solar panel AC main disconnect

hhsting

Senior Member
Attached sketch shows main service disconnect that has line side tap to main AC solar disconnect. However, main AC solar disco has no main bonding jumper. Is this acceptable by code under 2014?
No one knows? No takers?
 

ggunn

PE (Electrical), NABCEP certified
Attached sketch shows main service disconnect that has line side tap to main AC solar disconnect. However, main AC solar disco has no main bonding jumper. Is this acceptable by code under 2014?
The short answer is yes. The longer answer is yes, subject to interpretation. Most AHJs I deal with say yes, but a couple say no. I just ask the AHJ and do what they say.
 

shortcircuit2

Senior Member
The 2020 NEC added new section 250.25 which requires separate supply-side connected system enclosures to comply with 250.24(A)-(D)... which will not only require a Main Bonding Jumper per 250.24(B), you also need a Grounding Electrode Conductor per 250.24(D) and it must be connected in compliance with 250.64(D)

Given that the 2014 NEC gives no direction for installer or AHJ on your supply-side connected PV Disconnect...I agree ask the AHJ...but I would suggest to install the green screw because IMO this will be a safer installation by creating a ground-fault path to the utility source. I do not see how it would create a hazard by installing the main bonding jumper. Doing so is consistent with separate service disconnect enclosures as required by 250.24(B)
 

hhsting

Senior Member
The 2020 NEC added new section 250.25 which requires separate supply-side connected system enclosures to comply with 250.24(A)-(D)... which will not only require a Main Bonding Jumper per 250.24(B), you also need a Grounding Electrode Conductor per 250.24(D) and it must be connected in compliance with 250.64(D)

Given that the 2014 NEC gives no direction for installer or AHJ on your supply-side connected PV Disconnect...I agree ask the AHJ...but I would suggest to install the green screw because IMO this will be a safer installation by creating a ground-fault path to the utility source. I do not see how it would create a hazard by installing the main bonding jumper. Doing so is consistent with separate service disconnect enclosures as required by 250.24(B)
Post #1 attachment has bonding jumper and neutral from main service disconnect. The main service disconnect has main bonding jumper.

Let’s take a look at EGC fault path post #1 it would from PV AC main disco EGC to EGC of main service disco to main bonding jumper in service disco to neutral. It has main bonding jumper connection only in service disco.

Now if you add main bonding jumper in PV AC disconnect let’s take a look at EGC fault current it would go from EGC PV AC disco to main bonding jumper PV AC disco to neutral in AC PV disco to neutral in main service disco then to utility. It still has to go thru neutral in main service disco.

Between the two scenarios how is one safe the other not safe?

Now l am not sure how it would work if their was lightning strike utility line or tree fell on utility line side line and fault current flow. If anyone here know what exactly happens each of the scenarios and how one is safer than other?



Sent from my iPhone using Tapatalk
 

shortcircuit2

Senior Member
Proper sizing of the Main Bonding Jumper per 250.28 is important for each separate service enclosure on services with multiple switches. 250.28 does not have exception for situation of a Main Bonding Jumper installed in another switch. Each separate switch needs its own based on the conductors feeding that switch.

In your diagram, I agree there is a ground-fault path.

But, lets say we have a 800-amp service with 2 sets of 500MCM copper service entrance conductors feeding a trough with 4 separate 200-amp switches with their main bonding jumper sized for the 200-amp conductors feeding the switch. Now a 600-amp solar PV switch is added in its own enclosure connected to the SE conductors in the trough. Would that 600-amp PV switch have an effective ground-fault path through the other 200-amp main bonding jumper? No it will not.

There is no exception to 250.28(2) with regard to separate enclosures.
 

hhsting

Senior Member
Proper sizing of the Main Bonding Jumper per 250.28 is important for each separate service enclosure on services with multiple switches. 250.28 does not have exception for situation of a Main Bonding Jumper installed in another switch. Each separate switch needs its own based on the conductors feeding that switch.

In your diagram, I agree there is a ground-fault path.

But, lets say we have a 800-amp service with 2 sets of 500MCM copper service entrance conductors feeding a trough with 4 separate 200-amp switches with their main bonding jumper sized for the 200-amp conductors feeding the switch. Now a 600-amp solar PV switch is added in its own enclosure connected to the SE conductors in the trough. Would that 600-amp PV switch have an effective ground-fault path through the other 200-amp main bonding jumper? No it will not.

There is no exception to 250.28(2) with regard to separate enclosures.
Ok in your example so let’s say your PV disconnect SE conductors are now connected not to trough but to one of the 4 disconnects line side lug inside the disconnect and supply side bonding jumper and neutral are brought to PV disconnect separate enclosure from the disconnect then you would have clear ground fault clear path thru EGC to the disconnect main bonding jumper. What would be the problem be with this scenario?
 
