Grounding requirement in conduit

[Firebrick43]

I plan to run all wiring in emt for my house addition / garage. As far as I understand I really don't need a ground and could used the emt as the grounding conductor. But I would like to run one ground conductor to bond everything together. To size it, make it as large as the largest hot conductor in the raceway that has multiple circuits in it?

 

[alfredeneuman]

No.
If the largest size circuit is not over 15 amps, you can run #14
20 amps = #12
30 - 60 amps = #10
>60 - 100 amps = #8
>100 - 200 amps =#6
Sizes given are for copper conductors.

Properly installed EMT alone provides a better low impedance ground path than the ground wire in NM cable

 

[Firebrick43]

No.
If the largest size circuit is not over 15 amps, you can run #14
20 amps = #12
30 - 60 amps = #10
>60 - 100 amps = #8
>100 - 200 amps =#6
Sizes given are for copper conductors.

Properly installed EMT alone provides a better low impedance ground path than the ground wire in NM cable

Thanks. So is bonding everything with a seperate wire a bad thing? I understand the proper bonding of emt (didn't realize it was better than nm ground). The issue that concerns me is a lot of the emt will be buried in the foam of an icf wall. Just concerned that time might take its toll on the ground path and access to check couplings will be non exsistant.

 

[mm08822]

Thanks. So is bonding everything with a seperate wire a bad thing? I understand the proper bonding of emt (didn't realize it was better than nm ground). The issue that concerns me is a lot of the emt will be buried in the foam of an icf wall. Just concerned that time might take its toll on the ground path and access to check couplings will be non exsistant.

Its not a bad thing, but if conduit is made up properly, its not needed. If you are set on doing it, i wouldnt go larger than #12. 12 is easy to terminate and doesnt waste a lot of conduit fill or $.

 

[alfredeneuman]

If you're going to run a ground in EMT at all, it's required that you run the size that corresponds to the circuit.

 

[mm08822]

If you're going to run a ground in EMT at all, it's required that you run the size that corresponds to the circuit.

I dont think so since the emt covers the nec requirement. Extra ground is above and beyond nec.

Howeve, any jumpers from device to box have to be properly sized .

 

[alfredeneuman]

2011 NEC 250.122 (C)

(C) Multiple Circuits. Where a single equipment grounding
conductor is run with multiple circuits in the same
raceway, cable, or cable tray, it shall be sized for the largest
overcurrent device protecting conductors in the raceway,
cable, or cable tray.

 

[sberry]

Yes, it needs to be sized for the circuit. I did just happen to find one wher the lock nut was missed, full of wire. It was on a steel building so I added another clamp next to it to jamb it against the box it was connected to, the box is also screwed to the purlin in the ceiling.

 

[grantw]

I have some 3/4 EMT in my garage, and I ran a green #12 along with the circuits. The metal boxes I used have the green bonding screws, and I don't mind the redundant ground. Every box I have installed are the extra deep steel city boxes.

I Like running a dedicated ground path for each circuit as well as the EMT itself. Yeah, it's probably redundant, but I feel better knowing there is a direct path to ground from each outlet via #12 copper.

 

[mm08822]

2011 NEC 250.122 (C)

(C) Multiple Circuits. Where a single equipment grounding
conductor is run with multiple circuits in the same
raceway, cable, or cable tray, it shall be sized for the largest
overcurrent device protecting conductors in the raceway,
cable, or cable tray.

If the raceway were nonconductive then I agree. However 250.122c speaks of a SINGLE equipment grounding conductor run with current carrying conductors. 250.118 explains that emt alone meets the requirement of a single egc.

 

[mm08822]

I have some 3/4 EMT in my garage, and I ran a green #12 along with the circuits. The metal boxes I used have the green bonding screws, and I don't mind the redundant ground. Every box I have installed are the extra deep steel city boxes.

I Like running a dedicated ground path for each circuit as well as the EMT itself. Yeah, it's probably redundant, but I feel better knowing there is a direct path to ground from each outlet via #12 copper.

One conductor is more than enough in emt.

 

[grantw]

One conductor is more than enough in emt.

IIRC, ground doesn't count towards conductor count, but does count in fill tables?

 

[mm08822]

IIRC, ground doesn't count towards conductor count, but does count in fill tables?

Doesnt count towards current carrying conductors but does for fill count.

 

[alfredeneuman]

If the raceway were non-conductive then I agree. However 250.122c speaks of a SINGLE equipment grounding conductor run with current carrying conductors.

I don't see anywhere that the Code reference excludes ANY raceway, non-conductive or not. It says simply RACEWAY.

 

[mm08822]

I don't see anywhere that the Code reference excludes ANY raceway, non-conductive or not. It says simply RACEWAY.

Another NEC Easter egg hunt.

