electrofelon
Senior Member
- Location
- Cherry Valley NY, Seattle, WA
This is an interesting interpretation I had never considered. I think it is a very reasonable interpretation.
A/C wire size.
- Thread starter Daja7
- Start date
electrofelon
Senior Member
- Location
- Cherry Valley NY, Seattle, WA
wwhitney
Senior Member
- Location
- Berkeley, CA
- Occupation
- Retired
Correct, so let's do it via an exhaustive method:
Is #14 Cu compliant for my example? No, with NM cable, the 60C ampacity is 15A, and 15A < MCA = 28A.
Is #12 Cu compliant for my example? No, with NM cable, the 60C ampacity is 20A, and 20A < MCA = 28A.
Is #10 Cu compliant for my example? Yes, with NM cable, the 60C ampacity is 30A, and 30A >= MCA = 28A.
Is #8 Cu compliant for my example? Yes, with NM cable, the 60C ampacity is 40A, and 40A >= MCA = 28A.
So which one is the minimum? #10 Cu. There is no other reasonable answer.
Under the 2017 NEC, #8 NM cable with #10 EGC should fail for a branch circuit supplying HVAC equipment with an MCA of 28A and an MOCP of 40A and a 40A OCPD. Sorry if you've been doing it wrong. Although admittedly, that's a stupid rule other than the case of upsizing due to excessive branch circuit length.
Under the 2020 NEC, if you didn't pick #8 NM due to excessive branch circuit length, you get to say under the exception "well, I upsized from #10, but I didn't have to, the #10 NM would have worked fine, therefore a #10 EGC is an effective ground fault current path."
Cheers, Wayne
P.S. I guess you'd agree it should fail if a 30A OPCD is used?
The minimum-sized EGC is selected from Table 250.122 based on the rating or setting of the feeder or branch-circuit OCPD(s).
If that OCPD is the maximum allowed by that wiring method you have not increased anything in size.
334.80 limits NM to 60C, so #8 NM on a 40 is the max size. There are examples in 250.122 where they go by maximum OCPD size like (D)(2).
If you have a 30A OCPD on a 8/2 NM then yes you've increased in size.
If its a 10/2 on a 30A then no.
And on top of it I have a IEEE green book and I am a qualified person LOL.
Last edited:
wwhitney
Senior Member
- Location
- Berkeley, CA
- Occupation
- Retired
This is where we disagree. If a smaller size is NEC compliant, it is obvious to me the current size is increased. You're bringing in other criteria without any justification.
You seem to have this idea that we only need to compare the current installation with other installs that comply with 240.4 first paragraph, ignoring that 240.4 provides that 240.4(A) through (G) may modify that requirement, so it doesn't always hold.. This comes up in the context of 240.4(B): if I use #4 NM protected at 60A for a 53A load, have I upsized the conductor, because 240.4(B) would allow #6 NM? Or have I not, because 240.4(B) is optional, and #4 NM is the smallest size whose ampacity is at least 60A?
And moreover it comes up in the context of 240.4(G). Motors and HVAC/MCA is just the most frequent case where we run into 240.4(G); I'm not arguing it's special. Another case is apparently Article 630 on welders, although I'm not so familiar with that section. But I expect we can also use it to create an example where #8 NM on a 40A breaker has been increased in size, because the circuit only requires a 30A ampacity.
I'm saying there's nothing special about 240.4, and that all we have to look at is whether the smaller size would cause a violation or not, of any section.
I'm sorry if I've been too vehement about this. Hopefully my comments so far have sufficed, and I'll attempt to hold my peace on this henceforth.
Cheers, Wayne
- Occupation
- Licensed Electrician
Copy that and me too, My milkshake offer stands if you or anyone gets red tagged on 250.122(B) with a 8/2 NM on a 40A breaker, just don't spill it on my IEEE green book.
mtnelect
HVAC & Electrical Contractor
- Location
- Southern California
- Occupation
- Contractor, C10 & C20 - Semi Retired
I love this discussion ... HVAC is a different animal, and all of you are helping all of us to understand it.
mtnelect
HVAC & Electrical Contractor
- Location
- Southern California
- Occupation
- Contractor, C10 & C20 - Semi Retired
HVAC ... Is an "Anomaly".
LarryFine
Master Electrician Electric Contractor Richmond VA
- Location
- Henrico County, VA
- Occupation
- Electrical Contractor
It also shows just how complicated answering a seemingly simple question can become.
Not what I said.
I said I like seeing the wiring to a disconnect rated for the overcurrent protection ahead of it.
You said It'd be no different than 16ga lamp cord plugged into a receptacle.
