Soares & length of metal conduits used as EGC

[dana1028]

A discussion came up in class tonight.

In Soares 'Book on Grounding'...Chap. 20 - Tables, there is a table entitled "Maximum length of steel conduit or tubing that may safely be used as an equipment grounding circuit conductor"

Bottom line, there's been testing done by the Georgia Institute of Technology and they feel there is a length limit to various sizes and types of conduit that will safely clear a fault.

I know 250.118 lists these conduits as 'Types of EGCs' and makes no mention of length limitations.

However - 250.4(A)(3) & (5) establishes a performance criteria for any bonding system. That is, even if you do install a system correctly, using all the right listed components....installed per mfr's instructions, etc., there is still a performance criteria that must be met. To whit - the system must be able to clear a fault.

With Georgia Tech providing test results indicating there is a performance deficiency when using steel conduit as an equipment grounding conductor in lengths greater than those in the table, would it be fair for an inspector to ask for performance verification for an installed system when those table lengths are exceeded. [I know Georgia Tech is not the code book - but they have performed, I would think, credible testing that shows performance deficiencies].

I'm not talking inches here...those table indicate no conduit 'may safely be used' in lengths greater than 300 ft. I talking about looking at a job with lengths of 500+ feet.

An analogy: Table C.1 - EMT - says I can install 9 #12 THHN conductors in 1/2" EMT conduit; however 300.17 says the number of conductors in any raceway shall not be more than will permit...ready installation or withdrawal of the conductors without damage to the conductors or to their insulation. I know if you try to pull (9) solid #12s in 1/2" EMT with a couple 90s you're going to damage the insulation. So even if you were code compliant with respect to fill, you still have not met the performace criteria of 300.17.

I'm not saying to use the Soares tables to red tag a job, I'm asking would it be legitimate to ask for performance verification?

 

[iwire]

How is this different from a wire EGC?

There is a point where a NEC sized wire EGC will become less effective and that distance is likely less than a conduit EGC

 

[don_resqcapt19]

When you use the software that came out of the Georgia Tech study, you will find that in most cases the maximum length of metal conduit that is suitable for use as an EGC is actually longer than the maximum length of the copper EGC that is run with the circuit conductors. You can download a copy of the software here.

 

[ptonsparky]

I have proven the limits of EG conductors at least once.

What is meant by "Arc Voltage"?

 

[ptonsparky]

..and since we are checking EG, is the operating voltage phase to ground or phase to phase?

 

[dana1028]

When you use the software that came out of the Georgia Tech study, you will find that in most cases the maximum length of metal conduit that is suitable for use as an EGC is actually longer than the maximum length of the copper EGC that is run with the circuit conductors. You can download a copy of the software here.

OK - if a wire EGC is no more effective than metal conduit, how would you achieve the 'effective ground fault current path' as required by 250.4(A) when you have long runs?

 

[iwire]

OK - if a wire EGC is no more effective than metal conduit, how would you achieve the 'effective ground fault current path' as required by 250.4(A) when you have long runs?

In some cases you would have to increase the wire EGC size, but how do you do determine that?

What method would we use to decide if we have an effective fault path? How fast must the breaker open?

 

[bob]

I have proven the limits of EG conductors at least once.

You have?

What is meant by "Arc Voltage"?

The study referred to by the OP assumed a voltage drop across the arc of 50 volts.
This is a engineering problem an not covered by the NEC. As you have heard many times, the NEC is the minimum required.

Dana
Look at the bottom of that table and see the criteria used to arrive at these limits. If you had a very long run you could consider AL conduit to extend the length.

 

[ptonsparky]

You have?


The study referred to by the OP assumed a voltage drop across the arc of 50 volts.
This is a engineering problem an not covered by the NEC. As you have heard many times, the NEC is the minimum required.

Dana
Look at the bottom of that table and see the criteria used to arrive at these limits. If you had a very long run you could consider AL conduit to extend the length.

Yes, I went to a great deal of trouble to calculate VD to each motor on a feeder. Direct bury aluminum. Dropping wire size the whole 1200' +-. Results were that the bolted fault (a cattle stampeed) to ground caused a phase monitor to drop out the control contactor on low voltage prior to the fuse blowing. Reset would occur after ___ seconds then drop out again. PPD.

 

[bob]

Yes, I went to a great deal of trouble to calculate VD to each motor on a feeder. Direct bury aluminum. Dropping wire size the whole 1200' +-. Results were that the bolted fault (a cattle stampeed) to ground caused a phase monitor to drop out the control contactor on low voltage prior to the fuse blowing. Reset would occur after ___ seconds then drop out again. PPD.

Is this posted in this forum? Where? I would like to see it.
The Google info does not use a bolted fault for the calculations. The arcing
fault is a worse case scenario. Is not likely that 120 volts can maintain an arc.

 

[don_resqcapt19]

OK - if a wire EGC is no more effective than metal conduit, how would you achieve the 'effective ground fault current path' as required by 250.4(A) when you have long runs?

The comparisons are between the metal conduit and the standard EGC sized based on T250.122. You could upsize both your circuit conductors and EGC for longer lengths. My only point was that the conduit is often suitable for a longer distance than an EGC sized per T250.122.

 

[ptonsparky]

Is this posted in this forum? Where? I would like to see it.
The Google info does not use a bolted fault for the calculations. The arcing
fault is a worse case scenario. Is not likely that 120 volts can maintain an arc.

Yes, it is posted on here someplace. 3yrs ago? I think Iwire gave me hell for using poor design. It was.

 

[bob]

The comparisons are between the metal conduit and the standard EGC sized based on T250.122. You could upsize both your circuit conductors and EGC for longer lengths. My only point was that the conduit is often suitable for a longer distance than an EGC sized per T250.122.

The Ga Tech study showed that 90% of the faullt current returned via the conduit and 10% via the EGC.

 

[don_resqcapt19]

The Ga Tech study showed that 90% of the faullt current returned via the conduit and 10% via the EGC.

Bob,
I think that part was about a GEC within a metal raceway, not about an EGC in a metal raceway. (from memory, as i did not go back and read the study, so I could be wrong)

 

[bob]

Bob,
I think that part was about a GEC within a metal raceway, not about an EGC in a metal raceway. (from memory, as i did not go back and read the study, so I could be wrong)

Don
I am fairly certain that the test was for the EGC comparing the conduit, the green conductor and the conduit with the green conductor combined. I do not think there was a test for the GEC. This information is also in the Soares Manual.

 

[Cold Fusion]

...How fast must the breaker open?

Inside the damage curve of the conductors.

If the installation is that long and a prompt trip matters, maybe the application is outside of the Cookbook design of the NEC. The NEC is just a bunch of minimums - not necessarily safe or workable.
cf