250.122(B) Increasing Grounding Conductor Size

[1972Grady]

I understand that when we increase our ungrounded conductors for voltage drop that we also have to increase our grounding conductor proportionately. I know how to find my ratio between the smaller ungrounded conductors and the increased in size ungrounded conductors. I know how to size my new grounding conductor based off of the ratio the ungrounded conductors increased in size. What I struggle with is this.

If i have a feeder consisting of (11) sets of 750 KCMIL AL feeding from a 4000A OCP to a load that is far enough away from its source to warrant adding (1) more set due to voltage drop. Table 250.122 tells me I need a 750 KCMIL AL ground for a 4000A OCP. With the addition of another set of wires collectively my ungrounded conductors cir mills increased. So would i need to then make all my ground wires 800 KCMIL?

 

[Dennis Alwon]

I don't believe the equipment grounding conductor needs to ever be larger than the circuit conductors in each conduit

 

[1972Grady]

I don't believe the equipment grounding conductor needs to ever be larger than the circuit conductors in each conduit

Indeed the grounding conductor can and must be larger than the ungrounded conductor in some cases. In a parallel run each individual ground must be of sufficient size to open the OCP. So if I had (12) sets of 600 KCMIL feeding from a 4000A OCP then table 250.122 tells me I would ned a 750 KCMIL in each raceway. This would be a case where the grounding conductor would be larger than the ungrounded conductor in the individual raceways but collectivly my ungrounded conductor size is 600 X 12

 

[Dennis Alwon]

Indeed the grounding conductor can and must be larger than the ungrounded conductor in some cases. In a parallel run each individual ground must be of sufficient size to open the OCP. So if I had (12) sets of 600 KCMIL feeding from a 4000A OCP then table 250.122 tells me I would ned a 750 KCMIL in each raceway. This would be a case where the grounding conductor would be larger than the ungrounded conductor in the individual raceways but collectivly my ungrounded conductor size is 600 X 12

So what happens in a line to line short. The conductors are not larger than the equipment grounding conductor req.

 

[Dennis Alwon]

You are correct but I was wondering about the phase to phase

parallel in multiple raceways,
wire-type equipment grounding
conductors, where used, shall be installed in parallel in each
raceway. The equipment grounding conductor installed in
each raceway shall be sized in compliance with 250.122 based
on the overcurrent protective device for the feeder or branch
circuit. Metal raceways or auxiliary gutters in accordance with
250.118 or cable trays complying with 392.60(B) shall be
permitted as the equipment grounding conductor.

 

[jumper]

I understand that when we increase our ungrounded conductors for voltage drop that we also have to increase our grounding conductor proportionatly.

FYI: In the 2014 NEC, the rule was slightly changed. Now the EGC must be increased any time the ungrounded conductors are larger than the minimum size needed, not just for VD.

 

[1972Grady]

So what happens in a line to line short. The conductors are not larger than the equipment grounding conductor req.

In a phase to phase your not dealing with a ground fault. Your dealing with a short circuit. The current is never leaving the ungrounded conductors and returning on the grounding conductor. The breaker is designed with a trip curve to open in a short circuit scenario.

 

[Dennis Alwon]

In a phase to phase your not dealing with a ground fault. Your dealing with a short circuit. The current is never leaving the ungrounded conductors and returning on the grounding conductor. The breaker is designed with a trip curve to open in a short circuit scenario.

If there is too much resistance then the breaker will not trip on a phase to phase either.

 

[Grady_G]

If there is too much resistance then the breaker will not trip on a phase to phase either.

What kind of high restiance phase to phase short are you expecting to have. I guess if your going across a heating element then you would have a high resistance but that’s not the kind of phase to phase short circuit your talking about is it? Can you please describe a phase to phase high resistance short circuit that is in a parallel conduit installation you have in mind?

Maybe a direct burial cable where a hot wire is nicked up and then 100’ away another hot conductor is nicked. Both make contact to earth. The system is grounded so the currents going to try to get back to the source. I’m not sure how much of the current would try to go back to the source by returning on the other phase conductor compared to the current traveling through the dirt looking for the system grounding electrode conductor for its path back to the source. In this case I would tend to agree that no matter how large the ungrounded conductors are or the ECG is would make a difference because the breaker is not going to open. But it would be nice to see other people’s opinions.

