Sol-Ark Grid connection options

[wwhitney]

You could add a 200A SUSE enclosed circuit breaker in between the meter and your existing main panel. That circuit breaker becomes the service disconnect and overcurrent device and the only place with a neutral-ground bond. Ideally the load side has double lugs.

Then you run a 200A feeder to your existing main panel from one set of the lugs on the service disconnect. From the other set of lugs you run a 100A feeder tap to your 100A fused PV disconnect. You have a feeder interconnection instead of a supply side interconnection.

Now the 100A feeder for the bypass can run from the main panel to the manual transfer switch through the new service disconnect, the PV disconnect, and the Sol-Ark inverter. Ideally each of those cabinets would be within 24" of conduit run from its neighbors, so that no derating is required.

Cheers, Wayne

 

[PWDickerson]

The split should be after the fused disco for the line-side tap. See hastily sketched attachment. I have drawn a second disconnect (unfused) so you can still power the backup panel from the fused disco and have the inverter de-energized for service.

In that way you can carefully arrange the neutral path so that it follows both the normal path or the bypass path, and there is still only one path for neutral current. Note that if the system is in bypass, the neutral current flowing into and out of the inverter disco and the inverter will cancel and so not create any inductive heating in metal parts. To simplify, I have only shown one line conductor and the neutral conductor.

 

[wwhitney]

On that last diagram, the new "fused disconnect" could instead be a SUSE 100A enclosed circuit breaker. Not sure if that would have any space or cost advantage.

Also, if the POCO doesn't require a PV system disconnect in its own enclosure, the "fused disconnect" and the "inverter disconnect" could be combined into a single enclosure. E.g. a SUSE 200A 4-space panel with a 100A main breaker, a 100A PV disconnect breaker, and a 100A bypass breaker. Physics-wise you only need a 100A bus, but 705.12(B) doesn't have any allowances that recognizes that for busbars with only 3 connections.

Cheers, Wayne

 

[solarken]

On that last diagram, the new "fused disconnect" could instead be a SUSE 100A enclosed circuit breaker. Not sure if that would have any space or cost advantage.

Also, if the POCO doesn't require a PV system disconnect in its own enclosure, the "fused disconnect" and the "inverter disconnect" could be combined into a single enclosure. E.g. a SUSE 200A 4-space panel with a 100A main breaker, a 100A PV disconnect breaker, and a 100A bypass breaker. Physics-wise you only need a 100A bus, but 705.12(B) doesn't have any allowances that recognizes that for busbars with only 3 connections.

Cheers, Wayne

Thanks for the suggestion. Most of the enclosed breakers I am finding locally do not have barrier on the line terminals. Bought one that was 50% off on clearance, but may have to return it. Our POCO wants the outside PV disconnect to be lockable.

 

[solarken]

The split should be after the fused disco for the line-side tap. See hastily sketched attachment. I have drawn a second disconnect (unfused) so you can still power the backup panel from the fused disco and have the inverter de-energized for service.

In that way you can carefully arrange the neutral path so that it follows both the normal path or the bypass path, and there is still only one path for neutral current. Note that if the system is in bypass, the neutral current flowing into and out of the inverter disco and the inverter will cancel and so not create any inductive heating in metal parts. To simplify, I have only shown one line conductor and the neutral conductor.

Thank you for detailed explanation and sketch!

 

[solarken]

Thank you all for your input. This site has a lot of space and equipment location limitations that complicate things. I took all your suggestions and went thru a few iterations, and arrived at this solution, using the main load center as the wiring box that gets the switched grid connection back together with a bypass feeder, so they can be run in same conduit to a wiring trough near the inverter and critical loads subpanel.

The main load center is on the basement west wall, fairly centered in the room, and there is sump pump and well pump plumbing and equipment on either side, leaving no room for more equip. The PV Disconnect will just be outside, so I just need room for the 2 conduit paths next to the loadcenter. The inverter, battery, and subpanel will be on the north wall, with the battery in the NW corner, then a space because of a 32-inch basement window, and then the inverter with trough below, and then the subpanel. I may use a Leviton subpanel that I have onhand, and they make a mechanical interlock kit that will eliminate the need for the transfer switch.