I said sure it'd be different because the wiring to the receptacle outlet would be 20a rated.
Seems you'd be fine having 16ga wire from receptacle all the way back to the breaker in the panel that feeds it.
I'm not.
What you want to do with the wiring from the disconnect to the unit is your choice.
I'm just saying I pull wire from the OCPD to the disconnect sized for the OCPD, not what it's feeding.
JAP>
- Location
- New Jersey
- Occupation
- Journeyman Electrician
Yes, even non-fused disconnects.
If you follow me and there's a 40a breaker in the panel, my AC disconnect will have 40a rated wire to the line side of it.
You're welcome.
JAP>
- Location
- New Jersey
- Occupation
- Journeyman Electrician
kwired
Electron manager
- Location
- NE Nebraska
Or did not know what unit would be at rough in time.It's okay to exceed the code requirement but IMO it a waste of money, to each his own. When I see #8 to a disconnect with an MCA of 25 amps I assumed that the installer didn't know the code.
I get this sometimes with cooktops or ovens as well. Some may need 8 AWG others may only need 10 AWG. Often they don't have a specific model picked out yet and you run the larger cable just in case. May run three wire as well and end up not needing three wire.
It's okay to exceed the code requirement but IMO it a waste of money, to each his own. When I see #8 to a disconnect with an MCA of 25 amps I assumed that the installer didn't know the code.
Ac units come and go,,,, Machinery comes and goes,,,, most times the OCPD and the disconnect are of adequate size to service a large unit if that may be the case. (That is if the wiring to them is sized properly).
If I see a 50a Receptacle I assume it has 50a wiring to it. If I installed it, it will.
If I saw #14 on a 40a breaker in the panel feeding an outside AC disconnect I'd just have to assume they weren't brought up around industrial electricians as I was.
Resi installer are much more competetive and I can see why it may be done there.
That's fine.
It has nothing to do with knowing the code.
Sometimes even though you can, doesn't always mean you should.
Oh well.
JAP>
Elect117
Senior Member
- Location
- California
- Occupation
- Engineer E.E. P.E.
As Wayne pointed out, it is more than just the fault current return that matters. It is also the clearing time of the OCPD. I did some messed up math and then deleted my post because I really didn't like how many assumptions I made to explain it.
For a simple fault current calc, you take the available full load current from the source and divide that by the impedance. As the impedance gets smaller in larger cables, that value will go up. As that value increases, the OCPD's Current Vs Time graph, which hasn't changed, will probably clear faster. But the fault current the EGC has to carry back to the source in order for the OCPD to clear is higher. EGCs are sized to be large enough to return the fault current. A simple fault should be a fast high current value that gets interrupted very quickly. The insulation shouldn't fail prior to clearing. The ampacity of the wire needs to be large enough to carry it back or it could end up increasing the clearing time, which puts more strain on the insulation, and can lead to a failure. But the NEC table 250.122 groups OCPD devices from 20A to 60A in one wire size, which leads me to believe that the wire size is tested to be large enough for the available fault current's return back to the source for any load between 20A to 60A.
P.S. I am not saying anything about NM #8 vs #10. I was just explaining why EGC wire size would reasonably increase when the CCCs wire size increases. I have no idea if a simple change from #10 to #8 would make a difference, but in an instance where you used 1000kmcil instead of #10, one could argue it will.
Check out chapter 2 of the IEEE green book it all on equipment grounding, its a interesting read. One of the things it says is in AC systems of 60hz circuits rated 40A or less the circuit reactance is an insignificant part of the circuit impedance and its really not a factor till you go over 100A OCPD.
Everything begins at and builds on the OCPD you select.
That is what the equipment ground need to do, 'trip the breaker'.
Nothing about the branch circuit load calc to size that breaker matters, once your have chosen the breaker you need a equipment ground that can trip it in the correct time frame, and you need to be able to prove that.
wwhitney
Senior Member
- Location
- Berkeley, CA
- Occupation
- Retired
Does your reference specify the time frame?
1 second would be a lot different than 1 minute. Looking at a QO230 trip curve, if I'm reading it right, at 40C ambient you'd need to carry 350A to ensure tripping within 1 second, while you'd only need to carry 60A to ensure tripping within 1 minute (or 105A at 25C ambient). For the approximation that the impedance is dominated by the EGC contribution, that would be over a 5 fold difference in EGC area.
Cheers, Wayne
Basically it says "When a copper conductor is used, the design should be evaluated to ensure that the conductor thermal rating is not exceeded."
Since your a math guy you'll like this:
Source IEEE green book 2007