My first 320 residential install

Status
Not open for further replies.
This is my revision from what I am understanding from this thread.

This 320 residential service will have 2 circuit breaker disconnects feeding 2 indoor sub panels.
Each circuit breaker disconnect neutral bar is bonded and the sub panels are not.
  1. Installation of one continuous #6 bare gec from disconnect #1 to 2 ground rods 6’ apart. 250.66 A
  2. Installation of one #4 stranded conductor from disconnect #1 to rebar ufer. 250.64D & 250.66
  3. Each sub panel will have a #4 stranded green conductor from the ground bar to each disconnect. 250.102 C1
 
roger said:
My agreement with Bob is how it should be done.
Click to expand...
If by "bob's method" you mean where he said "you just run number six between the two circuit breakers and then to the two ground rods" that still creates a neutral loop. The only way to avoid a neutral loop is to land a single GEC on the service grounded conductor upstream of the point where it divides for the two service disconnects.

But as a metallic service raceway already can create a "neutral loop," it seems to me a sufficient objective is to keep the neutral loop in the vicinity of the service equipment. Which the drawing in post #12 doesn't do, but both "bob's method" and the drawing in post #17 do do.

Cheers, Wayne
 
wwhitney said:
If by "bob's method" you mean where he said "you just run number six between the two circuit breakers and then to the two ground rods" that still creates a neutral loop.

Cheers, Wayne
Click to expand...
Draw it out and post it here so we can see what you're trying to say.
 
roger said:
Draw it out and post it here so we can see what you're trying to say.
Click to expand...
It would look the same post #17, connectivity wise.

Anytime you have a grounded service conductor that splits to two (or more) service disconnects, and a GEC in each disconnect to a common GES, you get a loop that will carry neutral current.

Of course, even if you land a common GEC upstream (utility side) upstream of the point where the grounded service conductor splits, there's still the EGC systems, with the separate MBJs in each disconnect. Then if the different disconnects' EGC systems ever end up bonded to each other, you have a loop that will carry neutral current.

As the two EGC systems getting bonded to each other seems extraordinarily hard to avoid, it seems like the best way to minimize the resulting current loop is to immediately bond the EGC systems together at the two (or more) disconnects. That way even if there is a downstream bond between the EGC systems, current on the EGC downstream of the intentional bond should be small or zero.

Cheers, Wayne
 
Last edited:
Wayne, look at the illustration below and imagine the panel is the meter and there is panel beside the imaginary meter, now imagine the GEC jumping over to the panel shown from the imaginary meter, where do you see a loop?

1661730587322.png
 
roger said:
Wayne, look at the illustration below and imagine the panel is the meter and there is panel beside the imaginary meter, now imagine the GEC jumping over to the panel shown from the imaginary meter, where do you see a loop?
Click to expand...
If I'm following your description correctly, there is only one meter and one panel, and there is no loop. I've been discussing the case of two or more service disconnects in separate enclosures, on a single service.

Maybe this will help. In the marked up diagram below from the NEC Handbook, the loop is in red. Whenever you have two or more service disconnects in separate enclosures, and you don't have a common GEC terminated to the grounded service conductor before it splits (e.g. in the wireway above the panels), there will be such a loop from the separate GECs.

[Likewise, if the two separate EGC systems from the two separate MBJs in the two separate disconnects ever get bonded together anywhere, you end up with a loop.]

Cheers, Wayne

loop.jpg
 
wwhitney said:
As the two EGC systems getting bonded to each other seems extraordinarily hard to avoid, it seems like the best way to minimize the resulting current loop is to immediately bond the EGC systems together at the two (or more) disconnects. That way even if there is a downstream bond between the EGC systems, current on the EGC downstream of the intentional bond should be small or zero.
Click to expand...
That's exactly what I did here. A single length of #6, rod to disconnect to disconnect to rod. I bent the wire in a U in each disconnect so it was one piece, and each end went through the wall to one rod for minimal outside wire length.

If you open and zoom in on each pic, you can see the GEC. In the first one, it runs down each stud and horizontally a few inches above the bottom plate, then outside over 6' apart; and what I did in both discos is in the second pic.
 

Attachments

  • DSC00528.JPG
    DSC00528.JPG
    881.4 KB · Views: 13
  • Dsc00529.jpg
    Dsc00529.jpg
    232.9 KB · Views: 13
LarryFine said:
That's exactly what I did here. A single length of #6, rod to disconnect to disconnect to rod. I bent the wire in a U in each disconnect so it was one piece, and each end went through the wall to one rod for minimal outside wire length.
Click to expand...
Sure, if you are going to land GECs in each panel (or disconnect), keeping the loop at the panels as small as possible seems wise. And once you have that GEC bond between the two panels' combination ground/neutral bars, then that bond would minimize any current on the EGC systems you might get if you have an inadvertent downstream EGC-EGC bond between the two systems.

