Is this conductor big enough?
You bring the ampacity out of your own book. This does the derating chain around it — including the termination limit that turns 90°C wire into 75°C ampacity.
Derating a conductor, and the limit people forget
The ampacity comes out of your book. What happens to it afterwards is arithmetic, and that is where conductors get undersized.
Open your ampacity table (NEC 310.16 or the equivalent in your adopted code) and read the conductor's ampacity twice: once in the insulation's temperature column, and once in the column matching what the terminations are rated for. Those two numbers do different jobs below, and the second one is the whole point of this page.
The conductor
Conditions in the raceway
The terminations — 110.14(C)
The breaker and the lugs have their own temperature rating, and it caps the answer no matter what the insulation can take. Leave this blank only if you genuinely intend to ignore it.
The chain, step by step
How this is calculated
derated = base × ambient correction × bundling adjustment
final = min(derated, termination ampacity)
The ambient correction is derived, not looked up. The published correction table is itself computed from √((Tc − Ta) ÷ (Tc − 30)), where Tc is the insulation rating and 30°C is the table's basis. Deriving it reproduces every printed factor to the two decimals the table carries, and it also answers for ambients between the rows instead of making you round to the nearest one.
The bundling adjustment is a real table with no formula behind it, so it is included as a small built-in list you can override: 4–6 conductors 80%, 7–9 70%, 10–20 50%, 21–30 45%, 31–40 40%, 41 and over 35%.
The termination limit is a comparison, not a third multiplier. This matters more than it sounds. Multiplying by the termination rating instead of taking the smaller of the two is a slip that produces an absurdly small answer, and because the answer is conservative it can survive review unnoticed while the real conductor is chosen by guesswork.
Why 90°C wire does not give you 90°C ampacity
This is the single most misunderstood rule in conductor sizing, and it costs people either money or safety depending on which way they get it wrong. THHN is rated 90°C, the 90°C column is generous, and it is tempting to size straight from it. You cannot. The conductor terminates on a breaker lug, and that lug is listed for 60°C or 75°C. Heat does not respect the difference: run the conductor at its insulation rating and the termination sits at a temperature it was never listed for.
What the 90°C column is good for is derating. You are permitted to start from it, apply your ambient and bundling factors to that larger number, and keep whatever survives — capped at the termination column. In a hot, crowded raceway this is worth real capacity, which is exactly why 90°C wire is worth buying. It is a better starting point, not a bigger finish line.
The two factors are not interchangeable
Ambient correction and bundling adjustment answer different questions and both apply. Ambient asks how hot the air around the raceway is; bundling asks how much of its own heat the conductor can shed given its neighbours. A conductor in a hot attic and in a full conduit gets both, multiplied. Applying only the worse of the two is a common shortcut and it is wrong — the code does not offer a choice between them.
What counts as current-carrying
Equipment grounding conductors never count: they carry no current in normal operation. The neutral is the interesting case. On a straightforward three-phase, four-wire system where the neutral carries only the unbalance, it does not count. On a system feeding substantial non-linear load — electronic ballasts, drives, large numbers of switching supplies — the harmonic current on the neutral is real and continuous, and the neutral does count. Getting this wrong in a panel feeder is how a neutral ends up running hotter than any phase.