I welcome feedback. A few questions remain. Can I run the GEC from the Outside PV Disc in the conduit with other conductors back to the load center and bond usinf split-bolt to the GEC from the loadcenter as shown, or should I keep the GEC (and make it #6?) outside and go directly from the switch to the existing GEC going to the ground rod using a split-bolt? And does the GEC run as shown eliminate the need for an EGC between the PV Disc and the loadcenter? I am also considering eliminating the wire-type EGC's and using EMT as the EGC, any thoughts?

 

[wwhitney]

A couple comments on the latest diagram:

- You have two service disconnects (one in the main panel, one at the PV disconnect) that aren't grouped, which would violate 230.72. Although if you are under the 2017 NEC, maybe there's enough ambiguity as to whether the PV disconnect is a service disconnect that you could finesse this issue?

But a 2nd outdoor disconnect as the sole service disconnect would solve a bunch of issues for you including this one. You'd only have one N-G bond, and you could run your neutral as a tree.

- You still have a neutral routing issue. If the critical panel is set to "grid" rather than "inverter," the ungrounded conductors are powered through the main panel's 100A bypass breaker and through the service conductor powering that panel. But the grounded conductor loops through the main panel (without connection, as I understand it) and then to the PV disconnect and its service conductors. That's a different path. Connecting the grounded conductor to the main panel busbar doesn't help, as then all grounded conductor current gets splits between the two paths, direct to the meter and via the PV system disconnect.

Cheers, Wayne

 

[solarken]

A couple comments on the latest diagram:

- You have two service disconnects (one in the main panel, one at the PV disconnect) that aren't grouped, which would violate 230.72. Although if you are under the 2017 NEC, maybe there's enough ambiguity as to whether the PV disconnect is a service disconnect that you could finesse this issue?

But a 2nd outdoor disconnect as the sole service disconnect would solve a bunch of issues for you including this one. You'd only have one N-G bond, and you could run your neutral as a tree.

- You still have a neutral routing issue. If the critical panel is set to "grid" rather than "inverter," the ungrounded conductors are powered through the main panel's 100A bypass breaker and through the service conductor powering that panel. But the grounded conductor loops through the main panel (without connection, as I understand it) and then to the PV disconnect and its service conductors. That's a different path. Connecting the grounded conductor to the main panel busbar doesn't help, as then all grounded conductor current gets splits between the two paths, direct to the meter and via the PV system disconnect.

Cheers, Wayne

Ohio is on 2023 NEC. Solar PV is not a service, so I am not seeing where 230.72 applies. I will need to look closer at the neutral issue, I thought this iteration was not any different than the one with the transfer switch.

 

[wwhitney]

Ohio is on 2023 NEC. Solar PV is not a service, so I am not seeing where 230.72 applies.

Sure a 705.11 interconnection requires a service disconnect. 705.11(D) is titled "Service Disconnecting Means" and says the 705.11 interconnection requires a disconnecting means in accordance with Parts VI (Disconnecting Means) and VIII (Overcurrent Protection) of Article 230 (Services).

But I guess you have an out by calling the service conductor spur to the PV disconnect a separate set of service entrance conductors, as permitted by 705.11(A)(3) and 230.40 Exception 5. And 230.71 says that you can have two to six disconnects per service permitted by 230.2 or "for each set of service-entrance conductors permitted by 230.40, Exception No. 1, 3, 4, or 5." And then 230.72 says "The two to six disconnects, if permitted in 230.71, shall be grouped." So presumably that means the two to six disconnects permitted for the service shall be grouped with each other, and the two to six disconnects permitted for the 230.40 Ex 5 service entrance conductors shall be grouped with each other, without requiring the two groups to be in the same location.

At least if I'm reading this part of Article 230 correctly.

But would you have room for an outdoor 200A service disconnect next to the 100A PV disconnect if you desire? That would fix your neutral issues. From the 200A service disconnect, you'd run a 3-wire 200A feeder to the 100A PV disconnect, where you can make a tap to the 100A PV disconnect line side. Then from there to inside you run 5-wires, a single neutral conductor shared between the 200A feeder and the 100A PV feeder. As long as your interior 200A panel has a main breaker (which you were planning for anyway), that complies with the feeder interconnection rules. Only other change is your GECs have to go to the new exterior service disconnect instead of the interior 200A panel.