Last edited:

hhsting

Senior Member
Ok in your example so let’s say you have 3 - 200A separate PV disconnect enclosures (making same as 600A one PV disconnect enclosure) and SE conductors each one are now connected not to trough but to one of the 3 200A disconnects line side lug inside enclosure and supply side bonding jumper and neutral are brought to each PV disconnect separate enclosures from its respective disconnect then you would have clear ground fault clear path thru EGC to the respective disconnect main bonding jumper. What would be the problem be with this scenario?
 

hhsting

Senior Member
On post #1 let’s say main service disco is 400A and PV AC disconnect is 100A. Let’s assume simultaneous fault occurs on 400A side and 100A side then would the main bonding jumper in 400A main service disco be sufficient size to handle both fault same time? I think not since it’s sized for 400A main disco and so two options:

1. Either provide separate main bonding jumpers in each enclosure 400A and separate in 100A or

2. Re size the main bonding jumper in main service disconnect to handle both the 400A and 100A incoming service entrance conductor sizes.

Not sure. What are your thoughts about two options above and are both per code?


Sent from my iPhone using Tapatalk
 

shortcircuit2

Senior Member
Ok in your example so let’s say your PV disconnect SE conductors are now connected not to trough but to one of the 4 disconnects line side lug inside the disconnect and supply side bonding jumper and neutral are brought to PV disconnect separate enclosure from the disconnect then you would have clear ground fault clear path thru EGC to the disconnect main bonding jumper. What would be the problem be with this scenario?
Connecting 600-amp SE conductors to the line side of a 200-amp disconnect just wouldn't work...something would meltdown and burst into flames.
 

shortcircuit2

Senior Member
There are many rules in the NEC where you can have an installation that will work and not comply directly with the rule. But given that there is not an exception to the rule, the installation violates the rule. This Main Bonding Jumper rule in discussion here is a good example. The Code requires a MBJ in each separate disconnect per 250.28(2)

Your diagram in post #1 will require a MBJ in 2020. Installing a MBJ in your 2014 installation will not create a hazard, and will provide a properly sized MBJ thereby creating an effective ground-fault path back to the utility source which can provide high levels of fault current into the your installation under a fault condition.

Ask you AHJ what he requires and get back to us.

Thanks.
 

hhsting

Senior Member
Connecting 600-amp SE conductors to the line side of a 200-amp disconnect just wouldn't work...something would meltdown and burst into flames.
Ok but did you take a look at post #8 you have three separate 200A PV disco enclosures each connecting to 3 main service 200 disconnect line side terminal lug. What is wrong with this scenario?


Sent from my iPhone using Tapatalk
 

hhsting

Senior Member
On post #1 let’s say main service disco is 400A and PV AC disconnect is 100A. Let’s assume simultaneous fault occurs on 400A side and 100A side then would the main bonding jumper in 400A main service disco be sufficient size to handle both fault same time? I think not since it’s sized for 400A main disco and so two options:

1. Either provide separate main bonding jumpers in each enclosure 400A and separate in 100A or

2. Re size the main bonding jumper in main service disconnect to handle both the 400A and 100A incoming service entrance conductor sizes.

Not sure. What are your thoughts about two options above and are both per code?


Sent from my iPhone using Tapatalk
No takers. Nobody knows?


Sent from my iPhone using Tapatalk
 
Solar is not a service but a lot of people/ utilities treat us as one therefore having the MBJ or grounded neutral kit in PV disco.

There is no wrong or right answer in this case and there are arguments for both sides. I can think of plenty. Usually go with what the AHJ says.
 

hhsting

Senior Member
Unfortunately I am contractor to AHJ reviewer and AHJ does not have any rule about this.

Spoke with the engineer and he replaced and modified post #1 sketch with attached sketch this post. He provides main bonding jumper in main PV AC disconnect however supply side tap he is brining 4#3Awg + 1#6 AWG in 1 inch conduit from 125A main service disco to 80A PV AC disco. The grounding electrode system is still in main AC service disco.

Was wondering few questions and people who encountered how different AHJ handle or what are their opinion:

1. Is their anything wrong per code attached sketch?

2. Can you have EGC coming from supply side tap as shown sketch?If no then please which code section it does not comply and if yes then which code section it comply?

3.What are both arguements for both for and against NEC 2014 grounding electrode system and main bonding jumper at PV AC Disco?

4. Can you have grounding electrode system in main AC service disco and EGC going to PV AC disco as shown?
 

Attachments

Last edited:

shortcircuit2

Senior Member
1. The EGC may carry objectionable current in violation of 250.6(A) and it could be considered an installation of the grounded conductor in parallel in violation of 310.10(H) 2014/2017 NEC or 310.10(G) 2020 NEC

2. Same comment as 1.

3. Advocates against say that PV is not a "service"...
Advocates for say bonding creates a more effective low-impedance ground-fault path back to the utility source.

