Is it legal to land multiple solar backed breakers in one panel?

jaggedben

Senior Member
They don't? What about 705.12(B)(5). Fastening. Listed plug-in-type circuit breakers backfed from electric power sources that are listed and identified as interactive shall be permitted to omit the additional fastener normally required by 408.36(D) for such applications.

Maybe it is how you interpret the words... is the "that are listed and identified as interactive" modifying plug-in circuit breakers or electric power sources? Without any commas, you may be right that feeding a breaker from a listed and identified interactive power source is all you need to eliminate the hold down. I was reading it the other way (e.g. there could be plug-in circuit breaker that are listed as "interactive" types and only those don't need the hold down).
So yes, using parentheses in the mathematical sense, it is:

"Listed plug-in-type circuit breakers backfed from (electric power sources that are listed and identified as interactive) shall be..."

not

"Listed plug-in-type circuit breakers (backfed from electric power sources) that are listed and identified as interactive..."

In addition to what others said, I don't believe there is a product standard for listing circuit breakers as interactive. There is such a standard for inverters. Also the two instances of 'listed' are not redundant, one refers to the circuit breakers and one refers to the power sources.
 

ggunn

PE (Electrical), NABCEP certified
Location
Austin, TX, USA
Occupation
Electrical Engineer - Photovoltaic Systems
In addition to what others said, I don't believe there is a product standard for listing circuit breakers as interactive.
There is, however, or at least used to be, language in the code that says/said that if a breaker is stamped "LINE" and "LOAD" you cannot backfeed it. I have never seen such a breaker, but they may exist.
 

JoeNorm

Senior Member
Location
WA
We Just got called on this. Inspector is referencing 705.12(B)(1) Dedicated Overcurrent and Disconnect.

The way I interpret this is that each inverter would need its own breaker, but it wouldn't necessarily limit the amount of breakers.

What do you all think?
 

jaggedben

Senior Member
You are correct. You might try pointing out that the code language actually changed from the 2014 NEC to the 2017 NEC, precisely to preclude the misinterpretation he's making. Perhaps he just hasn't caught up to the change. Another point is that if you have multiple inverters there is no sense in which each doesn't satisfy the definition of a separate system and thus a separate source.

The intent of this section was never to limit all sources to a single overcurrent device, but rather to preclude loads from being on the same branch circuit. I don't know why the code-making-panel has never agreed on putting 'no loads' language into the code, but at least one person heavily involved (Bill Brooks) advocated doing that.
 

jaggedben

Senior Member
Not all battles with inspectors are worth fighting. You'll have to be the judge if it's worth it, and make a note about the jurisdiction if you decide it's not.
 

JoeNorm

Senior Member
Location
WA
Thank you for clarifying. You're right in that not all battles are worth it, but being able to eliminate an entire panel(combiner) when possible is worth it to me.

On further review of the code its obvious that the new words in the 2017 make multiple connections a clear option.

I have a hard time when inspectors are (1)stubborn interpreters of the code for no good reason and (2) fail you for something that has nothing to do with inherent safety of a system.
 

GoldDigger

Moderator
Staff member
The opposite end provision is there to make sure that there are no load branch breaker downstream of the PV breaker(s) which would therefore be able to draw more current from the bus than the bus current rating. A stack of PV breakers at the far end satisfies this condition just fine, but an inspector can freely argue that logic has no probative value where code interpretation is involved. :)

Sent from my XT1585 using Tapatalk
 
Every code section from 2014 705.12(D)(1), 2017 705.12(B)(1) to present 705.12(A) has the same requirement "Each source interconnection of one or more power sources installed in one system shall be made at a dedicated circuit breaker or fusible disconnecting means". Not trying to be "stubborn interpreter for no good reason", just trying to properly interpret the code fairly for all contractors.
 

jaggedben

Senior Member
Every code section from 2014 705.12(D)(1), 2017 705.12(B)(1) to present 705.12(A) has the same requirement "Each source interconnection of one or more power sources installed in one system shall be made at a dedicated circuit breaker or fusible disconnecting means". ...
Well, actually that's not quite true, and this speaks to the issue. The language changed in three successive code cycles in an apparent attempt to clarify the requirement. (It still isn't totally successfully clarified in my opinion.)

2011:
Each source interconnection shall be made...
2014:
The source interconnection of one or more inverters installed in one system shall be made...
2017
Each source interconnection of one or more power sources installed in one system shall be made...
(Emphasis mine.)

