MCB, MCCB, ACB and RCCB: what each protects, and where

Four devices turn up on every distribution drawing, and the differences between them are usually explained as a matter of size — MCB small, MCCB medium, ACB large. That is true and it is not the useful distinction. They differ in what they can see, how far they can be adjusted, and how much fault current they can break, and those three properties decide where each one belongs.

The RCCB is not on the same list at all. It protects against something the other three cannot detect.

The short version#

The one that is different#

An MCB, MCCB and ACB all measure current flowing through them, and they trip when there is too much of it. That protects the cable and the equipment. It does essentially nothing for a person touching a live part.

Current flowing through someone to earth is a small current — 30 mA can be lethal, and a 32 A MCB will not notice 30 mA any more than it notices a light bulb switching on. An RCCB compares the current going out with the current coming back. If they differ, current is leaving the circuit somewhere it should not be, and it trips.

An RCCB has no overcurrent protection of its own. It must always sit behind an MCB or MCCB, sized so the RCCB's own current rating is never exceeded. An RCBO combines the two functions in one module, which costs more per way and means a leakage fault trips only the affected circuit — usually the better arrangement on anything the user cares about.

RCD types: the mistake that VFDs cause#

This one is missed constantly, and it turns a safety device into an ornament.

The failure mode is specific and dangerous. A fault downstream of a rectifier can produce leakage with a DC component, and DC leakage can saturate the core of a Type AC device so that it no longer detects the AC leakage it was fitted for. The device does not fail obviously — it stops working while looking perfectly healthy.

If the circuit has a VFD, an EV charger or a solar inverter on it, the RCD must be Type B (or Type F for single-phase drives). Type AC on such a circuit is a false sense of protection.

Trip curves#

The trip curve says how fast the magnetic element operates as a multiple of the rated current:

A B-curve device on a motor circuit trips on every start. A D-curve device on a lighting circuit at the end of a long run may not trip on a genuine fault at all, because the fault current at that distance never reaches ten times the rating. That second failure is the more dangerous one, and it is why long final circuits sometimes need a B curve or an earth fault loop impedance check rather than an assumption.

For motors specifically, the right answer is usually not an MCB at all but an MPCB — a motor protection circuit breaker with an adjustable thermal element set to the motor's full load current and a fixed magnetic element around 13 × In that lets the starting inrush pass. The motor starter calculator picks one from the motor rating.

Icu, Ics and Icn — three breaking capacities#

Datasheets quote more than one, and the difference matters:

• Icu — ultimate breaking capacity. The device will interrupt this fault once, safely. It may not be usable afterwards.
• Ics — service breaking capacity. It will interrupt this fault and remain fit for service. Usually quoted as a percentage of Icu: 50 %, 75 % or 100 %.
• Icn — rated breaking capacity for MCBs to IS/IEC 60898, the domestic and similar standard. Typically 6 kA or 10 kA.

Specifying on Icu alone means accepting that after a serious fault the device is scrap and possibly no longer protecting anything. On main incomers and anywhere downtime is expensive, specify Ics = 100 % of Icu.

Get the fault level from the transformer with the transformer sizing calculator, and remember it falls as you move down the system — which is why 10 kA MCBs are acceptable at a final distribution board fed by a long submain, and would not be acceptable on the main bus.

Discrimination, and the cheat that goes with it#

Discrimination means a fault takes out only the device closest to it, leaving the rest of the installation running. It is built with the adjustable settings on MCCBs and ACBs: current thresholds that step up going upstream, and short time delays on the upstream devices so the downstream one gets the first chance to clear.

Those delays are not free. Every one of them extends the time the fault current flows, and the cable short-circuit withstand calculation scales with √t. A 0.6 s delay upstream can push a cable up two sizes.

Cascading (or back-up protection) is the related trick: an upstream current-limiting device is allowed to help a downstream one, so the downstream device can be specified below the prospective fault level at its position. It is cheaper, and it is only valid for tested combinations published by the manufacturer — you cannot derive it yourself, and you cannot mix brands.

Choosing, in order#

The MCB and MCCB sizing calculator runs the sequence below for a given circuit — rating, curve, breaking capacity and the Ib ≤ In ≤ Iz check against your derated cable, with the panel-ambient correction applied.

1. Fault level at this point in the system — that sets the minimum breaking capacity, and Ics rather than Icu where continuity matters.
2. Load current and the cable it protects — the device rating must sit between the design current and the derated cable capacity, Ib ≤ In ≤ Iz, which is step four of sizing a cable correctly.
3. Inrush behaviour of the load — which sets the curve, or sends you to an MPCB.
4. Whether anyone can touch it — which decides the RCD, and its sensitivity.
5. What electronics are downstream — which decides the RCD type.
6. Whether a fault here should take out anything else — which decides how much adjustability you need to pay for.