Cheers, Wayne

 

[solarken]

- You still have a neutral routing issue. If the critical panel is set to "grid" rather than "inverter," the ungrounded conductors are powered through the main panel's 100A bypass breaker and through the service conductor powering that panel. But the grounded conductor loops through the main panel (without connection, as I understand it) and then to the PV disconnect and its service conductors. That's a different path. Connecting the grounded conductor to the main panel busbar doesn't help, as then all grounded conductor current gets splits between the two paths, direct to the meter and via the PV system disconnect.

It seems to me the neutral current path will always be different that the ungrounded conductor current path. I thought the main issue was ensuring there was a neutral conductor in every conduit run. For example when the Bypass interlock in the critical loads panel is put in the bypass position, like your example, and the main panel bypass breaker is ON, the ungrounded conductor current path is Meter to Main panel, out Bypass breaker, to Critical bypass breaker, to load. The neutral return current path is load to critical neutral bar, up to inverter neutral, back down into trough, to Main panel, and then lopped out to PV disconnect and back as you stated. If the conduits to the PV disconnect are PVC, is there an issue? There is no way to prevent the neutral current from looping thru the inverter even though the ungrounded conductor current does not, right?

 

[wwhitney]

It seems to me the neutral current path will always be different that the ungrounded conductor current path.

No, the point of 300.3(B) is to ensure that within each metal-sheathed cable or metal raceway at every point the total current through that point is 0, i.e. the current one way is canceled by current returning the other way. If the neutral is routed differently, and is carrying any current, that is no longer true.

If the conduits to the PV disconnect are PVC, is there an issue?

If the PV disconnect enclosure is ferrous, then you'd still have a 300.20(A) issue.

There is no way to prevent the neutral current from looping thru the inverter even though the ungrounded conductor current does not, right?

PWDickerson's solution was to provide the "bypass" grid connection for the non-critical loads panel via the 100A PV service disconnect, rather than via the 200A main panel. So you'd run your 100A 3-wire feeder from the PV service disconnect to the inverter enclosure. Which hopefully has dual line side lugs, or you could use 3 port Polaris connectors. Then you'd run a single 5-wire conduit from the inverter to your mechanically-interlocked critical loads panel, two ungrounded wires from the inverter load side, two ungrounded wires from the inverter line side, and a single common grounded conductor. With the two sets of ungrounded wires landing on the two interlocked breakers.

I guess that configuration does raise the question as to whether your PV service disconnect is in fact supplied by a 230.40 Exception 5 set of service entrance conductors, as it is supplying loads in addition to the 230.82 Exception 6 equipment. Which could then raise the 230.72 grouping issue.

Cheers, Wayne

 

[solarken]

Sure a 705.11 interconnection requires a service disconnect. 705.11(D) is titled "Service Disconnecting Means" and says the 705.11 interconnection requires a disconnecting means in accordance with Parts VI (Disconnecting Means) and VIII (Overcurrent Protection) of Article 230 (Services).

But I guess you have an out by calling the service conductor spur to the PV disconnect a separate set of service entrance conductors, as permitted by 705.11(A)(3) and 230.40 Exception 5. And 230.71 says that you can have two to six disconnects per service permitted by 230.2 or "for each set of service-entrance conductors permitted by 230.40, Exception No. 1, 3, 4, or 5." And then 230.72 says "The two to six disconnects, if permitted in 230.71, shall be grouped." So presumably that means the two to six disconnects permitted for the service shall be grouped with each other, and the two to six disconnects permitted for the 230.40 Ex 5 service entrance conductors shall be grouped with each other, without requiring the two groups to be in the same location.

At least if I'm reading this part of Article 230 correctly.

But would you have room for an outdoor 200A service disconnect next to the 100A PV disconnect if you desire? That would fix your neutral issues. From the 200A service disconnect, you'd run a 3-wire 200A feeder to the 100A PV disconnect, where you can make a tap to the 100A PV disconnect line side. Then from there to inside you run 5-wires, a single neutral conductor shared between the 200A feeder and the 100A PV feeder. As long as your interior 200A panel has a main breaker (which you were planning for anyway), that complies with the feeder interconnection rules. Only other change is your GECs have to go to the new exterior service disconnect instead of the interior 200A panel.

