The Derating Table and When It Applies to You

Why this matters

A derating table is documentation that exists to take a number away from you. Every published rating is measured under test conditions, and a derating table is the manufacturer or the code telling you how much of that rating survives contact with your actual installation. Techs get burned two ways. They apply a table that does not govern their situation and undersize themselves out of a job they could have done, or they skip a correction that does govern and install something that works fine for four years and then fails in the first genuinely hot week. The second one is the callback that comes with an insurance adjuster.

What the table is a table of

A derating table has three parts, and you need all three to know whether it is yours.

• A base rating, which is a measured value under a stated reference condition. That condition is the part nobody reads, and it is the only thing that makes the correction meaningful.
• A varying condition, which is the one thing the table lets you change: ambient temperature, elevation, number of conductors bundled together, duty cycle.
• A multiplier, which is almost always less than 1.

What a derating table is not is a complete answer. It is one correction out of several, and the corrections that apply to a given install are usually printed in different documents by different authors. That is the single most common way a correctly-applied table produces a wrong installation.

Worked: one circuit, four corrections, and the one that governs

A feeder runs six current-carrying copper conductors, 8 AWG with 90 degrees C insulation, through a raceway in an equipment room that reaches 45 degrees C in summer. The load is 34 amperes, continuous. Terminals on both ends are listed for 75 degrees C.

Before you open any enclosure to verify what is actually installed, de-energize it. For work on or near exposed energized conductors, 29 CFR 1910.333(b)(2) requires the circuit be de-energized and locked and tagged unless de-energizing introduces additional or increased hazards or is infeasible, and the proving sequence is test the meter on a known live source, test the conductors, test the meter again, per NFPA 70E-2021, 120.5.

Start at the conductor's own insulation column. An 8 AWG copper conductor with 90 degrees C insulation carries a base ampacity of 55 amperes in the standard ampacity table.

Correction one, ambient. The ampacity table is referenced to a 30 degrees C ambient. At 41 to 45 degrees C, the correction factor in the 90 degrees C column is 0.87. So 55 times 0.87 is 47.85 amperes.

Correction two, bundling. With four to six current-carrying conductors in the same raceway, the adjustment factor is 80 percent. So 47.85 times 0.80 is 38.28 amperes. Note that these two multiply, they do not average, and they do not compete: whoever wrote each table assumed the other would also be applied.

The gate that is not a multiplier. Termination temperature caps the whole result. With 75 degrees C terminals, the conductor cannot be used above its 75 degrees C ampacity, which for 8 AWG copper is 50 amperes. Our corrected 38.28 is already below that, so the corrections govern and the terminal rating does not bind here. On a lightly-corrected circuit the terminal rating is often the binding number instead, and a tech who derates from the 90 degrees C column and stops has skipped the only limit that mattered.

The load side of the comparison. A continuous load requires the conductor's corrected ampacity to be at least 125 percent of it. That is 34 times 1.25, or 42.5 amperes. Our corrected conductor gives 38.28. It fails, by about 10 percent.

Step up and redo it, all of it. A 6 AWG copper conductor with 90 degrees C insulation has a base ampacity of 75 amperes. Corrected: 75 times 0.87 is 65.25, times 0.80 is 52.2 amperes. The 75 degrees C ampacity for 6 AWG is 65 amperes, so the terminal cap again does not bind. And 52.2 clears the 42.5 requirement with room. That is the install.

The failure mode of stopping early is specific and it looks fine on inspection day. Apply the ambient correction only and you land at 47.85, clear the 42.5 requirement, and pull 8 AWG. The circuit works every day it is under 40 degrees C in that room. It cooks its insulation on the days it is not, and the damage is cumulative and invisible until a termination discolors.

The corrections that live somewhere else

Nothing in the paragraph above came from one table. Ambient correction and conductor-count adjustment are two separate tables in the wiring code's conductor article. Termination temperature limitation is in the general-requirements article, hundreds of pages away. The continuous-load factor is in the branch-circuit and feeder articles. Four documents, one circuit.

Table numbering inside that conductor article has also moved between code editions, which is exactly the failure this whole group is about: a note in your phone that names a table number and not the rule is worthless the next cycle. Record the rule and the article, and look up the table number in the edition your jurisdiction has adopted.

What the table cannot see

This is where derating tables are most dangerous, because the omissions look like permissions.

