A potential code violation that seems to be too elusive and difficult to understand. Please help!

[jaggedben]

What I know is that the EGC from solar" which includes the MC conduit" since it is bonded on PV disconnect is
carrying current from "N" and looping it with the other home EGC wiring.

The MC conduit would only carry neutral current if it made contact with the neutral (not EGC) of another circuit. I.e some fluke like a screw from a strap holding the MC hit a cable in the wall.

Have an electrician amp clamp the neutral in the PV disconnect to see if it is actually carrying any current or not.

If the MC conduit is bonded, it would effectively be a neutral wire and if that Neutral wire( conduit) is then touching EGC-home (with less than 1ohm of resistance) then why isn't that viewed as double bonding. Which would be a clear code violation?

Bonding just means to bring diverse voltage of two different surfaces into equilibrium with one another. So in this case, so long as that MC conduit remains connected to the PV disconnected with the bonding screw then it acts as a neutral wire.

Only if the circuit is completed to a neutral wire somewhere else.

Now we ask the question? If that neutral wire touches a ground wires ( as evidenced by <1ohm) or resistance, whether you wrap around a metal or screw them in doesn't matter very much ( resistance is resistance). You would effectively be creating a second bond no matter how you do it. Why is this not a "clear" code violation?

The only place we have some reason to believe there is a loop is between the main panel and PV disconnect. But we have no reason to believe there's current flowing around that loop. And even if it were, it wouldn't feel any different to you than if it were wired normally, if you can indeed feel any of this stuff.

 

[don_resqcapt19]

What I know is that the EGC from solar" which includes the MC conduit" since it is bonded on PV disconnect is
carrying current from "N" and looping it with the other home EGC wiring. If the MC conduit is bonded, it would effectively be a neutral wire and if that Neutral wire( conduit) is then touching EGC-home (with less than 1ohm of resistance) then why isn't that viewed as double bonding. Which would be a clear code violation? Bonding just means to bring diverse voltage of two different surfaces into equilibrium with one another. So in this case, so long as that MC conduit remains connected to the PV disconnected with the bonding screw then it acts as a neutral wire. Now we ask the question? If that neutral wire touches a ground wires ( as evidenced by <1ohm) or resistance, whether you wrap around a metal or screw them in doesn't matter very much ( resistance is resistance). You would effectively be creating a second bond no matter how you do it. Why is this not a "clear" code violation?

That parallel path exists in every system that has more than one service disconnect. In general, while the 2017 and earlier codes did not say it like the newer codes do, every disconnect for a line side connected solar system is a service disconnet.

 

[wwhitney]

The MC conduit would only carry neutral current if it made contact with the neutral (not EGC) of another circuit.

That's not correct. If you have Service Disconnect 1 with N-G bond creating EGC1, and Service Disconnect 2 with N-G bond creating EGC2, and EGC1 is connected somewhere with EGC2, then you have the loop described in my first post in this thread (#13). Therefore all neutral current from both systems will be split between the intended path and the alternate path through the EGCs.

For the simple case that the GES connection is singular to a common neutral point upstream of both service disconnects (and therefore does not contribute a further loop), the current will divide in inverse proportion to the impedances of the intended path from the N-G bond to the common neutral point, and the path through both EGC systems, through the other service N-G bond to to the common neutral point.

Cheers, Wayne

 

[jaggedben]

That's not correct. If you have Service Disconnect 1 with N-G bond creating EGC1, and Service Disconnect 2 with N-G bond creating EGC2, and EGC1 is connected somewhere with EGC2, then you have the loop described in my first post in this thread (#13). Therefore all neutral current from both systems will be split between the intended path and the alternate path through the EGCs.

For the simple case that the GES connection is singular to a common neutral point upstream of both service disconnects (and therefore does not contribute a further loop), the current will divide in inverse proportion to the impedances of the intended path from the N-G bond to the common neutral point, and the path through both EGC systems, through the other service N-G bond to to the common neutral point.

Cheers, Wayne

I suppose your technically correct, but in most cases the amps taking the normal route will be one or several hundred times more than those taking the alternative route. So 'split' has a misleading connotation there.

And theoretically what you say is true without separate service disconnects, but so insignificant it might as well not be.

