Interlock kits and neutral ground bonding. When a main panel is sometimes a subpanel

Status
Not open for further replies.
J

jhalla

Member
Location
georgia
Occupation
solar installer
So I am looking to get some clarity around neutral ground bonding when a main panel becomes a subpanel. Specifically if there is a new "critical load panel" which is served by a solar hybrid inverter. Lets say this critical load subpanel has a breaker that feeds an interlocked breaker in the "main panel". That way the critical panel will have automatic backup on critical loads such as, fridge, freezer, lights etc.. however in an extended grid down situation the homeowner would have the ability to swap over to the interlocked breaker in the main panel, allowing for whole home backup (within the output limits of the inverter)

My question is is this setup possible if neutral and ground are already bonded in the "main panel" where the first means of disconnect is the main breaker ? Or for this setup to be possible must the neutral ground bond occur at a first means of disconnect that is further upstream?

I have attached a quick line diagram. I am guessing that option 1 would not be okay since the main panel becomes a subpanel so it is not allowed to have the neutral ground bond. But option 2 would be okay since the neutral ground bond does not take place in the main panel but further upstream? Any thoughts on this or similiar setups would be appreciated

.
https://imgur.com/380l24F
 
Sol-Ark? No biggie, just looks like it.

If so, they have some detailed diagrams in their monthly (or so) updated mess thing they call a manual ( joking, not joking).

Anyway -- It is a Solar "something." Reason I am starting with that is most everywhere we go, the UTILITY (not NEC, not local AHJ) also requires we have (a LOTO, exposed blade disconnect outside near the meter) . . . . a dedicated service rated disconnect between the GRID connection on the Inverter and the Utility. Comes from a long-past ponder that as a generator type source, they could go turn the backfeed from the site off if they wanted to work on the local grid. LOTO and all that. Reason THAT matters for your question -- IF that is done by a "line-side-tap" (above the First Means of Service Disconnect aka the Main Service Disconnect) the Solar PV disconnect can ALSO be considered a 2nd Main. Some AHJ see it that way, some do not.

Which brings up the question if the Neutral and Ground should be tied together there, as well? Since it is now also a MAIN. (standard answer -- set it up to do either, and let your local AHJ tell you what they want to see. Some even require a second ground rod and EGC tied to that Solar disconnect. We do not fight these things, but rather smile and wave and do what they say)

But for BOTH you #1 and #2, it appears your G-N tie is in the FIRST means of Disconnect? The Service Main as it were? IF SO -- they are both correct for the G-N tie -- which is your question, right?
 
Last edited:
btw -- this is also NOT your question -- but it also looks like your set-up could backfeed the LOAD connection on your inverter? From the sketch, I mean. That is generally a NO-GO. (This is generally an Inverter Manufacturer requirement -- Not NEC or Utility).
 
Phil Timmons said:
btw -- this is also NOT your question -- but it also looks like your set-up could backfeed the LOAD connection on your inverter? From the sketch, I mean. That is generally a NO-GO. (This is generally an Inverter Manufacturer requirement -- Not NEC or Utility).
Click to expand...
oh wait. I see you have a lock-out to the critical loads -- so they normally are fed from the inverter. (?)
 
On the OP, when using breaker interlock kits only the ungrounded conductors are switched. The grounded conductor and EGC connectivity remains the same in either state. So when both sources are on the same service, you need to come up with a grounded conductor routing that works for both states, and doesn't introduce any loops or multiple N-G bonds, and also use an ungrounded conductor routing that complies with 300.3(B) in both states.

With a supply side PV interconnect, it seems to mean that's not possible to do. You'd need a transfer switch that will switch the neutral. If you had a load side PV connection, then it's possible to do with breaker interlocks, as then you can route your alternate source ungrounded supply conductors back along the route of the grounded conductor to the panel with the interlocked breaker.

Cheers, Wayne
 
@jhalla
I do not think it is practically possible to do what you're proposing in a safe and code compliant manner. The showstopper is that a backfed breaker with an interlock kit does not switch the neutral, so therefore you are going to either have objectionable current on exposed non-current-carrying parts, or you're going to have unbalanced neutral current in a different cable or conduit, or you're going to have service and non-service conductors in the same raceway.

