

It's alternative interpretation bingo. Grab a card and play along.