4. The GEC generally needs to land in each separate enclosure per 250.64(D)


2020 changed the definition of "service" to "The conductors and equipment connecting the serving utility to the wiring of the premises served."
 

hhsting

Senior Member
I am hard time finding in NEC 2014 what the attached pictures shows: AC PV disconnect is not counted as service disconnect. I looked in NEC 2014 Section 230.71(A) but does not say AC PV disconnect not counted. Does anyone know exactly what section it is in NEC 2014? If none then please say none.
 

Attachments

shortcircuit2

Senior Member
Well you were asking for input on the Main Bonding Jumper in this thread and now Grouping.

2014 section 230.71(A) has what your asking about in that it says "or for each set of service entrance conductors permitted by 230.40, exception No. 1,3,4 or 5" and 230.40 ex # 5 specifies systems connected per 230.82 (5) & (6)

It is amusing in that here these conductors are regarded as Service Entrance Conductors through these Code references, but refer to them as the Inverter Output Circuit conductors in 690... service entrance conductors should terminate in a...Service Disconnect.

I'm happy to see that Mike Holt guides his readers to install the Main Bonding Jumper and line side bond in the photo you have uploaded.
 

synchro

Senior Member
Am I right in assuming that the PV inverters drive little or no current through their neutral connection if they have one? If so then there does not seem to be a good technical argument that a bonding jumper in the PV disconnect creates a problem, because it will be conducting essentially no current in normal operation. But during a ground fault on the PV side, as shortcircuit2 mentioned, it provides a more direct and lower impedance path for the fault current.
 

hhsting

Senior Member
Am I right in assuming that the PV inverters drive little or no current through their neutral connection if they have one? If so then there does not seem to be a good technical argument that a bonding jumper in the PV disconnect creates a problem, because it will be conducting essentially no current in normal operation. But during a ground fault on the PV side, as shortcircuit2 mentioned, it provides a more direct and lower impedance path for the fault current.
Not sure about Enphase Microinverters IQ7+. Do they have neutral current flow?


Sent from my iPhone using Tapatalk
 

hhsting

Senior Member
Since they have no neutral conductor, I would say no. :D
Not the inverter I guess but They have something called Enphase Envoy 2 which needs to be powered from AC solar sub panels which require neutral. Not sure how much neutral current it draws.


Sent from my iPhone using Tapatalk
 

synchro

Senior Member
Not the inverter I guess but They have something called Enphase Envoy 2 which needs to be powered from AC solar sub panels which require neutral. Not sure how much neutral current it draws.
With only a quick look I didn't see info on an Envoy 2, but the power draw of the IQ Envoy is spec'd at 3 watts typical. Seems negligible to me in the context of this discussion.
 

GeorgeB

ElectroHydraulics engineer (retired)
What IQ7 and Envoy 2 bring to this discussion beyond earlier Enphase inverters and the standard Envoy is a little beyond me. But their standard inverters supply power at 240; the transformer CT (center tap) is used to verify line voltages, as it was explained to me. They put a signal on both 120V APPARENTLY referenced to the CT.

The Envoy "reads" the modulated, imposed, signal and makes it available on "my" network. I understand the Envoy 2 is not compatible with older inverters, and the older Envoy is not compatible with newer inverters, the IQ7 in particular. Thus if one of my inverters fails, it is questionable if it can be replaced with the newer design ... I've not asked if it will work but not communicate; will when the need arises.

I do know that with my home's meter-main and feeder to my panel, and with the inverters connected to a breaker in the meter-main, my installer had a difficult time finding a receptacle with a good signal; he could not use one near (perhaps 2 feet) but maybe 60 feet from the panel. One 3 feet from the panel was marginal, but I did have wired internet available there via a 10BaseT powerline bridge, and now via a mesh node with 1000BaseT ports.
 

jaggedben

Senior Member
Whether or not the inverters or any other equipment utilizes a neutral is an irrelevant detail that is a distraction from the original question. You need an effective ground-fault current path at the disconnect, and to do this you need an adequately sized wire bonding the disconnect enclosure to the service neutral. One way to do this is, as advocated above by shortcircuit2, is to call the the new PV disconnect a service disconnect and to then comply with 250.24(C) and other sections that require bringing a neutral to the disconnect and installing a main bonding jumper. Another way is to bring a supply-side bonding jumper sized to 250.102. It is not really valid to use table 250.122 for sizing since there is no OCPD ahead of the disconnect, hence I strongly suggest using 250.102, especially the minimum size. However, as long as the wire is adequately sized and ultimately connects to the service neutral, it really does not matter if it is white or green. The code ultimately only requires one wire, not two. (This is leaving aside grounding electrode conductor stuff, which depends on other details not essential to the main question).

