Requirement for breaker feeding inverter

Status
Not open for further replies.

Grouch1980

Senior Member
Location
New York, NY
Looking at section 690.64(B)(2) in the 2008 NEC code cycle:

BUS OR CONDUCTOR RATING. The sum of the ampere..... .... shall not exceed 120 percent of the rating of the busbar or conductor. In systems with panel boards connected in series, the rating of the first overcorrect device directly connected to the output of a utility-interactive inverter(s) shall be used in the calculations for all busbars and conductors.

Where i highlighted in bold and underlined above... I have no idea what that statement means or is trying to imply. what do they mean by the 'first' device? and why does that matter?

Thanks again!
 

jaggedben

Senior Member
Location
Northern California
Occupation
Solar and Energy Storage Installer
Dude you are really causing me some deja vu or something with these old (to me) code questions. Anyway...

You are allowed to connect inverters at a subpanel. However both your subpanel and any panels 'upstream' must follow the rules as well.

So for example let's say you have an inverter on a 40A breaker in a 125A subpanel that is fed from a 100A breaker in the main service panel. The language you are talking about says that you get to use the 40A solar breaker when calculating the rules in the main panel. You do not have to use the 100A subpanel breaker, because the 40A solar breaker is the 'first device'. Therefore if your main panel is 200A breaker and busbar, you still meet the 120 percent limit.

In the 2014 NEC all of the parts referring to inverter breaker ratings were changed to '125 percent of inverter output', so the language in question was no longer necessary because breaker ratings were no longer used for such calculations.
 

Grouch1980

Senior Member
Location
New York, NY
So if that sub-panel has two breakers that each feed an inverter (just for argument sake), both those breakers have to be used in the calculation for the 120% limit on the main service panel busbar?

Ha, sorry man! NYC is still behind and stuck with the 2008 NEC code. It's ridiculous.
 

Carultch

Senior Member
Location
Massachusetts
So if that sub-panel has two breakers that each feed an inverter (just for argument sake), both those breakers have to be used in the calculation for the 120% limit on the main service panel busbar?

Ha, sorry man! NYC is still behind and stuck with the 2008 NEC code. It's ridiculous.

It depends on if that subpanel has a mix of sources and loads, or if that subpanel is dedicated to combining the outputs of the inverters.

If it is a mix of sources and loads, then yes, you would add up both those breakers, to govern compliance with interconnecting at the subpanel. You would then continue using that same sum, for the calculation propagating to the main panel.

If that subpanel is dedicated to combining those inverters, and contained no loads of any significant amps, then you'd treat them as if they were one larger inverter, connecting at the main panel, and use the rating of the interconnecting breaker where they connect at the main panel.

As of 2014, you no longer use the breaker for this calculation, but you use the calculation that sizes the breaker as the input to the 120% rule. Since we don't have a word for what "125% of Inverter FLA" is called, the wording of the code is particularly confusing and verbose, such that you loose track of what you are supposed to do with 125% and 120%. It is easier to see it written out as an equation. (1.25*I) + M <= 1.2*B, where I is the total of the inverter FLA, M is the main breaker trip rating, and B is the busbar ampacity. This makes it such that rounding errors are no longer a show stopper.
 

Grouch1980

Senior Member
Location
New York, NY
If it is a mix of sources and loads, then yes, you would add up both those breakers, to govern compliance with interconnecting at the subpanel. You would then continue using that same sum, for the calculation propagating to the main panel.

If that subpanel is dedicated to combining those inverters, and contained no loads of any significant amps, then you'd treat them as if they were one larger inverter, connecting at the main panel, and use the rating of the interconnecting breaker where they connect at the main panel.
So I have 3 questions based on the above...
1. In either case you pointed out, do you take into the consideration the bus rating of the sub-panel? or the calculation is always based on the bus of the main panel only?
2. Not sure i understand the difference. is the only difference the way i add up the breakers? In point 1, i just add the breakers. In point 2, i combine the inverters and calculate the full sized breaker.
3. Is this in the newer codes? I did not come across this in the 2008 NEC, where there's a difference if the sub-panel is a mix of sources and loads or not.
 

Carultch

Senior Member
Location
Massachusetts
So I have 3 questions based on the above...
1. In either case you pointed out, do you take into the consideration the bus rating of the sub-panel? or the calculation is always based on the bus of the main panel only?
2. Not sure i understand the difference. is the only difference the way i add up the breakers? In point 1, i just add the breakers. In point 2, i combine the inverters and calculate the full sized breaker.
3. Is this in the newer codes? I did not come across this in the 2008 NEC, where there's a difference if the sub-panel is a mix of sources and loads or not.


