OCDP Size And Extreme Circuit Length

It is a common issue on the utility side. High resistance return makes the fault look like load but the fault location with available energy is still enough to hurt or cause fire.

It typically requires zones of protection and monitoring of voltage sags or phase shifts to catch.

The return path is really how you are going to catch it. Whether that is a wire type EGC, conduit, neutral, bonded metal piping, etc.

I don't know if what Jim is saying relieves me of the the possibility of the long line impedance still being capable of harm at the fault location since it will be trying to get on it's return path.

Like a PVC run with wire type EGC vs RMC for 2000'. Does the pipe being a high impedance return create a difference in potential and capable of shock?

I guess it is two issues at once. The wire at the fault end might arc but not be destroyed but could the fault location (if somewhere between load and breaker) have enough available energy / difference in potential to shock someone touching the RMC? And should the breaker be able to protect against that.

I agree with the point that the breaker will protect the wire but is that enough for the NEC?
 
Elect117 said:
I agree with the point that the breaker will protect the wire but is that enough for the NEC?
Click to expand...
Yes, that is all the NEC cares about, except for faults to ground in some situations.
Read the definition of Overcurrent.

It is your design choice to add additional protection devices or functions. Some type of ground fault protection is probably a good idea if not already required by the NEC.
 
jim dungar said:
Yes, that is all the NEC cares about, except for faults to ground in some situations.
Click to expand...
110.7 to me reads like a catch all performance code requirement that would cover anything the prescriptive code requirements in 250 left out.
wwhitney said:
Ground fault is defined in the NEC, and if the 15A fault current is due to an ungrounded circuit conductor faulting to the EGC, that is definitely a ground fault, even when the conductors have 20A ampacity.
Click to expand...
Say the other end of this 2000' long #12 circuit energizes a some exposed metal say a lamp post, conduit or manhole cover at a bus stop, would that not result in a potentially fatal situation?
 
tortuga said:
110.7 to me reads like a catch all performance code requirement that would cover anything the prescriptive code requirements in 250 left out.
Click to expand...
110.7 is about the installer, it has nothing to do with faults that occur after the power is turned on.
 
tortuga said:
Say the other end of this 2000' long #12 circuit energizes a some exposed metal say a lamp post, conduit or manhole cover at a bus stop, would that not result in a potentially fatal situation?
Click to expand...
If the only fault path is through earth, a non-GFCI OCPD won't trip either way. And the presence of an extra 4 ohms of resistance in the fault path won't affect the operation of a GFCI.

If the fault path doesn't involve earth, then the fault is line to neutral (both conductors fault to the same metal object) or line to EGC (which could occur with an ungrounded conductor faulting to a bonded metal object). The OCPD won't trip in the 2000' long case we've been discussing This does seem like it could be a hazard, as that would put the faulted metal object at 60V to ground (at the originating panelboard), via the voltage divider of the two 4 ohm impedances of 2000' of #12 Cu. Which is above the common 50V threshold for considering voltages to be hazardous.

Cheers, Wayne
 
We don't know what the load actually is, but if it includes something like a grounded metal pole, or a grounded metal enclosure, it looks like 250.4A (3), (4), and (5) apply.

IMO, these would require either:
1. A larger ECG that reduces the enclosure voltage to below 50V, or
2. Either a larger EGC, or a smaller OCP so the OCP will trip during a ground fault at the far end.

Would an additional grounding electrode at the load also satisfy this requirement? Seems like that would limit any possible touch voltages ....

2000' of wire would have a lot of leakage, so a GFCI probably wouldn't hold.
 
steve66 said:
We don't know what the load actually is, but if it includes something like a grounded metal pole, or a grounded metal enclosure, it looks like 250.4A (3), (4), and (5) apply.

IMO, these would require either:
1. A larger ECG that reduces the enclosure voltage to below 50V, or
2. Either a larger EGC, or a smaller OCP so the OCP will trip during a ground fault at the far end.

Would an additional grounding electrode at the load also satisfy this requirement? Seems like that would limit any possible touch voltages ....

2000' of wire would have a lot of leakage, so a GFCI probably wouldn't hold.
Click to expand...
A GFCI at the load end would be a solution for a ground fault, of course, but do nothing for over current.
 
  • Like
Reactions: J2H
steve66 said:
We don't know what the load actually is, but if it includes something like a grounded metal pole, or a grounded metal enclosure, it looks like 250.4A (3), (4), and (5) apply.

