Checked against primary sources 2026-08-24
Two multiplications, and a limit that never gets multiplied
The arithmetic is two multiplications. The marks are lost before either of them, in picking the wrong correction table, and after both of them, in running the termination check on the wrong number.
On this page
- The sequence, start to finish
- Which correction table goes with which ampacity table
- The rooftop adder goes on the ambient
- The count is the question, the factor is the lookup
- Where the adjustment does not apply at all
- Why the 90 degree column is the right starting point
- A worked example, with the two tests kept apart
- The two limits that finish the problem
- What this page cites
The sequence, start to finish
Six moves, and only two of them are multiplications.
- Read the ampacity from the column matching the conductor insulation rating. For a raceway, a cable or direct burial that is Table 310.16, and for a 90 degree insulation it is the 90 degree column.
- If the ambient is anything other than the base temperature of that ampacity table, multiply by the ambient correction factor from the correction table that ampacity table names. Where a rooftop is involved, the adder goes on the ambient before this lookup, not on the answer after it.
- If more than three current-carrying conductors are bundled, multiply again by the adjustment percentage for the count.
- That product is the conductor ampacity. It has to cover the load.
- Separately, and using no correction and no adjustment at all, read the ampacity in the column that matches the termination temperature rating from 110.14(C). That number has to cover the load as well.
- If the conductor is 14, 12 or 10 AWG, apply the small conductor cap in 240.4(D) to the overcurrent device.
Steps two and three commute, because multiplication does. Nothing else on that list does. Step two has an adder inside it that has to happen first, step five uses a number that steps two and three never touched, and step six happens after both factors have been applied, which 240.4(D) says in its own opening sentence.
The single most expensive misunderstanding in this subject is treating step five as though it were step four continued. The termination limit is not the end of the multiplication chain. It is a second, parallel test on the raw table value, and the conductor has to pass both.
Which correction table goes with which ampacity table
Article 310 carries two ambient correction tables and they hold different numbers for the same ambient. One is built on a thirty degree Celsius base, the other on forty.
Each one serves a named set of ampacity tables, and each prints a note saying which. Table 310.16, the everyday one for conductors in a raceway, in a cable or directly buried, carries a note in the other direction that sends you to 310.15(B) for correction. Follow that pair of pointers and the error cannot happen.
It happens constantly anyway, because both answers look reasonable. Neither produces an absurd number. You get a wrong conductor size with no signal that anything went wrong, which is the worst kind of error to have on a timed paper.
The habit that fixes it costs four seconds. Before you write down a correction factor, look at the top of the correction table you are reading and confirm the base temperature is the one your ampacity table was built on.
The rooftop adder goes on the ambient
Where a raceway or cable is in direct sunlight on or above a rooftop and the gap from the roof surface to the bottom of the raceway or cable is under three quarters of an inch, 33 degrees Celsius, which is 60 degrees Fahrenheit, is added to the outdoor ambient. You then look up the correction factor at that higher figure (NEC 310.15(B)(2)).
Order matters here in a way it does not between the two multiplications. Adding the adder to the outdoor temperature and then reading the table gives a different answer from reading the table at the outdoor temperature and applying the adder afterward, which is not a defined operation at all.
The distance is the trigger and it is small. Three quarters of an inch of standoff removes the adder entirely, which is why you see raceways on strut over flat roofs.
One insulation type is excused by an exception printed under the subsection. Find it in your own code book and mark it, because an item that names that insulation type on a roof is testing whether you know the exception exists.
The count is the question, the factor is the lookup
Adjustment applies once more than three current-carrying conductors run in the same raceway or cable without maintaining spacing, for a continuous length longer than twenty four inches. The table then gives a percentage in bands by count.
Reading the band is trivial. Arriving at the number to look up is the whole item, and it turns on which conductors count. 310.15(E) decides the neutral, and the 2026 edition splits that section into a positive list and an exclusion that only reaches what the list left. 310.15(F) keeps grounding and bonding conductors out. Spare conductors stay in.
That subject has its own page on this site, and it is worth the visit before you drill derating, because a misread count produces a clean wrong answer that no later step contradicts.
Where the adjustment does not apply at all
Four conditions in 310.15(C)(1) switch the adjustment off, and two of them turn up in items regularly.
- A raceway no longer than twenty four inches. The bundle is too short to build heat, so the factors do not attach.
- Underground conductors entering or leaving an outdoor trench, where the physical protection is rigid metal conduit, intermediate metal conduit, PVC or RTRC, the protected length is no more than ten feet, and there are no more than four conductors.
The twenty four inch case has a companion in Chapter 9. The note that permits a nipple of that length to be filled to sixty percent says in the same breath that the adjustment factors need not apply to that condition. Two rules, one physical reason, and an item that puts a stuffed nipple between two cabinets is usually reaching for both.
