That table is NEC Table 310.16, and for developers it is genuinely interesting: it is a dense, multidimensional dataset with awkward column semantics, coupled constraints, and correction factors that stack multiplicatively. This article treats it the way I'd treat any other reference data — how to structure it, how to query it, and where the "gotchas" live that a naive implementation will get wrong.

A quick framing note before we start: this is educational material about how the code's data model works. It is not a design approval, not a substitute for your local electrical code, and not a replacement for a licensed electrician. If you are pulling permits or wiring a dwelling, the authority having jurisdiction (AHJ) has the final word, and it is often stricter than the book.

Why ampacity is a table and not a formula

Engineers love closed-form expressions. Ampacity resists one, because it is not fundamentally an electrical property — it is a thermal property. A conductor's current-carrying capacity is limited by how hot the copper (or aluminum) is allowed to get, and that depends on:

• the conductor's cross-sectional area (an AWG property),
• the resistivity and thermal conductivity of the metal,
• the insulation and jacket chemistry, which sets a maximum continuous conductor temperature,
• how the wire is bundled — free air, in a raceway, in a cable bundle, in insulation,
• the ambient temperature of the surrounding air,
• how many current-carrying conductors share the same enclosure.

So the "table" is really a family of empirical results, tabulated for a fixed set of assumptions (30 °C ambient, not more than three current-carrying conductors in a raceway or cable, 100% load factor). Everything else is a correction applied on top. If you think of it as a base case plus a set of multipliers, the structure becomes much easier to reason about — and much easier to encode.

The three columns and what they actually mean

Table 310.16 gives three ampacity columns per conductor size: 60 °C, 75 °C, and 90 °C. Beginners read these as "how hot the wire gets." That's backwards. The columns are indexed by the temperature rating of the termination and insulation system, not by the actual operating temperature.

The rule that trips people up: the 90 °C column is usually a derating ceiling, not a usable ampacity. You may use the 90 °C rating when you are applying adjustment and correction factors, but the final answer must still be compared against the 60 °C or 75 °C limit of the weakest termination in the circuit. In code language, the 90 °C column is an intermediate variable; the 75 °C column is almost always your return value.

A concrete example. #10 AWG copper is rated 30 A at 60 °C, 35 A at 75 °C, and 40 A at 90 °C. If your breaker, receptacle, and lugs are all 75 °C rated, a #10 copper conductor in a normal raceway tops out at 35 A — and if you're feeding a continuous load you'll derate from there. The 40 A number is a real number and it is almost never the answer you should ship.

Copper versus aluminum: not a flavor preference

Aluminum is not "cheaper copper." It is a different metal with roughly 61% of the conductivity of copper by volume, which is why an aluminum conductor must be about two AWG sizes larger to carry the same current. In Table 310.16 that shows up as a systematic offset: #6 copper is 65 A at 75 °C while #6 aluminum is 50 A at 75 °C.

The practical consequences for a hardware project are worth internalizing:

• Terminations matter more with aluminum. Aluminum oxidizes, creeps under thermal cycling, and needs terminals listed for AL conductors (often marked AL/CU). Using an AL conductor on a CU-only lug is a fire risk, not an inconvenience.
• Small aluminum sizes are effectively unavailable. Below #8 or so, aluminum branch wiring is a historical artifact (and a notorious one). For anything under 40 A, copper is the sane default.
• Ampacity is per-metal, so your data structure needs a metal axis. If you build a lookup keyed only by AWG, you have already introduced a bug. The correct key is (awg, metal, temp_column).

There's a fourth axis hiding in there too: insulation type, which determines whether a conductor is even allowed to use the 90 °C column. THHN/THWN-2 and XHHW-2 are 90 °C insulation. Older TW is 60 °C. If your dataclass doesn't carry an insulation rating, you'll eventually apply a 90 °C correction to a 60 °C wire.