Read 250.122(A) "General" paragraph before you let your eyes lock onto "Multiple Circuits" in 250.122(B).

 

[sberry]

I got a lot of respect for code wording. By the time one understands it they may understand it. I like to think I remember better than I really do.

 

[zmaxmotorsports]

2011 NEC 250.122 (C)

(C) Multiple Circuits. Where a single equipment grounding
conductor is run with multiple circuits in the same
raceway, cable, or cable tray, it shall be sized for the largest
overcurrent device protecting conductors in the raceway,
cable, or cable tray.

 

[wyliesdiesels]

ok i guess i need to join in on this debate.

let me grab my 2011 NEC and BRB

 

[wyliesdiesels]

I dont think so since the emt covers the nec requirement. Extra ground is above and beyond nec.

However, any jumpers from device to box have to be properly sized .

Another NEC Easter egg hunt.

Read 250.122(A) "General" paragraph before you let your eyes lock onto "Multiple Circuits" in 250.122(B).

I think Im siding with Alfred on this one.

250.122(A) backs him up on that.

Your stance is that the EMT covers the requirement automatically. But thats only IF its installed properly to meet the requirements to be an EGC eg. jumpers from box to device if not self grounding(u already mentioned this), bonding bushings on concentric knockouts etc.

How many existing installs have u come upon where the EMT was used as an EGC only to discover that a jumper was missing, bonding bushing lead not hooked up or someone decided to change from EMT to non metallic(read LTFNMC; it was cheaper than LTFMC, it was already on the truck and the contractor was already overbid) midway to get around something, then change back to EMT. And the inspector missed it...

I dont know about you but Ive seen my share of F*ck ups....

Just because EMT IS used doesnt automatically make it an EGC. Proper bonding of boxes, fittings and panel connections is required for EMT to meet the NEC requirements to make it an EGC...

Feel free to correct me if Im wrong...

 

[sberry]

Just like wire, if it isn't hooked up its no good. I opened a box today contractor installed, it never failed but wasn't done with too much care.

 

[wyliesdiesels]

Just like wire, if it isn't hooked up its no good. I opened a box today contractor installed, it never failed but wasn't done with too much care.

Exactly. Perfectly summed up what I said.

 

[sberry]

It was cable, just sort of semi careless strip lenght, done by guys do it every day, fast, spec built. It worked.

 

[mm08822]

I think Im siding with Alfred on this one.

250.122(A) backs him up on that.

Your stance is that the EMT covers the requirement automatically. But thats only IF its installed properly to meet the requirements to be an EGC eg. jumpers from box to device if not self grounding(u already mentioned this), bonding bushings on concentric knockouts etc.

How many existing installs have u come upon where the EMT was used as an EGC only to discover that a jumper was missing, bonding bushing lead not hooked up or someone decided to change from EMT to non metallic(read LTFNMC; it was cheaper, it was on the truck and the contractor was already overbid) midway to get around something, then change back to EMT.

I dont know about you but Ive seen my share of F*ck ups....

Just because EMT IS used doesnt automatically make it a properly connected EGC. Proper bonding of boxes, fittings and panel connections is required for EMT to meet the NEC requirements...

Feel free to correct me if Im wrong...

Hack work is another story and yes there is plenty out there. Even with the right materials nothing is guaranteed. Sure running an additional egc would most likely "safen" up a lot of hack work. No argument on the necessary treatment for concentric ko's.

The code is written assuming installation is as per the materials are listed and the rest of the code is complied with. And this answer is based on meeting the code minimums. You are choosing to go above and beyond. That is ok - nothings says you can't. Nothing says you have to exceed it.

250.122(A) clearly permits the raceway as egc and refers you to 250.118 and any of the methods listed in it. Emt is clearly one method permitted. Seperate egc are permitted as well. But it only requires one method as minimum and more are ok - emt by itself is ok, PVC would need the egc. EMT could also have an additional egc if you wanted (but not req'd.), etc.

250.118(B) gets into the reqs for ltfmc, ckt sizes, ltfmc lengths however the OP only stated emt so no need to bloat the reply.

If the OP decided to not run a separate egc with his single hot(s) and neutral(s) it would pass code. Regardless of the circuit ampacity.

Is running an egc safer yes. And that is why I said run if you really want to run one, just run a 12 and be done with it. He is exceeding the code with the extra conductor. Size doesn't matter here b/c emt meets the requirement. The 12 will provide an adequate return to trip the breaker even if the conduit "circuit" were broken.

If this were strictly a pvc job, then egc has to match req's for the largest ckt size.

The OP question was does he have to match the egc size for largest ckt and the code compliant answer is no b/c of the emt.

 

[sberry]

I think it still needs to be sized right. if we decided we need to use it for some reason couldnt.

