What Derating Is, and Why a Conductor Rating Is Conditional

Why this matters

An ampacity is not a property of a wire. It is the answer to a heat question asked about one particular installation, and the wire is only one of the terms in it. That distinction sounds academic until the day a run that was correctly sized, correctly installed and correctly inspected becomes non-compliant with nobody having touched a conductor. Another trade pulled three circuits into a spare raceway. A replacement unit arrived with terminals listed one temperature step lower. Neither of those is electrical work on your circuit, and both of them changed its rating.

The load-bearing idea here is that the rating belongs to the installation, not to the conductor, so it expires when the installation changes. Everything else on this card is a consequence of that.

What the number on the table is a number for

A conductor's rating is fundamentally a TEMPERATURE, and the amperes are derived from it. The insulation has a maximum continuous operating temperature - 60, 75 or 90 degrees C on the common building wire types - and exceeding it does not blow anything, it ages the insulation faster. The ampacity is simply the current at which the conductor's own heating brings it to that temperature and holds it there under an assumed set of surroundings.

Two things generate the answer. Heat goes IN at a rate of current squared times the conductor's resistance. Heat goes OUT by conduction through the insulation, the raceway and the air, at a rate set by the temperature difference between the conductor and its surroundings, divided by the thermal resistance of the path. The published ampacity is the current where those two balance exactly at the insulation's limit.

Look at what is in that balance. The conductor supplies its resistance. The INSTALLATION supplies the ambient temperature and the thermal resistance of the escape path. Two of the three terms are not the wire's, which is why the wire cannot have a rating on its own, and why the same copper appears three times in the ampacity table under three different columns. The 60, 75 and 90 degree C columns are not three grades of copper. They are the same balance solved against three different permitted insulation temperatures.

Derating is not a penalty, it is the same equation solved again

The word "derating" makes it sound like a fine levied against a conductor that did something wrong. It is not. When the ambient rises, the temperature difference available to push heat out of the conductor shrinks, so less current is permitted before the insulation reaches the same limit. When conductors are bundled, the escape path's thermal resistance rises and each conductor is also being heated by its neighbours, so again less current reaches the same limit. In both cases the insulation limit never moved. The correction is just the balance re-solved for the surroundings you actually have.

That framing matters in the field because it tells you what a correction factor CAN and cannot be traded against. You cannot buy back a bundling adjustment with a better terminal, because the terminal is not in that balance. You can buy it back by improving the escape path, by splitting the circuits into separate raceways or by dropping the ambient. The mechanics of the two corrections - why ambient has a closed-form relationship and bundling is a tabulated step, and how the count of current-carrying conductors is decided - are worked in the ambient-and-bundling card, and the full multi-correction sequence with its ordering is worked in the derating-table card. This one is about why the number is conditional at all.

The limits that are NOT corrections

Here is the distinction that produces most of the field errors: some limits multiply the conductor's ampacity, and some cap it. They do not combine the same way and they are not fixed the same way.

Corrections multiply. Ambient and conductor count act on the heat balance, so they scale the number, and they scale it together.

Termination temperature caps. In the NEC as adopted by your authority having jurisdiction, 110.14(C) ties the conductor to the temperature rating of the equipment terminals it lands on. That is a limit on the TERMINAL's ability to survive the heat coming down the conductor, not on the conductor's insulation, so it is not part of the balance and it is not a multiplier. It is a ceiling applied to whatever the corrections produced.

Small-conductor rules cap the protective device. Also in the adopted NEC, 240.4(D) limits the overcurrent device on 14, 12 and 10 AWG copper regardless of what the ampacity table says. It does not reach 8 AWG or larger, so it does not appear in the worked example below, and stating that scope is the point: a limit that does not apply to your conductor should not be in your arithmetic.

Three different kinds of limit, on the same conductor, from three different mechanisms. The lowest one governs, and which one is lowest changes as conditions change. A tech who can name which limit is currently binding can predict what will break the installation; a tech who only has one number cannot.

Worked case: a rating that expired twice

A branch circuit was installed for a 38 A continuous load: three current-carrying 8 AWG copper conductors with 90 degrees C insulation, in one raceway, in a mechanical room at the standard 30 degrees C reference ambient, landing on equipment terminals listed for 75 degrees C.

At commissioning. A continuous load requires the conductor's governing ampacity to be at least 125 percent of it, so 38 times 1.25 is 47.5 A. The 8 AWG conductor's base ampacity in the 90 degrees C column is 55 A, and with three current-carrying conductors at the reference ambient there are no corrections to apply, so the corrected ampacity stays 55 A. The 75 degrees C termination cap for 8 AWG copper is 50 A. The lowest of those is 50 A, which is the governing number, and 50 against 47.5 gives about 5 percent margin. Compliant, inspected, correct.