Your diagrams are incomplete because you appear to be showing a supply side connection for the inverter and backup loads without showing an additional neutral-ground bond. Once you put this in, and once you run a neutral along with your supply to the back-fed interlocked breaker, you have a neutral loop that will return partial current the other way. If you don't run a neutral with that supply, then when the backfed breaker is closed in an outage, neutral current won't be running in the same raceway as the ungrounded conductor current. One way or another you are violating 300.3 and/or 300.20 and creating inductive heating of any metal raceways.

As Wayne said, if you make your inverter connection load side instead of supply side, you have slightly better prospects but it still has to be done in a very particular way (backfed connection in same conduit that feeds inverter) which will make it more difficult to install. It's also still prone to user error which could be catastrophic for the inverter.

More legit ways to do this might be:
- put the 'main' panel on a manual transfer switch that switches the neutral.
- Use some kind of smart controls or a smart panel (e.g. SPAN) instead to control what's powered and not in an outage.
- give up on having backup power to anything but the critical loads panel.

Honestly my initial reaction to the backfed interlocked breaker was 'yuck', and in thinking it through in more detail that conclusion didn't change.

For general context, usually the devices that isolate the microgrid do not switch the neutral. As long as this is the case, the main bonding jumper in the service panel is always the place where neutral is bonded to ground, and the only place. Do not do anything to change this. Ground fault current always goes through the MBJ, whether it comes from the utility or the other sources.
 
jaggedben said:
Honestly my initial reaction to the backfed interlocked breaker was 'yuck', and in thinking it through in more detail that conclusion didn't change.
Click to expand...
FWIW, I actually did this at my house with a Tesla PW and Gateway. It was relatively easy as the Gateway, backed up MB panel, non-backed up MB panel, and generation panel are all next to each other. So all that was required was a breaker on the backed up side, the interlocked breaker on the non-backed up panel, and routing the ungrounded conductors between the two breakers along the same route the grounded conductor took.

Cheers, Wayne
 
Phil Timmons said:
Sol-Ark? No biggie, just looks like it.

If so, they have some detailed diagrams in their monthly (or so) updated mess thing they call a manual ( joking, not joking).

Anyway -- It is a Solar "something." Reason I am starting with that is most everywhere we go, the UTILITY (not NEC, not local AHJ) also requires we have (a LOTO, exposed blade disconnect outside near the meter) . . . . a dedicated service rated disconnect between the GRID connection on the Inverter and the Utility. Comes from a long-past ponder that as a generator type source, they could go turn the backfeed from the site off if they wanted to work on the local grid. LOTO and all that. Reason THAT matters for your question -- IF that is done by a "line-side-tap" (above the First Means of Service Disconnect aka the Main Service Disconnect) the Solar PV disconnect can ALSO be considered a 2nd Main. Some AHJ see it that way, some do not.

Which brings up the question if the Neutral and Ground should be tied together there, as well? Since it is now also a MAIN. (standard answer -- set it up to do either, and let your local AHJ tell you what they want to see. Some even require a second ground rod and EGC tied to that Solar disconnect. We do not fight these things, but rather smile and wave and do what they say)

But for BOTH you #1 and #2, it appears your G-N tie is in the FIRST means of Disconnect? The Service Main as it were? IF SO -- they are both correct for the G-N tie -- which is your question, right?
Click to expand...
Yes solark, I did not draw it into the line diagram but yes there would be a fused disconnect switch in between the tap connectors and the inverter. We typically bond g-n at the fused disconnect. No additional ground rods though.

My question pertains to the interlock in the 200a panel. In line diagram 1, n-g is bonded in the main panel. However due to the interlocked breaker this main panel becomes a subpanel when it is running off of the inverter and not off the grid. Once it becomes a subpanel does the existing n-g become an issue?
 
wwhitney said:
FWIW, I actually did this at my house with a Tesla PW and Gateway. It was relatively easy as the Gateway, backed up MB panel, non-backed up MB panel, and generation panel are all next to each other. So all that was required was a breaker on the backed up side, the interlocked breaker on the non-backed up panel, and routing the ungrounded conductors between the two breakers along the same route the grounded conductor took.

Cheers, Wayne
Click to expand...
jaggedben said:
@jhalla
I do not think it is practically possible to do what you're proposing in a safe and code compliant manner. The showstopper is that a backfed breaker with an interlock kit does not switch the neutral, so therefore you are going to either have objectionable current on exposed non-current-carrying parts, or you're going to have unbalanced neutral current in a different cable or conduit, or you're going to have service and non-service conductors in the same raceway.

