Ampacity Is Not a Table Number: The Derating Stack in Article 310

October 6, 2026

Every conductor size enquiry we get arrives with the same assumption buried in it: that a size has an ampacity, and the ampacity is a number in a table.

It has a number in a table, and then it has conditions. Article 310 says so in three places, and the third one is a prohibition rather than a table note. If you have ever sized a feeder off the 90 °C column and wondered why the inspector disagreed, this is where the disagreement lives.

The prohibition that makes derating necessary

310.15(A)(3) is the sentence the rest of the section exists to serve:

"No conductor shall be used in such a manner that its operating temperature exceeds that designated for the type of insulated conductor involved. In no case shall conductors be associated together in such a way, with respect to type of circuit, the wiring method employed, or the number of conductors, that the limiting temperature of any conductor is exceeded."

The second sentence names three things that can put a conductor over its limit: the type of circuit, the wiring method, and the number of conductors. Not the load. The number of conductors. That is the legal basis for the adjustment factor table, and it is why the count matters as much as the current.

The informational note that follows the section lists the four things that actually determine how hot a conductor runs, and it is a better mental model than any table.

Determinant	What it covers
Ambient temperature	Which may vary along the conductor length as well as from time to time
Internally generated heat	From load current, including fundamental and harmonic currents
Dissipation rate into the ambient medium	Thermal insulation that covers or surrounds conductors affects the rate
Adjacent load-carrying conductors	Which both raise the ambient and impede heat dissipation

Two of those four are decided by the installing contractor rather than by the load, and one by the cable construction. That is why a catalog ampacity figure is a starting point rather than an answer.

The base table is a two-condition number

The ampacity tables carry their conditions in the title. Table 310.15(B)(17), the free-air table, is titled as the allowable ampacities of single-insulated conductors rated up to and including 2000 volts in free air, based on ambient temperature of 30 °C (86 °F).

That is one condition stated openly: the table’s numbers assume a 30 °C surrounding. The second condition is in the adjustment table’s own column heading, which reads as a percentage of values in Table 310.15(B)(16) through (B)(19) — and the table it sits under applies, in its own words, "where the number of current-carrying conductors in a raceway or cable exceeds three."

So the everyday number is a two-condition number: three or fewer current-carrying conductors, and 30 °C ambient. Move either condition and the number moves with it.

How a base ampacity becomes a usable ampacity

There is a third condition too, in 310.15(B)(4): where bare or covered conductors are installed with insulated conductors, the temperature rating used is that of the lowest-rated insulated conductor in the group. One bare conductor in a bundle with 75 °C insulation does not get to behave like a bare conductor.

What counts as a current-carrying conductor

This is where most counting arguments start, and the code has already answered them.

Conductor	Counted?	Reference
A neutral carrying only the unbalanced current of the same circuit	No	310.15(B)(5)(a)
A neutral where the major portion of the load is nonlinear	Yes	310.15(B)(5)(c)
A grounding or bonding conductor	No	310.15(B)(6)
Each conductor of a paralleled set	Yes, each one	310.15(B)(3)(a)

The nonlinear-load clause is worth pausing on. A neutral in a wye system is normally not a current-carrying conductor, because it carries only the difference between the phase currents. Put electronic loads on that system and the triplen harmonics add in the neutral instead of canceling, so the neutral becomes a loaded conductor, and the code stops pretending otherwise.

The adjustment table

Table 310.15(B)(3)(a) is short and worth knowing by heart, because it is the one that bites.

Number of current-carrying conductors	Percent of table value
4 to 6	80
7 to 9	70
10 to 20	50
21 to 30	45
31 to 40	40
41 and above	35

The adjustment factors and the step that costs you

The table’s own footnote settles what goes into the count: the total number of conductors in the raceway or cable, adjusted per 310.15(B)(5) and (6). The counting rules and the table are one provision, not two.

Look at the step from 9 to 10. One more conductor costs 20 percentage points. A ten-conductor raceway runs at half of the table number, and the tenth conductor is often a neutral somebody decided to count, or a phase somebody added without re-running the calculation.

