PV Outside Disconnect Wiring Detail

[solarken]

Is there any code issue with running a single conductor for Neutral out to the PV system disconnect outside as shown in yellow highlight, instead of running the in and out neutral conductors to the disconnect? The Jbox is directly next to the 100A svc disconnect to it's left, via a chase nipple and the conduit to the outside disconnect is longer with bends and this would be easier to install and would save a bit of wire.

 

[wwhitney]

Why do you need a neutral in the "PV System Disconnect" at all? You don't show any load side (away from the utility) neutral at all, so providing a line side (utility side) neutral seems pointless.

I can't see the whole drawing--are you complying with 300.3(B) and 310.10(G) everywhere? Meaning no loops in the neutral conductor, and if a single neutral is shared between two separate sets of ungrounded conductors, all of them are in the same cable/conduit. I think everything in excerpt looks OK in that regard, but as it's a bit busy and incomplete, not sure.

Cheers, Wayne

 

[Elect117]

300.3(B) would basically mean no. It would be unique to meet those requirements.

 

[solarken]

Why do you need a neutral in the "PV System Disconnect" at all? You don't show any load side (away from the utility) neutral at all, so providing a line side (utility side) neutral seems pointless.

I can't see the whole drawing--are you complying with 300.3(B) and 310.10(G) everywhere? Meaning no loops in the neutral conductor, and if a single neutral is shared between two separate sets of ungrounded conductors, all of them are in the same cable/conduit. I think everything in excerpt looks OK in that regard, but as it's a bit busy and incomplete, not sure.

Cheers, Wayne

Thanks for responding. The load side ends in a critical loads panel, so the neutral is used there and will carry current. However the PV System disconnect has no use for the neutral. Even so, it is my understanding that the "where used" wording in 300.3(B) means it must be run to the disconnect even though it is not used at the disconnect, because it exists and is used elsewhere. I agree It would be much easier not to run it to the PV System disconnect.
More context: The load side of PV Disconnect feeds the grid input of inverter, the Critical loads panel main breaker is fed by the Load output of the inverter. The Critical Loads panel branch breaker is for bypassing the Inverter.
And yes, there are no neutral loops, the sum of currents in phase and neutral in all enclosure penetrations adds to zero, etc. The neutral conductor running to the PV disconnect obviously carries no current, but is there for 300.3(B).

 

[solarken]

300.3(B) would basically mean no. It would be unique to meet those requirements.

I am not sure I understand what you mean here. Are you saying it does not meet 300.3(B) as shown?

 

[wwhitney]

Thanks for responding. The load side ends in a critical loads panel, so the neutral is used there and will carry current. However the PV System disconnect has no use for the neutral. Even so, it is my understanding that the "where used" wording in 300.3(B) means it must be run to the disconnect even though it is not used at the disconnect, because it exists and is used elsewhere.

No. The neutral is not used in the PV System Disconnect, there is no 300.3(B) requirement to run one there at all. Just delete the yellow highlighted neutral.

Thumbs up on the rest of your post about neutral loops, etc.

Cheers, Wayne

 

[Elect117]

I am not sure I understand what you mean here. Are you saying it does not meet 300.3(B) as shown?

Where run, the neutral is one of the circuit conductors and therefore should be contained within the same conduit, gutters, etc.

I understood your questions to be running the conductor in a separate raceway or outside of the conduit and outside disconnect box. You have two EGCs that appear to need to be 1.

It is not the yellow highlighted grey neutral that I am concerned with but the grey and green conductors that do not enter that outside disconnect but are a apart of that circuit.

The grey and green should follow the phase conductors to the outside disconnect and back in order to meet 300.3(B).

 

[solarken]

Where run, the neutral is one of the circuit conductors and therefore should be contained within the same conduit, gutters, etc.

That seems to be the key point. The neutral is not used in the PV disconnect, i.e. running it in the conduit, out to the switch and back, with no termination, serves no purpose functionally or electrically, it just takes up space and wire. So what does "where used" mean in this case?

I understood your questions to be running the conductor in a separate raceway or outside of the conduit and outside disconnect box.

No, that was not my question.

You have two EGCs that appear to need to be 1.

Can you explain what you mean by this?

It is not the yellow highlighted grey neutral that I am concerned with but the grey and green conductors that do not enter that outside disconnect but are a apart of that circuit.

