Isolated Ground Size When Fed From Separately Derived System

[jrohe]

I know, I know; everyone has their opinion on isolated ground systems. Unfortunately, isolated ground systems are sometimes demanded by a client. Putting opinions aside, I am hoping you can help me settle a debate I am having with another engineer in my office.

See the attached file for a very simplified sketch I would like to use for this exercise. Also, lets assume:

1. Everything is completely code compliant with the only question centering around how to size the isolated ground.
2. All connections from the transformer to the GEC are made at the transformer, not the first disconnecting means.

My contention is that an IG is clearly considered by the NEC as an EGC because of where it falls within the NEC (under Part VII of Article 250). As such, the IG conductor would be sized based on T250.122 based on the rating of the 200 amp OCPD protecting the IG panel (#6 AWG CU). This IG would be routed with the SSBJ and neutral to the transformer where it is also bonded with X0, the SBJ, and the GEC.

His contention is that an IG conductor sized per T250.122 is too small to be protected from a ground fault by the primary OCPD and that the IG conductor should be the same size as the SSBJ, which is sized per T250.102(C)(1).

While his argument has merit, I cannot get around the fact that the NEC permits (not requires) IG conductor(s) to pass through panelboards, boxes, etc. all the way back to the source without making connections within the enclosures. Therefore, if the IG panel is full of 20 amp circuits, we would be permitted to run a separate #12 AWG IG conductor from each 20 amp circuit back to the transformer grounding connection point without any intermediate connections. We would never do this but the NEC would permit it.

I am hoping someone will be able to solve this debate. Thank you all in advance for your expertise!

Jason Rohe, P.E.

 

[infinity]

His contention is that an IG conductor sized per T250.122 is too small to be protected from a ground fault by the primary OCPD and that the IG conductor should be the same size as the SSBJ, which is sized per T250.102(C)(1).

Why would the IG even come into play when opening the primary OCPD? The SSBJ already has that covered. The IG is for the circuits that come after the OCPD in the panel.

 

[ron]

This graphic is from the NEC Handbook and refers to the condition when there is a non-metallic connection of the raceway to the box, but they refer to the conductors as an isolated EGC. We would size them based on 250.122 but there needs to be a SSBJ sized based on 250.102.

 

[infinity]

I think of it this way, if the IG begins at the panel it is sized according to 250.122 as Ron stated. If you extend it further upstream to the transformer its use does not change it's still an IG and not a SSBJ. That SSBJ would be a completely separate conductor, sized according to derived condcutors and T250.102(C).

 

[jrohe]

Why would the IG even come into play when opening the primary OCPD? The SSBJ already has that covered. The IG is for the circuits that come after the OCPD in the panel.

If there were a fault between a phase conductor and the IG between the transformer secondary and the first disconnecting means, the IG would play a role in carrying fault current.

 

[infinity]

If there were a fault between a phase conductor and the IG between the transformer secondary and the first disconnecting means, the IG would play a role in carrying fault current.

Good point, but what code section would require the IG to be sized like a SSBJ to carry that fault current?

 

[jrohe]

Good point, but what code section would require the IG to be sized like a SSBJ to carry that fault current?

That's where I am coming from. My counterpart disagrees, although there are no code requirements that back it up, electrical theory supports his point. This is what caused me to ask this group. He is proposing running two SSBJs, one from the bonded equipment ground bar and one from the isolated ground bar. What he is proposing is definitely feasible, but it seems pointless to me without code basis.

 

[infinity]

That's where I am coming from. My counterpart disagrees, although there are no code requirements that back it up, electrical theory supports his point. This is what caused me to ask this group. He is proposing running two SSBJs, one from the bonded equipment ground bar and one from the isolated ground bar. What he is proposing is definitely feasible, but it seems pointless to me without code basis.

Since there is no code basis as you've stated then it's merely a design issue. Design conflicts happen all of the time, ultimately in the end it's the guy who's paying for something that he didn't need that get hosed.

 

[Tulsa Electrician]

This is the one time I would move the SBJ and GEC to the panle in lieu of the transformer. Then it no longer becomes an issue of consideration.