1. Draw an imaginary box around the panelboard you are studying. Ignore all other panels, until we get to them. Determine its main supply OCPD trip rating and call it M, whether inside the panel, or external. Examples of an external main supply OCPD would include (but not limited to) the following: a fused disconnect, an enclosed circuit breaker, the breaker in a meter/main combination device, the breaker in a higher ranking panel that supplies an MLO subpanel. Then determine the busbar rating (which we'll call B), and the total rating of the interconnection circuit breaker(s) that we'll call I.

NEC2014 and later, we no longer use breaker ratings to determine I, but rather we use the calculation that sizes that breaker prior to rounding up to a breaker rating. An example where this makes a difference, is given a 600A main panelboard with a 600A main breaker. Instead of limiting you to a 110A breaker (88A of current), it now allows up to 96A of current, and it is OK that it is interconnected on a 125A breaker. Even though 125A+600A exceeds 120% of 600A, it isn't the 125A that matters. It's the 125% of 96A that matters instead.

Your condition is as follows, to use the 120% rule, assuming you feed the panel on opposite ends of the busbar:
I + M <= 1.2*B

Once the first panelboard at the point of interconnection passes, move on to studying the higher ranking panelboards, until you get to the service point. Each panelboard from point of interconnection to the service point must qualify under one of the busbar protection rules in 705.12. You do not necessarily need to select the same rule each time.

2. The difference is that when you mix sources and loads, you have more total amps among all the breakers. The reason we don't get to use a 200% rule for feeding panelboard busbars from opposite ends, as Kirchhoff's current law alone would imply, is that it is an industry compromise to limit you to 120% of the busbar rating. The underlying issue is that the 200A panelboard is not tested for 400A worth of current among all the breakers, even when no more than 200A are present at any given cross section of the busbar. The industry compromise is that it is OK to have 20% extra current that is potentially feeding the breakers, provided the second source is strategically located to not add up with the main supply.

When you only have inverter sources in a panelboard, these devices don't draw current, they supply current. This is what installers would call an AC combining panelboard. You aren't at the point of interconnection yet, where there is a mixture of sources and loads. The most the current can add up to, both on the busbar and among the branch breakers, is the sum of the inverter output current ratings. You'd still apply the 125% factor for continuous loads when sizing OCPD's along the main supply, and that OCPD would still have to protect the busbar. But you don't get the situation of current travelling in both directions on the busbar, and feeding all the branch breakers with more than the busbar rating worth of current.

Suppose you use an AC combiner to add up two 14A inverters, each behind a 20A breaker. If both those 20A breakers are in a mixed use panelboard, you'd use 40A as the input for your 705.12 calculation. However, if you first combine them in a dedicated load center, then you get to interconnect them at a single 35A breaker. Now you get to use 35A in your calculation as your inverter interconnection breaker, instead of 40A. This is per the wording of the 120% rule in 2011 and earlier.

3. Article 705 is about interconnected power sources. In a dedicated AC combining panelboard, you aren't interconnected yet. You only have one type of power source and you are paralleling your inverter outputs, for an aggregate feed to the general electrical distribution of the service. The total amps on a dedicated AC combining panelboard is not going to exceed the sum of the inverter output currents, because there aren't any loads to also draw current. It is a continuous load, so you still have to apply the 125% factor for sizing the bus and main breaker, but that's really all it will be. You aren't going to have load branch breakers that add up to far greater than the main supply or busbar rating, as is common for load panelboards in general.

Load panelboards in general are routinely filled with loads that can add up to a lot more than the main supply. The underlying assumption is that it is an unlikely scenario that they all draw their full load at once. If they do, the main supply will trip before the busbar gets overloaded. If you add a second source to the panelboard, now that second source can support the main supply in supplying these loads, and potential overload of the panelboard becomes a blindspot to the main breaker. That's why we are restricted in the 705.12 rules, when mixing sources and loads in the same panelboard.
 
Last edited:

jaggedben

Senior Member
Location
Northern California
Occupation
Solar and Energy Storage Installer
...
3. Is this in the newer codes? I did not come across this in the 2008 NEC, where there's a difference if the sub-panel is a mix of sources and loads or not.

That stuff is all in the newer codes, 2014 or later.

Part of the reason that the code was changed is that the old language you're dealing with was imprecise or overly restrictive when it came to 'rounding errors' and such.