 

[mm08822]

I think it still needs to be sized right. if we decided we need to use it for some reason couldnt.

If he chose not to run any egc it would not be wrong and meets code which is only the minimum requirement.

Don't confuse required with nice to have.

 

[Dagny]

Don't need it but put it anyway.

 

[wyliesdiesels]

Hack work is another story and yes there is plenty out there. Even with the right materials nothing is guaranteed. Sure running an additional egc would most likely "safen" up a lot of hack work. No argument on the necessary treatment for concentric ko's.

The code is written assuming installation is as per the materials are listed and the rest of the code is complied with. And this answer is based on meeting the code minimums. You are choosing to go above and beyond. That is ok - nothings says you can't. Nothing says you have to exceed it.

250.122(A) clearly permits the raceway as egc and refers you to 250.118 and any of the methods listed in it. Emt is clearly one method permitted. Seperate egc are permitted as well. But it only requires one method as minimum and more are ok - emt by itself is ok, PVC would need the egc. EMT could also have an additional egc if you wanted (but not req'd.), etc.

250.118(B) gets into the reqs for ltfmc, ckt sizes, ltfmc lengths however the OP only stated emt so no need to bloat the reply.

If the OP decided to not run a separate egc with his single hot(s) and neutral(s) it would pass code. Regardless of the circuit ampacity.

Is running an egc safer yes. And that is why I said run if you really want to run one, just run a 12 and be done with it. He is exceeding the code with the extra conductor. Size doesn't matter here b/c emt meets the requirement. The 12 will provide an adequate return to trip the breaker even if the conduit "circuit" were broken.

If this were strictly a pvc job, then egc has to match req's for the largest ckt size.

The OP question was does he have to match the egc size for largest ckt and the code compliant answer is no b/c of the emt.

I get what youre saying but

IMHPO, you are adding to the code. 250.122(A) simply says that an EGC cannot be smaller than the sizes in Table 250.122.

So yes, if someone chose to run an EGC in addition to the EMT, it would need to be sized per T250.122. There is no exception that says if a supplemental EGC is ran in EMT then it can be sized smaller than the requirements of Table 250.122....

Im gonna see if theres anything on this in the mike holt and electrician talk forums...

 

[Daveco]

I get what youre saying but

IMHPO, you are adding to the code. 250.122(A) simply says that an EGC cannot be smaller than the sizes in Table 250.122.

So yes, if someone chose to run an EGC in addition to the EMT, it would need to be sized per T250.122. There is no exception that says if a supplemental EGC is ran in EMT that it can be sized smaller

And this is why code books are thick and incomprehensible.




Sent from my iPad using Tapatalk

 

[sberry]

So, you run a wire in case there is an interuption, but if there is its too small? I agree, it needs to be sized right. No mystery or confusion here.

 

[sberry]

Kind of like having the wrong undersized spare tire on a pickup. Its ok till you need it.

 

[alfredeneuman]

I spent the majority of my career doing commercial electrical construction and service work.
Included in that are plenty of "grounding conductors supplemental to EMT" engineered jobs. NONE of the electrical engineers specified anything less than what's listed in 250.122. (US govt engineers specified grounding conductors larger than those in 250.122)

I've run into a few #12 grounds used with 30Amp circuits on restaurants, but restaurants are notorious for electrical violations and bartender/customer that knows electric/handyman repairs

sberry had a good analogy with his spare tire

 

[mm08822]

I spent the majority of my career doing commercial electrical construction and service work.
Included in that are plenty of "grounding conductors supplemental to EMT" engineered jobs. NONE of the electrical engineers specified anything less than what's listed in 250.122. (US govt engineers specified grounding conductors larger than those in 250.122)

I've run into a few #12 grounds used with 30Amp circuits on restaurants, but restaurants are notorious for electrical violations and bartender/customer that knows electric/handyman repairs

sberry had a good analogy with his spare tire

Alfred,

The code is not a design guide and it admits that is at best a safety code and not necessarily adequate for current or future user needs. The code is usually updated to address previous issues and not always modified with non-conflicting statements. And never mind the commercial influence/approvals for the latest and greatest allowed products.

It is easy to meet and exceed the NEC when you are in the industrial and governmental sectors. Typically there are deep pockets and greater dependencies on system reliability as it affects many people or has significant loss on manufacturing/income.

The NEC was spit on where I used to work as many called it a fire code – meaning nothing better than a way not to start a fire. It had minimal best practices and certainly little use to enable future expansion. We typically went way beyond in our designs and purchases. All grc, min ¾”, with grounds, NEMA 1 size starters minimum, all copper XHHW, double-ended switchgear, tie breakers, 2x rated feeders to switchgear, dual utility co feeds, all plc based control, rebuilds every 3 years, blah, blah, etc.