Your diagrams are incomplete because you appear to be showing a supply side connection for the inverter and backup loads without showing an additional neutral-ground bond. Once you put this in, and once you run a neutral along with your supply to the back-fed interlocked breaker, you have a neutral loop that will return partial current the other way. If you don't run a neutral with that supply, then when the backfed breaker is closed in an outage, neutral current won't be running in the same raceway as the ungrounded conductor current. One way or another you are violating 300.3 and/or 300.20 and creating inductive heating of any metal raceways.

As Wayne said, if you make your inverter connection load side instead of supply side, you have slightly better prospects but it still has to be done in a very particular way (backfed connection in same conduit that feeds inverter) which will make it more difficult to install. It's also still prone to user error which could be catastrophic for the inverter.

More legit ways to do this might be:
- put the 'main' panel on a manual transfer switch that switches the neutral.
- Use some kind of smart controls or a smart panel (e.g. SPAN) instead to control what's powered and not in an outage.
- give up on having backup power to anything but the critical loads panel.

Honestly my initial reaction to the backfed interlocked breaker was 'yuck', and in thinking it through in more detail that conclusion didn't change.

For general context, usually the devices that isolate the microgrid do not switch the neutral. As long as this is the case, the main bonding jumper in the service panel is always the place where neutral is bonded to ground, and the only place. Do not do anything to change this. Ground fault current always goes through the MBJ, whether it comes from the utility or the other sources.
Click to expand...
I neglected to show this in the line diagram. But yes this would be a supply side connection with pierce tap connectors and a fused disconnect in between the inverter and taps. N-G bond would take place in this disconnect.

On a 200a panel a load side connection would not be possible for this inverter (solark 12k) without first derating the main breaker. Assuming the main breaker was derated you could then tie in the inverter with a 60a backfed breaker however this would present a new issue. As the interlocked breaker and the backfed breaker cannot be on at the same time as this could damage the grid relay in the inverter.

Both the smart panel or the 3 pole 3 position transfer switch may be a better way to get comparable functionality but unfortunately neither of those options are cheap and will add significant cost to the project
 
wwhitney said:
FWIW, I actually did this at my house with a Tesla PW and Gateway. It was relatively easy as the Gateway, backed up MB panel, non-backed up MB panel, and generation panel are all next to each other. So all that was required was a breaker on the backed up side, the interlocked breaker on the non-backed up panel, and routing the ungrounded conductors between the two breakers along the same route the grounded conductor took.

Cheers, Wayne
Click to expand...
Wayne, for this install did the grid output of the gateway connect with a back fed beaker in the main panel or with a supply side connection?
 
jhalla said:
Wayne, for this install did the grid output of the gateway connect with a back fed beaker in the main panel or with a supply side connection?
Click to expand...
I assume you mean the grid side connection to the Gateway. In which case the answer is that the single service OCPD supplies a feeder to the Gateway and adjacent main breaker non-backed up panel.

Cheers, Wayne
 
jhalla said:
Yes solark, I did not draw it into the line diagram but yes there would be a fused disconnect switch in between the tap connectors and the inverter. We typically bond g-n at the fused disconnect. No additional ground rods though.

My question pertains to the interlock in the 200a panel. In line diagram 1, n-g is bonded in the main panel. However due to the interlocked breaker this main panel becomes a subpanel when it is running off of the inverter and not off the grid. Once it becomes a subpanel does the existing n-g become an issue?
Click to expand...
With a "Sharp" Inspector -- yes (generally) they want us to take the N-G connection APART in the prior Main-Service-Panel . . . which as you noted is now a Sub-Panel.

That is maybe 5 out 10 AHJ. Another 4 out of 10 inspectors would nod in agreement if we told them what we did. And the last Inspector -- maybe 1 out of 10 . . . he is really the Plumbing Inspector, but he got sent out to look at the job because the Electrical Guy is busy. He will ask you if you think he should get Solar for his house. ;P
 
Phil Timmons said:
With a "Sharp" Inspector -- yes (generally) they want us to take the N-G connection APART in the prior Main-Service-Panel . . . which as you noted is now a Sub-Panel.

That is maybe 5 out 10 AHJ. Another 4 out of 10 inspectors would nod in agreement if we told them what we did. And the last Inspector -- maybe 1 out of 10 . . . he is really the Plumbing Inspector, but he got sent out to look at the job because the Electrical Guy is busy. He will ask you if you think he should get Solar for his house. ;P
Click to expand...
Not many sharp inspectors around here, they have the few things they look for but doubt most would have any thoughts on this so Im sure it would pass 9/10. The post was more or so to have the line diagram picked apart by others that are more knowledgeable (than most inspectors or myself) about any potential problems with either line diagram.
 
