NEC 690.8

[Cal Elliott]

I have just had an interesting interpretation of NEC 690.8 and would appreciate comments. I designed a solar system with four arrays/strings using 240 watt panels, each with a micro inverter. Three of the four arrays on a circuit have 16 panels and the fourth 15. running the math, using the manufacturers maximum AC output of 0.9 amps per inverter, the math is 16 x 0.9 or 14.4 amps. All four circuit run to a Solar Sub- Panel with four 20A-2P circuit breakers. The inspector and I agree that the Solar main breaker in the existing main panel is 75 amps. Where we disagree is the size of the wire between the sub-panel and the main panel. his interpretation of this section of the Code is that one must take the four breakers add them up (80amps), add 75 amps and divide by 1.2 equaling 129.19 amps minimum size for the wire. Where does this come from? this is not standard electrical design practice and I cannot find in this section of the Code justification for this. Where does the 75 amps come from? Interesting, we agree on the size of the solar sub-panel as also being 100 amps.

 

[SolarPro]

See 705.12(D)(2) [or 690.64(B)(2) if you're still on NEC 2008]. Per this section, the sum of the OCPD ratings supplying power to a busbar or conductor cannot exceed 120% of the rating of the busbar or conductor. Based on your description, it's hard to say why the AHJ is only applying this section to the conductor and not the subpanel.

 

[GoldDigger]

See 705.12(D)(2) [or 690.64(B)(2) if you're still on NEC 2008]. Per this section, the sum of the OCPD ratings supplying power to a busbar or conductor cannot exceed 120% of the rating of the busbar or conductor. Based on your description, it's hard to say why the AHJ is only applying this section to the conductor and not the subpanel.

Adding up all of the power source breakers (PV-end and feeder-breaker at the main panel) to determine the required ampacity of the feeder conductors only really makes sense if there is the possibility of a load being connected to the wire at some point along the way. If there is, then that load could utilize up to the sum of the currents supplied from both ends.

This has been discussed a few times before, with the consensus that it does not make any sense, but that the inspector probably can justify doing it if he holds firm. In theory there could be a stable high resistance short anywhere along the route of the feeder which could then overload the wires without tripping any of the breakers. Likely? No. Credible? Maybe. Perhaps an AFCI would help.

You are in an exposed position because the PV feed into the sub-panel is such a large fraction of the total rating of the panel. (BTW, how did you meet the 120% rule for the sub-panel?)

 

[Cal Elliott]

Adding up all of the power source breakers (PV-end and feeder-breaker at the main panel) to determine the required ampacity of the feeder conductors only really makes sense if there is the possibility of a load being connected to the wire at some point along the way. If there is, then that load could utilize up to the sum of the currents supplied from both ends.

This has been discussed a few times before, with the consensus that it does not make any sense, but that the inspector probably can justify doing it if he holds firm. In theory there could be a stable high resistance short anywhere along the route of the feeder which could then overload the wires without tripping any of the breakers. Likely? No. Credible? Maybe. Perhaps an AFCI would help.

You are in an exposed position because the PV feed into the sub-panel is such a large fraction of the total rating of the panel. (BTW, how did you meet the 120% rule for the sub-panel?)

Let me understand, four 20A-2P circuit breakers in the panel is 80 amps at 1.25% this is 100 amps. Using the 120% rule we are allowed 120 amps? But where in the Code is the inspector getting 75 amps additional load?

 

[GoldDigger]

Let me understand, four 20A-2P circuit breakers in the panel is 80 amps at 1.25% this is 100 amps. Using the 120% rule we are allowed 120 amps? But where in the Code is the inspector getting 75 amps additional load?

The 120% rule as applied to the sub-panel adds up the total of the PV breakers (80, with no 125% factor added. That is already in the breaker size) and the main breaker if any (if there is no main breaker use the feeder breaker at the main panel (75).) That would be 80 + 75= 155. That must not be more than 120% of the bus rating of the panel. If it is a 100A panel, then you would only be allowed 120A in the sum.
If the bus is rated at 150 (you are using an oversized panel or the panel maker uses an oversized bus set) then you are allowed a total of 180.

With 125A bus bars, often found in 100A panels, you are allowed exactly 150.

Again, the theory is that if you put 150A of load in the same panel, it could overload the bus by drawing from both sources. You only get the extra 20% allowance on the condition that the POCO feed and the PV feed are at opposite ends of the panel so that any load breakers would have to be between them and neither end of the bus would be overloaded.