Cheers, Wayne

I acknowledge that 705 does not use terminology that clearly states intentions, but I was always taught through my NABEP training and through other means is that Solar PV generation systems are not a service, even though many aspects of a supply side connection are treated as such. Art 100 definition of a service explicitly states "The conductors and equipment connecting the servicing utility to the wiring system of the premises served."
As you referenced 705.11 discusses source connections to a service, and points to 230.82(6) which permits Solar PV systems to be connected to the supply side of a service disconnecting means, if provided with a disconnecting means listed as SUSE, and OCPD. I interpret this as solar is a source, not a service but you connect it to the supply side using a service disconnect.
So to me, 230.72 does not apply to such sources, only to services. It is different because with solar and grid, they are two sources. An example of a service with required grouping per 230.72 in my opinion would be like a 200-Amp service that feeds a meter-main with 6 breakers in it, each breaker feeding a different load in the building. You can't add a 7th breaker or the max 6 would be exceeded.
I would have room for an outside new service disconnect, but I have not been able to find enclosed breakers with barriers on the line side, and a 200A service rated disconnect with fuses is pretty expensive, and also, I would have to separate all the damn grounds and neutrals in the main loadcenter because the previous electrician mixed them all across the ground and neutral bars. Also, the SER cable between the main loadcenter and the outside meter does not have an EGC. So if I can avoid it, it would be nice.

 

[wwhitney]

I acknowledge that 705 does not use terminology that clearly states intentions, but I was always taught through my NABEP training and through other means is that Solar PV generation systems are not a service, even though many aspects of a supply side connection are treated as such.

The NEC's stance on this question has changed between the 2017 NEC, when it was arguable either way, to the 2023 NEC, where it is clearly a service. So depending on when your NABEP training was, it may be out of date.

is that Solar PV generation systems are not a service, even though many aspects of a supply side connection are treated as such. Art 100 definition of a service explicitly states "The conductors and equipment connecting the servicing utility to the wiring system of the premises served."

Right, and so with a line side interconnection, the conductors between the line side of the PV system disconnect and the existing service conductors are themselves service conductors, as they connect the utility to the premises wiring. They need a disconnect from the wiring system of the premises served. That disconnect is the PV system disconnect, which makes it a service disconnect.

Cheers, Wayne

 

[solarken]

The NEC's stance on this question has changed between the 2017 NEC, when it was arguable either way, to the 2023 NEC, where it is clearly a service. So depending on when your NABEP training was, it may be out of date.


Right, and so with a line side interconnection, the conductors between the line side of the PV system disconnect and the existing service conductors are themselves service conductors, as they connect the utility to the premises wiring. They need a disconnect from the wiring system of the premises served. That disconnect is the PV system disconnect, which makes it a service disconnect.

Where does the 2023 NEC state clearly that solar PV system is a service?

 

[solarken]

If the PV disconnect enclosure is ferrous, then you'd still have a 300.20(A) issue.

So if I changed the supply side conductors and the load side conductors to SEU and SER cables, respectively and brought them into the PV disconnect in a single service entrance cable clamp connector? I will have to look at the diameter of the combined cables.

 

[wwhitney]

So if I changed the supply side conductors and the load side conductors to SEU and SER cables, respectively and brought them into the PV disconnect in a single service entrance cable clamp connector? I will have to look at the diameter of the combined cables.

Can you really get a cable clamp that would be listed for two cables of that size? Of course you could get non-ferrous cable clamps and cut a slot between the holes with just one cable per clamp.

Seems like PWDickerson's solution is simpler.

Cheers, Wayne

 

[wwhitney]

Where does the 2023 NEC state clearly that solar PV system is a service?

705.12(B)(2) mentions it in passing:

"The service conductors connected to the power production source service disconnecting means shall be sized in accordance with 705.28 and not be smaller than 6 AWG copper or 4 AWG aluminum or copper-clad aluminum."

It tells you that the power production source disconnect is a service disconnecting means, and that it is connected on the line side to service conductors. Sounds like a service to me.

Cheers, Wayne

 

[jaggedben]

705.12(B)(2) mentions it in passing:

I think you mean 705.11.

Where does the 2023 NEC state clearly that solar PV system is a service?

A PV system is not a service. But a disconnect for a supply side connection is arguably a service disconnect. The code has been revised over the last couple cycles to require almost all the same things of a supply-side PV disconnect as would be required of a service disconnect, while stopping short of declaring them to be the same. (NABCEP training from 5 years ago is likely to be out of date on this.) The application of grouping requirements is still either ambiguous or very between-the-lines; I would treat the PV disconnect as a service disconnect unless I knew the AHJ would allow otherwise.