13.8 kV Industrial Grounding

P

PWR_ENG

Member
Location
Tomball, TX
Occupation
Electrical Engineer
Hey All,

I have been a long-time reader and user of this forum but have never posted anything. I appreciate any information and feedback that you provide.

Summary:
I am designing an industrial power system for multiple 13.8 kV generators to run in parallel. The grounding scheme currently in place at other locations (not by my design) does not have the equipment ground tied through to the switchgear and downstream loads. The generators have individual ground grids that are bonded together (3 ground rods in a triangle each) and then the switchgear and loads have individual ground grids (same ground rod configuration). The generation GEC and the switchgear GEC are not tied together. That means that any ground fault that occurs would be returning through the earth back to the generators.

Here are some of the high-level details:
- The generators have 13.8kV Wye output and each has a separate high-resistance ground.
- The cable assembly from the generator to the switchgear is a 3-wire, 15kV heavy-duty portable power cable WITHOUT a ground terminated.
- The cable assembly from the switchgear to the loads is a 3-wire, 15kV heavy-duty portable power cable WITH ground terminated.
- The loads each have separate low-resistance grounds.
- Here is the link to a simple drawing that I have made: https://ibb.co/PvV1RhWg

My thoughts:
The big problem that I see here is the path back to the source is unknown and therefore the resistance in the path through earth is unknown. This could affect the generator relay ground pickup elements since we could be adding resistance in the path back to the source. I strongly believe that the switchgear GE's should be tied back to the generation GE's. NEC 250.4(A)(5) clearly states that the earth shall not be considered an effective ground-fault return path, and NEC 250.54 allows auxiliary GE's but references 250.4(A)(5), so I take that as they need to be bonded to the GEC. From a safety perspective, I cannot definitively say this is unsafe since I do not know if the step/touch potential exceeds 50VAC, but I would not be surprised if there was a difference in potential. Mike Holt had a great article that I found in research that talks about how ground rods do not reduce touch potential here.

My questions I have:
I would like to know if anyone has seen something similar to this and can point to other standards that would allow the isolation of the GE's. I have looked through the NEC, IEEE-142, and IEEE-80 and have yet to find an exception to this. Regarding safety, I would like to know your thoughts and experience in situations like this. On the auxiliary grounding for the switchgear and loads, I have not done a lot of multi-grounded schemes before, but I do not see a risk here and would like to know your thoughts on keeping/removing those and making it a single-point grounded system. Again, I believe that the grounding from generation to switchgear is incorrect and should be addressed but would like to use the knowledge base of this forum to shoot holes in my analysis lol.
 
250.190(A) exception?

I would think they were terminated or landed on both sides. But maybe the switchgear side is good enough?
 
Elect117 said:
250.190(A) exception?

I would think they were terminated or landed on both sides. But maybe the switchgear side is good enough?
Click to expand...
I stand corrected, they are terminated on both ends. Just confirmed with some of the field guys. I was not aware that you could use the shield as GEC on impedance-grounded systems. Do you see a benefit in using multi-point grounding in this application?
 
PWR_ENG said:
I was not aware that you could use the shield as GEC on impedance-grounded systems.
Click to expand...

EGC** Equipment grounding conductor. It would not be your GEC (grounding electrode conductor).


PWR_ENG said:
Do you see a benefit in using multi-point grounding in this application?
Click to expand...

It is more of a pros and cons list. It can also be situational. Like planned expansion, existing infrastructure being upgraded in sections or if all of this equipment is in a substation. I am not convinced that your system is multipoint grounded. To me it looks like a single point grounded with earth bonding for voltage stabilization or equipotential purposes.

There are times where you rederive a grounding system and intentionally have a "multi-point" grounded system. For example, separately derived systems and services where the transformer is outdoors.

The bonding on the EGC should remain a effective fault path and the earthing should stabilize voltage. Protection schemes can be used to mitigate potential risks associated with it.
 
1. What is the ground fault current for each of these "high resistance" grounding systems? Normally high resistance grounding at 13.8 kV would involve a neutral side grounding transformer.