To my understanding, 2014 was changed to make it permitted to install more than one inverter on a single dedicated overcurrent and disconnect (e.g. micro-inverters), not to require that all of them on a premise be connected to one dedicated disconnect. In 2017, 'the' was changed back to 'each' so as to make it clear that there could be multiple dedicated disconnects. At least that's the best sense I can make of why 'the' would be changed to 'each.' Other code sections allow multiple systems.

If you read the NFPA code making documents it's clear that the CMP has not ever intended the NEC to require that all sources on a premise go through a single disconnect.
 
The 2014 Handbook explains that each source can include solar, generator and other power sources, etc. That is why they have sources in 705.12(A). It is not the intent to have all solar, generators and other sources connected together. But 705.12(D) makes it clear that each system connection of one or more inverters shall be made at a dedicated breaker or fusible disconnect on the load side of the main. Line side connection is addressed in 705.11 as an (meaning one IMHO) electric power production source.
 

jaggedben

Senior Member
The 2014 Handbook explains that each source can include solar, generator and other power sources, etc. That is why they have sources in 705.12(A). It is not the intent to have all solar, generators and other sources connected together. But 705.12(D) makes it clear that each system connection of one or more inverters shall be made at a dedicated breaker or fusible disconnect on the load side of the main. Line side connection is addressed in 705.11 as an (meaning one IMHO) electric power production source.
But look at the definition of a system. For example a PV system (nowadays in article 100). That definition in no way determines that if I have multiple solar inverters on a site that all the inverters belong to one system. If I have multiple microverters connected to a single disconnecting means then that is one system (most clearly in the 2020 NEC). But if I have multiple string inverters each with their own disconnecting means, or multiple micro-inverter circuits, then they simply become multiple systems.

The intention of the language in question is to keep loads off the power source circuit. The Handbook is not the code.
 

tortuga

Senior Member
Location
(44.057116, -123.103394)
Occupation
field supervisor
The 2014 Handbook explains that each source can include solar, generator and other power sources, etc. That is why they have sources in 705.12(A). It is not the intent to have all solar, generators and other sources connected together. But 705.12(D) makes it clear that each system connection of one or more inverters shall be made at a dedicated breaker or fusible disconnect on the load side of the main. Line side connection is addressed in 705.11 as an (meaning one IMHO) electric power production source.
In my opinion "interconnected electric power source" is a poor choice of words and occurs noplace else in the code, it should read
'Separately Derived System' as any UL listed grid interactive inverter is.

NEC Art 100 said:
Separately Derived System. An electrical source, other than
a service, having no direct connection(s) to circuit conductors
of any other electrical source other than those established by
grounding and bonding connections.
There was a time in the 70's when PV could have gone for its own chapter like 800 where the rest of the code did not apply, unless mentioned.
705.3 was written in a similar way and got deleted from the 2020 code, because the entire code applies to 705 unless 705 modifies it.
I think allot of people still have that perspective in this industry.
The definition of 'PV system' is a echo of that perspective. I could see things continuing to move out of 690 and getting integrated with the rest of the code. Then the definition of PV system could eventually get deleted, and requirements for inverters should move to phase converters, or a new article in chapter 4. And PV conductors are either branch circuits or feeders, with tap rules and additional requirements from 690.
Just my 2 cents
 

jaggedben

Senior Member
In my opinion "interconnected electric power source" is a poor choice of words and occurs noplace else in the code, it should read
'Separately Derived System' as any UL listed grid interactive inverter is.
Dude, this is wrong. I think it's not the first time you've said this: you should stop saying it. Because it's really not a matter of debate. Most inverters these days do not contain a transformer. They are no more a separately derived system than is any load with a switched power supply. Switched power supplies qualify as direction connections. An example of what this means in practice: on a transformerless inverter, a ground fault on the DC side can trip the AC inverter breaker through the main bonding jumper at the service. (I've seen it happen.) That does not happen on a separately derived system (where fault current goes through the system bonding jumper). Moreover, from a code standpoint, the CMP (to their credit) removed most of the superstitious grounding requirements for PV systems in the 2017 cycle, and it would be awful to bring that back through any implied requirement regarding Article 250 sections for separately derived systems. Sometimes power source systems contain a separately derived system (transformer, or if they are not interconnected), but "Interconnected electric power source" was developed to refer to a specific animal that is not the same as 'separately derived system'. The Venn diagram is two circles partially overlapping.

