310.15(b)(7) and derating

[Tainted]

3 apartments share a 2-inch conduit, 9 conductors total. Apartments are 120/208V single phase. A raceway is used as an equipment grounding conductor so there are no non-current carrying conductors in this raceway.

If all 3 apartments are upgraded to 100Amps, would it be legal to use (9) #2 THWN-2, 90 degrees C copper conductors in the shared 2-inch conduit?

This is how I did my calculation:

100A*83% = 83A as per 310.15(b)(7)

83A/70% = 118.6A wire required as per table 310.15(B)(3)(a) because there are 9 current carrying conductors in a raceway

Since 90 degrees wire is used, we derate from 90 degree column and the next size up to 118.6A is #2 wires

Is this correct?

 

[wwhitney]

I believe that is correct, but I have two comments:

1) First, there is to me a fairly convincing argument that the 2017 NEC erred in adding 120/208 3-wire single phase residential services to 310.15(B)(7), although the CMP has subsequently rejected PIs intended to undo the change based on that argument. The argument is that the 83% factor for residential 120/240 3-wire single phase services is in fact a form of "anti-derating" due to the fact that such a service has only 2 CCCs in it. That is analogous to the difference between Columns A and B of Table 400.5(A)(1). Whereas a 120/208 3-wire single service would have 3 CCCs, so the 83% factor should not apply.

So if you buy that argument, you should not take the 83% factor in the proposed configuration, even though it is allowed. If, however, you consider the 83% factor to be an additional diversity factor on top of those allowed in Article 220, there's no problem taking the 83% factor.

2) 215.4 would in fact let you run the 3 feeders with a common neutral conductor, and then you'd only have 6 CCCs in your conduit. Then if you take the allowed 83% factor, your base ampacity would need to be 83/.8 = 104a. So you could use #3 Cu 90C conductors for the 6 ungrounded conductors. The common neutral would need a base ampacity of twice that and could be 3/0 Cu, or perhaps smaller if there are additional reduction factors that could be applied to the neutral.

Of course, when the 3 feeders ultimately separate to go to different locations, you'd need a way to splice a #3 Cu neutral conductor to the 3/0 conductor.

Cheers, Wayne

 

[Tainted]

I believe that is correct, but I have two comments:

1) First, there is to me a fairly convincing argument that the 2017 NEC erred in adding 120/208 3-wire single phase residential services to 310.15(B)(7), although the CMP has subsequently rejected PIs intended to undo the change based on that argument. The argument is that the 83% factor for residential 120/240 3-wire single phase services is in fact a form of "anti-derating" due to the fact that such a service has only 2 CCCs in it. That is analogous to the difference between Columns A and B of Table 400.5(A)(1). Whereas a 120/208 3-wire single service would have 3 CCCs, so the 83% factor should not apply.

So if you buy that argument, you should not take the 83% factor in the proposed configuration, even though it is allowed. If, however, you consider the 83% factor to be an additional diversity factor on top of those allowed in Article 220, there's no problem taking the 83% factor.

2) 215.4 would in fact let you run the 3 feeders with a common neutral conductor, and then you'd only have 6 CCCs in your conduit. Then if you take the allowed 83% factor, your base ampacity would need to be 83/.8 = 104a. So you could use #3 Cu 90C conductors for the 6 ungrounded conductors. The common neutral would need a base ampacity of twice that and could be 3/0 Cu, or perhaps smaller if there are additional reduction factors that could be applied to the neutral.

Of course, when the 3 feeders ultimately separate to go to different locations, you'd need a way to splice a #3 Cu neutral conductor to the 3/0 conductor.

Cheers, Wayne

I actually would agree with that argument, quite interesting, but still, if it ain't broke don't fix. But suppose the 120/208V 3-wire systems weren't added to the 83% rule, does that mean all the wire strictly needs to be 100% rather than 83%?

 

[wwhitney]

I actually would agree with that argument, quite interesting, but still, if it ain't broke don't fix.

If you agree with that argument, it would seem to me your engineering judgement would require you not to take advantage of the 83% allowance.

But suppose the 120/208V 3-wire systems weren't added to the 83% rule, does that mean all the wire strictly needs to be 100% rather than 83%?

Of course? I don't understand why there would be any question given the hypothesis.

Cheers, Wayne

 

[Tainted]

Of course? I don't understand why there would be any question given the hypothesis.

Cheers, Wayne

I hate new york city, they are still stuck with modified 2011 electrical code. I cannot use that generous 83% generous factor for 208V 3-wire system then... darn

 

[Tainted]

If you agree with that argument, it would seem to me your engineering judgement would require you not to take advantage of the 83% allowance.

I agree with the logic, but to be fair, what could go wrong if you take 83% of a 208V 3-wire system? Sure wires will get hotter, but I doubt it'll be enough to burn down a building... If test are performed with practical data rather than a book telling me what I should do, then I would be satisfied

 

[electrofelon]

I hate new york city, they are still stuck with modified 2011 electrical code. I cannot use that generous 83% generous factor for 208V 3-wire system then... darn

A couple points:.
1. Even without the 83% rule, you end up with 91 amps and can next size up to 100. Calculated load would of course have to be 91 amps or less.
2. It appears you could go up to number one and still fit nine number one conductors in a 2 inch conduit. Perhaps you could go with reduced neutrals depending on the load specifics and what the calcs show.
3. Wayne's idea of using a common neutral for the multiple feeders is intriguing. If you had a common neutral for 3 120/208 single phase feeders, then you would be balanced and could reduce the neutral quite a bit.