• Voltage drop is not in any derating table, and in the United States it is generally not an enforceable ampacity limit at all, only an informational recommendation. A conductor can pass every correction and still deliver too little voltage at the end of a long run, and the equipment at the far end will show you a fault that has nothing to do with the equipment.
• Solar heat gain is a separate adder, not part of ambient. A raceway exposed to sunlight on a roof runs hotter than the shaded air around it, and the correction for that is its own table keyed to height above the roof surface.
• A neutral that carries harmonic current counts as a current-carrying conductor for bundling purposes even though a balanced neutral on a three-phase circuit normally does not. Nonlinear loads are the trigger, and the count changes the adjustment factor, which changes the answer.
• The table assumes a sound termination. Every ampacity in the book assumes the connection is made to the right torque on a clean, correctly-sized lug. A loose lug produces heat the table never modeled, and no amount of correct derating protects against it.
• The table assumes the condition is the worst it gets. Techs read summer ambient off the day they are standing there. Use the design maximum for that space, and remember that a room housing heat-producing equipment gets hotter after the equipment is fully loaded, not on install day.

Where derating does not apply at all

Equally important, and equally missed in the other direction.

• Conductors that do not carry current. Grounding and bonding conductors do not count in the bundling number.
• Short raceway sections. The bundling adjustment does not apply to a raceway short enough to act as a nipple rather than a thermal enclosure; the code's threshold has long been raceways not exceeding 24 inches in length. That exception has saved a lot of unnecessary upsizing at gutter-to-panel transitions.
• Loads that cannot run at the same time. If the design prevents two loads from operating simultaneously, only one of them counts as current-carrying, but this depends on the interlock actually existing in hardware or listed control, not on the customer promising not to run both.
• Motor service factor. A motor nameplate service factor above 1.0 is a permitted overload allowance, not a derating. Continuous operation in that band shortens insulation life, and under NEMA MG 1 the service factor is defined for the nameplate ambient, commonly 40 degrees C, and up to roughly 3,300 feet of elevation. Outside those, the extra margin is not there to spend.

The same idea in three other trades

Gas appliance input at elevation. In the United States, altitude derating is generally addressed above 2,000 feet, and the appliance manufacturer's own altitude table governs where one exists. Where none does, the National Fuel Gas Code (NFPA 54 / ANSI Z223.1) has long carried a reduction on the order of 4 percent per 1,000 feet above that threshold, applied by orifice change rather than by manifold pressure alone. Do not derate by guessing at a gas valve.

Engine and generator output at elevation and temperature. Naturally aspirated engines lose output as air density falls, and a commonly cited field rule of thumb is on the order of 3 percent per 1,000 feet, with a further reduction for high inlet air temperature. Turbocharged units behave differently and hold output further up. The manufacturer's derate table is the only one worth using for a sizing decision, because the rule of thumb is a sanity check, not a spec.

Rigging and sling angle. A sling's rated capacity applies to a vertical pull. Angle it and the tension rises by the reciprocal of the sine of the angle from horizontal: at 60 degrees the factor is about 1.155, at 45 degrees about 1.414, at 30 degrees exactly 2.0. Two legs at 30 degrees put the full load on each leg, not half. This is the same structure as an ampacity correction, applied to steel instead of copper, and it is the one where getting it wrong drops equipment.

How to verify you applied it right

Four checks, in this order, before the number goes on a quote or into a panel.

1. Name the reference condition out loud. "The 55 is a 30 degrees C number." If you cannot state what the base rating assumed, you do not yet know whether the correction is needed.
2. List every correction and multiply them in one line. Write it as one expression, base times factor times factor, so a skipped step is visible rather than buried in a mental running total.
3. Check the cap separately from the multipliers. Terminal ratings, relief settings and structural limits are ceilings, not factors. They do not multiply and they never raise a number.
4. Compare against the load requirement, not the load. Continuous loads, motor loads and duty-cycle loads each carry their own factor on the demand side. A corrected ampacity that clears the raw load and fails the required load is the most common near-miss on this whole procedure.

References

• NFPA 70 (National Electrical Code), Article 310 for conductor ampacity, ambient correction and conductor-count adjustment, and Article 110 for termination temperature limitation; table numbering within Article 310 has changed across editions, so use the edition your jurisdiction has adopted
• 29 CFR 1910.333(b)(2), de-energizing and lockout or tagging before work on or near exposed energized parts
• NFPA 70E-2021, 120.5, the test-before-and-after voltage verification sequence
• NFPA 54 / ANSI Z223.1 (National Fuel Gas Code) for altitude adjustment of gas appliance input
• NEMA MG 1 for motor service factor and its rated ambient and elevation basis