I would repeat the OP should have an electrican amp clamp the neutral in the PV disconnect to see what it actually is.

 

[wwhitney]

And theoretically what you say is true without separate service disconnects, but so insignificant it might as well not be.

How's that? The loop involving the EGC and the neutral requires multiple N-G bonds.

Something similar could happen on a service with only one disconnect if there are multiple GECs and one lands at, say, the meter, while another lands at the service disconnect. But that loop would only put neutral current on the GECs and associated bonding jumpers, not on the EGC.

I suppose your technically correct, but in most cases the amps taking the normal route will be one or several hundred times more than those taking the alternative route. So 'split' has a misleading connotation there.

The OP seems to be concerned about 50 ma, so a ratio of 500 would clearly be of concern to them. That would just take 25A of neutral current.

But I've been wondering about what the order of magnitude the ratio might be, so below I'll make up an example and calculate it.

Cheers, Wayne


Say you have a 200A residential service with a 2/0 Al grounded conductor. There's a line side PV interconnection in a trough under the meter enclosure, with 4 ft of 2/0 Al grounded conductor going to the primary service disconnect/distribution panel, and 4 ft of #10 Cu grounded conductor going to a 60A PV disconnect. Then the PV side has 50 ft of #10 Cu EGC to the EGC interconnection, while the circuit on the primary panel has 50 ft of #12 Cu EGC to that point.

Ignoring connection resistances for simplicity (how inaccurate is that?) , and assuming the PV inverter doesn't put out any neutral current, we need to compare the impedances of (a) 4 ft of 2/0 Al grounded conductor with (b) 54 ft of #10 Cu plus 50 ft of #12 Cu.

Coopting the Southwire Voltage Drop Calculator as an Impedance Calculator (using a PF of 1.0, a voltage of 100V, a current of 1000A, single phase and half the lengths, the % VD will be the drop in volts, which is the impedance in milliohms), I get that (a) is 0.71 milliohms, while (b) is 68 milliohms plus 100 milliohms = 168 milliohms. So yep, a ratio of 168/0.71 = 240.

Of course, varying the wire sizes and lengths could change that ratio an order of magnitude either way.

 

[jaggedben]

How's that? The loop involving the EGC and the neutral requires multiple N-G bonds.

Something similar could happen on a service with only one disconnect if there are multiple GECs and one lands at, say, the meter, while another lands at the service disconnect. But that loop would only put neutral current on the GECs and associated bonding jumpers, not on the EGC.

It's common to land all neutrals and grounds on the same terminal bar, without any regard for the order, in a service panel with one main disconnect. That also allows for neutral current to follow ground loops created by EGCs.

 

[wwhitney]

It's common to land all neutrals and grounds on the same terminal bar, without any regard for the order, in a service panel with one main disconnect. That also allows for neutral current to follow ground loops created by EGCs.

So for a common terminal bar in a service panelboard, does anyone bother to spatially segregate the neutrals and the grounds, with the neutral service conductor landed in the neutral section, to eliminate this possibility?

Cheers, Wayne

 

[jaggedben]

So for a common terminal bar in a service panelboard, does anyone bother to spatially segregate the neutrals and the grounds, with the neutral service conductor landed in the neutral section, to eliminate this possibility?

Cheers, Wayne

I'm sure a few do. I think I've seen discussions on this forum of the general issue of having them on the bar, but don't recall them bringing up this exact point. But in any case, generally in my experience in residential .... No, they're landed willy nilly.

 

[jaggedben]

I've also seen plenty of neutral-ground faults, as well as neutrals shared or swapped improperly between circuits, to ever make the assumption that a ground loop would be the main reason for neutral current in an odd place.

 

[GuruLal]

Providing they are not being very naughty and using the conduit shell to carry neutral current. That would be a nono

I think I have read some regulation regarding ligjtning protection that requies the grounding of the solar panel frsme and rails to go directly to main system ground. It can even have its own ground rod but you have to run a buried grounding wire from that ground rod the the house ground rod. The solar grounding conductors must contact other house grounding conductors as a lightning strike on tje panels vould then damage other household equipment

What I am describing to you is exactly that. The conduit "shell" is what's carrying the current in this setup, in addition to the ground wire. I can measure magnetic field when I test the shell in isolation with no other wires connected (L1,L2, N and G). Testing the shell from the outside PV disconnect to inner runs, there is field on that shell. That is what I'm trying to get resolved.

