Grounded conductor size for ATS

electrofelon

electrofelon

Senior Member
Location
Cherry Valley NY, Seattle, WA
Occupation
Electrician
I have a 1600A feeder leaving a switchboard, going thru an ATS, then back to the switchboard. The generator/ATS is NOT an SDS. The generator output breaker is 1200A. The conductors to (and back from) the ATS and switchboard are 6 sets of 500 AL, 2 sets in each pipe. The conductors from the gen to the ATS are 6 total sets of 350 AL, in 3 pipes. Max demand over 3 years is 380 amps. What size grounded conductors should I use?

My thoughts: Its temping to go real small, maybe minimum size. Under normal operation, the neutral to/from the ATS wont be carrying anything. Minimum size per 215.2(A)(2) is 350MCM AL, which would be a 2/0 in each pipe. Also was kinda thinking that (3) 2/0 exceeds the demand figure (assuming perfectly balanced), but not sure that makes any logical sense to compare those two that way. I guess I could say absurd worst case is the 136 KW all on one phase which would be like 1100A This is large facility, im not doing a neutral calc, it would take half a week.
 
electrofelon said:
Under normal operation, the neutral to/from the ATS wont be carrying anything.
Click to expand...

Theoretically, it could be "forced" to carry its normal amount of current, if the imbalance is large. I don't know the proper terminology, but the wires all run together act as a transformer of sorts that would like to remain a zero sum. If you have a worst case, all imbalanced current on one phase, then the neutral would sort of automatically flow to cancel that out.

I would still size it to a minimum size. Whatever minimum size for the whole building is, or could reasonably be calculated or measured per code. The max it could carry would be the total 120 volt imbalance, which is usually not all that much.
 
Isn't the minimum size based on the calculation in 220.61, for the case that the generator is powering the loads? (I.e. so any L-N loads automatically shed when on generator could be ignored). Since when the loads are on grid power, there will be no current on this grounded conductor.

Cheers, Wayne
 
wwhitney said:
Isn't the minimum size based on the calculation in 220.61, for the case that the generator is powering the loads? (I.e. so any L-N loads automatically shed when on generator could be ignored). Since when the loads are on grid power, there will be no current on this grounded conductor.

Cheers, Wayne
Click to expand...
220.87 isnt really clear on if I can use it to size the grounded conductor. Now every utility demand figure I have seen doesnt cover imbalance or neutral current, but 220.87 also doesnt get specific about what parameters are measured.
 
You said under normal operation, the neutral to/from the ATS wont be carrying anything. If there are 0 L-N loads, then leave the grounded (neutral) conductor up at the generator and the switchboard. Save on conductor and just bond the N-G at the generator and skip the neutral all together.
 
electrofelon said:
220.87 isnt really clear on if I can use it to size the grounded conductor.
Click to expand...
Hmm, 220.61 says "The feeder or service neutral load shall be the maximum unbalance of the load determined by this article." Since it says "article" rather than "Part III of this article," that still applies when using a Part IV method. So if you are using 220.87 and have demand data on the grounded conductor current, I see no reason you can't use that to size the grounded conductor.

If you only have demand data for the ungrounded conductors, then obviously you can't use 220.87 for sizing the grounded conductor. You can do a Part III calculation and apply the provisions of 220.61 to that calculation.

Cheers, Wayne
 
ron said:
You said under normal operation, the neutral to/from the ATS wont be carrying anything. If there are 0 L-N loads . . .
Click to expand...
The OP also says the generator is not an SDS, so the ATS is not switching the grounded conductor. So I'm envisioning one conduit between the switchboard and the ATS which contains ungrounded conductors for the utility supply to the ATS, ungrounded conductors for the supply to the load, and a grounded conductor from the generator. That grounded conductor would only carry current while on generator.

Cheers, Wayne
 
wwhitney said:
The OP also says the generator is not an SDS, so the ATS is not switching the grounded conductor. So I'm envisioning one conduit between the switchboard and the ATS which contains ungrounded conductors for the utility supply to the ATS, ungrounded conductors for the supply to the load, and a grounded conductor from the generator. That grounded conductor would only carry current while on generator.

Cheers, Wayne
Click to expand...

Correct there are line to neutral loads but the ATS is more of a "switch loop".
 
electrofelon said:
Correct there are line to neutral loads but the ATS is more of a "switch loop".
Click to expand...
So are you physically putting all the conductors in a single raceway?