However, most applications discussed here are residential (with residential budgets) and helping/getting people to comply with the NEC minimums is frequently the challenge.

 

[Norcal]

I do not care what mm08822 says, the NEC does not give permission to use what ever grabs you because it's a conduit or tubing recognized as a grounding conductor, done wrong is worse then no EGC in EMT.

 

[sberry]

Becase they are code minimums does not mean they are not adequate. Wire sizing is a good example of this. I tend to agree that some comments in this thread are good examples of smart people not having a great grasp of some code principles.

 

[mm08822]

Becase they are code minimums does not mean they are not adequate. Wire sizing is a good example of this. I tend to agree that some comments in this thread are good examples of smart people not having a great grasp of some code principles.

"and not necessarily adequate for current or future user needs." is different than "not adequate." NFPA states that on page 1 in NEC.

I agree, like people who still can’t get article 210.21(B)(3) understood after 7+ years.

 

[mm08822]

I do not care what mm08822 says, the NEC does not give permission to use what ever grabs you because it's a conduit or tubing recognized as a grounding conductor, done wrong is worse then no EGC in EMT.

A little vague - can you explain "done wrong" and the results?

Code says if a wire-type egc is installed in a metal raceway, the raceway is in parallel with the wire, and the combination of the metal raceway and the wire is the egc for the enclosed circuits.

If the OP puts in a 30a ckt in his new conduit network, he runs a #10 egc for it and then it’s required all throughout the remainder of that home-runned conduit network. This then includes any other 15a and 20a ckts in this same conduit network vs not having to pull any egc in the first place?

There’s a gap in the code then. How can you go from 0 egc requirement using metallic conduit and meeting code to being required to do full 100% egc size requirement as if you didn’t have a metallic conduit at all when you do?
Also be required to upsize the egc if your supply conductors have to be upsized. Seems like they forgot a scenario in 250.118/122.

Maybe I’m reading something more into it but there is some screwy wording going on and only part of the scenario considered. I will find the full intent.

Here’s an example that shows my suggestion of how going to #12 doesn’t have an impact on forcing ocpd to operate:
Assume the OP had a 50A 240v circuit with #8’s in that conduit and the recept was 100’ from the cb.
Then a ground fault occurred at the recept.
For a #10 wire grd, the fault current would be ~600A. A QO 50A cb would open in 17.5 msec or less (1 cycle) according to the trip curve time range at this fault value.
For a #12 wire grd, the fault current would be ~575A (more resistance in egc – 2 ohms vs. 1.24 ohms/ 1000ft). A QO 50A cb would open in 20 msec or less(1.1 cycle) according to the trip curve time range at this fault value.
The conductivity of the parallel path of conduit isn’t even factored in here. It doesn’t matter b/c the breaker trip time can’t get any faster. Only difference would be ~ 2x the above fault values due to lower impedance with conduit.

Kinda like a spare tire inflated to 64PSI instead of 65PSI. Still gets the job done.

 

[sberry]

It does explain that it's not for every future concern. But if you are installing it so it's adequate due to disconnection then it's not. It doesn't need extra if the pipe is in place. While you study this you will find in the end the wording is accurate, the problem is the understanding most of the time but you did strike one of the most confusing sections. This is why it's important to simply follow the directions.
Done wrong is simple here, wrong wire. As per example, the tire isn't a pound low on air but it's the wrong size or load range.

 

[penright]

If the OP puts in a 30a ckt in his new conduit network, he runs a #10 egc for it and then it’s required all throughout the remainder of that home-runned conduit network. This then includes any other 15a and 20a ckts in this same conduit network vs not having to pull any egc in the first place?

As someone on the outside looking in and not an electrician, my guess is "what would someone assume looking at the box. If you have not wire and some where the EMT pulls loose, then you have an open ground no question. Meters and lights will show it as open. If you pull a wire that is too small for the ground and the EMT pull loose, then meters and lights will show a ground but not big enough. Just thinking out loud.

 

[sberry]

That is pretty good logic.

 

[mm08822]

As someone on the outside looking in and not an electrician, my guess is "what would someone assume looking at the box. If you have not wire and some where the EMT pulls loose, then you have an open ground no question. Meters and lights will show it as open. If you pull a wire that is too small for the ground and the EMT pull loose, then meters and lights will show a ground but not big enough. Just thinking out loud.

You are right but i also showed in this example that the 12 wire opens the ocpd in virtually the same time as #10. Fault cleared, purpose accomplished.

If emt as an equivalent egc is such a poor grounding method, it should be removed from the code to accomplish ground path. But what i'm reading is poor installation and maintenance is root cause.

How much emt gets beat up buried behind sheetrock as i assume this is the OP's need.