 

[GuruLal]

Providing they are not being very naughty and using the conduit shell to carry neutral current. That would be a nono

I think I have read some regulation regarding ligjtning protection that requies the grounding of the solar panel frsme and rails to go directly to main system ground. It can even have its own ground rod but you have to run a buried grounding wire from that ground rod the the house ground rod. The solar grounding conductors must contact other house grounding conductors as a lightning strike on tje panels vould then damage other household equipment

"Must contact".. or must not? You meant must not contact other house grounding conductor right? Could you please bring that code book here, that is what's critical.

 

[GuruLal]

They tapped on the supply side which is an allowed way to connect a PV system. It looks like they pulled a ground wire but did not connect it, instead using the neutral as the fault current return path from the disconnect (since you said they bonded neutral to ground in the PV disconnect). This does not result in anything meaningfully different from what is allowed for multiple service disconnect enclosures under the NEC.

Arguably under the 2017 code they should land the green wire in the main panel where your picture is taken and not have the neutral-ground bond in the PV disconnect. HOWEVER your solar system almost certainly does not put any current on the neutral in any case (The neutral is only used as a voltage reference.) So that change would make no real difference to what is happening.

Again, if there is neutral current on the exposed parts of the PV system - (and you haven't shown us exactly why we should be believe that) - then it's due to something else. Some other accident or mistake somewhere else.

Some corrections, That ground wire is pulled through, it is connected on the main panel and it feeds into my PV disconnect also and bonds with the disconnect. Not sure what you were looking at.

There is neutral current on the exposed parts of the PV system. When the PV system is energized, that neutral current grows further and it dumps that on the conduit shell which then dumps it into the EGC of my home systems. See picture. I've seen that neutral current go up as much as .5Amps when the PV system is engaged and now considering that the EMC conduits are directly bonded to the panel, the "outer shell" of the conduit is carrying that neutral current and looping it with my home EMC. Which code can I site for violation? Common guys!

 

[GuruLal]

The MC conduit would only carry neutral current if it made contact with the neutral (not EGC) of another circuit. I.e some fluke like a screw from a strap holding the MC hit a cable in the wall.

Have an electrician amp clamp the neutral in the PV disconnect to see if it is actually carrying any current or not.





Only if the circuit is completed to a neutral wire somewhere else.


The only place we have some reason to believe there is a loop is between the main panel and PV disconnect. But we have no reason to believe there's current flowing around that loop. And even if it were, it wouldn't feel any different to you than if it were wired normally, if you can indeed feel any of this stuff.

The circuit is completed with the EMC ground with my home. The conduit outside makes contact with the disconnect panel that is bonded to neutral.

 

[GuruLal]

That parallel path exists in every system that has more than one service disconnect. In general, while the 2017 and earlier codes did not say it like the newer codes do, every disconnect for a line side connected solar system is a service disconnet.

Please explain what you are trying to convey. I'm bit confused. Pardon my ignorance.

 

[GuruLal]

That's not correct. If you have Service Disconnect 1 with N-G bond creating EGC1, and Service Disconnect 2 with N-G bond creating EGC2, and EGC1 is connected somewhere with EGC2, then you have the loop described in my first post in this thread (#13). Therefore all neutral current from both systems will be split between the intended path and the alternate path through the EGCs.

For the simple case that the GES connection is singular to a common neutral point upstream of both service disconnects (and therefore does not contribute a further loop), the current will divide in inverse proportion to the impedances of the intended path from the N-G bond to the common neutral point, and the path through both EGC systems, through the other service N-G bond to to the common neutral point.

Cheers, Wayne

Why isn't this a code violation? I'm trying to get my county inspector to understand that this is a code violation that they need to object to. Help me with how I bring this up to them?

 

[GuruLal]

How's that? The loop involving the EGC and the neutral requires multiple N-G bonds.

Something similar could happen on a service with only one disconnect if there are multiple GECs and one lands at, say, the meter, while another lands at the service disconnect. But that loop would only put neutral current on the GECs and associated bonding jumpers, not on the EGC.


The OP seems to be concerned about 50 ma, so a ratio of 500 would clearly be of concern to them. That would just take 25A of neutral current.