 

[Carultch]

FYI: In the 2014 NEC, the rule was slightly changed. Now the EGC must be increased any time the ungrounded conductors are larger than the minimum size needed, not just for VD.

Specifically, "the minimum size that has sufficient ampacity for the intended installation".

So if you increase it in size because you are compensating for bundling or high ambient temperature, you don't need to increase the EGC. But if you are increasing it in size because you are salvaging the next size larger wire that you had left over from a previous job (and have no operational reason to do so), then you do need to increase the EGC. Another non-VD reason would be if a manufacture specifies a larger than necessary wire that fits their terminals, and disallows the NEC minimum.

 

[Carultch]

I understand that when we increase our ungrounded conductors for voltage drop that we also have to increase our grounding conductor proportionately. I know how to find my ratio between the smaller ungrounded conductors and the increased in size ungrounded conductors. I know how to size my new grounding conductor based off of the ratio the ungrounded conductors increased in size. What I struggle with is this.

If i have a feeder consisting of (11) sets of 750 KCMIL AL feeding from a 4000A OCP to a load that is far enough away from its source to warrant adding (1) more set due to voltage drop. Table 250.122 tells me I need a 750 KCMIL AL ground for a 4000A OCP. With the addition of another set of wires collectively my ungrounded conductors cir mills increased. So would i need to then make all my ground wires 800 KCMIL?

One thing the NEC fails to specify, is what the starting point should be/could be for this calculation, when you change the number of parallel sets.

11 sets of 700 kcmil AL could be interpreted as the minimum size (7700 kcmil total), but so could 12 sets of 600 kcmil AL (7200 kcmil total). So which number is the starting point for this calculation? The fail-proof choice is to assume the minimum code-compliant size in the intended quantity of sets.

 

[GoldDigger]

If there is too much resistance then the breaker will not trip on a phase to phase either.

Unfortunately the two ways to mitigate the risk you call out would be to
1. require the size of each conductor in a parallel set to be the same size as the sum of the conductors for that phase in the set, limiting you forever to only one set. Or
2. Have OCPD at the ampacity of the single conductors for each individual member of the set. That causes other problems when one breaker trips and cascades the set offline.

IMHO there is not the same risk to safety associated with a phase to phase short because it will not try to elevate the potential of metal surfaces connected to the EGC. If the EGC in one member of the set were only the size of the hot conductor(s) in that member, then the voltage divider effect would raise the voltage of the EGC to near full source voltage instead of just half the source voltage.

 

[Grady_G]

Unfortunately the two ways to mitigate the risk you call out would be to
1. require the size of each conductor in a parallel set to be the same size as the sum of the conductors for that phase in the set, limiting you forever to only one set. Or
2. Have OCPD at the ampacity of the single conductors for each individual member of the set. That causes other problems when one breaker trips and cascades the set offline.

IMHO there is not the same risk to safety associated with a phase to phase short because it will not try to elevate the potential of metal surfaces connected to the EGC. If the EGC in one member of the set were only the size of the hot conductor(s) in that member, then the voltage divider effect would raise the voltage of the EGC to near full source voltage instead of just half the source voltage.

I would respectfully disagree with booth 1 and 2. A low resistant phase to phase short will have no issues opening the OCP if they are sized correctly. If your talking about a high resistive short then it wont matter how large the phase conductors are the short is not going to open the OCP. Think about a heating element. It is a phase to phase short through a high resistance.

 

[Grady_G]

One thing the NEC fails to specify, is what the starting point should be/could be for this calculation, when you change the number of parallel sets.

11 sets of 700 kcmil AL could be interpreted as the minimum size (7700 kcmil total), but so could 12 sets of 600 kcmil AL (7200 kcmil total). So which number is the starting point for this calculation? The fail-proof choice is to assume the minimum code-compliant size in the intended quantity of sets.

I agree that the proper way to increase for voltage drop is to figure out how many sets you will wind up with. Take that number of sets and find the smallest wire size that will carry the load then upsize from that starting point and this is why.
You can have a single run of 500 KCMIL that might fail in voltage drop so you choose (2) sets of 250 KCMIL. The total circular mills has remained the same, but now you can run the conductors farther than if using 1 set of 500's and thus you will increase the resistance of the ground in a fault situation if you keep the origional size ground. So to keep the low resistance you would certainly need to increase the ground size. Therefore you would have to find the smallest conductors that would size appropriately for the OCP for 2 sets and increase the phase conductors from that starting point up to the 250's to get the ratio you need to solve for the increased ground size.