In the past I've posted a list of code sections that can be used to argue 'for' and 'against' considering a supply-side disconnect to be a service disconnect. (But I don't have that list on the device I'm typing this on and I'm having trouble finding it with searches.) It is notable, as shortcircuit mentioned above, that the 2020 NEC changes the definition of a service. Thus the strongest argument that it is 'not a service disconnect' is no longer supported by the code. On the other hand, there was proposed language for the 2020 NEC that would have told us exactly how we could ground the disconnect in the new 705.11, but it got deleted from the final version. I don't have the inside scoop on how and why it got left out, but I do know there were negative comments filed against the first draft. I hypothesize that the CMP couldn't come to enough of an agreement on the matter, so they just punted the issue as they always have in the past. So really, not all that much has changed. It's still up to the AHJ.

hhisting as a contractor to the AHJ I think you are free to make your own recommendation as to whether to consider the disconnect a service disconnect, if they do not already have a clear policy in place.
 

synchro

Senior Member
Whether or not the inverters or any other equipment utilizes a neutral is an irrelevant detail that is a distraction from the original question. You need an effective ground-fault current path at the disconnect, and to do this you need an adequately sized wire bonding the disconnect enclosure to the service neutral. One way to do this is, as advocated above by shortcircuit2, is to call the the new PV disconnect a service disconnect and to then comply with 250.24(C) and other sections that require bringing a neutral to the disconnect and installing a main bonding jumper.
When I mentioned a neutral not being utilized in the PV disconnect it was to eliminate any concerns (and which in the OP's case would be distractions) about objectionable current that a bonding jumper in the PV disconnect could introduce. Such objectionable current from paralleled ECG and grounded conductors was cited as one of the possible code issues by shortcircuit2 in his reply to the OP's question #1 below. Obviously without any equipment connected to the grounded conductor it will be conducting no current and so there would also be no current in a parallel ECG. Therefore in the OP's case not having any utilization equipment connected to neutral means that a full sized neutral can be brought to the PV disconnect and a bonding jumper installed without any technical basis for concerns about objectionable current.

OP posted question:
Spoke with the engineer and he replaced and modified post #1 sketch with attached sketch this post. He provides main bonding jumper in main PV AC disconnect ......
Was wondering few questions and people who encountered how different AHJ handle or what are their opinion:
1. Is their anything wrong per code attached sketch?
shortcircuit2 responded:
1. The EGC may carry objectionable current in violation of 250.6(A) and it could be considered an installation of the grounded conductor in parallel in violation of 310.10(H) 2014/2017 NEC or 310.10(G) 2020 NEC
 

shortcircuit2

Senior Member
Of interest here and something I want to point out, posts 21-24 discuss how much power consumption the PV system uses. As with most systems that are Utility Interactive, there is some electric energy consumed. Although minimal, power is consumed. (Nighttime & Standby Mode for example)

IMO, and something that has long been my counter argument against the advocates argument that the PV system is not a Service as in post #14...the Serving Utility does deliver Electric Energy to the Interconnected Electric Power Production Source wiring system. This is essential for the operation of the associated equipment.

2014-2017 Service: The conductors and equipment for delivering electric energy from the serving utility to the wiring system of the premises served.
 

jaggedben

Senior Member
When I mentioned a neutral not being utilized in the PV disconnect it was to eliminate any concerns (and which in the OP's case would be distractions) about objectionable current that a bonding jumper in the PV disconnect could introduce. Such objectionable current from paralleled ECG and grounded conductors was cited as one of the possible code issues by shortcircuit2 in his reply to the OP's question #1 below. Obviously without any equipment connected to the grounded conductor it will be conducting no current and so there would also be no current in a parallel ECG. Therefore in the OP's case not having any utilization equipment connected to neutral means that a full sized neutral can be brought to the PV disconnect and a bonding jumper installed without any technical basis for concerns about objectionable current.

OP posted question:

shortcircuit2 responded:
Thanks for the clarification and I understand the point you were making in that context. But I would just expand my comment to say that the whole kaboodle - including shortcircuit's point about objectionable current - is ultimately still irrelevant to the original question. Nobody is arguing for a parallel neutral and EGC/SSBJ with neutral bonded at both ends. Or at least no one should be. (And again, let's leave aside grounding electrode system, which the code allows to create a parallel path for service disconnects.)

Once one decides whether the PV disconnect is a service disconnect or some other kind of disconnect not included in that category, the code becomes fairly clear what to do with it and how to make it both code compliant and safe. (I mean, there's still lot's of relevant sections to become familiar with, but their application becomes relatively clear.) The problem is that the code doesn't offer solid guidance on making that decision.
 
Top