The section text also reaches outside raceways. Conductors or cables installed without maintaining spacing for a continuous length longer than 600 mm (24 in.) and not in a raceway are subject to the same reduction. A bundle of cables strapped together across a long cable run is not exempt just because it is not in conduit. Spacing is the alternative to derating, which is why support and spacing decisions belong in the same conversation as conductor sizing.

Two counting rules inside the adjustment section

Two sentences in 310.15(B)(3)(a) are easy to miss and change the answer.

Paralleled sets count individually. "Each current-carrying conductor of a paralleled set of conductors shall be counted as a current-carrying conductor." A parallel run of three phases in two sets is six current-carrying conductors, not three. If the designer sized the parallel set to make the ampacity work and then left the count at three, the count is wrong and the conductor is undersized.

Another note points at a different set of numbers. Informational Note No. 1 to the section refers the reader to Annex B, Table B.310.15(B)(2)(11), "for adjustment factors for more than three current-carrying conductors in a raceway or cable with load diversity." Load diversity means the conductors are not all at full load at the same time, which is the normal condition on a panel feeder and almost never the condition the everyday table assumes. It is a route worth knowing exists before accepting a 50 percent factor as the final word on a large raceway.

Mixed systems count differently. Where conductors of different systems are installed in a common raceway or cable, per 300.3, "the adjustment factors shown in Table 310.15(B)(3)(a) shall apply only to the number of power and lighting conductors (Articles 210, 215, 220, and 230)."

That is a narrower rule than most people assume. A raceway carrying power conductors alongside Article 725 Class 2 circuits does not put every conductor in the pipe into the count. The adjustment applies to the power and lighting conductors; the others are present, and they are not part of the fraction.

The exceptions, and the one that sets the factor to 60 percent

The adjustment factors do not always apply, and the exceptions list is, in order:

Exception	Effect
Cables in cable trays	392.80 governs instead
Raceways 600 mm (24 in.) or less in length	Adjustment factors do not apply
Underground conductors entering or leaving a trench, in RMC, IMC, PVC or RTRC protection not over 3.05 m (10 ft), four or fewer conductors	Adjustment factors do not apply
Type AC or MC cable without an overall outer jacket, three or fewer current-carrying conductors per cable, 12 AWG copper, 20 or fewer conductors stacked or bundled	Adjustment factors do not apply
Type AC or MC cable without an overall outer jacket, more than 20 conductors stacked or bundled over 600 mm (24 in.)	A 60 percent factor applies

The last row is the one worth flagging. It is the only place in the list where the exception is not a relief but a substitute number — a stack of more than twenty single-conductor-armored cables gets 60 percent, which is better than the 35 percent the table would give and worse than nothing. If you are buying or specifying metal-clad cable for a bundled run, that row decides the size.

Ambient correction, and the rooftop adder

Article 310 carries two ambient correction tables, and the difference between them is the point: Table 310.15(B)(2)(a) is based on 30 °C (86 °F) and Table 310.15(B)(2)(b) is based on 40 °C (104 °F). You use the one that matches the basis of the ampacity table you started from. Picking the wrong one is a silent error — the arithmetic still works, and the answer is wrong.

The 90 °C column of the 40 °C-based table looks like this.

Ambient °C	Correction factor
10 or less	1.26
11 to 15	1.22
16 to 20	1.18
21 to 25	1.14
26 to 30	1.10
31 to 35	1.05
36 to 40	1.00
41 to 45	0.95
46 to 50	0.89
51 to 55	0.84
56 to 60	0.77
61 to 65	0.71
66 to 70	0.63
71 to 75	0.55
76 to 80	0.45

Note that the factor is 1.00 in the 36 to 40 °C row, which is what "based on 40 °C" means, and that the table pays a bonus below 30 °C. The same table punishes a mechanical room hard. At 51 to 55 °C a 90 °C conductor is down to 0.84, and at 66 to 70 °C it is at 0.63.

Then there is the roof. For raceways exposed to direct sunlight on or above rooftops, the adjustments in Table 310.15(B)(3)(c) are added to the outdoor temperature to determine the ambient used for the correction factor.