Change one, eighteen months later. Another trade pulls three more current-carrying conductors through the same raceway for an unrelated circuit. Nobody touches the original conductors. With four to six current-carrying conductors in the raceway, the adjustment factor is 80 percent, so the corrected ampacity becomes 55 times 0.80, or 44 A. The termination cap is still 50 A. The lowest is now 44 A against a required 47.5 A, so the circuit is short by 3.5 A, or about 7 percent. The conductor did not change, the load did not change, and the installation is no longer compliant.

Change two, at the next equipment replacement. The unit is replaced with one whose terminals are listed for 60 degrees C only. The 60 degrees C ampacity for 8 AWG copper is 40 A, so the termination cap drops from 50 A to 40 A. Now the corrected ampacity is 44 A and the cap is 40 A, so the governing number is 40 A against a required 47.5 A, short by 7.5 A, or about 16 percent.

Read the two failures against each other, because they need different work. The first was a heat-balance failure, and the fixes are all about the escape path: pull the added circuits into their own raceway, or upsize, or reduce the load. The second was a terminal failure, and none of those fixes touch it - a bigger conductor still lands on a 60 degrees C terminal. That one is answered at the equipment, by a unit or a lug kit listed for a higher terminal temperature, or by sizing the conductor from the 60 degrees C column and accepting the size that follows.

Same circuit, same symptom on paper, two unrelated repairs. Anyone carrying "8 AWG is good for 50" as a fact about the wire gets both of them wrong.

Opening the raceway to find out

Everything in the worked case above was established by counting conductors and reading terminal markings, which means opening enclosures. Do that de-energized: 29 CFR 1910.333(a)(1) requires the circuit be de-energized before an employee works on or near exposed live parts unless the employer can demonstrate that de-energizing introduces additional or increased hazards or is infeasible due to equipment design or operational limitations, and a conductor count is not one of the cases that qualifies, because it can wait. Lock and tag under 29 CFR 1910.333(b)(2), which is the standard that owns electrical work - 1910.147 excludes exposure to electrical hazards from work on conductors and equipment in electric utilization installations at (a)(1)(ii)(C), and the construction counterpart for electrical lockout and tagging is 29 CFR 1926.417. Prove the conductors dead live-dead-live per NFPA 70E-2021 120.5, in the edition your employer's electrical safety program or your authority having jurisdiction has adopted.

If you need to know whether an existing raceway is running hot before you shut it down, read it with a non-contact infrared instrument from outside the restricted approach boundary rather than putting a hand on it. Insulation at its rated temperature will burn you, and a conductor that is over its rating will burn you badly.

What the rating never knew about

The published number is derived from an assumed installation, and there are conditions it does not model at all. These are not corrections you apply; they are reasons to be conservative or to ask an engineer:

• Sunlight on a rooftop raceway, which raises the effective ambient well above the shaded air temperature and has its own treatment in the adopted code.
• Thermal insulation over a cable, which is the clearest case of the escape path being destroyed without anyone touching the conductor. A cable buried in blown insulation or foamed into a wall cavity is in a different thermal world than the one that produced its rating.
• Harmonic current in a neutral, which is current the ampacity model did not expect to be there at all (the neutral card owns this).
• A connection with resistance. The balance assumes the conductor's own resistance. A loose lug adds a heat source the table has never heard of, in the one place with the worst escape path.

How to verify you got this right

The verification is a record rather than a reading, and it is the practical payload of this card. On any run you size or inspect, write down the four things the rating depends on, not just the wire size: the insulation temperature rating, the number of current-carrying conductors in that raceway, the ambient the run passes through, and the terminal temperature rating at both ends. Leave it in the panel or in the job file.

Then the check is whether the installation still matches the record, and any tech can run it in two minutes without recomputing anything. Count the conductors in the raceway and compare. Read the terminal markings on whatever equipment is there now and compare. If either has moved, the rating has moved with it and the calculation has to be redone. Without that record, the next person has a wire size and no way to know what it was ever conditional on, which is precisely how a compliant circuit becomes a non-compliant one that nobody can date.

References

• NEC Article 310 (conductor ampacities and the correction and adjustment provisions), 110.14(C) (temperature limitations at terminations) and 240.4(D) (small-conductor overcurrent protection), in the edition your authority having jurisdiction has adopted
• 29 CFR 1910.333(a)(1) and 1910.333(b)(2), general industry electrical safety-related work practices, with 1910.147(a)(1)(ii)(C) marking the exclusion; 29 CFR 1926.417 for the construction counterpart
• NFPA 70E-2021, 120.5, as adopted by your employer's electrical safety program or your authority having jurisdiction
• See related: The Derating Table and When It Applies to You; What Ambient and Bundling Do to a Conductor; How Heat and Current Relate in a Conductor