For example let's say you have three 5kW single phase inverters, each outputting 21A. You're combining them in a subpanel that you'll then connect to the main. Individually at 125% they each need a 30A breaker, but combined they only need an 80A breaker. The 2008 code would have you add up the three 30A breakers to determine you subpanel breaker and feeder. The 2014 and later codes let you do the math on the actual combined inverter output and use an 80A.

Also, by the letter of the 2008, it doesn't matter if the subpanel also has loads or not, when it comes to busbar rating. But some AHJs were seeing reason and making allowances.
 

Grouch1980

Senior Member
Location
New York, NY
3. Article 705 is about interconnected power sources. In a dedicated AC combining panelboard, you aren't interconnected yet. You only have one type of power source and you are paralleling your inverter outputs, for an aggregate feed to the general electrical distribution of the service. The total amps on a dedicated AC combining panelboard is not going to exceed the sum of the inverter output currents, because there aren't any loads to also draw current. It is a continuous load, so you still have to apply the 125% factor for sizing the bus and main breaker, but that's really all it will be. You aren't going to have load branch breakers that add up to far greater than the main supply or busbar rating, as is common for load panelboards in general.

Load panelboards in general are routinely filled with loads that can add up to a lot more than the main supply. The underlying assumption is that it is an unlikely scenario that they all draw their full load at once. If they do, the main supply will trip before the busbar gets overloaded. If you add a second source to the panelboard, now that second source can support the main supply in supplying these loads, and potential overload of the panelboard becomes a blindspot to the main breaker. That's why we are restricted in the 705.12 rules, when mixing sources and loads in the same panelboard.
wow.... lot of good stuff here. I have to read up on Article 705. Thanks for the examples... I just have to take some time to go through everything.
 

Grouch1980

Senior Member
Location
New York, NY
That stuff is all in the newer codes, 2014 or later.

Part of the reason that the code was changed is that the old language you're dealing with was imprecise or overly restrictive when it came to 'rounding errors' and such.

For example let's say you have three 5kW single phase inverters, each outputting 21A. You're combining them in a subpanel that you'll then connect to the main. Individually at 125% they each need a 30A breaker, but combined they only need an 80A breaker. The 2008 code would have you add up the three 30A breakers to determine you subpanel breaker and feeder. The 2014 and later codes let you do the math on the actual combined inverter output and use an 80A.

Also, by the letter of the 2008, it doesn't matter if the subpanel also has loads or not, when it comes to busbar rating. But some AHJs were seeing reason and making allowances.
Thanks for this as well. I have to go through these last 2 posts and read everything.
 

Grouch1980

Senior Member
Location
New York, NY
Hey guys... I ready through everything... and it all made sense. I appreciate the explanation on how the 120% limit came about, and why the inverter breaker would be located on the opposite side of the panel / sub-panel bus bar.

I do have a few follow-up questions.
1. jaggedben, you mention in your example the 90 amp breaker versus the 80 amp breaker, depending on the applicable NEC version. I still follow the 2008 code, so it appears that I would have to use the 90 amp breaker. Would you know what section implies adding up the actual breaker ratings?
2. from jaggedben also: "Also, by the letter of the 2008, it doesn't matter if the subpanel also has loads or not, when it comes to busbar rating. But some AHJs were seeing reason and making allowances."... this i did not follow... could you elaborate a little?
3. from Carultch: "Now you get to use 35A in your calculation as your inverter interconnection breaker, instead of 40A. This is per the wording of the 120% rule in 2011 and earlier.".... do you mean 2011 and 'later'?
4. General question: when you have a circuit breaker that's back-fed rated, can that breaker still operate in forward-fashion? or it can only strictly operate in reverse?

Thanks again!
 

Carultch

Senior Member
Location
Massachusetts
"Now you get to use 35A in your calculation as your inverter interconnection breaker, instead of 40A. This is per the wording of the 120% rule in 2011 and earlier.".... do you mean 2011 and 'later'?

No, I meant 2011 and earlier. The 120% rule specified to use the interconnecting breaker rating. The fact that there are inverter breakers on the load side of the 35A interconnection breaker that add up to a greater rating, is irrelevant, when studying the main panelboard where you interconnect.

In 2014 and later, you could use a 100A breaker in that example to connect the AC combiner to the point of interconnection, and you'd still get to use 35A in your calculation. Because it is no longer determined by the breaker rating, but rather by the calculation that sizes the breaker, prior to selecting a breaker.

4. General question: when you have a circuit breaker that's back-fed rated, can that breaker still operate in forward-fashion? or it can only strictly operate in reverse?

