I need to run a feeder from my panel to subpanel. It will be 3x #3 THWN copper and and 1x #8 bare copper. The only knockouts in my existing panel box that unused are 3x 3/4", no one of which would accommodate those 4 wires alone.

My inspector will not allow me to drill out those knockouts (e.g. to 1"), nor to drill elsewhere in the panel box. (Though it seems common for people to do this, he does appear to be correct, as doing so technically violates the UL Certification).

The best solution I have so far is to use two of the 3/4" knockouts

entering panel box using two knockouts

I'll bring the 4 wires in 1-1/4" PVC conduit to a 4x4 exterior rated PVC box just below the panel box. Then I'll connect from the top of the 4x4 into two of the 3/4" knockouts and run two wires through each (within fill factor). The Box will be Carlon E987NR or similar.


  1. Anything about this that would not be compliant with NEC 2019?
  2. Any other reasons not to do it?
  3. Any better suggestions?
  4. I'm not sure yet how I'll waterproof the nipples where they enter the boxes. Suggestions?


  • Have you run your solution by your inspector? He's saying what you can't do, ask him what you can do.
    – JACK
    Oct 8, 2022 at 0:21
  • @JACK I do and I will. I'm asking here to get ahead of any issues, check whether I'm missing a better option, and share this example with the Internet.
    – tom
    Oct 8, 2022 at 0:29
  • Ick. You have potential for heating the section of (steel - not the PVC) box between the two nipples due to eddy currents caused by the EMF which is one reason we run all wires in a circuit together. Pretty sure there is something in code about that, but probably under the "paralleling" section which does not apply here...
    – Ecnerwal
    Oct 8, 2022 at 0:39
  • @Ecnerwal good point - hadnt thought about the induced current between the nipples. Maybe I can avoid this by going with your suggestion below.
    – tom
    Oct 8, 2022 at 1:48
  • 1
    Regarding 'why copper': Good question, and apologies for only marking it with the 'CU' in the drawing (text now updated). I picked copper because I assume it's going to be a lot easier to work with and pull through conduit. I need ~150ft total, and my local store sells #3 Copper for $2/ft. In Aluminium I'd need #1 (I think, for 100A 75C rating), which is maybe $1.25/ft (not checked locally), plus I'll need to splice to copper before squeezing through the 3/4" nipple. So copper is maybe $100 more in parts, but less effort. Would you use aluminium in that situation?
    – tom
    Oct 8, 2022 at 16:10

1 Answer 1


The conduit between boxes is short - less than 24" - that makes it a nipple, and a nipple is allowed 60% fill, not 40%. Use an IMC nipple, (bigger hole than PVC or EMT or rigid and it's conductive) and I get ~49% fill for the 3x 3AWG wires (and you don't even need a ground, technically.) You are still under 60% with an insulated 8 AWG so you're even safer with a bare one.

I think rigid will also squeak by just under the limit with bare 8 AWG, since it's barely over with insulated 8 AWG (60.03%). A shortcoming of my default fill calculator is that it doesn't do bare wires.

Right - forgot that IMC and rigid's thicker wall also starts with a larger outside diameter than EMT so the holes are bigger than EMT before I went and double-checked. Always learning...

  • I like it. So I'd run my 4 wires in 1-1/4 conduit to a Type-C conduit body, then run a short nipple from there into just one of the 3/4" knockouts. Is that what you're thinking? According to the fill calculator on Southwire's website EMT, IMT and RMC are all under 60% fill, IMC by the most. (The Type C will allow me to wiggle them through the nipple if needed). Thanks!
    – tom
    Oct 8, 2022 at 1:53
  • Sounds about right, yes.
    – Ecnerwal
    Oct 8, 2022 at 1:57
  • The ground does not need to go in the same conduit as the conductors. NEC 300.3. Oct 8, 2022 at 2:01
  • 1
    But it fits, so why complicate the matter?
    – Ecnerwal
    Oct 8, 2022 at 2:04
  • I calculated the fill factors using table 9 from the NEC. 3*0.0973 + 1*0.013 = 0.3049, which is fill factors of... EMT: 57.2%, IMT: 52.0%, RMC: 55.5%. My choice of nipple will probably come down to fit in the knockout and fit in the conduit body. Thanks again!
    – tom
    Oct 8, 2022 at 2:56

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.