The Green are EGC's, so they are not part of any circuit. There is an EGC run in every conduit, including the conduit feeding the outside disconnect, and that EGC is bonding the outside disconnect enclosure. The jbox is PVC.

The grey and green should follow the phase conductors to the outside disconnect and back in order to meet 300.3(B).

There is an EGC in the conduit with the phase conductors to the outside PV Disconnect. The question is: If the neutral serves no purpose (so is not used) in the PV Disconnect, then why run it there? And if that is the case, 300.3(B) does not require N run where not used. However if the interpretation is that the N needs to be run there because it is used in the circuit between the 100A Service Disconnect and the Inverter, then is it sufficient just to run a single conductor from the 100A Svc Disc N to the outside PV Disc in that conduit, and not loop it back? It all routes thru the PVC Jbox, so no Neutral loops and no inductive heating effects.

 

[Elect117]

That seems to be the key point. The neutral is not used in the PV disconnect, i.e. running it in the conduit, out to the switch and back, with no termination, serves no purpose functionally or electrically, it just takes up space and wire. So what does "where used" mean in this case?

If it serves no purpose then remove it from the feeder all together. No neutral in the feeder or to any of the other boards that are being disconnected by that switch (or breaker). The issue is you do use the neutral for something and as such, it needs to follow the circuit.

The Green are EGC's, so they are not part of any circuit. There is an EGC run in every conduit, including the conduit feeding the outside disconnect, and that EGC is bonding the outside disconnect enclosure. The jbox is PVC.

That is incorrect. They are apart of the circuit's ground fault return path and are sized by the OCPD. How else would you size the EGC? That EGC (where wire type) should be contained within the same conduit, enclosures, etc. and bonded to the metal boxes (unless the phases are just pulled straight through). That includes where splices exist (like your J-box if metal).

There is an EGC in the conduit with the phase conductors to the outside PV Disconnect.

Where is the one for the conductors come back to the j-box? What is the fault path for the section of conductor with no EGC?

The question is: If the neutral serves no purpose (so is not used) in the PV Disconnect, then why run it there?

I can't answer the question on if it is used or not. It may not be used in the disconnect but it is certainly used upstream or downstream. I just can't tell because I only have a small snapshot. The neutral, when carrying the return current of a line to neutral load, is needed to cancel out the magnetic field so that there is no inductive heating. That can occur in raceways or in boxes.

And if that is the case, 300.3(B) does not require N run where not used.

Then remove it from the panel board and j-box for the circuit. I can't see the whole drawing and there are no feeder Ids so all I can say is if it needs to exist for the feeder. If it does than it needs to be contained within the same ferrous metal conduit and ferrous metal boxes.


I am only getting a minimal amount of information from the plans. I can't provide a clear answer on the EGCs or the neutral. You are probably better off asking the engineer of record (PE) or a friend to review your plans.

 

[wwhitney]

Where is the one for the conductors come back to the j-box? What is the fault path for the section of conductor with no EGC?

The horizontal dotted ellipse pointed to by the triangle 7 symbol encompasses both those conductors and the two ungrounded conductors supplying the PV System Disconnect, as well as an EGC and the unused neutral that may be deleted. So I infer the diagram calls for all 6 wires in a single conduit or cable. No need for a second EGC in that conduit or cable.

Cheers, Wayne

 

[solarken]

If it serves no purpose then remove it from the feeder all together. No neutral in the feeder or to any of the other boards that are being disconnected by that switch (or breaker).

The neutral is used in the critical loads panel downstream.

That is incorrect. They are apart of the circuit's ground fault return path and are sized by the OCPD. How else would you size the EGC?

I understand that, sorry my wording was bad. Not normally current carrying. There is a correctly sized EGC running in each conduit with the 2AWG conductors, as shown.

Where is the one for the conductors come back to the j-box? What is the fault path for the section of conductor with no EGC?

There is no section of conductor run without an EGC. I am attaching a more complete photo to give more context.

I can't answer the question on if it is used or not. It may not be used in the disconnect but it is certainly used upstream or downstream.

Correct, not used in the disconnect, but used downstream.

The neutral, when carrying the return current of a line to neutral load, is needed to cancel out the magnetic field so that there is no inductive heating. That can occur in raceways or in boxes.

There is no inductive heating whether the neutral is run to the outside disconnect or not, because of the pathway chosen and the hub and spoke arrangements. The sun of current in every conduit penetration is zero, even if the neutral is not brought into the outside disconnect.