The same argument, of a load in the middle and feed from both ends, leads to the application of the 120% rule to the feeder wires themselves. This is harder to justify if there is nowhere that a load could be connected along the feeder, but the code does not explicitly recognize this as an exception.

 

[jaggedben]

Using the 120% rule we are allowed 120 amps?

No, that is not what 705.12(D)(2) says.

It says that ratings of the breakers - utility plus solar - that can supply the busbar or conductor can be no more than 120% of the rating of that busbar or conductor.

For example, on a 100A busbar or conductor you can have 120A, say 100A utility and 20A solar, or 60A utility and 60A solar.

Thus his math is correct, (75+80)/1.2 = 129.2A or higher rating that you are required to have.

As GoldDigger alluded to, do not expect it to make sense according to electrical theory, it doesn't really.

 

[Smart $]

No, that is not what 705.12(D)(2) says.

It says that ratings of the breakers - utility plus solar - that can supply the busbar or conductor can be no more than 120% of the rating of that busbar or conductor.

For example, on a 100A busbar or conductor you can have 120A, say 100A utility and 20A solar, or 60A utility and 60A solar.

Thus his math is correct, (75+80)/1.2 = 129.2A or higher rating that you are required to have.

As GoldDigger alluded to, do not expect it to make sense according to electrical theory, it doesn't really.

The inspector's math may be correct, but his interpretation may be off. 705.12(D)(2) doesn't apply if 705.12(D) general condition doesn't exist... specifically, "...where this distribution equipment is capable of supplying multiple branch circuits or feeders or both..."

It appears the equipment on the PV side of the 75A breaker in the main panel does not supply any branch circuits or feeders. However, the key word in the general condition is "capable". There is no criteria established to make a definitive capable or incapable determination. Also, the OP'er hasn't supplied enough detail to sway my opinion one way or the other...

Are there any available spaces in the PV subpanel which can be used to supply loads?

Details of feeder wiring method: raceway, cable, underground, distance, inside, outside, etc...? IOW, need to establish the likelihood of it being tapped into to supply loads.

As an aside, the 705.12(D) general condition says multiple branch circuits or feeders... so is one is okay?

 

[jaggedben]

The inspector's math may be correct, but his interpretation may be off. 705.12(D)(2) doesn't apply if 705.12(D) general condition doesn't exist... specifically, "...where this distribution equipment is capable of supplying multiple branch circuits or feeders or both..."

Pretty clear to me from the OP's post that the language you are quoting applies.

It appears the equipment on the PV side of the 75A breaker in the main panel does not supply any branch circuits or feeders.

The inverter output circuits mentioned are branch circuits and feeders.

However, the key word in the general condition is "capable". There is no criteria established to make a definitive capable or incapable determination. Also, the OP'er hasn't supplied enough detail to sway my opinion one way or the other...

I think it's fairly obvious most of the time. Examples: A switchboard or panelboard is capable. A fused disco is not.

Are there any available spaces in the PV subpanel which can be used to supply loads?

As John Wiles has pointed out, the CMP has rejected proposals wherein this sort of thing would be taken into account. It's pretty clear that the CMP has thought about these issues and decided to remain very conservative in their approach. Some AHJ's may make allowances, but if they don't, that's pretty much their prerogative.

Details of feeder wiring method: raceway, cable, underground, distance, inside, outside, etc...? IOW, need to establish the likelihood of it being tapped into to supply loads.

The current code doesn't require anything to mitigate concerns about taps. The 2014 code will introduce the question of taps into 705 for the first time.

As an aside, the 705.12(D) general condition says multiple branch circuits or feeders... so is one is okay?

Again, the solar is a branch circuit or feeder, so if you've got solar plus a load then the equipment is "capable." Personally I would wager that the CMP intentionally included the word 'capable', as well as the phrase "including switchboards and panelboards" so as to be inclusive of distribution equipment in general. It is not a situational requirement; I think they pretty much intended 705.12(D) to apply generally to all load side connections. 2014 will update this section to be even more inclusive of various unaddressed situations, such as taps and center-fed panels.

 

[Smart $]

Pretty clear to me from the OP's post that the language you are quoting applies.

Yes, the language I'm quoting from 705.12(D) general statement applies... but it is a conditional statement which determines whether or not 705.12(D)(2) applies... and I'll recant my earlier statement that the key word is "capable" for the installation type under consideration here. Read on to see what the real key term is...

The inverter output circuits mentioned are branch circuits and feeders.

While many view them as such, they technically by Code definition are not branch circuit conductors.

Feeder. All circuit conductors between the service equipment,
the source of a separately derived system, or other
power supply source and the final branch-circuit overcurrent
device.

