Neutral & Ground Bonding

[jaggedben]

I think a lot of this is due to hangups about the meaning of words and ignorance of the physics and engineering. But yeah, AHJs will ask for stuff.

 

[ggunn]

I think a lot of this is due to hangups about the meaning of words and ignorance of the physics and engineering. But yeah, AHJs will ask for stuff.

Personally I don't see a physics or engineering problem either way with this issue, but as you have seen from me before...

Ours is not to reason why; ours is to build it like they want or fail the inspection.

 

[Carultch]

I think a lot of this is due to hangups about the meaning of words and ignorance of the physics and engineering. But yeah, AHJs will ask for stuff.

What exactly is the Physics behind having multiple N-G bonds when you have multiple service disconnects? Is this just an oversight of the code, that we end up having to follow to make a compliant install? Or is there a physical basis for this to happen?

I understand the Physics behind restricting the number of places an N-G bond occurs, to avoid letting the EGC and other electrically inactive metal be in parallel with the neutral and carry objectionable current. However, this doesn't explain the double standard about why you get to / have to install multiple N-G bonds with multiple (separately-installed) service disconnects.

 

[BradPV]

The AHJs that direct us to leave G and N unbonded in the PV AC disco also require us to run a grounding conductor through the disco back to the main distribution panel from the PV system, so they do not consider it to be a service.

I get it, since you are running 4 wire they don’t want you to bond again which I do agree with. That is why it is so hit and miss in each jurisdiction we go to. IMO for a line (supply) side tap there should only be a three wire connection to the panel you are tapping and then the main PV disco should tap the existing GEC. Of course they don’t often ask for my opinion……

 

[BradPV]

What exactly is the Physics behind having multiple N-G bonds when you have multiple service disconnects? Is this just an oversight of the code, that we end up having to follow to make a compliant install? Or is there a physical basis for this to happen?

I understand the Physics behind restricting the number of places an N-G bond occurs, to avoid letting the EGC and other electrically inactive metal be in parallel with the neutral and carry objectionable current. However, this doesn't explain the double standard about why you get to / have to install multiple N-G bonds with multiple (separately-installed) service disconnects.

I was taught by a person I totally respect to look at a tap as if the panel the tap was in was non-existent. So by looking at it in that light there is no neutral/ground bond in the newly created service panel. Of course as stated earlier, this really only works with a three wire connection in the tapped panel. Code allows you to ground the neutral at any accessible point 250.24 which could be the original service disco as well as the new service disco. The main objective is to not have a parallel path which theoretically could allow fault current to flow in a circular motion without tripping the OCPD, theoretically.

 

[pv_n00b]

Texas adopted the 2020 NEC in 11/2020 and the 2020 NEC requires the N-G bond in the PV disconnect.

 

[pv_n00b]

The main objective is to not have a parallel path which theoretically could allow fault current to flow in a circular motion without tripping the OCPD, theoretically.

You can have parallel current paths in the neutral and equipment ground in systems with a grounded conductor that can desensitize the OCPD. But that's not circulating current.

 

[ggunn]

Texas adopted the 2020 NEC in 11/2020 and the 2020 NEC requires the N-G bond in the PV disconnect.

Where does it say that? Is it in so many words specific to PV or is it an interpretation based on what it says about other things? None of the AHJs we deal with here in Texas have changed what they expect to see; a couple of them want a N-G bond in the PV AC disco and all that goes with that, but the rest do not.

BTW, although as a state Texas has adopted the 2020 NEC, jurisdictions within the state can opt out; we deal with several that are still on the 2017 code and a couple that are on older versions than that. That said, there are several that have adopted the 2020 code and still want N and G isolated in the PV disco.

Truth be told, I don't see where it really makes a whole lot of difference one way or the other.

Also BTW, I am in no sense claiming that AHJ's always know what they are doing. One in particular flatly states that "PV is a service" even when it is load side interconnected AND they are one of those that wants N and G isolated in the disco. Go figure.

 

[jaggedben]

What exactly is the Physics behind having multiple N-G bonds when you have multiple service disconnects? Is this just an oversight of the code, that we end up having to follow to make a compliant install? Or is there a physical basis for this to happen?
...

My point about physics there is no meaningful difference to the safety of people or property between a service disconnect and a supply side PV disconnect. And since the purpose of the code is to ensure safety, I do not see, and have never seen, why the code should allow or prohibit different installation practices, regardless of what those practices might be. And yet, by accident (in my estimation), because of the way certain things are defined and described in the code, they get treated differently. (In my opinion the only disconnect that is connected directly to the service that should be treated differently than a service disconnect is the fire pump. But the code as written is a long way from that. There are literally a dozen or so sections that would have to be revised.)

Where does it say that?

See the new section 250.25

250.25 Grounding Systems Permitted to Be Connected on the Supply Side of the Disconnect.

Many line side connections terminate in an enclosure that is not considered a service disconnect.

www.electricallicenserenewal.com

 

[pv_n00b]

Where does it say that? Is it in so many words specific to PV or is it an interpretation based on what it says about other things? None of the AHJs we deal with here in Texas have changed what they expect to see; a couple of them want a N-G bond in the PV AC disco and all that goes with that, but the rest do not.

250.25

BTW, although as a state Texas has adopted the 2020 NEC, jurisdictions within the state can opt out; we deal with several that are still on the 2017 code and a couple that are on older versions than that.

They can do that. I remember in Hawaii before they got everyone on the same train that there were some AHJs with an NEC that was five revisions old. Made PV really hard.

