# Do I need to derate wire ampacity for two or more sets of wires in the same raceway for subpanel wiring?

I was planning to pull three sets of four conductors (3 AWG wire) through 2" EMT conduit for a 30' length that is common to all three sets of wires, and then branch off at a listed metal junction box (10"x10"x4") to three 1 1/4" EMT conduits. I was planning to do this as part of running wiring for three 100-amp subpanels from a 200-amp main.

However, I was doing some re-reading of the NEC (2011 version), and recalled the need to de-rate ampacity of conductors where spacing is not maintained (such as in an EMT raceway) for runs greater than 24" per clause 310.15(B)

Per 310.15(B), it appears that the max ampacity for the wires is now only 92A and that I will not be able to use the single run of 2" conduit I had planned and instead will need to run three completely separate runs of EMT to allow for the three 100A subpanels with the 3 AWG wire I was planning to use -- is this correct or is there something else I am missing that would make my original plans okay?

Calculations and Code References:

• 3 AWG THHN with 90 degree rating = 115 A allowable ampacity per Table 310.15(B)(16)

• 6 conductors in the same conduit = 80% allowable ampacity per 310.15(B) (relevant sections quoted below)

• +6 conductors => 3 sets of 2 out-of-phase hot wires
• +0 conductors => 3 sets of 1 neutral carrying the unbalanced load (per 310.15(B)(5)(a)

310.15(B)(5)(a) A neutral conductor that carries only the unbalanced current from other conductors of the same circuit shall not be required to be counted when applying the provisions of 310.15(B)(3)(a).

• +0 conductors => 3 sets of 1 ground wire (per 310.15(B)(6)

310.15(B)(6) Grounding or Bonding Conductor. A grounding or bonding conductor shall not be counted when applying the provisions of 310.15(B)(3)(a). (b) In a 3-wire

• 115 A * 80% = 92 A

• Looks like you probably have it right that you'll need 3 runs. A word to the wise - you may not want to stick with the minimum size that passes code (40% fill) - it can be rather a pain to do the pulling at that fill, and a bit more room is often not terribly expensive. For one thing, use the 2" for one run if you have it in place already. Commented Jun 2, 2015 at 21:58
• What are the temperature ratings for the terminals of the breaker in the main panel, and the lugs in the second panel? Seems like your conductors may have been undersized to begin with. Commented Jun 3, 2015 at 2:39
• @Tester You got me, the terminals in both the main and subpanels are 75 degrees C. So, I should be derating from the 75 degree column (100 A) to 80 A. 3 separate conduit runs should still be fine at 30 degree C ambient, no? Commented Jun 3, 2015 at 3:40
• @statueuphemism No, you should be sizing from the 75°C column. If the conductors are rated at 90°C, then you can derate using the 90°C column. In either case, you'll end up using #2 CU conductors I think. Commented Jun 3, 2015 at 3:45
• @Tester101 Why would the wire size differ from your recommendation here: diy.stackexchange.com/a/29077/36011 Commented Jun 3, 2015 at 4:10

Your final wire size selection for the ungrounded conductors was exactly right at #2 copper. Your neutrals, however, could have been much smaller. You are allowed to calculate the total unbalanced load and in turn size the neutral accordingly. You still must apply the ambient temperature derating factor and you are required to at least match the size of the grounding conductor. For example, if your total calculated unbalanced load on the neutral was 25 amps and the ambient temperature was negligible, the ampacity table would allow for only a #10 AWG wire to carryn the neutral current, but because the neutral cannot be smaller than the grounding conductor you would still be required to use a #6 AWG. I have the tables underlined in my codebook but don't have it on hand , sorry. The minimum size grounding conductor chart chart is in article 250. Just found this on Mike Holt's website: { The grounded (neutral) service conductor must be sized to carry the maximum unbalanced load in accordance with 220.22 and must not be sized smaller than required by 250.24(B).}

## What you're missing is you get to round up on breaker size

92A. They don't make a 92A breaker.

95A. Oh, they don't make one of those either.

OK, 100A breaker. Done.

What remains is my stream-of-consciousness thoughts on sizing wire.

It's an interesting puzzler, to be sure.

Indeed, when computing the derate under 310.15b3a, you disregard neutrals. But why? Because there is no configuration where the 3 conductors will be worse than 2 conductors - usually they'll be better. It's all about heat, and heat is the square of current. Say 100% usage makes 10,000 heat units (100^2).

• If power is split 100% on L1 and L2 and 0% neutral, that's 10,000 heat units x2 for 20,000.
• If power is split 100% on L1 and neutral, and 0% on L2, that's 10,000 heat units x2 for 20,000.
• If power is split 100% on L1 and 50% on L2 and neutral, that's 10,000 plus 2x 2,500 heat units, for 15,000.

In a similar vein, since your supply is only 200A, your worst case load is actually four conductors carrying 100A and all others carrying 0. There's a case to make to your AHJ to let you rate for 4 conductors instead of 6. Of course this doesn't help here, but it would if you added a fourth or fifth 100A circuit.

Now with the ratings, it's important not to get tangled up. The 310.15b3a multi-wire derate always comes off the highest thermal rating of the wire type in the ampacity chart, even if you have 60C terminations. Then separately, the wire/termination temperature columns come into play. The lowest number applies. And then even separate from that, is any upsizing of the wire for long distances. And separate from that any statutory limits like "20A on 12AWG". So suppose you have 12AWG wire going 200' but traveling with 1 other circuit in a conduit:

• 24A based on 80% derate due to 4 active conductors off the 30A in the 90C column
• 25A based on wire and terminations are all good For 75C
• 15A recommended to limit voltage drop given the long length
• 20A as a hard rule since it is 12 AWG wire

The final answer is 20A mandatory, with 15A recommended to limit voltage drop.

Now if you're using EMT as a wiring method rather than just a damage shield, you don't need to wire grounds. EMT is the ground. I mean it doesn't hurt, belt and suspenders, but it doesn't help either.

• @statueuphemism de nada. But you are only thinking about the utility for you... and it's not for you. It's for others with the same problem. That is SE's format here, less a "custom for you" Q&A and more an encyclopedic reference that aims to answer similar questions too. Commented Nov 12, 2017 at 17:27
• @statueuphemism this forum is not competitive like that. I'm way past caring about rep, I answer questions for fun and to improve my own thinking. I wish I was better at searching but the question you linked is not that similar. I think you are equating questions and answers with large overlap in the shards of info they provide. This disregards structure and the brainpower needed to organize structure. For instance I had never thought of structuring the idea in the "maximum of" bullet list you see above. Commented Nov 14, 2017 at 20:57