2. No utility or transformer connections at 13.8 kV? A solidly-grounded wye winding could greatly increase the fault current.

If it was me, I'd install a full sized GEC from each generator to the switchgear or use a conduit that qualified as a GEC. If these are truly HRG units, the ground fault current from each will be probably less than 5 amps, but at some point in the future, someone could easily make a terrible decision to convert the generators to solidly-grounded. All the generators should be bonded together and to the building grounding system, imo.
And yes I am one of those conservative engineers who wears a belt and suspenders.
 
D. Castor said:
1. What is the ground fault current for each of these "high resistance" grounding systems? Normally high resistance grounding at 13.8 kV would involve a neutral side grounding transformer.

2. No utility or transformer connections at 13.8 kV? A solidly-grounded wye winding could greatly increase the fault current.

If it was me, I'd install a full sized GEC from each generator to the switchgear or use a conduit that qualified as a GEC. If these are truly HRG units, the ground fault current from each will be probably less than 5 amps, but at some point in the future, someone could easily make a terrible decision to convert the generators to solidly-grounded. All the generators should be bonded together and to the building grounding system, imo.
And yes I am one of those conservative engineers who wears a belt and suspenders.
Click to expand...
We are less than 10A and this is a "microgrid" application without a utility interconnect.

I am leaning the same direction as you. Full size EGC would also provide a lower resistance path than the small shielded cables. And there is nothing wrong with conservative engineering or suspenders lol. I am trying to understand the minimum so that I can design above that, but also have rational behind my decision.
 
Elect117 said:
EGC** Equipment grounding conductor. It would not be your GEC (grounding electrode conductor).




It is more of a pros and cons list. It can also be situational. Like planned expansion, existing infrastructure being upgraded in sections or if all of this equipment is in a substation. I am not convinced that your system is multipoint grounded. To me it looks like a single point grounded with earth bonding for voltage stabilization or equipotential purposes.

There are times where you rederive a grounding system and intentionally have a "multi-point" grounded system. For example, separately derived systems and services where the transformer is outdoors.

The bonding on the EGC should remain a effective fault path and the earthing should stabilize voltage. Protection schemes can be used to mitigate potential risks associated with it.
Click to expand...
Thank you for the correction and the explanation. I think the utility would be an example of multi-point grounded system since they drive a ground rod every so often, right?

What I got so far is that there is nothing wrong with the grounding design from the NEC perspective (which is what I was mostly trying to achieve). I I will likely have them terminate the EGC inside of the generator to switchgear cable assembly, but will perform a grounding analysis beforehand.
 
PWR_ENG said:
Thank you for the correction and the explanation. I think the utility would be an example of multi-point grounded system since they drive a ground rod every so often, right?

What I got so far is that there is nothing wrong with the grounding design from the NEC perspective (which is what I was mostly trying to achieve). I I will likely have them terminate the EGC inside of the generator to switchgear cable assembly, but will perform a grounding analysis beforehand.
Click to expand...

I would first perform a fault study from the generators. You most likely already have that. I would then perform the grounding analysis and make adjustments from there.

Multi-point grounding, the way I understand it, is the re-earthing of a neutral conductor at multiple points. Since you don't have a neutral, but instead have an equipment grounding conductor, I would have called it a single point ground that is bonded to earth at more than one location.

Depending on where your equipment is located, you might not want to re-earth the "EGC" or "fault path". You typically only do it once but you can do it more than once depending on the protection scheme and the location of the equipment.

As an example, you re-earth transformers to derive a grounded system and service equipment to create a fault path back to the main on possible electrodes in the building. Though, those would be considered bonded to the electrode system and not exactly re-earthing.

Since you are trying to make a ungrounded system with using HRGs, your real function for earthing, as far as I can tell, is voltage stabilization and ground fault monitoring. You wouldn't necessarily have a trip on the first fault anyway.

But Maybe I am just not understanding it entirely either.
 
You must log in or register to reply here.
Top