 

[Tainted]

A couple points:.
1. Even without the 83% rule, you end up with 91 amps and can next size up to 100. Calculated load would of course have to be 91 amps or less.
2. It appears you could go up to number one and still fit nine number one conductors in a 2 inch conduit. Perhaps you could go with reduced neutrals depending on the load specifics and what the calcs show.
3. Wayne's idea of using a common neutral for the multiple feeders is intriguing. If you had a common neutral for 3 120/208 single phase feeders, then you would be balanced and could reduce the neutral quite a bit.

1) So if load is 91Amps, with 70% derate factor I would use #2 wire 90 degrees copper? correct?
2) Correct me if I'm wrong, wouldn't nine #1 THWN-2 copper conductors exceed 40% conduit fill if installed in 2 inch conduit?
3) I kind of want to be safe and not worry about unbalanced loads going through the neutral.

 

[infinity]

I agree with the logic, but to be fair, what could go wrong if you take 83% of a 208V 3-wire system? Sure wires will get hotter, but I doubt it'll be enough to burn down a building... If test are performed with practical data rather than a book telling me what I should do, then I would be satisfied

Actually it's worse it's the 2008 NEC. And you're correct you cannot use the 83% rule.

 

[Tainted]

Actually it's worse it's the 2008 NEC. And you're correct you cannot use the 83% rule.

Do you know when there will be a new new york city electrical code, I really hope they include 83% in the future

 

[electrofelon]

1) So if load is 91Amps, with 70% derate factor I would use #2 wire 90 degrees copper? correct?
2) Correct me if I'm wrong, wouldn't nine #1 THWN-2 copper conductors exceed 40% conduit fill if installed in 2 inch conduit?
3) I kind of want to be safe and not worry about unbalanced loads going through the neutral.

Well you didn't specify the type of conduit. EMT and rigid metallic come out just under 40%. Interestingly, PVC schedule 40 comes out just a hair over 40% - I thought rigid and PVC were the same physical dimensions. If you are using pvc, perhaps you can use a one size reduced neutral. Once you take off your line to line loads, I doubt you will see much neutral current even considering it's a 120/208 system

 

[Tainted]

Well you didn't specify the type of conduit. EMT and rigid metallic come out just under 40%. Interestingly, PVC schedule 40 comes out just a hair over 40% - I thought rigid and PVC were the same physical dimensions. If you are using pvc, perhaps you can use a one size reduced neutral. Once you take off your line to line loads, I doubt you will see much neutral current even considering it's a 120/208 system

Sorry, yes it's EMT. If you are using southwire calculator to calculate conduit fill, I don't trust that.

NEC table tells me you cannot fit nine #1 THWN-2 copper in 2-inch EMT, annex C. To be honest not sure if they are compact or not.

 

[electrofelon]

Sorry, yes it's EMT. If you are using southwire calculator to calculate conduit fill, I don't trust that.

NEC table tells me you cannot fit nine #1 THWN-2 copper in 2-inch EMT, annex C. To be honest not sure if they are compact or not.

I just checked four spec sheets from four different manufacturers for number one THHN. Allied wire and cable, southwire and encore all said 435 diameter. Priority wire and cable was the outlier saying 450 (I would be inclined to trust priority the least, as I have had some issues with them and I don't think they manufacture many of the products they sell) If you run the numbers 40% fill works with 435 but not with 450.

 

[Tainted]

I just checked four spec sheets from four different manufacturers for number one THHN. Allied wire and cable, southwire and encore all said 435 diameter. Priority wire and cable was the outlier saying 450 (I would be inclined to trust priority the least, as I have had some issues with them and I don't think they manufacture many of the products they sell) If you run the numbers 40% fill works with 435 but not with 450.

This brings up an interesting question... What takes priority over conduit fill requirements? NEC tables or actual conduit fill?

for non-compact conductors, only (8) THWN-2 conductors can fit in 2 inch EMT according to NEC, weird:

 

[infinity]

Actually it's worse it's the 2008 NEC. And you're correct you cannot use the 83% rule.

They've been talking about it for several years. I believe that they were supposed to do something on 2018 well it 2022 and still the 2008 NEC.

For these installations were typically run a common neutral for every three panels.

 

[electrofelon]

This brings up an interesting question... What takes priority over conduit fill requirements? NEC tables or actual conduit fill?

for non-compact conductors, only (8) THWN-2 conductors can fit in 2 inch EMT according to NEC, weird:

View attachment 2560247

I would not trust those tables. Those are just there for convenience they are not what you are necessarily supposed to follow. Yuu are supposed to go by actual dimensions. I have seen entries in the table that are not correct.

 

[electrofelon]

For these installations were typically run a common neutral for every three panels.

So would that common neutral be sized for the max theoretical imbalance on the same phase for all three units together? It seems like that is wildly overly conservative but I think it's what the code would require. What size do you typically see used for the situation then?

 

[Tainted]

For these installations were typically run a common neutral for every three panels.

I thought by code if loads are sharing neutrals, they would have to be interlocked to break the circuit if a fault occurs. I'm probably thinking of multi-wire branch circuits but I guess it doesn't apply to feeders feedings apartments.

 

[Tainted]

I would not trust those tables. Those are just there for convenience they are not what you are necessarily supposed to follow. Yuu are supposed to go by actual dimensions. I have seen entries in the table that are not correct.

Hard to find a devils advocate for this answer... But is there anything in the code that tells you to only use values directed by the code rather than actual dimensions?

 

[electrofelon]

Hard to find a devils advocate for this answer... But is there anything in the code that tells you to only use values directed by the code rather than actual dimensions?

For certain things, like motor full load currents, we are specifically directed to use the tables provided. I see nothing directing us to specifically use the tables for conduit fill.