Short answer: an MCB is calibrated to carry its rated current at a reference ambient of 30 °C under IEC/EN 60898-1, and at 40 °C under IEC 60947-2. Above that reference, the thermal element has less margin before it reaches its trip point, so the current it will actually carry falls — roughly 6–8% at 40 °C, 13–19% at 50 °C and 19–24% at 60 °C on typical miniature devices. The fix is not a bigger breaker chosen by feel: it is to take the manufacturer’s derating table for the exact device, work out the real temperature inside the enclosure rather than the temperature of the room, and then check that the derated breaker still protects the cable.
This is the quiet failure mode in a lot of export boards. A 40 A breaker on a 40 A cable, in a metal enclosure on a roof in a 45 °C climate next to three contactors, is a 40 A breaker carrying maybe 33 A and a cable expecting protection at 40 A. Nothing looks wrong on the schedule. The breaker trips on a hot afternoon at 85% load and the buyer concludes the breaker is faulty. It is not faulty; it is hot.
We write as an independent trading company — not a manufacturer, and not an appointed distributor of any brand — so what follows is selection logic, not a badge.
Why the thermal element cares about the room
An MCB trips thermally when a bimetal strip, heated by the current passing through it, bends far enough to release the latch. The strip does not measure current directly; it measures its own temperature, which is ambient plus the rise caused by the current. Two consequences follow: the trip current is a curve that shifts with ambient, and the shift is not linear in current, because heating goes with the square of the current while heat lost to the surroundings goes with the temperature difference.
Write it as I²R heating balanced against a fixed temperature ceiling. If the device reaches its trip temperature at rated current with a rise of ΔT above 30 °C, then at ambient θ the current that produces the same rise satisfies I² ∝ (ΔT + 30 − θ), which gives the working approximation:
I(θ) ≈ In × √(ΔT / (ΔT + θ − 30))
With a rise of about 60 K at rated current — reasonable for a small miniature breaker — that gives the numbers below. Treat them as the shape of the curve, not a substitute for the datasheet: the bimetal, housing material and calibration tolerance move the result by a few percent either way, and the manufacturer publishes the real table for the exact catalogue number.
| Ambient at the device | Approximate carry current (% of In) | 40 A device carries | 63 A device carries |
|---|---|---|---|
| 30 °C (IEC 60898-1 reference) | 100% | 40.0 A | 63.0 A |
| 40 °C | 93% | 37.1 A | 58.5 A |
| 50 °C | 87% | 34.6 A | 54.6 A |
| 55 °C | 83% | 33.3 A | 52.5 A |
| 60 °C | 82% | 32.7 A | 51.4 A |
| 70 °C | 74% | 29.7 A | 46.8 A |
Note the last two rows. A device in a metal enclosure in direct sun, or in a plant room above a furnace, can easily see 60–70 °C even when the “ambient” in the project specification says 40 °C. That is a 20–26% loss of usable current, which is far more than most schedules allow for.
The temperature that matters is inside the enclosure
Project documents usually quote the site ambient. The breaker sees the air inside the board, which is hotter by the heat dissipated by everything in it: the breakers themselves, the RCDs, contactors, terminal blocks, the busbar, and any transformer or power supply mounted in the same enclosure. A typical figure for a reasonably loaded enclosure is 10–15 K above room temperature, and considerably more for a densely packed board or one with a sealed IP65 cover and no ventilation.
Three multipliers stack, and they are usually applied in the wrong order or not at all:
| Effect | Typical allowance | How to handle it |
|---|---|---|
| Site ambient above 30 °C | Per the curve above | Take the maximum design ambient for the site, not the annual average |
| Rise inside the enclosure | +10 to +15 K typical; more if sealed or densely packed | Use the internal temperature; measure it on a built board if the design is critical |
| Grouping — breakers side by side, all loaded | About 0.8 on a fully loaded bank | Apply to the bank, not to a single device with an empty slot beside it |
| Altitude above 2,000 m | Reduced cooling; check the datasheet | Air density falls, convection falls, the same watts give a bigger rise |
The enclosure is part of the answer, not just the breaker. Vented enclosures, metal heat-sinking, spacing devices apart, and keeping contactors and power supplies in a separate compartment all buy back current. So does choosing a board with spare ways rather than filling every module — the same argument as spare capacity, seen from the thermal side.
Derating and cable protection have to be done together
This is where most mistakes survive review. The cable’s capacity Iz is also derated for ambient and grouping, and the breaker’s In must sit below the derated Iz while above the design current Ib. Derate the breaker for a hot enclosure and forget the cable, and the breaker now protects a cable that is also running hot and has less capacity than the project table says.
