Riser sizing for ungrounded systems or neutral grounded systems

[qman5]

Hello all,

In the case of ungrounded system, where the SLG fault current is sensibly 0, or in the case of a NGR, limited by the NGR current, do you typically size the grounding pigtail risers for this limited current, or the maximum fault current (e.g. LLL, LLL)? Consider an ungrounded delta system inside a substation.

Would a scenario like LL-G not bypass the NGR, but still, would we see an appreciable fault current in the riser (assuming flashover from LL to ground)? I don't see this.

Is there any scenario where the riser could see more than the ~0A or NGR-limited current? As it stands, I would presently size the riser for 500 MCM going by IEEE 80, but I really question if it's necessary to have such a cross sectional area, of if less can be argued. The cautious approach would be to go with the full 500MCM, but may be unoptimized.

 

[waross]

Refer to section 10 of the Canadian Electrical Code(CEC).

I really question if it's necessary to have such a cross sectional area, of if less can be argued.

If you are subject to an AHJ, he will settle the argument based on the CEC.
Even if your are not subject to the CEC, then compliance with the section on grounding is a good Cover Your ASSets approach.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[wcaseyharman]

Perhaps this might be useful to you:
All of our large generators are high resistance grounded. Regardless, the generator enclosure and everything else in the plant is grounded to the ground grid by two 4/0 grounds minimum. This includes plants built in the last 60 years by Black and Veatch, stone and Webster, GE, Westinghouse, Patch, to name a few of the engineering firms involved.

 

[cuky2000]

In general, cable size for grounding risers should be designed to withstand the maximum full current during the clearing time of the protective device without fusing, annealing the cable or creating a fire hazard.
Determine the above parameters requires a good modeling technique considering that the maximum fault current is associated with the zero sequence impedance including the return path to the source either through a metallic, ground return or a combination of both requiring to calculate a current division and other factors.
If there is not resource available to modeling the system, a common conservative approach is over-sizing the riser using the maximum fault current LLL in the understanding that current LG<LLL or sometime using Cu 4/0 AWG if recommended by the utility checking that against the assumed fault for 0.5 sec for substations or 3 second for MV and smaller applications.
It should be noted that even an ungrounded delta system, the fault current is >0 since the fault current travel back to the source using the earth or other paths.
Check the IEEE STD 80 section 15.3 and figs 28 to 31 for additional details.