Distance above roof to bottom of conduit	Adder °C	Adder °F
0 to 13 mm (½ in.)	33	60
Above 13 mm (½ in.) to 90 mm (3½ in.)	22	40
Above 90 mm (3½ in.) to 300 mm (12 in.)	17	30
Above 300 mm (12 in.) to 900 mm (36 in.)	14	25

The rooftop adder is a stand-off decision

The first row is the one that gets forgotten. A conduit laid essentially on the roof surface carries a 33 °C adder, so a 35 °C summer afternoon becomes a 68 °C ambient for the purpose of choosing a correction factor. Raising the same conduit 100 mm above the roof drops the adder to 17 °C. Stand-off height is a derating decision, and it is decided by a strut clamp rather than by a cable.

One number, walked through

Take a conductor with a base ampacity of 100 A from the table.

Eight current-carrying conductors in the raceway: that is the 7 to 9 row, so 0.70.

Ambient 45 °C: on the 40 °C-based correction table that is the 41 to 45 row, so 0.95.

100 × 0.70 × 0.95 = 66.5 A

Now put that raceway on a roof, 20 mm above the surface. The adder for a conduit between 13 mm and 90 mm above the roof is 22 °C, so the ambient to use is 45 + 22 = 67 °C. That is the 66 to 70 row, so 0.63.

100 × 0.70 × 0.63 = 44.1 A

Two conditions, one of which was created by where the conduit sits, and a 100 A conductor is now good for 44 A. The termination limit under 110.14(C) still sits on top of that, and it caps the result at the ampacity of the temperature rating of the terminations. That half of the problem is the subject of our voltage drop and conductor sizing guide, and it is worth reading next rather than after the inspection.

The example shows the shape of the whole thing. The corrections multiply. The adjustment is a table lookup on a count that people get wrong. The rooftop adder changes the ambient before the lookup happens. None of it is difficult, and all of it happens before anyone looks at the load.

FAQ

Question	Answer
Does a neutral count as a current-carrying conductor?	Not where it carries only the unbalanced current of the same circuit. It does count where the major portion of the load is nonlinear, because triplen harmonics do not cancel in the neutral.
Does a paralleled set count as one conductor?	No. Each current-carrying conductor of a paralleled set is counted separately.
Do conductors of other systems in the same raceway count?	The adjustment applies only to the number of power and lighting conductors, per Articles 210, 215, 220 and 230.
Do cables outside a raceway get adjusted?	Yes, where they are installed without maintaining spacing for a continuous length longer than 600 mm (24 in.).
Which ambient correction table applies?	The one matching the basis of the ampacity table you started from: 310.15(B)(2)(a) is based on 30 °C and 310.15(B)(2)(b) on 40 °C.

What to specify

Most of this article is about the installation, and the part a cable supplier controls is narrower but real.

The conductor material decides which column you start in, and a copper-clad aluminum conductor that is presented as copper puts the whole calculation on the wrong row. Our guide to verifying conductor material covers how that is checked.

The insulation system decides which temperature column is available at all, so a 90 °C-rated product is only useful if the terminations and the environment allow 90 °C to be used. Our guide to CM, CMR and CMP selection covers the listing side of that, and spacing decisions that avoid the adjustment entirely are in the pathway support guide.

Our cable range covers the constructions this calculation is normally run against, and OEM and ODM projects are where a specific temperature rating or conductor configuration gets built.

Sources

NFPA 70, National Electrical Code, 2011 edition — Article 310 (Conductors for General Wiring): 310.15(A)(3) with its Informational Note No. 1, 310.15(B)(2)(a) and (b), 310.15(B)(3)(a) with Table 310.15(B)(3)(a) and its exceptions, 310.15(B)(3)(c) with Table 310.15(B)(3)(c), 310.15(B)(4), 310.15(B)(5), 310.15(B)(6), and Table 310.15(B)(17). These are the sections the article quotes or cites. Article 110: 110.14(C). Article 300: 300.3. Article 392: 392.80. Public-domain copy consulted at Internet Archive, gov.law.nfpa.nec.2011.