That said, there are several that have adopted the 2020 code and still want N and G isolated in the PV disco.

They can ignore the NEC if they want, but they are in violation of state law if they don't adopt a local ordinance that specifically overrides 250.25. But hey it's Texas, "NEC? We don't need no stinkin' NEC!"

 

[ggunn]

250.25

They can do that. I remember in Hawaii before they got everyone on the same train that there were some AHJs with an NEC that was five revisions old. Made PV really hard.

They can ignore the NEC if they want, but they are in violation of state law if they don't adopt a local ordinance that specifically overrides 250.25. But hey it's Texas, "NEC? We don't need no stinkin' NEC!"

To a cursory reading it looks to me that it hinges on whether or not a supply side PV interconnection is considered a service as defined by the NEC. But you're not going to trap me into defending the AHJs' interpretations of the NEC.

 

[jaggedben]

To a cursory reading it looks to me that it hinges on whether or not a supply side PV interconnection is considered a service as defined by the NEC. ...

It doesn't. 250.24 applies to service disconnects. 250.25 applies to anything anyone tries to say is not a service disconnect, and refers back to 250.24, essentially requiring they be treated the same.

 

[ggunn]

It doesn't. 250.24 applies to service disconnects. 250.25 applies to anything anyone tries to say is not a service disconnect, and refers back to 250.24, essentially requiring they be treated the same.

All I can say is "whatever"; that part of the code is a convoluted mess, IMO. I really don't care; I will design it the way the AHJ tells me to. It's not my job to play code nazi to the AHJ, especially in this case where it isn't a safety issue for people or equipment.

 

[pv_n00b]

All I can say is "whatever"; that part of the code is a convoluted mess, IMO. I really don't care; I will design it the way the AHJ tells me to. It's not my job to play code nazi to the AHJ, especially in this case where it isn't a safety issue for people or equipment.

Grounding is always a safety issue. I would not say, "the AHJ told me to do it wrong so I did it wrong, it's not my fault if something bad happened" but that's just me.

 

[ggunn]

Grounding is always a safety issue. I would not say, "the AHJ told me to do it wrong so I did it wrong, it's not my fault if something bad happened" but that's just me.

In this case I don't agree, but although I know a lot I don't know everything. Can you tell me what could happen that would be a safety hazard when a line side PV interconnection is not treated as a service, i.e., ground and neutral are separated in the PV AC disco and an EGC is run back to the MDP from the inverter(s) through the disco?

 

[CrazyWabbit]

I'm in Texas and had solar (enphase) installed October 2020 (tri-county electric) and my neutral/ground is NOT bonded in the first disconnect. i have two 200 amp loads panels that are lugged to the meter base (320 amp service) as well as the solar (tri lugged meter base), and both loads panels have the neutral/ground bonded but not the solar.
then in 2021 i got hit by lightning! primary strike missed the house, perhaps by inches, but a tentacle did hit the house. strike hit the end of the ridge, thru roofing material (hole) into flashing, to gutters, down the down spouts, to flashing above ledger, thru 3.5 inch screw and arced to a electrical junction box! also the lightning departed the end of the first section of roofing flashing and struck the same point on the ground. i visually saw this from inside the house.
energy entered the house from many directions, underground cable coax, sprinkler system, POE camera which was directly under roof strike, and house electrical system. result was 30K in damages. no visible signs of solar panels being hit, ~ 2 feet away, but seven panels had the bypass diodes shorted out. I replaced the Shockey diodes they they are perfectly functional. pattern was the closest three then a bit random from there. IQ7+ inverters were functional. Inverters are only 2 wire, racking system is grounded to same electrode as house ground rod, and house is also grounded to cold water system.
personally if only considering lightning strike, i don't think the ground and neutral should be bonded at on the solar system and the system should be on its own ground seven feet from house ground if possible. being metal on the roof, it is asking to be struck! but we all know then a fault on the 240v wiring to ground would not clear...
my ground to the solar is bonded to the neutral indirectly thru the loads panels.
i was on my ipad, while charging, and got zapped.
both service panels have ge branded surge devices, and they indicate to still be functional. yet lots of damage. my belief is that everything (ground, neutral and both hots) were at the same potential, thus they survived.

 

[LarryFine]

personally if only considering lightning strike, i don't think . . . the system should be on its own ground seven feet from house ground if possible.

my belief is that everything (ground, neutral and both hots) were at the same potential, thus they survived.

You seem to be contradicting yourself here.

 

[CrazyWabbit]

You seem to be contradicting yourself here.

no, i'm talking i think things should be wired and how my house was wired. my solar is connected to my only ground rod and my house is grounded to same ground rod, thus my solar is bonded to the neutral thru the neutral/ground bond in my two service panels. exactly as ggunn is describing above. so neutral and ground at same potential. ask my neighbors about their damages that are on the same transformer!

 

[CrazyWabbit]

nec allows things which should be illegal, like quick connect receptacles and switches. my house almost burned down three times do to these.

 

[ggunn]

no, i'm talking i think things should be wired and how my house was wired. my solar is connected to my only ground rod and my house is grounded to same ground rod, thus my solar is bonded to the neutral thru the neutral/ground bond in my two service panels. exactly as ggunn is describing above. so neutral and ground at same potential. ask my neighbors about their damages that are on the same transformer!

I don't see how bonding ground to neutral in your PV AC disco would have made any difference. I am assuming that your PV system is supply side interconnected; if it were connected through a backfed breaker bonding N to G would have been a code violation.