However, the text you quoted was referring to the case where a single GEC is connected to the grounded service conductor upstream of any split, so you don't have that GEC bond between the two panels' combination ground/neutral bars. Then if you are able to keep the EGC systems separate, you get no loops at all (other than possibly from metallic service conduit). But if the EGC systems do end up bonded together somewhere else, you get neutral current from both services on both EGC systems. So there I was suggesting to have an explicit, short EGC-EGC bond between the two service panels.

[Or better, if each disconnect just supplies a single feeder, route both feeders immediately to a common enclosure and tie the EGCs together there, i.e. single point bonding. But that's more trouble.]

Cheers, Wayne
 
wwhitney said:
Whenever you have two or more service disconnects in separate enclosures, and you don't have a common GEC terminated to the grounded service conductor before it splits (e.g. in the wireway above the panels), there will be such a loop from the separate GECs.
Click to expand...
Just to clarify, the above assumes that all available GEs are bonded together to form a GES, independent of the GECs. If it were allowed to segregate the GEs into two groups, with a GEC from one group going to one disconnect, and a separate GEC from the other group going to the other disconnect, so that only the grounded service conductor provides the bond between the two groups, then you'd eliminate the GEC loop. But you'd still have a risk of an EGC loop.

However, my understanding is that it is not allowed to depend on the grounded service conductor to bond together the GEs. Or maybe I'm wrong: On a service with a single disconnect, if the only GEs are two ground rods, can I run one GEC from one ground rod to the meter, and another GEC from the other ground rod to the service disconnect? If I can, then for the case of two service disconnects, what's wrong with just running separate GECs from separate ground rods to each service disconnect? Both examples rely on the grounded service conductor to complete the bonding between ground rods.

Cheers, Wayne
 
hotblueelectric said:
Look at my revised method post #22. Will that work? Or should I stay with my original post?
Click to expand...
In my opinion, you should do what I did, or what is shown in post 17, or whatever is easiest, and not worry about it. The important thing is that both discos "see" both rods (and any other electrodes)

When I converted a 200a service to a 2x 200a service, I simply ran a second #4 to a second clamp on the interior water piping (plastic water service) and a second #6 to a second clamp on the nearer rod.
 
Last edited:
hotblueelectric said:
This 320 residential service will have 2 circuit breaker disconnects feeding 2 indoor sub panels.
Each circuit breaker disconnect neutral bar is bonded and the sub panels are not.
  1. Installation of one continuous #6 bare gec from disconnect #1 to 2 ground rods 6’ apart. 250.66 A
  2. Installation of one #4 stranded conductor from disconnect #1 to rebar ufer. 250.64D & 250.66
  3. Each sub panel will have a #4 stranded green conductor from the ground bar to each disconnect. 250.102 C1
Click to expand...
The above list doesn't show any GEC going to disconnect #2. So I think you'd need to, say, extend the #4 conductor so it goes from the ufer to disconnect 1 to disconnect 2 (no splices). And then once you do that, the #6 from the ground rods could connect to that #4 anywhere between the ufer and disconnect #1, your choice (doesn't have to be in disconnect #1). Caveat: I didn't check your sizing on the GECs.

Texas appears to be on the 2020 NEC, so FYI the relevant code section is here: https://up.codes/viewer/texas/nfpa-70-2020/chapter/2/wiring-and-protection#250.64_(D) I believe the above would comply with (2) there.

Cheers, Wayne
 
Little Bill said:
You have not said why you are using ground rods when you have a Ufer.
Click to expand...
We don’t usually do new construction, but we took on this customers custom home. I have been looking at new construction in the area and similar large homes all have 320 Meterbase, (2) disconnects, (2) 200 amp panels. All have rebar coming up through the garage by the indoor panels and they are tied into the ground system. Didn't notice ground rods.
 
hotblueelectric said:
We don’t usually do new construction, but we took on this customers custom home. I have been looking at new construction in the area and similar large homes all have 320 Meterbase, (2) disconnects, (2) 200 amp panels. All have rebar coming up through the garage by the indoor panels and they are tied into the ground system. Didn't notice ground rods.
Click to expand...
You didn't notice probably because they aren't required if using a Ufer. If the POCO will allow the GEC in the meterbase, put it there and you will save a ton of time and headache.
 
Status
Not open for further replies.
Top