Being over 60A, your charge station requires a disconnect switch. I don't make the rules, but this will save its cost on wire, because a disconnect is a cheap way to splice from aluminum to copper wire.
The textbook answer is #1 aluminum from panel to disconnect, and #3 copper from disconnect to EVSE. ("charger"). Source: Table 310.15(B)(16). 2025 edit: Let's expand that. A lot.
Determining correct wire size per NEC
Your station is 80 amps actual. NEC 625.41 says "[circuit breakers] shall have a current rating of not less than 125 percent..." So people go OK 100A breaker but the wire can still be 80A, right? Right?
Nope. Next article, 625.42, says "EV loads shall be considered continuous loads". What does that mean? 210.19(A) addresses this.
(A) General
Branch-circuit conductors shall have an ampacity not less than the larger of the following and comply with 110.14(C) for equipment terminations:
(1) Where a branch circuit supplies continuous loads or any combination of continuous and noncontinuous loads, the minimum branch-circuit conductor size shall have an ampacity not less than the noncontinuous load plus 125 percent of the continuous load in accordance with 310.14.
Exception to (1): If the assembly, including the overcurrent devices protecting the branch circuits, is listed for operation at 100 percent of its rating, the ampacity of the branch-circuit conductors shall be permitted to be not less than the sum of the continuous load plus the noncontinuous load in accordance with 110.14(C).
So yeah. The branch circuit conductors need to be sized for 125% of the actual load, meaning you need to use real, honest 100A conductors fit for 100A in the application you are using. So how do we determine what size wire that is? We get that out of Table 310.16 but that is indexed by tempearture. How do we determine temperature? The article mentions 110.14(C) twice, let's go there:
110.14(C) The temperature rating associated with the ampacity of a conductor shall be selected and coordinated so as not to exceed the lowest temperature rating of any connected termination, conductor, or device.
So "A convoy goes the speed of the slowest ship". The temperature rating of the wire is decided by the lowest temperature terminal or enclosure traversed by that wire. This is affirmed in 110.14(C)(1)(a) which applies to less than OR EQUAL TO 100A circuits:
(a) Termination provisions of equipment for circuits rated 100 amperes or less, or marked for 14 AWG through 1 AWG conductors, shall be used only for one of the following:
(1) Conductors rated 60°C (140°F) [being NM or UF cable]
(2) Conductors with higher temperature ratings, provided the ampacity of such conductors is determined based on the 60°C (140°F) ampacity of the conductor size used. [so other wires if we run them at 60C - not useful]
(3) Conductors with higher temperature ratings if the equipment is listed and identified for use with such conductors. [so just like the preamble says: limited by the equipment]
"The equipment" in this context is the EV "charger", the breaker enclosure, and the circuit breaker in question. And here's the gotcha you can't escape: the circuit breaker is rated 75 degrees C unless you are using very expensive industrial panelboards and breakers. So in residential we are always constrained to 75C in Table 310.16.
"Whoa, whoa, it said something about 310.12. I like that table a LOT better. Can't we just use that?" No. 310.12 is also a text rule, and the text rule states WHEN the table can be used. It can only be a service or feeder, it must supply the entire load to a dwelling. See
310.12(B) For a feeder rated 100 amperes through 400 amperes, the feeder conductors supplying the entire load associated with a one-family dwelling, or the feeder conductors supplying the entire load associated with an individual dwelling unit in a two-family or multifamily dwelling, shall be permitted to have an ampacity not less than 83 percent of the feeder rating. If no adjustment or correction factors are required, Table 310.12(A) shall be permitted to be applied.
Misapplication of 310.12 happens a LOT - if you surveyed 100 electricians and asked them to size a 100A subpanel feeder, probably 40% of them would use 310.12 incorrectly. It is relying on the "diversity of loads" found in a dwelling. Using them for an EV station would be a disaster, because an EV station is the exact opposite of a dwelling - a hard continuous load pegged right at wire rating.
So, we are back to 310.16 and the insufferable 75C column. That gives us
- #4 copper at 85 amps
- #3 copper at 100 amps
- #2 copper at 115 amps
- #1 copper at 130 amps
- #2 aluminum at 90 amps
- #1 aluminum at 100 amps
- #1/0 aluminum at 120 amps
- #2/0 aluminum at 135 amps
I'm mentioning aluminum because it can be used panel to disconnect.
Avoiding meltdowns
Now, over on Reddit and other EV forums, we see a lot of meltdowns. Boy howdy! The triggering event of these meltdowns is that the terminals are running at their thermal limits already merely by the continuous load right at thermal max rating. And then, a flaw or blemish creates a hotspot, which oxidizes, making it worse, causing a vicious cycle of destruction.
Therefore -- especially because aluminum wire is CHEAP -- I recommend bumping 2 numerical wire sizes, to 2/0 aluminum and #1 copper, or whatever the largest wire the terminals on both ends will accept. (this may be a limitation at the 100A breaker or EVSE). This will keep things significantly cooler, and very significantly reduce the risk of meltdowns.
Note that quality of work is absolutely essential, and you MUST use torque wrenches on all terminals per instructions and per NEC 110.14. "Gud-n-Tite" is scientifically proven to be unreliable.