Structuring the data

Here is a minimal, honest model in Python-ish pseudocode. The point is not the syntax — it's which fields are load-bearing.

from dataclasses import dataclass
from typing import Literal

Metal = Literal["cu", "al"]
Column = Literal[60, 75, 90]

@dataclass(frozen=True)
class Conductor:
    awg: str                 # "14", "12", "10", "8", "6", "4", "3", "2", "1", "1/0" ...
    metal: Metal
    ampacity: dict[Column, float]   # raw table values, no corrections
    insulation_max_c: Column        # 60, 75 or 90

@dataclass(frozen=True)
class Termination:
    max_c: Column            # weakest link: breaker, lug, receptacle

@dataclass(frozen=True)
class Circuit:
    conductor: Conductor
    termination: Termination
    continuous: bool
    ambient_c: float
    conductors_in_bundle: int

Two design decisions in that snippet are worth defending.

First, ampacity is keyed by column and stores the unmodified table value. Keep correction factors out of the base data. If you bake derating into the table you lose the ability to explain why a number changed, and you lose the ability to reuse the same row for a different ambient.

Second, Termination is a separate object from Conductor. That mirrors the physical reality: the wire's insulation rating and the lug's temperature rating are independent facts, and the usable ampacity is the minimum of the two. Collapsing them into one field is the single most common modeling error I see in hobbyist calculators.

The decision flow

For a branch circuit, the query resolves roughly like this:

1. Compute the design load. Sum continuous loads at 125% and noncontinuous loads at 100%. A 16 A continuous EV charger plus a 4 A noncontinuous light is a 24 A design load, not 20 A.
2. Pick the base ampacity. Look up the conductor in the column matching the weakest termination (60 °C or 75 °C in practice).
3. Apply ambient correction. The table assumes 30 °C. Above that, multiply by the appropriate correction factor — roughly 0.91 at 35 °C, 0.82 at 40 °C, 0.71 at 45 °C, 0.58 at 50 °C for the 75 °C column. These are approximations; use the published table for real work.
4. Apply bundle adjustment. More than three current-carrying conductors in a raceway or cable triggers an adjustment factor — 80% for 4–6 conductors, 70% for 7–9, and so on.
5. Take the minimum. The result must be ≥ the design load. If not, step up an AWG size and repeat.
6. Check the overcurrent device. The breaker must protect the conductor, and there are specific rules for the next-size-up allowance. This step is where DIY calculations most often go wrong, because it's the one with the most exceptions.

Steps 3 and 4 are multiplicative, and order doesn't matter — which is convenient, because it means you can implement them as a fold over a list of factors and unit-test each factor in isolation.

Worked examples

Example A — a 20 A kitchen circuit, copper, 75 °C terminations. #12 Cu is 25 A at 75 °C. Noncontinuous load: fine. If it were a continuous 16 A load, the design load is 16 × 1.25 = 20 A, still under 25 A. #12 works, which is why it's the residential default.

Example B — a 48 A continuous EVSE on a 60 A breaker. Design load is 48 × 1.25 = 60 A. #6 Cu at 75 °C is 65 A, so #6 copper is the minimum — and in practice most installs go to #6 Cu or #4 AL. Note that if the terminations were only 60 °C rated, #6 Cu drops to 55 A and you'd need #4.

Example C — the same load in a hot garage at 45 °C ambient. Take #6 Cu at 75 °C = 65 A, apply the 45 °C correction of about 0.71 → 46 A. That no longer covers a 48 A continuous load. You step up to #4 Cu (85 A at 75 °C) → 85 × 0.71 ≈ 60 A, which works. This is the single most under-appreciated correction in DIY work: a hot attic or garage can change your wire size by two AWG steps.

For the raw numbers behind these examples, a decent wire ampacity chart is faster than flipping through the code book, and it will show you the full NEC Table 310.16 ampacity rows side by side so you can see the Cu/Al offset directly rather than reconstructing it from memory.

