250.122(B) requires EGCs to be increased proportionally when circuit conductors are increased for voltage drop.
So, although a #10 EGC is good enough for up to a 60 amp ckt., per Table 250.122, if you run a particularly long, 20-amp ckt. and increase the ungrounded conductors to, say #8, you will need to increase the EGC to #8 also.
Any comments, please?

You summed it up nicely.

I'm wondering why a #10 ground is good enough for a 60-amp ckt with #6 circuit conductors, but not good enough for a 20-amp ckt. with #8 circuit conductors.

Effective Ground Fault Current Path

PCB,

If #10 is effective for a 60 amp OCPD, why not for a 20 amp OCPD. I don't believe that the increased resistance due to length will inhibit the operation of the CB.
And if the EGC and circuit conductors need to be matched in size for some concept related to balanced, or equal impedances between circuit conductors and EGCs, why can the EGC be smaller than the circuit conductors in circuits above 30-amp?

Redsy,
How would you write the rule? The EGC does have to increase with the size of the ungrounded conductors, and I don't know of a way to write a rule to accomplish this other than how it is currently written. If you do, it is not too late to submit an online proposal .
Don

Thanks, guys.

I am just wondering if there is some kind of theoretical concept that I am overlooking?

From a theoretical point of view you want the impedance of the PCBelarge's effective ground fault current path to be low enough to trip the breaker within a reasonable time.

If the 20A breaker is designed to be very forgiving to momentary surges in current while the 60A breaker so to speak is "trigger happy", you could get this counterintuitive result. The breaker manufacturers or their websites can no doubt tell.

[This message has been edited by C-H (edited 11-02-2005).]

[This message has been edited by C-H (edited 11-02-2005).]

I understand the concept but the reality is 2000 feet of #12 would still trip a 20a breaker with a bolted fault, even if it takes 150% (30a) to operate it.

gfretwell: Yes, the breaker would trip. But in the meantime, the voltage at the point of fault would be higher than if the EGC was sized per the rule, so the shock threat is greater.

Redsy: I know it sounds inconsistent, but I don't think it is. Remember, the reason the EGC was sized up was the long run. If that 60 amp circuit was the same length, it would need a #4 EGC.

redsy, I understand your confusion, but there is a purpose for the increase in size. The EGC has two serve two critical functions during a fault.

1. Is to be able to clear the fault quickly. Obviously even if the EGC were not increased in size, the standard EGC would be able to do that up to a point.

2. Limit the voltage at the point of the fault. This is where the size becomes important. If the supply conductor and EGC are the same size, then the voltage at the fault is roughly half the supply voltage. If the EGC is smaller, then the voltage at the fault is proportional higher.

Notice that Table 250.122 permits a #6 EGC for a 200-amp ckt.

A 3/0 conductor has an area of 0.173 sq.in.
A #6 has an area of 0.027 sq.in., which is about 15% the area of the 3/0.

So much for equal sizing. I still don't get it.

I think that increasing the EGC is a good idea, but some proportional value could be assigned for the conductors below a certain size.

Don, you got me thinking!

How about, call it a 40A circuit, and use #10 for the EGC?

This code section can really be a costly mistake to those who run cables.

Use a larger cable assembly on a 15, 20 or 30 amp circuit and the factory EGC will be to small.

Run 8/2 UF for a 20 amp circuit to some far away load and the 10 AWG EGC will be a violation.

iwire,

That is why I wonder why it would truly be unsafe to run a "proportionally increased" EGC for smaller circuits that have been upsized for voltage drop.
Unless my math is off, further calculations reveal that:

60 amp ckt-- EGC = 40% circuit conductor size
100 amp ckt. 32%
200 amp ckt. 16%
300 amp ckt. 12%

I would suggest a conservative 40% ratio of increased EGC comared to increased circuit conductors, for smaller circuits, up to 60 amps.
Could this possibly be based on whatever engineering data was used to generate Table 250.122?

Redsy all I can say is put together a proposal for the 2011 NEC.

Personally I think making allowances for 15, 20 and 30 amp circuits, (the only circuits that are in issue) would only add confusion.

I do not think it is a hardship to run a full size ground on these size circuits.

I run into this requirement often with site lighting and we may run 10, 8, 6, 4 even 2 AWG for use with 20 or 30 amp circuits we know when we do that we will be running a full sized EGC.

I will also point again that a 10 AWG is only acceptable for a 60 amp breaker if you use 6 AWG CU for the ungrounded. If you bump up the ungrounded conductors to say 2 AWG for voltage drop the EGC will need to be 6 AWG on that 60 amp circuit. (a guess, I did not do the calacs).

My point is 250.122(B) is not about sizing the EGC to the OCP but about sizing the EGC for the VD.

Guys,

Thanks for the input.
In case you haven't figured it out, I did a project relevant to this issue. And I did run a properly increased EGC.
The topic is interesting, and took on a life of it's own.

iwire,

Interesting point comes up when you say...

Use a larger cable assembly on a 15, 20 or 30 amp circuit and the factory EGC will be to small.

Run 8/2 UF for a 20 amp circuit to some far away load and the 10 AWG EGC will be a violation.

The point being that if you start with a short section of #14, 12, or 10 cable, (which have full size grounds) and then transition to your larger conductors, no violation would seem to occur because all the total line resistances would be equal, correct?

BTW,

How does one "clip a quote" to paste into a reply?

Thanks for everything guys!