NEC 250.122 - Size of EGC for 4,000A feeder

[Jeff Taylor]

NEC Table 250.122 calls for a 500 CU or 750 ALU EGC for a 4,000A feeder. Language in 250.122(A) states the EGC "in no case shall they be required to be smaller than the circuit conductor"

Plans call for the feeder to consist of (12) sets of 600 ALU conductors and a 600 ALU ground. The inspector is saying the ground must be 750 ALU in accordance to the Table. Which is correct?

BTW - I've been in the industry for 35+ years and have used this forum on numerous occasions to solve issues or answer code questions. In this case, I could not find this question, so I had to register and make my first post

 

[jumper]

750 kcmil AL is correct. See 250.122(F).

 

[jaggedben]

NEC Table 250.122 calls for a 500 CU or 750 ALU EGC for a 4,000A feeder. Language in 250.122(A) states the EGC "in no case shall they be required to be larger than the circuit conductor"

Fixing the quote would help.

However, since you have parallel conductors, "larger than the circuit conductor" would mean "larger than all 12 circuit conductors."

See 250.122(F).

 

[Jeff Taylor]

Fixing the quote would help.

However, since you have parallel conductors, "larger than the circuit conductor" would mean "larger than all 12 circuit conductors."

See 250.122(F).

Thanks for the correction - my bad.

 

[infinity]

I agree 750 Kcmil according to T250.122 as the OP stated. Inspector was correct on this one.

 

[petersonra]

Use rigid. no wire type EGC required at all.

 

[Carultch]

Fixing the quote would help.

However, since you have parallel conductors, "larger than the circuit conductor" would mean "larger than all 12 circuit conductors."

See 250.122(F).

This concept could use some clarification. You need multiple EGC's depending on how you distribute it among raceways/wiring methods.

How exactly do you get to "add this up" on both sides of this comparison?

Assign the following letters as such:
N: number of raceways & number of EGCs
n: number of each phase in parallel
S: kcmil size of each individual set
G: kcmil size of ground

Are you saying that:
1*G need not be larger than n*S?
or
N*G need not be larger than n*S?

 

[packersparky]

This concept could use some clarification. You need multiple EGC's depending on how you distribute it among raceways/wiring methods.

How exactly do you get to "add this up" on both sides of this comparison?

Assign the following letters as such:
N: number of raceways & number of EGCs
n: number of each phase in parallel
S: kcmil size of each individual set
G: kcmil size of ground

Are you saying that:
1*G need not be larger than n*S?
or
N*G need not be larger than n*S?

He is saying that in this case the size of the "circuit conductor" is 12 x 600 kcmil which is 7200 kcmil. So the EGC does not have to be larger than 7200 kcmil

 

[infinity]

He is saying that in this case the size of the "circuit conductor" is 12 x 600 kcmil which is 7200 kcmil. So the EGC does not have to be larger than 7200 kcmil

That's it, if there are multiple sets the EGC does not need to be larger than the sum of the sets as in this example 7200 kcmil.

 

[texie]

EGC in parallel

EGC in parallel

The OP's 4000 amp feeder and the odd results it produces for the EGCs is one of the reasons behind the proposal to totally revamp the way EGCs are sized for the 2020 edition. I think we all agree that under the present NEC the correct answer is 500 CU or 750 AL. To add insult to injury, if there are service conductors on the line side of the 4000 amp OCPD in the OP's application, the supply side bonding (if needed) would only need to be a 1/0 in each raceway. Talk about a lopsided result!

 

[Carultch]

That's it, if there are multiple sets the EGC does not need to be larger than the sum of the sets as in this example 7200 kcmil.

To add clarity, I take it what you mean is:
That's it, if there are multiple sets, each individual EGC does not need to be larger than the sum of the sets as in this example 7200 kcmil.

Also, how exactly does the language of 250.122(A) specify this? "In no case shall the EGC be required to be larger than the circuit conductors supplying the equipment" does not say "In no case shall each individual EGC be required to be larger than the combined cross sectional area of the circuit conductors supplying each phase/polarity of the equipment."

 

[infinity]

1)To add clarity, I take it what you mean is:
That's it, if there are multiple sets, each individual EGC does not need to be larger than the sum of the sets as in this example 7200 kcmil.

2)Also, how exactly does the language of 250.122(A) specify this? "In no case shall the EGC be required to be larger than the circuit conductors supplying the equipment" does not say "In no case shall each individual EGC be required to be larger than the combined cross sectional area of the circuit conductors supplying each phase/polarity of the equipment."

1) Yes and no. I stated "if there are multiple sets the EGC does not need to be larger than the sum of the sets as in this example 7200 kcmil." It could be parallel conductors in a single raceway such as a wireway with only one EGC. For multiple raceways you're correct by adding the word individual.