Where implementations go wrong

A short list, drawn from watching people build these tools:

• Treating 90 °C as usable ampacity. It is a derating ceiling. If your calculator returns the 90 °C column as a final answer, it is wrong for nearly every real circuit.
• Forgetting the 125% continuous multiplier. Continuous means three hours or more of maximum current. EV charging, water heating, and most server loads qualify.
• Applying derating to the wrong column. You apply ambient and bundle factors starting from the conductor's insulation column (often 90 °C), then compare the result to the termination column. Applying them to the 75 °C number and then comparing to 75 °C double-penalizes the design.
• Ignoring the neutral. In some configurations the neutral counts as a current-carrying conductor for bundle adjustment. It's a real edge case and it changes the count.
• Mixing units. Circular mils, mm², AWG, and kcmil all appear in the same conversation. Pick one internal representation (I use AWG string plus a derived area) and convert at the boundary.
• Assuming the table is the whole answer. Voltage drop is a separate constraint with its own limits, and it frequently forces a larger conductor than ampacity alone on long runs.

Building a checklist instead of a calculator

If you're doing this for a real project, a calculator that silently returns a number is worse than a checklist that forces you to look at each constraint. The version I use:

• Load list written down, with continuous vs noncontinuous marked
• Design load = continuous × 1.25 + noncontinuous × 1.0
• Weakest termination temperature identified (breaker, lug, receptacle)
• Conductor metal chosen (Cu default; Al only with AL-rated terminations)
• Base ampacity read from the correct column for that metal
• Ambient correction applied (table assumes 30 °C)
• Bundle adjustment applied (table assumes ≤3 current-carrying conductors)
• Result ≥ design load
• Overcurrent device sized to protect the conductor
• Voltage drop checked separately over the actual run length
• Local amendments checked — the AHJ often differs from the model code
• Permit/inspection requirements checked before any wall is opened

That last pair of items is not boilerplate. Local amendments are real: some jurisdictions restrict aluminum branch wiring, some require AFCI/GFCI in more locations than the model code, some have different rules for EVSE circuits. A table lookup cannot know your zip code.

Closing thought

Table 310.16 is a good example of a dataset that looks trivial and isn't. Three columns, a dozen or so rows, and yet the semantics — which column is an intermediate value, which axis is a constraint, which factors stack — carry all the actual engineering. Treating it as structured data with an explicit metal axis, an explicit termination object, and a fold over correction factors gets you most of the way to a correct implementation. What it won't give you is judgment about your specific installation, and that's the part where a licensed electrician and your local code still earn their keep.

If you're sizing a run right now and want the numbers in one place, the copper wire ampacity by AWG tables are a reasonable starting reference — then verify against your adopted code edition and your AHJ before you buy wire.

Start with the temperature columns

NEC ampacity tables list three columns: 60°C, 75°C and 90°C. These refer to the insulation rating of the wire. A 90°C-rated wire such as THHN can carry more current on paper than a 60°C-rated one of the same size.

Terminal ratings usually limit you to 75°C

Breakers, lugs and devices are also rated, and most equipment above 100 A is rated for 75°C. In practice you may use the 90°C column only for derating calculations, but the final ampacity cannot exceed the 75°C column value unless the terminals are listed for 90°C. For small circuits with 60°C terminals, you are limited to the 60°C column.

Derate for heat and bundling

Hot environments reduce capacity, so apply the ambient temperature correction factor when the surroundings are above 30°C (86°F). Also apply an adjustment factor when a conduit holds more than 3 current-carrying conductors: for example 80% for 4 to 6 conductors and 70% for 7 to 9.

Check voltage drop on long runs

A wire that meets ampacity can still waste power over distance. As a rule of thumb, keep voltage drop to about 3% on a branch circuit and 5% total. On long runs, step up one or two AWG sizes to compensate.

Quick reference

For the full AWG ampacity tables and the 60/75/90°C values side by side, see the Wire Ampacity Chart. Always confirm the final sizing against the current NEC and your local code.