EVSE stands for Electric Vehicle Supply Equipment (aka wall "charger").
So you've heard of 80% and you've heard of 125%. You are also aware of the NEC 310.15(B)(16) table (I'm guessing?) but haven't heard the good news about aluminum.
There's only one load derate - 125%. Apply it once.
This is always in the disfavorable direction.
The rule is if you have an
X amp load, you must provision
1.25 X power to it or 125%. That's the rule. So if you have a 32A EVSE, you must "provision" 32 x 125% = 40A - and provision means the wire is at least that large, and the breaker is at least that large.
This 125% rule is repeated over and over in Code for a huge variety of applications. It's repeated so often that many people simply apply it to everything - and then, you get this weird inversion that isn't in Code at all - the so-called 80% derate. By this logic, if you have a 40A circuit/breaker, you can put 80% of that or 32A of any load (such as a tankless water heater or an EVSE). This is technically wrong, but since it only errs on the side of safety and overcompliance to Code, no one cares and no one will correct you. This sets the stage for...
...inadvertently double-applying the derate. For instance you colloquially call something a "40A EV charger" because its spec calls out a 40A circuit and breaker. You know it charges at 80% of that or 32A. But then you go "we need to apply the 125% rule, and 125% of 40A is 50A". No, that's a goof, you unnecessarily applied the derate twice. It derates once, from 32A actual x 125% to 40A.
Where does 125% apply? Only certain loads. But many. First, it applies to all "continuous loads" (anything that will run > 3 hours at a stretch) e.g. "not a coffee maker". It also applies to cord-and-plug connected loads (so yes, a coffee maker lol). Second, for certain loads, they remove your right to argue by imposing 125% - just a few examples are space heaters (baseboard heaters) and EVSE's.
You don't get to use 90C rating
In order to make use of the 90C thermal rating on wire, four things need to be true.
- The wire itself has to be rated for it - THHN is, UF is not.
- The conduit or wiring route has to be rated for it.
- The source terminals and enclosure must be rated for it - breakers and panels are only rated 75C.
- The destination enclosure and terminals must be rated for it.
Getting all four of those stars to align is hard. So as a practical matter, we invariably end up no better than 75C thermal - just plan on it. With NM and UF, we're also stuck at 60C thermal, but don't use that stuff anyway, it's made of copper. Also, older panels may be 60C rated (thanks Ed).
At large sizes, aluminum is your friend
America has lots of experience with aluminum wire, and it's always worked great for large feeder. The cheat-sheet is #1 use terminals properly rated for aluminum. #2 set torques with a torque driver. There were some problems with small branch circuit aluminum wiring in the 1970s, but that failed because of what I put in italics. Note that on larger feeder, electricians have always torqued those, and terminals have always been legitimately AL-rated (often made of aluminum).
We see people who are skittish of aluminum because of "folk knowledge" about the 1970s problems, and want to spend on copper "out of an abundance of caution" - but that makes no sense. If we are looking to turn dollars into safety, a rational study of risk would give 100 better places to spend that money - just in electrical, making sure everyhing is GFCI protected that should be.
By the way, EVSE's should not be GFCI protected. Why? Because EVSE's already contain a vastly superior GFCI, and feeding them from a run-of-the-mill GFCI defeats that improved protection. In particular, the EVSE's onboard computer will reset its own GFCI several times, to see if the problem has cleared, which it usually has. That way, you awaken to a recharged car and not a tripped GFCI. UL has approved this and this is part of the spec.