3-pole 3-wire trans. switch with not a separately derived system

[elec_eng]

The service is bonded, therefore the current flows on the service neutral to an(any)(all) ATS(s) where it transfers to the Gen neutral. (Note, the fault current will take a slightlyaltered path than the normal current, but it will get there...)

From the gen neutral thru the gen coils to the breaker - wala - complete circuit.

I didn't think of that before..I think you are right..you can complete the circuit that way!!!

One question though...how can you be sure that the neutral to other ATS is sized adquately to handle the fault current?

ATS to MCC is 1600A but ATS that has neutral is only 200A-400A.

 

[elec_eng]

Do the transfer switches share a common utility transfomer? If yes, then if the neutral is solid to one transfer switch (and back to the utility transformer), then I don't see a problem.

If there are separate utility transformers, or if there are step down transformers before the transfer switches, then it gets more complicated.

Steve

No..this hospital is fed from (3) 13.8KV lines and there are mutiple substaion with main tie.

Also, there are a lot of step down transformers before the transfer switches as i mentioned in the post #8.

 

[elec_eng]

First off, let's clarify the multi-generator, parallelling switchgear scheme...

Are all generators single output and parallelled under a service outage condition?

Yes.

Another thing you should check is that the neutrals are adequately sized out to the switchgear and on to the generators...

I can see that now there is another path to complete the circuit as wasasparky mentioned...but how can you be sure the neutrals at other ATS(s) are adequately sized to carry the fault current back to generator?

 

[elec_eng]

No..this hospital is fed from (3) 13.8KV lines and there are mutiple substaion with main tie.

Also, there are a lot of step down transformers before the transfer switches as i mentioned in the post #8.

[correction: Actually, the step down transformers are not before the ATSs..ATSs are feeding the step down ATSs)

 

[Smart $]

...but how can you be sure the neutrals at other ATS(s) are adequately sized to carry the fault current back to generator?

Good question...and there is not clear cut answer, as the NEC contains very little mention of non-separately-derived, alternate-power-source systems. I have a feeling there are no explicit requirements to cover your installation in this regard.

First stop, 445.13 says the ampacity of [generator] conductors which must carry ground-fault current shall not be smaller than required by 250.24(C) per 2005 NEC or 250.30(A) per 2008 NEC.

Going by the 2005 NEC is quite easy for the first part. Essentially you'd have to run conductors no smaller sized than per Table 250.66 (GEC sizing) for each generator to the switchgear's neutral bus. However, this leaves us unknowing of the size from the switchgear to the ATS(s)

Going by the 2008 NEC leaves me a bit perplexed because it throws one back to grounding of separately derived systems and one has to sort out which of the requirements is the neutral conductor size to be based on. My quess is 250.30(A)(4)(b) for each generator to the switchgear neutral bus (also per Table 250.66 GEC sizing), and 250.30(A)(4)(a), no smaller than 3/0 Cu or 250 kcmil Al, from switchgear to ATS(s) and from there back to the service bond point.

...but I'm not certain, so I'm open to other viewpoints

 

[elec_eng]

Good question...and there is not clear cut answer, as the NEC contains very little mention of non-separately-derived, alternate-power-source systems. I have a feeling there are no explicit requirements to cover your installation in this regard.

First stop, 445.13 says the ampacity of [generator] conductors which must carry ground-fault current shall not be smaller than required by 250.24(C) per 2005 NEC or 250.30(A) per 2008 NEC.

Going by the 2005 NEC is quite easy for the first part. Essentially you'd have to run conductors no smaller sized than per Table 250.66 (GEC sizing) for each generator to the switchgear's neutral bus. However, this leaves us unknowing of the size from the switchgear to the ATS(s)

Going by the 2008 NEC leaves me a bit perplexed because it throws one back to grounding of separately derived systems and one has to sort out which of the requirements is the neutral conductor size to be based on. My quess is 250.30(A)(4)(b) for each generator to the switchgear neutral bus (also per Table 250.66 GEC sizing), and 250.30(A)(4)(a), no smaller than 3/0 Cu or 250 kcmil Al, from switchgear to ATS(s) and from there back to the service bond point.

...but I'm not certain, so I'm open to other viewpoints

I think you are right..445.13 & 250.30(A) only address the conductor size from the generator to the switchgear neutral bus...Since the nuetral from the switchgear to ATS(s) also has to carry the fault current, shouldn't that be sized per 250.122?

 

[Smart $]

I think you are right..445.13 & 250.30(A) only address the conductor size from the generator to the switchgear neutral bus...Since the nuetral from the switchgear to ATS(s) also has to carry the fault current, shouldn't that be sized per 250.122?

...as long as the size is 3/0 Cu or 250 kcmil Al or larger per 250.30(A)(4)(a). Read on...

One issue is 250.122 only requires the size to be based on the OCPD protecting the conductors run with the fault-current conductor—the neutral(s) in this case. In an earlier statement you said the ATS feeders with neutrals were 200-400A. Sizing the neutral per 250.122 for even the 400A OCPD would still be too small for fault-current on the say the 1600A feeder circuit that you mentioned.

IMO, you should size per 250.122 based on the feeder leaving the parallelling switchgear having the highest rated OCPD, even though the neutral(s) do not run with those feeders. There appears to be no direct requirement to do so. The only hint is the Note to Table 250.122, which takes us to 250.4(A)(5)