This question was asked on another forum. I would like to hear your response.

A 240-volt, single-phase, 250-ampere load is supplied from a 300-ampere breaker located in a panelboard 500 ft away. The conductors are 250-kcmil copper, installed in rigid nonmetallic conduit, with a 4 AWG copper equipment grounding conductor. If the conductors are increased to 350 kcmil, what is the minimum size for the equipment grounding conductor based on the proportional-increase requirement?

The caculations were made correctly and #2 copper was given as the proper size.

There was nothing said about the 500 ft and its impact on the EGC. You comments please.

[This message has been edited by Bob (edited 11-30-2005).]

If the upsize of the EGC was in proportion to the upsize of the conductor for voltage it satisfies 250.122(B). In fact this is the example used in the handbook.

Greg Fretwell

#96483 - 12/01/0507:22 AMRe: Sizing the EGC when feeder is increased for VD

Voltage drop in the EGC is never _explicitly_ dealt with, but 250.122(B) is a crude and imprecise requirement that does deal with voltage drop. Essentially what 250.122(B) says is that if you do anything to deal with voltage drop in your circuit conductors, then do the same thing to you EGC. Of course, low voltage drop is _not_ a requirement of the code, and thus you could do something silly like run a 2000 foot long 120V 20A circuit using #12 conductors along with a #12 EGC; a bolted short between hot and ground would _never_ trip the breaker.

-Jon

#96484 - 12/01/0510:16 AMRe: Sizing the EGC when feeder is increased for VD

I'm just being letter of the law snotty here I agree that my 'silly' example is something that _should_ violate something in 250.4(A)... but: 250.4(A)(1) Not relevant to the branch circuit. 250.4(A)(2)-(4) simply say that all of the conductive materials need to be connected to the EGC, and that the impedance must be low enough to limit the voltage of these matierials. 2000 feet of #12 has a resistance of 3 ohms; which is certainly better than the 25 ohms mandated for grounding electrodes (when you only have one, when you test, la la la). The inductance of 2000 feet of wire needs to be considered for things like lightning imposed voltage...but this will be essentially unchanged when the conductor gets larger. The voltage imposed by a bolted fault would be 60V, which sounds limited to me. 250.4(A)(5) says that the ground fault current path must be able to safely carry the ground fault current. #12 can certainly carry the expected ground fault current (20A)

There is no definition of 'Effective Ground-Fault Current Path'. If an 'Effective Ground Fault Current Path' means 'must trip a breaker', then 250.4 implicitly limits voltage drop in _both_ the supply and egc. Consider a branch circuit with #12 hot and #12 neutral and #0 EGC. You still won't have a ground fault current path that would reliably trip the breaker.

-Jon

#96486 - 12/01/0501:55 PMRe: Sizing the EGC when feeder is increased for VD

There is no definition of 'Effective Ground-Fault Current Path'. If an 'Effective Ground Fault Current Path' means 'must trip a breaker', then 250.4 implicitly limits voltage drop in _both_ the supply and egc. Consider a branch circuit with #12 hot and #12 neutral and #0 EGC. You still won't have a ground fault current path that would reliably trip the breaker.

Actually 250.2 does define it.

Ryan Jackson, Salt Lake City

#96487 - 12/01/0505:15 PMRe: Sizing the EGC when feeder is increased for VD

Damn I'm being slow today Silly me for checking Article 100 (Definitions) and not checking the 'local definitions' in article 250. Thanks for keeping me on my toes.

I still think that 2000 feet of #12 would qualify as an 'Effective Ground Fault Current Path', but this is really stretching the point

Interesting side note that I noted recently: the size of GEC conductors scales linearly with the associated amp rating (constant circular mils per amp), but the size of the current carrying conductors scales as a power of ampacity. In other words, the number of circular mils per amp in a 650A feeder is far greater than the number of circular mils per amp in a 50A feeder. (50A feeder using #8 conductors, 16510 circular mils, 330 CM/amp, #10 GEC, 10380 CM, 208 CM/amp) (650A feeder using 1750 kcmil conductors, 2700 CM/amp, #0 GEC 105600 CM, 162 CM/amp)

Does funny things to the allowed voltage drop in the GEC as feeder size increases.

*grin* Jon

#96488 - 12/05/0506:22 PMRe: Sizing the EGC when feeder is increased for VD

Thanks for your responses. The NEC provide 3 requirements for effective grounding:

1. provide a permanent path. 2. Conductor must have the capacity to withstand the fault current and 3. must have a sufficiently low impedance to limit the voltage to ground and to facilitate the operation of the OC device.

Numbers 1 and 2 are straight forward. However #3 leaves some questions to be answered. What does "limit the voltage to ground" mean and what does " facilitate the operation of the OC device" mean.

At the time this requirement was entered into the code I suspect there was not a great deal of information regarding the effects of electrical shock on the human body. We now have information that has provided the use of GFI devices to limit the effects on the human body. In the IAEI Soares Book on Grounding there is a chart that list voltage level and and the human body response.

For body resistance of 500, 1500 and 3000 ohms, the let go voltage is 4.5, 13.5 and 27 volts. For fibrillation the voltages were 11.5, 34.5 and 69 volts. Soares suggests using 40 volts to ground as the limit. This voltage is the result of the fault current x the impedance of the EGC.

Facilitating the operation of the OC device means opening the OC as fast a possible. Therefore we must have a fault current that will open the device instantly or in 1 cycle. Soares shows 5 times the OC rating as the fault current required to open the OC instantly however tests have shown that 4 times is adequate.

First the NEC is not an engineering manual. It provides safety guidelines to be used for the installation of safe electrical systems. Table 250.122 provides the MINIMUM SIZE EGC. I think everyone realizes that distance plays a part in selecting the EGC. A general rule is that over 100 ft the EGC selected from table 122 is increased one size for each 100 ft. The given EGC size for a 300 amp breaker is #4 copper which is good for a circuit of 100 ft. We need to increase the EGC 4 times to meet the general rule. That gives us 3/0 copper. This has nothing to do with the increasing of the feeder for voltage drop. We will need 4 time the OD device rating to insure an instant opening. Given a 300 amp breaker we need a 1200 amp fault. In order to check the voltage to ground we simply multiply the 3/0 impedance by the 1200 amps. 3/0 at 25C x 1200 = Voltage to ground = 1200 amps x .037 = 45 volts. That's very close to the suggested value. If you are interesed in more information visit this site http://www.steelconduit.org/gemi.htm.

[This message has been edited by Bob (edited 12-05-2005).]

#96489 - 12/05/0506:33 PMRe: Sizing the EGC when feeder is increased for VD

Are we talking about what the NEC requires us to do or what is good design?

The only EGC / VD requirement presently is 250.122(B)

Beyond that any increase in EGC is a design decision.

The increase 1 size per 100' rule of thumb is a very rough way to go as it does not take into account the current, voltage or impedance of the circuit.

So I guess my question is what is it your seeking?

Bob

Bob Badger Construction & Maintenance Electrician Massachusetts

#96490 - 12/06/0512:34 AMRe: Sizing the EGC when feeder is increased for VD

iwire Glad to hear from you. YES this is a design decision. The code gives us the minimum requirements in table 250.122. It seems very few engineers and electricians are aware of the limits of table 250.122. I wanted to point out that while the table lists the EGC for a circuit, there are other considerations that make the installation a safe one. One is that the OC device open quickly to remove the potential and the other is to limit the voltage to ground. It is a design problem. I just wanted to try and put the information out there and raise some questions if needed. I hope I answered your questions.

[This message has been edited by Bob (edited 12-05-2005).]