PV disco GEC connection

[electrofelon]

Say i have a service with a single 200A disconnect which is a typical 200A main breaker panelboard. Say i tap the service conductors in the panel, ahead of the main breaker and run them to another service . disconnect. Lets say the GES consists of 2 ground rods and the GEC lands in the original panel on the neutral bar. Can i use the "common location " provision in 250.64 and thus NOT run a GEC tap the the second disconnect? Would it make any difference if this for PV vs a regular 230.40 exception #2 install?

 

[jaggedben]

I think the common location argument ought to be fine but if the AHJ wanted to quibble they could say that the inside of one service disconnect isn't common to both service disconnects. Would you rather install a second GEC (or GEC tap) to the PV disconnect, and be ready to remove it if the inspector says "objectionable current"? Or not install it but be ready to install it if he doesn't buy the common location idea?

I don't think it makes any difference if it's PV or not, but I'm biased. Just keep calling it a new service disconnect. As you know, some AHJs have weird ideas about this.

 

[electrofelon]

I think the common location argument ought to be fine but if the AHJ wanted to quibble they could say that the inside of one service disconnect isn't common to both service disconnects. Would you rather install a second GEC (or GEC tap) to the PV disconnect, and be ready to remove it if the inspector says "objectionable current"? Or not install it but be ready to install it if he doesn't buy the common location idea?

I don't think it makes any difference if it's PV or not, but I'm biased. Just keep calling it a new service disconnect. As you know, some AHJs have weird ideas about this.

K yeah i was just curious, for some reason it just occurred to me today for the first time that in the aforementioned situation, it seems like you are already set up for the common location so why run a gec tap. In the past ive always run the tap without really thinking about it. Have you used that strategy or do you run a tap?

 

[wwhitney]

I think the common location argument ought to be fine but if the AHJ wanted to quibble they could say that the inside of one service disconnect isn't common to both service disconnects.

250.64(D)(3) says "A grounding electrode conductor shall be connected in a wireway or other accessible enclosure on the supply side of the disconnecting means . . ."

From the definitions of "disconnecting means" and "device," the main breaker is the disconnecting means. So on the supply side of that breaker complies with 250.64(D)(3).

I assume the idea is that if you intercept the service conductors with say 3 port Polaris connectors, you use a 4 port connector on the grounded conductor, so that you can land a GEC there?

Cheers, Wayne

 

[jaggedben]

K yeah i was just curious, for some reason it just occurred to me today for the first time that in the aforementioned situation, it seems like you are already set up for the common location so why run a gec tap. In the past ive always run the tap without really thinking about it. Have you used that strategy or do you run a tap?

Honestly it's been over 5 years since I've done one that way that and I don't really remember. I don't recall running a separate GEC (or tap) on the later ones. I think my strategy circa 2017-2020 (on the 2014 code here) was usually to run both a green and white wire to the PV disconnect and *not* bond them, on the theory I could put the green screw in the disco and remove the green wire if the inspector asked. So then I was prepared to argue the 'PV disco is not a service disconnect' way of looking at things that could fly back then. The supply side taps I did this way were few and far enough between (it's mostly meter/main combos around here) that I don't think I ever ran into an inspector who knew these parts of the code well enough to argue anyway. So yeah, I just relied on the existing GEC to the existing panel neutral bar.

Back even earlier (pre 2014) I probably did run a separate GEC or GEC tap at least once or twice, but that was because the old inverter(s) needed a DC GEC with the irreversible splicing and all that. So back then I was thinking about the old awful 690.47 rather than 250.64(D).

 

[electrofelon]

Honestly it's been over 5 years since I've done one that way that and I don't really remember. I don't recall running a separate GEC (or tap) on the later ones. I think my strategy circa 2017-2020 (on the 2014 code here) was usually to run both a green and white wire to the PV disconnect and *not* bond them, on the theory I could put the green screw in the disco and remove the green wire if the inspector asked. So then I was prepared to argue the 'PV disco is not a service disconnect' way of looking at things that could fly back then. The supply side taps I did this way were few and far enough between (it's mostly meter/main combos around here) that I don't think I ever ran into an inspector who knew these parts of the code well enough to argue anyway. So yeah, I just relied on the existing GEC to the existing panel neutral bar.

Back even earlier (pre 2014) I probably did run a separate GEC or GEC tap at least once or twice, but that was because the old inverter(s) needed a DC GEC with the irreversible splicing and all that. So back then I was thinking about the old awful 690.47 rather than 250.64(D).

I just really hate running the GEC in those situations because you are just (if you ignore the green color) making an parallel grounded conductor that likely violates the parallel conductor rules, so it just seems dumb.

 

[electrofelon]

250.64(D)(3) says "A grounding electrode conductor shall be connected in a wireway or other accessible enclosure on the supply side of the disconnecting means . . ."

From the definitions of "disconnecting means" and "device," the main breaker is the disconnecting means. So on the supply side of that breaker complies with 250.64(D)(3).

I suppose it could be argued that the common location enclosure needs to be ahead of the disconnecting means' enclosure.

I assume the idea is that if you intercept the service conductors with say 3 port Polaris connectors, you use a 4 port connector on the grounded conductor, so that you can land a GEC there?

Cheers, Wayne

I would have to review the precise wording, but couldn't you just land the GEC tap on the neutral bar of the original service disconnect?

I usually make the ungrounded conductor tap with crimp H taps, then land the grounded and GEC tap on the neutral bar - that's in the case of the gec tap being run with the other conductors. Sometimes I'll run a gec tap not in the raceway with the other conductors and H tap it on to the existing grounding electrode conductor

 

[tortuga]

I most often see it done like this:

Sometimes I'll run a gec tap not in the raceway with the other conductors and H tap it on to the existing grounding electrode conductor

My understanding of the theory end and just talking theory not NEC
An ideal TN-C (MGN) system you have no less than 4 grounding electrode systems per mile (NESC Rule 96C) and exactly one grounding electrode system (GES) at each end and at each transformer.
First at each structure tie all electrodes that qualify together along with all required inter-system bonding / parallel metal paths back to the MGN system; coax, landline phone, communications cable protectors etc.
Also tower / antenna grounds, metal pipe, lighting protection bond etc.

Then run exactly one Grounding Electrode Conductor (GEC) from that single grounding electrode system to the service neutral (grounded conductor to be precise) at the structure no later than the service disconnect(s).
So anywhere from the Service point - service disconnect ( CT can / meter pack / main distribution switch gear etc)
what I think your proposing the common location 250.64(D)(3) is good -closest to the upstream MGN service point, especially if the disconnects are not closely grouped.
After the first OCPD or as soon as I can I'd completely switch to a TN-C-S system.

I think my strategy circa 2017-2020 (on the 2014 code here) was usually to run both a green and white wire to the PV disconnect and *not* bond them,

I have seen that done also thats effectively making a TN-C-S system also a good approach.