Breakers are always capable of being operated in the forward direction. A backfed rated breaker means that it can operate with power flowing in either direction. Most breakers made today are backfed rated, provided that they don't have extra functionality such as AFCI or GFCI. For breakers with extra functionality, you have to be careful with product selection.
 

jaggedben

Senior Member
Location
Northern California
Occupation
Solar and Energy Storage Installer
...

I do have a few follow-up questions.
1. jaggedben, you mention in your example the 90 amp breaker versus the 80 amp breaker, depending on the applicable NEC version. I still follow the 2008 code, so it appears that I would have to use the 90 amp breaker. Would you know what section implies adding up the actual breaker ratings?

It's really just the section you started this thread with. "The sum of ampere ratings of overcurrent devices in circuits supplying power to a busbar or conductor.." That includes all three of your inverters in the example above. The part about the " first overcorrect device directly connected to the output of a utility-interactive inverter(s)" (note the possibility of plural inverters) implies that the 'sum' that you add up includes the three 30A breakers.

2. from jaggedben also: "Also, by the letter of the 2008, it doesn't matter if the subpanel also has loads or not, when it comes to busbar rating. But some AHJs were seeing reason and making allowances."... this i did not follow... could you elaborate a little?
...

The part about loads is actually a little imprecise, but what I was referring to is that in 2014 and later codes there is a new rule that allows busbars to be rated less, if the sum of the breakers in the panel (not including the main breaker protecting it) does not exceed the busbar. It was put in to allow AC combiner panels to be rated less, although it also works in some cases where there are also loads in the panel.

Going back to our example of three inverters with 21A output being combined in a subpanel and then connecting to the main. Let's add the stipulation that there'd be no other breakers in that panel. In the 2008 code, your breakers cannot exceed 120 percent of the busbar rating. Your breaker sum (both inverters and main) is 30+30+30+90=180. 180A/1.2=150A So you need a 150A panel. In the 2014 code, you get to take 30A*3=90A. So you can use a panel rated 90A, or in practice a 100A panel instead of a 150A panel.

Some AHJs saw that a lower rated panel was fine and made allowances, sometimes requiring a label to state that no loads be added. 2014 NEC legitimized using a smaller panel in that example.

This is all a bit useless to you until you get on a later code, but you seem to want to know. ;)
 

Grouch1980

Senior Member
Location
New York, NY
So to wrap up, let's see if i have this right:

2008 / 2011 codes:
1. On an AC combiner sub-panel, the main breaker feeding it (located on the main panel) would be the sum of the inverter breakers located on the combiner sub-panel. So three 30 amp inverter breakers on the combiner panel would require a 90 amp main breaker on the main panel to feed that sub-panel.
2. The busbar of the AC combiner panel cannot be less than 120% of the sum of the main 90 amp plus the three 30 amp breakers. Based on the 2008 / 2011 codes, it doesn't matter if the sub-panel is purely sources or sources and loads, you still have to follow the 120% rule.

2014 code (or later):
1. Using the same numbers... the main breaker feeding the AC combiner sub-panel can be less than 90 amps... it's based on the actual inverter ratings, times 125%, rather than just adding up the three breakers. However i think this contradicts what jaggedben wrote directly above: "In the 2014 code, you get to take 30A*3=90A. So you can use a panel rated 90A, or in practice a 100A panel instead of a 150A panel."... can't you use less than a 90 amp breaker?
2. The AC combiner sub-panel, if it doesn't feed any loads, and only contains inverter sources, now doesn't have to follow the 120% rule.
 

thekwhour

Member
Location
California
Occupation
Solar, battery storage, and EV charging
why are you not doing it to the latest code? Isn't the latest code there for the best possible scenario for protection. So a new solar system will go in New York City and be updated to a 2008 or 2014 code. hmm seems like bad place to invest.
 

Grouch1980

Senior Member
Location
New York, NY
I was able to view the 2014 NEC code online, while comparing it to the 2008 code. Having them side by side, and reading carultch and jaggebens posts, I was able to finally understand it. It's amazing how much more the 2014 code added, and how it broke down the requirements.
 

Grouch1980

Senior Member
Location
New York, NY
why are you not doing it to the latest code? Isn't the latest code there for the best possible scenario for protection. So a new solar system will go in New York City and be updated to a 2008 or 2014 code. hmm seems like bad place to invest.
I think this opens the door to violations from the inspector. you have to follow the current code your jurisdiction follows, not the latest NEC. If i do something that the latest NEC permits, but wasn't part of the 2008 NEC, I will get screwed with a violation and a potential change order that the client would have to pay extra for. guess who gets yelled at?! Me.
 
Status
Not open for further replies.
Top