Branch Circuit. The circuit conductors between the final
overcurrent device protecting the circuit and the outlet(s).

Outlet. A point on the wiring system at which current is
taken to supply utilization equipment.

They are feeders if they do not otherwise fall under an Article 690 definition...

Inverter Output Circuit. Conductors between the inverter
and an ac panelboard for stand-alone systems or the conductors
between the inverter and the service equipment or
another electric power production source, such as a utility,
for electrical production and distribution network.

You can easily verify by looking at the Figure 690.1(A) & (B).

So starting from the microinverters, we're to the AC-combiner panel now with no feeders and no branch circuits. This is where the confusion starts because the combiner panel can easily be capable of supplying multiple branch circuits or feeders or both (again the question, is one okay? :blink. In the case of the OP, he did not state whether this panel is MCB or MLO, so I'll cover both. For this part, I refer you to 690.15 second paragraph, which says...

A single disconnecting means in accordance with 690.17
shall be permitted for the combined ac output of one or more
inverters or ac modules in an interactive system.

So we now have a single disconnect (either the MCB or the main panel breaker if MLO) under 690 that serves as a single disconnecting means for the Photovoltaic Equipment. Now refer to 705.12(D)...

The output of a utility interactive
inverter shall be permitted to be connected to
the load side of the service disconnecting means of the
other source(s) at any distribution equipment on the premises.

The term "distribution equipment" is not specifically defined by the NEC. However, by contextual usage, we consider it to be electrical equipment which distributes power from other sources to utilization equipment, and subject to articles other than 690.

Getting back to the single disconnecting means of 690.15, we have yet to get to any distribution equipment if MCB, and if MLO, the main panel breaker would be that connection point. In either case, the equipment on the inverter side of the disconnect is Photovoltaic Equipment... not distribution equipment. Thus 705.12(D)(2) is not applicable to the AC-combiner panel.... unless a "branch circuit" or "feeder" is added to the panel, which would then make it distribution equipment. It does not matter whether the panel is capable of having these added, because until they are added, the panel is not distribution equipment.

Now in the case of an MCB combiner panel, we have the question whether the conductors between it and the main panel breaker is subject to 705.12(D)(2). For this, I refer you to 690.14(C). In short, it requires a means to disconnect all conductors in a building or structure from photovoltaic system conductors, near the point of entrance. So again, we have yet to encounter any distribution equipment. 690.14(D)(3) and (4) further support this interpretation, as it requires disconnecting means in sight of the inverters and an additional means complying with 690.14(C)(1).

I think it's fairly obvious most of the time. Examples: A switchboard or panelboard is capable. A fused disco is not.


As John Wiles has pointed out, the CMP has rejected proposals wherein this sort of thing would be taken into account. It's pretty clear that the CMP has thought about these issues and decided to remain very conservative in their approach. Some AHJ's may make allowances, but if they don't, that's pretty much their prerogative.

The current code doesn't require anything to mitigate concerns about taps. The 2014 code will introduce the question of taps into 705 for the first time.

Again, the solar is a branch circuit or feeder, so if you've got solar plus a load then the equipment is "capable." Personally I would wager that the CMP intentionally included the word 'capable', as well as the phrase "including switchboards and panelboards" so as to be inclusive of distribution equipment in general. It is not a situational requirement; I think they pretty much intended 705.12(D) to apply generally to all load side connections. 2014 will update this section to be even more inclusive of various unaddressed situations, such as taps and center-fed panels.

Well I think I already covered this enough that no comment is required here.

I'll wait until 2014 edition is issued to comment on it...

 

[Zee]

Inspector is right per LETTER of the Code.

However, in practice this is what i do:
If you use....... say a 4/8 panel................ and use two @ 20/20 quads as your 4 @ 2p20A breakers......thus ensuring no physical way one could add loads in the future, then there is no danger of overloading either the busbar or the feeder conductors.

And go the extra mile, and label it DO NOT ADD LOADS! SOLAR ONLY!
He may smile, relent, and sign your permit.
Then again, he may not.

 

[SolarPro]

f you use....... say a 4/8 panel................ and use two @ 20/20 quads as your 4 @ 2p20A breakers......thus ensuring no physical way one could add loads in the future, then there is no danger of overloading either the busbar or the feeder conductors.

And go the extra mile, and label it DO NOT ADD LOADS! SOLAR ONLY!

Exactly. I've had AHJs and PEs approve strategies like this.

 

[Zee]

Solarpro,
Good to know that works elsewhere.