I think anyone on the CMP or with UL will agree with me.
 

tortuga

Senior Member
Location
(44.057116, -123.103394)
Occupation
field supervisor
Ben allow me to ask you this;
If we have for example a 2 wire PV electrical source of 20kW wired at 800V connected to a (red) 30A PV conductor.
This is 'directly connected' with a wirenut or tap block to some (blue) #8 THHN .
At the beginning of the circuit I take out my voltmeter and measure 800 VDC.
At the end of the blue conductor I measure 800VDC.
Say I have a 18kW 800VDC load that I can connect on the end, I think we can all agree that the connection between red and blue at the wirenut or tap block is a 'direct connection'?
sds-pv2.png
 

tortuga

Senior Member
Location
(44.057116, -123.103394)
Occupation
field supervisor
Given the NEC definition:
NEC art 100 said:
Separately Derived System. An electrical source, other than
a service, having no direct connection(s) to circuit conductors
of any other electrical source other than those established by
grounding and bonding connections.
The definition of direct:
direct
[ dih-rekt, dahy- ]
adjective
proceeding in a straight line or by the shortest course; straight; undeviating; not oblique:
Then if I replace the wirenut or tap block with a inverter that has a direct connection from the red to the blue what voltage should I be able to measure on both blue leads after the direct connection?

sds-pv1.png
 

Carultch

Senior Member
Location
Massachusetts
Given the NEC definition:

The definition of direct:

Then if I replace the wirenut or tap block with a inverter that has a direct connection from the red to the blue what voltage should I be able to measure on both blue leads after the direct connection?
If you have an inverter not tied to an AC grid, its contactors (or solid-state relays) will disconnect its core from its DC source. You would measure 0V on the output in your setup. Inverters work by monitoring the AC grid, and producing a waveform of slightly greater voltage in order to push current onto the line. The AC grid needs to be within the voltage and frequency specs outlined by the IEEE/UL standards that are pre-programmed into them in order to pass the safety testing.

The reason why a non-isolated inverter is considered enough of a "direct connection" to make the DC not be a "separately-derived system", is that the path between the DC and AC sides never leaves the realm of electricity. There are capacitors, inductors, and transistors "in the way", so it will be difficult to measure the continuity. It is due to the fact that there is no isolation transformer in a non-isolated inverter, that makes the DC side not "separately derived". The isolation functionality of a transformer, is what causes the NEC to define "separately derived systems". If you were using autotransformers, even though the two sides are at different voltages, the two sides would not be "separately derived".

The DC side is inherently dependent on the AC side, of a non-isolated inverter. The two polarities have to be equal and opposite to one another, once you connect to a grounded grid. The positive polarity is used to make the positive half of the sine wave, and the negative polarity for the negative half. They need to be equal and opposite when the unit is operating, to generate the waveform that is symmetric relative to ground. Unlike in an inverter with a transformer, where the inverter electronics generate a waveform between the positive and grounded negative, and then use the isolation property of the transformer to remove the DC offset.
 

jaggedben

Senior Member
Given the NEC definition:


The definition of direct:


Then if I replace the wirenut or tap block with a inverter that has a direct connection from the red to the blue what voltage should I be able to measure on both blue leads after the direct connection?

View attachment 2552258
Of course the blue leads will measure the grid voltage, but that's not really the answer to your question.

In order to 'see' the direct connection by measuring AC voltage, you measure from either of the DC terminals to the AC neutral. I have done this, on an operating transformerless inverter connected to a split phase grid connection. If I'm recalling correctly, the voltage was about 138V when the AC side was 120. (But I could be misremembering and I think it may have been higher.)

If you try to do this on an older inverter with a transformer and ground fault fuse, you will get zero from DC conductors to AC hots, whether you measure for DC or AC.

I hope that this, along with Carultch's explanation, is definitive for you.
 

tortuga

Senior Member
Location
(44.057116, -123.103394)
Occupation
field supervisor
Yes thank you gentleman well said. So there can be measurable AC between the AC neutral (or ground) and the DC side:
sds-pv0.png
This just brings up more questions for another time.

That must be what UL is referring to:
UL 1741 said:
47.6.2 In addition to compliance with 47.1.3, during the test described in 47.6.1 for a utility-interactive
inverter or converter, the maximum backfeed current that flows from the simulated utility source, see 41.3,
into the input source as a result of a faulted component shall not exceed the marked maximum input
source backfeed current. See Table 62.1, Item (e).
 
Top