I am only getting a minimal amount of information from the plans

Adding more details with these photos, I appreciate your feedback.

 

[solarken]

And here is version with neutral not brought to outside disconnect

 

[Elect117]

The horizontal dotted ellipse pointed to by the triangle 7 symbol encompasses both those conductors and the two ungrounded conductors supplying the PV System Disconnect, as well as an EGC and the unused neutral that may be deleted. So I infer the diagram calls for all 6 wires in a single conduit or cable. No need for a second EGC in that conduit or cable.

Cheers, Wayne

I didn't think they meant same conduit. I was mostly ignoring those since we don't have the callout detail. Those can say anything from required conductor type for wet or damp locations or random stuff about weatherproofing.

In response to the Updated photos and assuming it is the same conduit:

The way it is show on the first drawing is the way I would have designed the outside disconnect. With the neutral coming in and out. Not landed on anything and not cut.

But to the original questions, if they are in the same conduit then I don't know if you need to run the "second" neutral wire in that conduit. It isn't so much as the second wire but the same wire then. I was operating under the assumption each feeder/circuit had it's own conduit.

 

[wwhitney]

And here is version with neutral not brought to outside disconnect View attachment 2574770

Seems to me you do have a 300.3(B) issue at the critical loads subpanel. Its sole neutral connection is via the path: 100A service disconnect, J Box, Wiring Trough, Sol Ark, Critical Loads Subpanel. And when the critical loads subpanel is powered by the top "Solar" breaker, then the ungrounded supply follows the same route, in compliance with 300.3(B).

But when the critical loads subpanel is powered by the bottom "Bypass" breaker, then the ungrounded supply goes straight from the Wiring Trough to the Critical Loads Subpanel, bypassing the Sol Ark. So the ungrounded conductors and the neutral take different paths. If the two neutrals going to the Sol-Ark (the one from the Wiring Trough and the one to the Critical Loads Subpanel) are in different conduits with different entries into a steel enclosure at the Sol-Ark, then you will have (primarily theoretical) inductive heating of the enclosure between those two different entries whenever the critical loads subpanel is powered by the "Bypass" breaker and the neutral is carrying current due to unbalanced loading.

The simplest way to avoid this is, once you've chosen your loop-free path for the grounded conductor, to route all ungrounded conductors that are associated with that grounded conductor in any state of operation of the system all the same path as the grounded conductor. E.g. the grounded conductor could go from the 100A service disconnect to the critical loads subpanel and then to the Sol-Ark, and the "grid" ungrounded conductors for the Sol-Ark would be routed through the critical loads subpanel as well.

Cheers, Wayne

 

[solarken]

The way it is show on the first drawing is the way I would have designed the outside disconnect. With the neutral coming in and out. Not landed on anything and not cut.

The question is, why is that any better than not bringing the neutral into the outside disconnect and instead leaving the neutral conductor to pass thru the PVC Box from the 100A SVC disconnect on left to the conduit going to the trough?

Either way, the sun of currents is zero in any conduit. And from the perspective of the outside disconnect if the neutral loops into it and back out, it is not used there.

Consider the neutral ran to outside disconnect as you say is the way it should be done. Okay, then if the breaker interlock in the critical loads panel is in the Solar position, and so the main in that panel is on and the bypass breaker is off, then current to a 120v load will flow from the L1 SVC conductor tap thru 100A SVC disconnect breaker, upgrade thru PV disc, to inverter L1 Grid terminal, out inverter L1 load term, to load off the critical panel, with return back on neutral from panel to inverter to jbox up thru pv disconnect and back down (single conduit) thru SVC disconnect to main load center on the same physical path.

Now, if the interlock is in Bypass position, 120v load current will follow the same path from SVC tap to the jbox, but instead of going up thru the PVC disconnect will flow on the tap conductor to the trough (thru same conduit as before, but different conductor) thru the 100A branch breaker, to the load, and the return current flows from N of panel up thru inverter N, back down to trough, to jbox, and then up thru the PVC disconnect loop back down and then back to main load center.

Neither of the above scenarios results in neutral current following the L1 current side by side completely in the same exact physical path. But also the sum of currents in any conduit or thru any enclosure penetration is always zero so no inductive/magnetic heating effects.

So why is routing the neutral thru the PV disconnect necessary?