The other two conditions in 310.15(C)(1) are narrower, and one of them carries four separate qualifying tests on Type AC and Type MC cable. Read them in the book rather than from memory.
Why the 90 degree column is the right starting point
The correction tables are built from an equation the code prints in 310.15(B)(1), and the equation takes the conductor temperature rating, the new ambient and the ambient the table was computed at. Conductor size is not in it.
So the correction factor depends on the insulation rating and nothing else, and above the table base temperature a higher rating always takes the smaller reduction. That is the entire argument for starting at the 90 degree column even when the terminations are 75 degree: you have more ampacity to spend before the correction and adjustment have finished with it.
The relationship flips below the base ambient. In a cold ambient the factors are greater than one and it is the 60 degree column that gains the most, proportionally. That case shows up rarely and it is worth knowing about only so you do not carry "90 is always gentler" as a rule and get caught by the one item that inverts it.
None of this raises the ceiling. Starting high gives you room to derate. It does not move the termination limit, which is read from a different column and never multiplied.
A worked example, with the two tests kept apart
Twelve current-carrying conductors of one size and one insulation share a raceway. Ambient is 45 degrees Celsius. The insulation is rated 90 degrees. The terminations are ordinary 75 degree lugs.
- Table 310.16, 90 degree column, the size in question. Write it down.
- Note 1 under that table sends you to 310.15(B). Confirm you are in the thirty degree base correction table, then read the 45 degree row in the 90 degree column.
- Twelve conductors falls in one band of the adjustment table. Read that percentage.
- Multiply the table ampacity by the correction factor, then by the adjustment percentage. That is the conductor ampacity, and it is the number the load has to fit under.
- Now go back to Table 310.16 and read the same conductor size in the 75 degree column. No factors. That is the termination limit under 110.14(C).
- The answer is whichever of those two numbers is smaller.
Run that pair of numbers on any size you like and you will find the derated 90 degree figure comes out well under the untouched 75 degree figure, because twelve conductors at 45 degrees is a punishing combination. The termination limit is not binding here. Change the stem to three conductors at 30 degrees and it becomes the only thing that binds. Both stems are common, and the reason candidates get one of them wrong is that they only ever practiced the other.
The two limits that finish the problem
The small conductor cap
For 14, 12 and 10 AWG the overcurrent device is capped by 240.4(D) regardless of what the derating produced, and the section says outright that the cap applies after correction and adjustment have been applied. A derated ampacity above the cap does not lift it.
That is why an item can walk you through a full derating and then ask for the breaker size. The derating is there to see whether you stop at the right place.
When two ampacities apply to one circuit
A run that changes conditions partway, through a hot attic, out into free air, has more than one ampacity along its length. 310.14(A)(2) says the lowest of them governs, and then carries an exception for a short transition: the higher ampacity is permitted to continue past the point of change for the lesser of ten feet or ten percent of the circuit length figured at the higher ampacity.
Items on this are recognizable. They give you two environments and a distance, and the distance is there to be measured against that exception.
What this page cites
- NEC 310.15(B) Ambient temperature correction. (B)(1) General carries the equation the correction tables are computed from and the two correction tables themselves. (B)(2) Rooftop carries the three quarter inch distance, the 33 degree Celsius adder and the insulation exception. Verified against NFPA Code-Making Panel 6 first draft report for the A2025 cycle, which produced the 2026 edition, and against a direct read of the 2026 text. source
- NEC 310.15(C) Adjustment for more than three current-carrying conductors, including the twenty four inch spacing threshold and the conditions under which the adjustment does not apply. source
- NEC 310.14(A) Selection of ampacity. (A)(2) is the rule that the lowest applicable ampacity governs over a circuit length, with the exception for a transition of ten feet or ten percent of the circuit length, whichever is less. source
- NEC Table 310.16 Ampacities for insulated conductors in raceway, cable or earth. Note 1 beneath it names 310.15(B) as the source of correction factors, which is the pointer that identifies the correct correction table.
- NEC 310.15(E) When a neutral conductor counts as current-carrying. Restructured for the 2026 edition into a positive list at (E)(1) and an exclusion at (E)(2) that reaches only what the list did not take.
- NEC 310.15(F) Grounding and bonding conductors are not counted when the adjustment factors are applied.
- NEC 110.14(C) Temperature limitations at terminations. Read alongside this page rather than inside it: the termination check uses an ampacity column, not a derated figure.
- NEC 240.4(D) Small conductors. The opening sentence places the cap after ambient and conductor count factors have been applied, which is what fixes its position in the sequence.
- NEC Chapter 9, Table 1 Note 4 permits a nipple of twenty four inches or less to be filled to sixty percent and states that the 310.15(C)(1) adjustment factors need not apply to that condition.