But I've been wondering about what the order of magnitude the ratio might be, so below I'll make up an example and calculate it.

Cheers, Wayne


Say you have a 200A residential service with a 2/0 Al grounded conductor. There's a line side PV interconnection in a trough under the meter enclosure, with 4 ft of 2/0 Al grounded conductor going to the primary service disconnect/distribution panel, and 4 ft of #10 Cu grounded conductor going to a 60A PV disconnect. Then the PV side has 50 ft of #10 Cu EGC to the EGC interconnection, while the circuit on the primary panel has 50 ft of #12 Cu EGC to that point.

Ignoring connection resistances for simplicity (how inaccurate is that?) , and assuming the PV inverter doesn't put out any neutral current, we need to compare the impedances of (a) 4 ft of 2/0 Al grounded conductor with (b) 54 ft of #10 Cu plus 50 ft of #12 Cu.

Coopting the Southwire Voltage Drop Calculator as an Impedance Calculator (using a PF of 1.0, a voltage of 100V, a current of 1000A, single phase and half the lengths, the % VD will be the drop in volts, which is the impedance in milliohms), I get that (a) is 0.71 milliohms, while (b) is 68 milliohms plus 100 milliohms = 168 milliohms. So yep, a ratio of 168/0.71 = 240.

Of course, varying the wire sizes and lengths could change that ratio an order of magnitude either way.

My PV inverters puts about neutral current. Good bit! about 500 mA or so.

 

[jim dungar]

My PV inverters puts about neutral current. Good bit! about 500 mA or so.

Have you checked the specifications of your meter? The instruction manual says its accuracy only applies in the range from 5%(1A) to 100% (20A).

Based on the info you have provided, there are no readily apparent NEC violations.
Non-current carrying metallic paths must be bonded together, even if they have separate sources. This bonding is allowed to occur in multiple places, effectively creating ground loops. Neutral to ground bonding is only allowed at the service entrance equipment unless the system is separately derived.

You may be interested in a book called "Tracing EMFs in Building Wiring and Grounding" by Karl Riley.

 

[Fred B]

There is neutral current on the exposed parts of the PV system.

Exactly how are you determining this? Unless something is drastically mis-wired, the "system" wouldn't be culprit. Based on pictures everything looks to be code compliant.

N/G voltages can be present in normally operating and compliant systems in the home coming from the utility side. Current readings on neutral can certainly be within the nominal range seen from a utility source and it can be much higher. (Typical of a multi-grounded utility system.) Also, are you using a low impedance meter. (False voltage reading can happen if not) These types of voltages (NEV) will fluctuate depending on normal neutral currents from neighbors into the utilities system, this is present and flowing thru the earth from the utility transformer connections.
Points of connections will create an anomalies in your readings between points due to voltage gradients, same reason for pool Equipotential bonding putting everything at same potential.

I am debating with county officials and the licensed contractor on why this is not a code violation which I strongly believe it is based on many other electricians that I've consulted with. The contractor doesn't want to admit cause it's more work for them to rectify it and the county officials seem to be bit confused about this.

If the "electricians" you are consulting with are unfamiliar with solar installations they likely are giving false or at least uninformed advice. Solar installations create unique installations created by parallel power production.

Well this issue got escalated quite quickly after my body sensed a huge electrical field. It manifested in tinnitus pressure type headaches. This would persist even after I completely turned off the solar panels, the inverters etc.

This is another indicator the issue is resulting from NEV from utility not directly related to the Solar installation.

Your issues indicating an EMF sensitivity would actually be benefited by the use of EMT and MC wiring methods. Typically these wiring methods will substantially "shield" any EMF resulting from the wiring.

 

[jaggedben]

What I am describing to you is exactly that. The conduit "shell" is what's carrying the current in this setup, in addition to the ground wire. I can measure magnetic field when I test the shell in isolation with no other wires connected (L1,L2, N and G). Testing the shell from the outside PV disconnect to inner runs, there is field on that shell. That is what I'm trying to get resolved.

What exact device are you using to make these measurements, and how?

 

[jaggedben]

The circuit is completed with the EMC ground with my home. The conduit outside makes contact with the disconnect panel that is bonded to neutral.

That is not completing a circuit.