IMO, these would require either:
1. A larger ECG that reduces the enclosure voltage to below 50V, or
2. Either a larger EGC, or a smaller OCP so the OCP will trip during a ground fault at the far end.

Would an additional grounding electrode at the load also satisfy this requirement? Seems like that would limit any possible touch voltages ....

2000' of wire would have a lot of leakage, so a GFCI probably wouldn't hold.
Click to expand...
All that would do is to energize a small area of earth around the grounding electrode. If you are standing right on top of the electrode there would be a reduction in the touch voltage...more than a few feet away from the electrode there would be no change in the touch voltage.
 
Dennis Alwon said:
2023 Nec
Short Circuit.


An abnormal connection (including an arc) of relatively low impedance, whether made accidentally or intentionally, between two or more points of different potential. (CMP-10)
Click to expand...
If this definition is now in the NEC, I will change my position and say the condition in the OP needs to be protected against, even though it does not represent an overcurrent condition as I have been arguing.

Smaller OCPD or larger conductors are the only reasonable options.
 
jim dungar said:
No.
Ground/earth/dirt is not a good conductor at low voltages, especially 120V, so it could not be counted on to carry enough current to open a standard OCPD.
Click to expand...
Yes, the earth is obviously going to be a much worse conductor than the #12 ground wire. But a ground rod would reduce the touch voltage to the surrounding earth. So I was thinking a ground electrode might eliminate the "likely to become energized" (at least with regard to surrounding earth), and thus tripping the OCP might not be required in that case.

Again, we don't know what the load is, but if its any kind of remote structure, isn't a grounding electrode required anyway?

At any rate, I still believe the wire needs to be increased in size, or the OCP needs to be decreased so the OCPD will trip on a remote short. Isn't that basically what 250.4(A)(5) says?
 
steve66 said:
But a ground rod would reduce the touch voltage to the surrounding earth. So I was thinking a ground electrode might eliminate the "likely to become energized" (at least with regard to surrounding earth), and thus tripping the OCP might not be required in that case.
Click to expand...
Ground rods eliminate very few hazards at 120V. This is why the NEC requires multiple ground rods to be connected with an actual wire.
 
steve66 said:
Yes, the earth is obviously going to be a much worse conductor than the #12 ground wire. But a ground rod would reduce the touch voltage to the surrounding earth. So I was thinking a ground electrode might eliminate the "likely to become energized" (at least with regard to surrounding earth), and thus tripping the OCP might not be required in that case.

Again, we don't know what the load is, but if its any kind of remote structure, isn't a grounding electrode required anyway?

At any rate, I still believe the wire needs to be increased in size, or the OCP needs to be decreased so the OCPD will trip on a remote short. Isn't that basically what 250.4(A)(5) says?
Click to expand...
Where in the time/current curve do you want the OCPD to open? Increase the conductor until you get the results you want. You will need to coordinate with the device you pick.
How long is too long?
 
jim dungar said:
If this definition is now in the NEC, I will change my position and say the condition in the OP needs to be protected against, even though it does not represent an overcurrent condition as I have been arguing.
Click to expand...
So to be pedantic, what NEC section requires protecting against a non-overcurrent short circuit?

Thanks,
Wayne
 
ptonsparky said:
How can it be a short circuit if there is no over current? It's just a load.
Click to expand...
Uh, see post #28. It's an abnormal connection. If you intend the connection, then it's not a short-circuit. Otherwise, it is.

In other words, the definition of "short-circuit" doesn't depend on the source impedance. A short circuit on a 100' long branch circuit doesn't stop being a short circuit when you extend the branch circuit length to 2000'.

Cheers, Wayne
 
wwhitney said:
Uh, see post #28. It's an abnormal connection. If you intend the connection, then it's not a short-circuit. Otherwise, it is.

In other words, the definition of "short-circuit" doesn't depend on the source impedance. A short circuit on a 100' long branch circuit doesn't stop being a short circuit when you extend the branch circuit length to 2000'.

Cheers, Wayne
Click to expand...

Then it becomes a "Long-Circuit" 😅 ;)
 
wwhitney said:
So to be pedantic, what NEC section requires protecting against a non-overcurrent short circuit?
Click to expand...

For those locations on the NEC 2023, short circuits are defined by voltage not by amount of current.
We have not adopted the 2023 edition in Wisconsin so my original position was a valid response.
 
You must log in or register to reply here.
Top