 

[solarken]

Seems to me you do have a 300.3(B) issue at the critical loads subpanel. Its sole neutral connection is via the path: 100A service disconnect, J Box, Wiring Trough, Sol Ark, Critical Loads Subpanel. And when the critical loads subpanel is powered by the top "Solar" breaker, then the ungrounded supply follows the same route, in compliance with 300.3(B).

But when the critical loads subpanel is powered by the bottom "Bypass" breaker, then the ungrounded supply goes straight from the Wiring Trough to the Critical Loads Subpanel, bypassing the Sol Ark. So the ungrounded conductors and the neutral take different paths.

You are right, I did miss that, the neutrals are in separate conduits at the inverter. Hmmm, need to read the rest of your post and adjust

 

[Elect117]

I probably am just not understanding this drawing. The overlapping runs and red/black wires for DC and AC is making it hard to follow. Plus I would have no way of knowing what material the conduit or box was from looking at the drawing. It is just not clear to me that it isn't needed.

All I can say is the neutral should follow the circuit.

Ill leave it up to you and whomever to decide if it is needed in the breaker disconnect box.

 

[wwhitney]

Schematically, I would propose a linear arrangement:

1) Existing Service Disconnect
2) New 100A Service Disconnect
3) PV System Disconnect (if necessary in this location)
4) Critical Loads Panel
5) Hybrid Inverter
6) Batteries/PV (DC Connections)

There would be a single ungrounded conductor 1 - 2 - 3 - 4 - 5. Service ungrounded conductors from (1) to (2). Feeder ungrounded conductors from (2) to (5) that do not pass through the PV System Disconnect. Additional PV ungrounded conductors to (5) that do pass through the PV System Disconnect; these could originate in (2) or via a splice in (3) to the first feeder. And lastly a feeder from (5) back to (4) as the alternate supply to the critical loads panel.

Cheers, Wayne

 

[wwhitney]

There would be a single ungrounded conductor 1 - 2 - 3 - 4 - 5.

That should be "single *grounded* conductor."

Cheers, Wayne

 

[solarken]

Schematically, I would propose a linear arrangement:

1) Existing Service Disconnect
2) New 100A Service Disconnect
3) PV System Disconnect (if necessary in this location)
4) Critical Loads Panel
5) Hybrid Inverter
6) Batteries/PV (DC Connections)

There would be a single ungrounded conductor 1 - 2 - 3 - 4 - 5. Service ungrounded conductors from (1) to (2). Feeder ungrounded conductors from (2) to (5) that do not pass through the PV System Disconnect. Additional PV ungrounded conductors to (5) that do pass through the PV System Disconnect; these could originate in (2) or via a splice in (3) to the first feeder. And lastly a feeder from (5) back to (4) as the alternate supply to the critical loads panel.

Cheers, Wayne

I put together a few simpler drawings, one with the original arrangement, and one with the 2 thru 5 series order you have there, passing all conductors through the Critical Loadcenter on the way to the inverter. I left the Existing main loadcenter and the DC conductors to the inverter out since they are irrelevant to this discussion. The dotted ovals group conductors in a single conduit.

I really don't think whether the Neutral is brought to the PV Disconnect or not is much of an issue, either way the sum of currents is zero and no inductive heating. The bigger issue is I don't see a way to have the neutral current always flow in the neutral conductor that is associated with it's line conductor in all instances. The issue in the first arrangement that Wayne found is that neutral current always flowed in the neutrals up thru the inverter, when grid power was the source and the Interlock was in Bypass position.

The issue in this second case is that the neutral to the inverter will never be used when the energy is coming from the grid (i.e. not from PV or Battery), since there is a neutral bus in the Critical Loads Panel. All neutral current from loads connected to that panel will flow directly from 4 to 2, even if the line current came from 5. If the Inverter sources the energy using PV or Battery, then there will be current flowing back on the neutral between 4 and 5, but even then, it will likely split between the load and the grid neutrals, that are in different conduits.

I am starting to think there is no way to use separate Load and Grid conduit paths in this Sol-Ark and keep the sum of currents to zero in each path. Maybe I need to go down to just one conduit for both Grid and Load, which will be tight, but they are less than 24 inches so no derating. Going to a single conduit seems to keep the sum of currents in each conduit to zero, but the neutral path issue remains. I am scratching my head on this. I must be missing something.

Drawing of the two arrangements is attached.