2) It's one circuit, IMO if there are parallel conductors the combined cross sectional area is already implied.

 

[Wire-Smith]

I don't think i'm understanding you guys completely.

I used to think you weren't allowed to parallel EGC's even over 1/0 (except in cable assemblies). but i don't believe that anymore, i believe you are allowed to parallel EGC'c 310.10(H)(5)

i know 250.122(F)
i don't think it is saying each wire has to be sized by that. 1/0 and larger can be paralleled.

it says size off 122 which you are, just paralleling it like you do your ungrounded conductors. otherwise why does 310.10(H) say smaller than 1/0 may be paralleled in cables, you wouldn't have to say that if you weren't allowed to parallel above 1/0 outside of cables, you would just say cable grounds can be segmented.

this is one item i have been wanting to send in a proposal on to clarify this or myself be corrected.
am i missing something?

Thanks

 

[infinity]

You cannot parallel two smaller conductors to make a bigger EGC. Look at table 250.122 those are the minimum EGC sizes and there are no provisions for paralleling. The last sentence in 250.122(F) takes you back to the table 250.122.

 

[jaggedben]

...

it says size off 122 which you are, just paralleling it like you do your ungrounded conductors. ..

...

Which you do how? By calculating ampacities through 310.15? Seems like a stretch given that to my knowledge there is nowhere in the code where ampacities are used to determine EGC size. Everything refers just to size. You're reading a lot into the code that it doesn't explicitly say.

 

[Wire-Smith]

You cannot parallel two smaller conductors to make a bigger EGC. Look at table 250.122 those are the minimum EGC sizes and there are no provisions for paralleling. The last sentence in 250.122(F) takes you back to the table 250.122.

i don't see where it says that, although like i stated i used to believe the way you do so i do understand why someone would think that, but after studying it i believe you are for 1/0 and over just like ungrounded conductors. its referring you to 250.122 because you do have to size off that, but your still allowed to parallel like with ungrounded with minimum of 1/0 being used (besides in cables).

310.10(H)(5)


why does 310.10(H)(5) say smaller than 1/0 may be paralleled in cables, you wouldn't have to say that if you weren't allowed to parallel above 1/0 outside of cables, you would just say cable grounds can be segmented.


size in accordance with 250.122

250.122(F)(1)(b)

i don't see it saying you cannot parallel grounds, i believe it is misinterpreted because it refers you to 250.122, which it would have to even if it explicitly said something like "you can use multiple wires to add up to ampacity of what would be required if not paralleled"

im looking at 2017 by the way, f was changed.

 

[Wire-Smith]

250.122(F)(1)(a)
"a single wire-type shall be permitted" it does not say it has to be one full sized ground. if you weren't allowed to parallel them then using the word "single" is very odd and does not make sense. and it says "permitted" not required

 

[Wire-Smith]

Which you do how? By calculating ampacities through 310.15? Seems like a stretch given that to my knowledge there is nowhere in the code where ampacities are used to determine EGC size. Everything refers just to size. You're reading a lot into the code that it doesn't explicitly say.

exactly, if 750mcm is required then look at ampacity of 750mcm in 310 and select your paralleled grounds by that ampacity.

 

[infinity]

i don't see where it says that, although like i stated i used to believe the way you do so i do understand why someone would think that, but after studying it i believe you are for 1/0 and over just like ungrounded conductors. its referring you to 250.122 because you do have to size off that, but your still allowed to parallel like with ungrounded with minimum of 1/0 being used (besides in cables).

310.10(H)(5)


why does 310.10(H)(5) say smaller than 1/0 may be paralleled in cables, you wouldn't have to say that if you weren't allowed to parallel above 1/0 outside of cables, you would just say cable grounds can be segmented.


size in accordance with 250.122

250.122(F)(1)(b)

i don't see it saying you cannot parallel grounds, i believe it is misinterpreted because it refers you to 250.122, which it would have to even if it explicitly said something like "you can use multiple wires to add up to ampacity of what would be required if not paralleled"

im looking at 2017 by the way, f was changed.

You're not reading 310.10(H) correctly. Look specifically at (1). No EGC's on that list.

(H) Conductors in Parallel.
(1) General. Aluminum, copper-clad aluminum, or copper
conductors, for each phase, polarity, neutral, or grounded
circuit shall be permitted to be connected in parallel (elec-
trically joined at both ends) only in sizes 1/0 AWG and
larger where installed in accordance with 310.10(H)(2)
through (H)(6).

 

[Wire-Smith]

You're not reading 310.10(H) correctly. Look specifically at (1). No EGC's on that list.

(H)(5) EGC

(H)(1) neutral or grounded circuit