The check is a chain, not a single number: Ib ≤ In(θ) ≤ Iz(θ) ≤ 1.45 × Iz(θ) for the conventional non-tripping current. Do it at the real temperatures at both ends of the run — a cable in a hot roof void and a breaker in a hot board are the same problem in two places. Where the derated breaker can no longer carry the load, the answer is usually one frame size up plus a cable size up, not a device with a higher calibration temperature; the heat has to go somewhere.
Where the ranges sit
The CHINT miniature ranges are calibrated to the same IEC/EN 60898-1 reference, so the derating logic applies to all of them; what changes is the breaking capacity and the frame, not the thermal physics. NB1-63 covers In 1–63 A with B, C and D curves and Icn 6 kA; NB1-63H is the 10 kA version of the same frame. NXB-63 covers the same current band with 6 kA, and its H variant reaches 10 kA at 220–240 V on two poles. All are IP20 front, with mechanical life around 20,000 operations and electrical life around 4,000 for the NB1-63 frame.
For higher currents the thermal question changes shape. NM1 and NM8 moulded-case breakers use thermal-magnetic or electronic trip units, and the electronic ones are specified over an operating temperature range — NM8 electronic trip units are quoted for −35 to +70 °C on the basic version — which is a different kind of statement from a bimetal curve: the electronics are compensated, but the current path and the terminals still heat. Read the specific trip unit’s temperature data rather than assuming the frame’s headline range covers your panel.
110–127 V and 60 Hz sites feel this first
A 3 kW load at 120 V draws 25 A; the same load at 230 V draws 13 A. Twice the current means four times the I²R heating in the same device, in the same enclosure, at the same ambient. That is why boards in 110–127 V markets run hotter for the same delivered power, and why a device comfortable at 80% loading on a 230 V board may be marginal on a 120 V board with the same physical layout. The derating table does not change; the starting current does.
Frequency matters less for the thermal element and more for everything around it: at 60 Hz, eddy-current and skin-effect losses in busbars and terminals rise slightly, and motor loads on 60 Hz systems often run at higher shaft speed and higher current than the same nameplate suggests. Specify 50/60 Hz rated devices and state the frequency on the enquiry.
What to send with the enquiry
Design current and load type (continuous, motor, mixed); system voltage, frequency and phases; the maximum ambient at the installation site and whether the board is indoors, outdoors, sealed, or in direct sun; the enclosure type and how many devices will be loaded side by side; the cable type, size, installation method and the ambient along the route; the required breaking capacity at the board position; and the destination country for the certificate requirement. With those, the derating can be done against the real condition and the order code fixed — including whether the 6 kA or the 10 kA frame is the honest choice at that temperature.
FAQ
At what temperature is an MCB rated?
IEC/EN 60898-1 calibrates miniature circuit breakers at a 30 °C reference ambient; IEC 60947-2 uses 40 °C. That reference is why a device rated 40 A carries 40 A in a laboratory and less in a real enclosure. Always check which standard the datasheet quotes before comparing two products on current alone.
How much does a 40 A MCB carry at 50 °C?
On typical miniature devices, roughly 87% of rated current, so about 35 A, and less if other loaded breakers are grouped beside it. The exact figure comes from the manufacturer’s derating table for that catalogue number, because the bimetal and housing design move the curve by a few percent. Do not size from a rule of thumb when the load is close to the rating.
Can I just fit a bigger breaker for a hot location?
Only if the cable can take it. Raising In to compensate for derating also raises the current the cable must survive, and the cable is in the same hot environment with its own derating. The correct move is usually one frame size up and one cable size up together, or reducing the heat load in the enclosure.
Does a D-curve breaker derate the same way?
The thermal element is the same kind of bimetal, so the thermal derating is broadly similar; the difference between B, C and D is in the magnetic trip threshold, which is largely independent of ambient. A D-curve device helps with motor inrush, not with a hot enclosure.
Is 40 °C ambient realistic for a rooftop board?
Frequently it is optimistic. A metal enclosure in direct sun in a hot climate can see internal temperatures well above the site air temperature, and the board’s own losses add to it. If the design depends on the ambient figure, measure the internal temperature of a comparable installed board rather than trusting the specification.
Related: Breaker and cable coordination · MCB trip curves B, C and D · How to read MCB marking · 6 kA vs 10 kA breaking capacity · How to size a distribution board · MCB range overview · Voltage and frequency by country · Request a quotation