One way to avoid the undesired current flows mentioned above is to insert a 240V to 240/120V isolation transformer between a breaker in the critical loads panel and the interlocked back-fed breaker of the main panel. The neutral of the transformer secondary should not be bonded to equipment ground in order to make it a non-separately derived system. This would be similar to having a portable generator that's non-bonded when it's back-feeding an interlocked breaker.

A 5kVA transformer can be had for around $600, and it goes up from there depending on how much load you want to support in the main panel.
https://www.zoro.com/acme-transformer-1ph-50kva-240x480-120240-t2530144s/i/G3716470/#description

Of course the breaker feeding the transformer could be shut off when it's not necessary to back-feed the main panel.
 
synchro said:
One way to avoid the undesired current flows mentioned above is to insert a 240V to 240/120V isolation transformer between a breaker in the critical loads panel and the interlocked back-fed breaker of the main panel. The neutral of the transformer secondary should not be bonded to equipment ground in order to make it a non-separately derived system.
Click to expand...
I don't believe the transformer secondary neutral can be left floating; per 250.20(B)(1) it needs to be grounded. If that grounding is not done via a connection to a GEC to make an SDS, then it would have to be done via the primary neutral. In which case I don't see how the above has any benefit as far as avoiding neutral loops.

Cheers, Wayne
 
wwhitney said:
I don't believe the transformer secondary neutral can be left floating; per 250.20(B)(1) it needs to be grounded. If that grounding is not done via a connection to a GEC to make an SDS, then it would have to be done via the primary neutral. In which case I don't see how the above has any benefit as far as avoiding neutral loops.

Cheers, Wayne
Click to expand...

It appears my wording did not convey my intent. I meant that the neutral from the 120/240V transformer secondary would not be bonded to the EGC/enclosure at the transformer itself. But a neutral conductor would still be run along with the two phase conductors from the transformer outputs to the neutral bus in the main panel. There it would be grounded by the main bonding jumper that's located there and also connected to the GES. This neutral conductor would be necessary to supply the L-N loads in the main panel. The transformer secondary winding's center tap would only be connected to this one neutral conductor that goes to the main panel's neutral bus, and therefore no neutral loops should be formed.

I guess I was relying on my analogy to the transformer of a portable generator where any bond from neutral to the generator case is removed because of the parallel current paths that it would create.
 
Last edited:
synchro said:
It appears my wording did not convey my intent. I meant that the neutral from the 120/240V transformer secondary would not be bonded to the EGC/enclosure at the transformer itself. But a neutral conductor would still be run along with the two phase conductors from the transformer outputs to the neutral bus in the main panel. There it would be grounded by the main bonding jumper that's located there and also connected to the GES.
Click to expand...
OK, thanks, I see how that has the potential to work, the neutral conductors end up in almost a loop, but the transformer serves as a break in the loop.

So the interlocked backfed breaker is the necessary secondary protection given the 3-wire secondary / 2-wire primary, and if that in the main panel, the GEC is available for grounding the secondary neutral. You set the transformer near the main panel.

If the GEC, primary neutral, and secondary neutral all land on the same neutral bar, then it's a bit unclear whether to consider that an SDS. On the one hand the secondary neutral is grounded to the GEC. On the other hand it's directly connected to the primary neutral. It's basically the limiting case of an SDS where the length of the conducive path between primary neutral and secondary neutral through the GES has gone down to zero.

Cheers, Wayne
 
wwhitney said:
It's basically the limiting case of an SDS where the length of the conducive path between primary neutral and secondary neutral through the GES has gone down to zero.

Cheers, Wayne
Click to expand...

Now you can tell that's a real mathematician speaking there. :)
I have a nephew with a PhD in mathematics from Boston University.
 
Phil Timmons said:
With a "Sharp" Inspector -- yes (generally) they want us to take the N-G connection APART in the prior Main-Service-Panel . . . which as you noted is now a Sub-Panel.

That is maybe 5 out 10 AHJ. Another 4 out of 10 inspectors would nod in agreement if we told them what we did. And the last Inspector -- maybe 1 out of 10 . . . he is really the Plumbing Inspector, but he got sent out to look at the job because the Electrical Guy is busy. He will ask you if you think he should get Solar for his house. ;P
Click to expand...
That last paragraph is 100% true…🤣
 
Status
Not open for further replies.
Top