If not, and you only have one grounded conductor on the non-generator side of the ATS, you will need to use non-ferrous cables/raceways and a non-ferrous plate for where the cables/raceways enter the ATS and the switchboard. Because while on utility power, the currents on the ungrounded conductors will not balance if there is current on the grounded conductor.

Cheers, Wayne
 
wwhitney said:
So are you physically putting all the conductors in a single raceway?

If not, and you only have one grounded conductor on the non-generator side of the ATS, you will need to use non-ferrous cables/raceways and a non-ferrous plate for where the cables/raceways enter the ATS and the switchboard. Because while on utility power, the currents on the ungrounded conductors will not balance if there is current on the grounded conductor.

Cheers, Wayne
Click to expand...
No, there are three line, and three load conduits, 4" IMC, each with 2 sets of ungrounded conductors. There would be a grounded conductor in each conduit.
 
electrofelon said:
No, there are three line, and three load conduits, 4" IMC, each with 2 sets of ungrounded conductors. There would be a grounded conductor in each conduit.
Click to expand...
Then I don't understand your statement that "Under normal operation, the neutral to/from the ATS wont be carrying anything". In fact each one would be carrying all the neutral current.

You'd be interrupting the neutral conductor that would exist if there were no ATS/generator. You certainly can't have a loop in the grounded conductor (other than within a parallel set).

Cheers, Wayne
 
Last edited:
electrofelon said:
... I guess I could say absurd worst case is the 136 KW all on one phase which would be like 1100A This is large facility, im not doing a neutral calc, it would take half a week.
Click to expand...

I would say you pretty much answered your own question there. What level of wild-guess at the real max imbalance are you comfortable with, without spending half a week on gathering more info?
 
electrofelon said:
Hmmm, no i wasnt planning on interrupting the neutrals in the switchboard :unsure: . Must I?
Click to expand...
Strictly speaking no, imho. But if you do it that way I think you must stick to sets of 7 conductors (or multiples thereof) in each pipe: three line 'to', three line 'from', and the grounded conductor. Well, definitely if the pipes are ferrous, that is.
 
electrofelon said:
Hmmm, no i wasnt planning on interrupting the neutrals in the switchboard :unsure: . Must I?
Click to expand...
What you need to do is ensure that the net current on the conductors within any ferrous conduit, or passing through a hole in a ferrous enclosure, is 0. For parallel sets apparently we get to treat the currents as dividing perfectly, which is why the lengths are supposed to be identical within each set.

If you have a grounded conductor in the existing switchboard, and you connect to it twice without interrupting it, and run grounded conductors both "to" and "from" the ATS, then as the grounded conductor is not switched, those two grounded conductors will be solidly connected within the ATS. That means that the roundtrip to the ATS will be in parallel to the direct path within the switchboard between the two different grounded conductor connections. Which means that while on utility power, almost all of the grounded conductor current will take the direct path through the switchboard. And that means that the currents will not sum to zero for any of the conduits "to" the ATS or"from" the ATS when the utility grounded conductor current is non-zero.

jaggedben said:
Strictly speaking no, imho. But if you do it that way I think you must stick to sets of 7 conductors (or multiples thereof) in each pipe: three line 'to', three line 'from', and the grounded conductor. Well, definitely if the pipes are ferrous, that is.
Click to expand...
Since post #10 calls for 6 conduits total, and 2 sets within each conduit, that works out to no change in the number and size of wires in each conduit as far as the ungrounded conductors, just a change in arrangement of terminations at each end.

Cheers, Wayne
 
As Wayne & Jaggedben covered, unless you use all non ferrous raceways and walls of the enclosures you will need to run the grounded conductors in and out of the transfer switch. Or use 1 really big raceway.
 
electrofelon said:
Ok matching ins and outs in the same pipe makes sense, but if there is a grounded conductor in each pipe, and the bond in the switchboard remains, wont you still have an imbalance? 🤔
Click to expand...
If your (unswitched) grounded conductor is a tree (no loops), and your ungrounded conductors always are in the same pipe as their associated grounded conductor (except for switch loops), then there is never an imbalance.

If you leave the grounded conductor continuous in the switchboard, and run just one pipe (ignoring parallel sets) from the switchboard to the ATS, which has one grounded conductor and the utility side ungrounded conductors and the load side ungrounded conductors, all is well. When on utility power, there will be no current on the grounded conductor, but it will be a switch loop with no net current. When on generator power, only the load set of ungrounded conductors will carry current, and the grounded conductor will carry the unbalance on those back to the generator, so again the net current is zero.

Cheers, Wayne
 
You must log in or register to reply here.
Top