Long time reader and first time poster here and want to make sure I'm running power to my detached garage properly.

I built a detached garage next to my house and just got done running the wires from the main meter panel to the subpanel inside the garage and the setup is a follows:

Main panel is approximately 125 feet from the subpanel. I have a total of 4 wires coming from the main: 2x 2awg aluminum wires connected to a double pole 125 amp breaker for hot, 1x 2awg aluminum wire connected to ground bar for neutral, 1x 4awg aluminum wire connected to ground bar for ground.

I have also installed 2 8' ground rods, hammered in 12 feet apart to provide a local ground for the garage with a solid 6awg copper wire in addition to the 4awg wire from the main panel.

I'm not going to bond the neutral bar to the subpanel chassis. And there will be another 125 amp breaker in the subpanel as the mail breaker for quick, local disconnect of all power.

This is my first time doing any wiring at this scope and level and want to make sure I'm not missing anything.

Thanks in advance,


  • 1
    I'm pretty sure you'll need bigger than 2 AWG Aluminum for 125A. On the other hand, do you actually need 125A service in the garage? You can have a smaller feed (e.g., 80A or 100A) with a smaller breaker in the main panel and smaller wire, while still using 125A breaker (as disconnect) and big panel in the garage. Also not sure about putting the big subpanel neutral feed on the ground bar, rather than the neutral bar. Jan 17, 2021 at 18:55
  • 1
    My house was built in '84 and the mail panel doesn't have a separate neutral bar.
    – Kash
    Jan 17, 2021 at 20:08
  • @Kash -- no, it doesn't have a separate grounding bar -- every panel has a neutral bar Jan 17, 2021 at 20:53
  • Gotcha, guess I was backwards on that one!
    – Kash
    Jan 17, 2021 at 22:21
  • Hello, and welcome to Home Improvement. If an answer is helpful, please click the large check mark next to it to accept. And, please take our tour so you'll know the details of contributing here. Jan 18, 2021 at 3:33

3 Answers 3


A 125A "main breaker" in the subpanel is fine.

However, the breaker feeding this, which is in the main panel, must be 80A or 90A depending on the insulation of your wire.

Actually, I'd recommend a "larger" subpanel than 125A. The thing you really want is spaces, because a 90A subpanel can support a whole lot more circuits than you think it can. A whole lot more! A 30-space sub would not be excessive, and ones with that many spaces will tend to be 200A.

The old "16 spaces 32 circuits" trick doesn't work anymore, since almost all circuits these days require AFCI or GFCI breakers. NEC 2020 now requires it for 240V loads.

If you thought "125A breaker in the main panel is OK", then the source of your information is extremely poor - that's a hard mistake to make since every internet source will tell you otherwise. Revisit every one of your decisions which came from that data source.

Other than that, everything sounds OK, but 125A on #2Al is such a blunder that you should really go through the rest with a fine-tooth comb. Feel free to make full use of us!

  • I will definitely look over everything. I pulled my data from many places which is why I wanted to make sure everything was good before I terminated the wires and flipped the breakers on. Thank you for the information.
    – Kash
    Jan 17, 2021 at 22:28
  • NEC 110.14.C states wire sizes below #1 use the 60 degree table. 75 amp is the correct maximum for #2 aluminum on the 60 degree table. So 90 would be wrong code allows the next larger to be used since 75 is not a standard rating but 80 is.
    – Ed Beal
    Jan 18, 2021 at 14:37
  • @EdBeal My thinking is per 110.14.C.1.a.3, "Conductors with higher temperature ratings if the equipment is listed and identified for use with such conductors", which any modern breakers and subpanel lugs are listed for 75C. The next question is if the cable is good for 75C. NM and UF aren't, but I doubt it would be those if it's #2 aluminum, so it's a mystery. Jan 18, 2021 at 19:11

#2 aluminum is only rated for 75 amps below #1 the 60 degree table is used and even if the 90 degree table could be used (it can’t except for derating) that table tops out at 100 amps, So you need to change the feeder breaker in the main panel but other than that what you have will meet code. Make sure to use a anti oxide compound like noalox or deox on the aluminum and torque the conductors , release torque 3x as this will help seat the stranded wires.

  • I'll swap out the breaker. I did go way overboard with what I think I'll need!
    – Kash
    Jan 17, 2021 at 20:10
  • @EdBeal can you clarify what you mean about “release torque 3x”? Do you mean to torque the bolts, loosen them, and re-torque them 3 times?
    – Ivan G.
    Jan 18, 2021 at 2:32
  • 1
    If you only torque a large bundle of stranded wire it will lock in place in the shape it was, releasing the tension and then torque again you will notice you may get a half turn more release the tension and again torque the wires move around and have a better connection, large conductors should also be wiggled back and forth this will also loosen them up similar to torquing multiple times. This procedure creates a solid connection that won’t start arcing and sparking with a few heat cycles and is really critical on aluminum feeders.
    – Ed Beal
    Jan 18, 2021 at 14:10
  • So the goal is to basically make sure the strands get pressed together and form a stronger connection under the bolt. Good tip, thanks
    – Ivan G.
    Jan 18, 2021 at 14:28
  • @ivan g , Yes. This is something many electricians do not do I have been called in quite a few times to burned off wires and damaged terminals. A few years back we had a company come in and do some installs (copper) I inspected the feeders and was able to push back pull forward and then pull up and every feeder pulled right out, a couple of thermal cycles and these would have been arcing. Aluminum is worse but within the load diversity in a residential environment it is rarely cycled beyond 50% capacity so the problem is not as bad in residential plus the voltage is less.
    – Ed Beal
    Jan 18, 2021 at 14:50

Your breaker is too big for your wires

Your 2AWG Al wire is limited to 90A due to the fact that you're limited to 75°C ampacity by your terminations, at least with breakers and panels that are newer than the 1970s or so. Otherwise, you should be good to go, provided you haven't shorted yourself spaces in the subpanel that is! And yes, do make sure you torque your lugs down correctly; mistorque is a common cause of connection failures, after all.

  • Thank you, I'll change out the breaker in the main panel. So far I'm only planning on 9 circuits, 3 of which are 240v. That still leaves me room for a couple of more spaces if I need them in the future. I can also combine circuits in the future if need be since I'm running 3 separate circuits for lighting, 2 for 120v outlets and a separate one for each of the 240v outlets.
    – Kash
    Jan 17, 2021 at 20:14
  • @Kash -- why the 3 lighting circuits? Jan 17, 2021 at 20:52
  • Just for ease of working on it as needed. Half on one circuit, half on another with a separate one for a room I'm building in the corner. I may still put all on one circuit but I was also thinking of having a few outlet drops from the ceiling which would use the same circuit as the lighting.
    – Kash
    Jan 18, 2021 at 0:22
  • NEC 110.14.C states wire sizes below #1 use the 60 degree table. 75 amp is the correct maximum for #2 aluminum on the 60 degree table. So 90 would be wrong code allows the next larger to be used since 75 is not a standard rating but 80 is.
    – Ed Beal
    Jan 18, 2021 at 14:39
  • @EdBeal -- not if the equipment is listed for 60/75deg terminations -- see 110.14(C)(1) point 3 Jan 18, 2021 at 14:40

Your Answer

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

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