 

[GoldDigger]

I would respectfully disagree with booth 1 and 2. A low resistant phase to phase short will have no issues opening the OCP if they are sized correctly. If your talking about a high resistive short then it wont matter how large the phase conductors are the short is not going to open the OCP. Think about a heating element. It is a phase to phase short through a high resistance.

I was not arguing that it was a real hazard, but rather that if it ever could be a real hazard there would be no practical way to prevent it!

Sent from my XT1585 using Tapatalk

 

[Grady_G]

I was not arguing that it was a real hazard, but rather that if it ever could be a real hazard there would be no practical way to prevent it!

Sent from my XT1585 using Tapatalk

Perhaps with a GFI breaker. If the high resistive short circuit were to be bleeding of elsewhere as well as returning phase to phase. That's really the best option for prevention. I'm not sure that an arcfault breaker would be of any use if the high resistive short circuit were to be constant.

 

[GoldDigger]

Perhaps with a GFI breaker. If the high resistive short circuit were to be bleeding of elsewhere as well as returning phase to phase. That's really the best option for prevention. I'm not sure that an arcfault breaker would be of any use if the high resistive short circuit were to be constant.

GFI can do a lot for a lot of situations. But it cannot do anything better or worse for a parallel conductor set situation than for a single conductor set situation. This thread got into specifics of what mitigations might be needed for various faults in a parallel set.

 

[Carultch]

I agree that the proper way to increase for voltage drop is to figure out how many sets you will wind up with. Take that number of sets and find the smallest wire size that will carry the load then upsize from that starting point and this is why.
You can have a single run of 500 KCMIL that might fail in voltage drop so you choose (2) sets of 250 KCMIL. The total circular mills has remained the same, but now you can run the conductors farther than if using 1 set of 500's and thus you will increase the resistance of the ground in a fault situation if you keep the origional size ground. So to keep the low resistance you would certainly need to increase the ground size. Therefore you would have to find the smallest conductors that would size appropriately for the OCP for 2 sets and increase the phase conductors from that starting point up to the 250's to get the ratio you need to solve for the increased ground size.

Approximately speaking, the ability to curtail voltage drop is independent of how the total current-carrying KCMIL is divided among the number of sets. There is a slight advantage of 2x250 kcmil over 1x 500 kcmil for voltage drop curtailment, but it is only a 6% difference in metal conduit or a 1% difference in PVC. So 2x250 can't really "go father" by any significant amount compared to 1x500. If you need the conductors to go father and maintain the same voltage drop, you are going to add more KCMIL of conductor.

Ampacity, by contrast, has a significant advantage to paralleling more sets.

 

[kwired]

What kind of high restiance phase to phase short are you expecting to have. I guess if your going across a heating element then you would have a high resistance but that’s not the kind of phase to phase short circuit your talking about is it? Can you please describe a phase to phase high resistance short circuit that is in a parallel conduit installation you have in mind?

Maybe a direct burial cable where a hot wire is nicked up and then 100’ away another hot conductor is nicked. Both make contact to earth. The system is grounded so the currents going to try to get back to the source. I’m not sure how much of the current would try to go back to the source by returning on the other phase conductor compared to the current traveling through the dirt looking for the system grounding electrode conductor for its path back to the source. In this case I would tend to agree that no matter how large the ungrounded conductors are or the ECG is would make a difference because the breaker is not going to open. But it would be nice to see other people’s opinions.

Basic overcurrent protection devices are for protecting the conductors from overheating from higher currents then the conductor was designed for.

A high resistance fault whether ungrounded to ungrounded or to a grounded conductor will not trip overcurrent protection unless perhaps it is low enough resistance it allows more current flow then the normal "load" of the circuit. for such a thing to happen with say a heating element, you would somehow have to shunt across a portion of the element, effectively making an overall lower resistance in the alternate current path. Even AFCI/GFCI won't respond to this unless it results in increased current, as long as there isn't any ground fault current or arc signature that the AFCI is programmed to reject.