EGC with High Resistance Ground

[beanland]

2500kVA 480Y/277V industrial transformer with high-resistance ground (HRG) (neutral grounding resistor NGR) that limits ground fault current to 10A. For 1600AF/1200AT OCPD, I am planning on 3 sets of 600kcmil copper to support the 1000A maximum loading with no neutral plus EGC in each 3" conduit. With 1200A OCPD, 250.122 would require a 3/0 copper EGC in each conduit. But, with HRG there is never more than 10A total ground fault current. Is there an NEC section that addresses sizing the EGC with HRG?

 

[winnie]

Okay, I am really not supposed to be posting right now....

The HRG system will limit any _single phase_ fault to ground, but does nothing to limit the current is a second phase also faults to ground. The EGC thus needs to be sized to open OCPD in the event of multiple faults.

HRG systems are generally used for continuity of service, meaning that you _don't_ open the OCPD in the event of a single fault. Thus there is a period of time when a second distant fault could occur put lots of current on the EGC system.

-Jon

 

[publicgood]

This is not an EGC. It is an equipment bonding jumper. The size depends on the location of the GEC for the system. See 250.36. Specially, 250.36(G). It will be minimum #8CU/#6AL or otherwise sized according to 250.66.

Furthermore, there must be an equipment bonding jumper in each conduit of the parallel set not smaller than 1/0. 250.24(C)(2) is one of many example references. Arguably the total area needs to be minimum 1/0 (or by factors noted above at the smallest), not a 1/0 in each conduit.

 

[Smart $]

.... But, with HRG there is never more than 10A total ground fault current. Is there an NEC section that addresses sizing the EGC with HRG?

Publicgood cited the appropriate section, 250.36(G).

Essentially you do not get to downsize even though the HRG limits ground fault current to 10A. For argument's sake, consider an ungrounded system. Also no downsizing on the grounding conductors. I believe the reasoning is just in case there are multiple line-to-ground faults... which amount to line-to-line faults. (Actually just read Jon's post in its entirety and I see he said the same thing. :blink

 

[Smart $]

Okay, I am really not supposed to be posting right now...

???

Date or significant other issues? :huh:

 

[Smart $]

...
Furthermore, there must be an equipment bonding jumper in each conduit of the parallel set not smaller than 1/0. 250.24(C)(2) is one of many example references. Arguably the total area needs to be minimum 1/0 (or by factors noted above at the smallest), not a 1/0 in each conduit.

It is debatable whether or not the 1/0 AWG minimum applies to these. See 310.10(H)(5) & (6). Of the two, which applies? One sends us to 250.122, the other to 250.102.

 

[publicgood]

It is debatable whether or not the 1/0 AWG minimum applies to these. See 310.10(H)(5) & (6). Of the two, which applies? One sends us to 250.122, the other to 250.102.

As you describe would be the case for a regular feeder, but this is for a separately derived AC system. 250.30(A)(3)(b)

 

[publicgood]

As you describe would be the case for a regular feeder, but this is for a separately derived AC system. 250.30(A)(3)(b)

I take back my "typical feeder" comment back, but the NEC reference holds. I see there is an informational reference to your noted 310. The code may conflict itself here. Otherwise, there is the 250.102 12.5% note that should be mentioned here.

 

[Smart $]

As you describe would be the case for a regular feeder, but this is for a separately derived AC system. 250.30(A)(3)(b)

I thought we were talking per se about the HRG system "equivalent" to a supply-side bonding jumper. 250.30(A)(3) is about the grounded conductor. 250.30(A)(2) sends us to 250.102(C) for sizing... but I believe 250.36 is intended to completely supersede 250.30 but offers us no direction when sizing the EBJ for parallel conductors in multiple raceways. 250.102 when dealing with parallel conductors in two or more raceways says an SSBJ only need be sized based on the conductors in the raceway and not the whole lot en masse.

 

[Smart $]

I take back my "typical feeder" comment back, but the NEC reference holds. I see there is an informational reference to your noted 310. The code may conflict itself here. Otherwise, there is the 250.102 12.5% note that should be mentioned here.

250.36(G) tells us how to size the EBJ in general. The only question I see is whether the EBJ in each raceway must meet the 1/0 AWG minimum placed on other grounding conductors in a parallel [circuit] conductors, two or more raceways situation.

 

[publicgood]

I reviewed noted references. I agree now that 250.30(A) exception points us to 250.36.

We still don't know 250.36(G)(1) or (2). These would provide minimum size and not have us looking anywhere else in the code. 250.66 or 250.36(B). You are right, no direction for if parallel raceways...which spawns further discussion.

If we install in parallel...

Your 310.10(H)(5) makes a point to say it is okay for the equipment bonding jumper to be less than 1/0 for multiconductor cables, even though it points you to 250.122. Why bother mentioning about the 1/0 size? 310.10(H)(1) wins - 1/0 is the minimum size.

We are in raceway, so 310.10(H)(6) is the right reference pointing us to 250.102 to see if the table or 12.5% drive us larger than 1/0.

The only question for me is: Do we have to be in parallel just because our ungrounded conductors are? Likely we would identify a requirement if we had a grounded conductor, but maybe not for the ground conductor?

The statement in 250.102(B)(2) has me wondering..."A single supply-side bonding jumper installed for bonding two or more raceways or cables shall be sized in accordance with 250.102(C)(1)."

 

[beanland]

HRG is monitored

HRG is monitored

I agree that if there is one phase-ground fault then the HRG will limit fault current. If there is a second phase-ground fault involving a different phase, then the fault current could go through the EGC. However, the HRG is monitored in an industrial application so that once the first fault occurs, there is an alarm. The HRG is meant to reduce the damage caused by a phase-ground fault.

Also, I am not concerned about the size of the bonding jumper from a metal enclosure to the EGC but rather the size of the EGC used in parallel PVC underground conduits on a feeder from the switchgear to the MCC. See attached.

In this case there is a 480Y/277V transformer with 10A HRG as part of a unit substation located several hundred feet from the MCC. There are three or four parallel PVC conduits from the unit substation switchgear to the MCC. Since I have a 1200A trip breaker, am I required to install a 3/0 copper EGC in each PVC conduit? Reality is that there is no fault that will ever require the rating of the 3/0 parallel conductors to be needed.

The argument that if there is a 2nd phase-ground fault in the MCC that the EGC back to the unit substation comes into play does not work because the fault would become phase-phase in the MCC and there would not be high current back to the unit substation.

This is a great learning process and I appreciate all the input.