Protection · 21 July 2026 · For contractors and project engineers
An RCD trips when it sees an imbalance between the current going out on the line and coming back on the neutral — the "residual" current that must be leaking to earth. But not every RCD can see every kind of leakage. As electronics, EV chargers and inverter-driven equipment have spread, the type of RCD has become a design decision, not a default — and fitting the wrong one can leave a circuit that looks protected but isn't.
| Type | Detects | Typical use |
|---|---|---|
| AC | Sinusoidal AC only | Restricted — only fixed loads known to have no DC component |
| A | AC + pulsating DC (up to 6 mA smooth) | The general minimum — sockets, lighting, most circuits |
| F | Type A + mixed frequencies | Single-phase inverter / VFD loads — heat pumps, some appliances |
| B | All of the above + smooth DC | Three-phase drives, some EV charging, PV, medical |
Each type is a superset of the one before it: a Type B detects everything a Type A does, plus more. The cost rises steeply from A to B, which is why you fit the lowest type that genuinely covers the load rather than defaulting to B everywhere.
The move away from Type AC wasn't fashion. Modern equipment with rectifiers and switch-mode power supplies can leak smooth DC to earth. That DC can magnetically saturate the core of a Type AC RCD and stop it responding to a later AC fault — the device is still there, still looks fine, but has been "blinded". A Type A tolerates pulsating DC and a small amount of smooth DC; beyond that you need Type F or B. This is why the standard now restricts Type AC to a narrow set of loads.
Keep two ideas separate. The type (AC/A/F/B) is which waveforms it detects. The rated residual current IΔn is the leakage at which it trips: 30 mA for additional protection against shock (most socket outlets and outdoor mobile equipment), and 100 mA or 300 mA for fire protection and for achieving selectivity between RCDs. A 30 mA Type A and a 300 mA Type A are both "Type A" but do very different jobs.
RCD selection sits alongside disconnection times, earth-fault loop impedance and discrimination in the protection design. On a TT system an RCD is essential for automatic disconnection; on any system it's how you provide additional shock protection. Getting the schedule of RCD types and ratings right, board by board, is part of a coordinated protection design.
Type A. BS 7671 now restricts Type AC to specific fixed equipment known to have no DC components, so for general new work — socket outlets and most circuits — Type A is the practical minimum, because it detects the pulsating DC leakage that modern electronic loads produce.
It depends on the charge point. If the EVSE has built-in protection against smooth (6 mA) DC fault current, a Type A RCD upstream is acceptable; if it does not, a Type B RCD is required, because a DC fault from the vehicle could otherwise blind an ordinary RCD. Always follow the charge-point manufacturer's instructions.
Heat pumps use variable-frequency (inverter) drives, which can produce mixed-frequency and smooth DC leakage. That usually calls for a Type F or Type B RCD depending on the drive — check the manufacturer's stated requirement rather than fitting a Type A by default.
Yes. Smooth DC leakage from electronic equipment can magnetically saturate a Type AC (or even Type A) RCD's core and stop it tripping on a subsequent AC fault — the "blinding" effect. This is the core reason the standard moved away from Type AC and introduced the higher types.
The rating (IΔn) is the residual current at which it trips, separate from the type. 30 mA is the level for additional protection against electric shock (required on most socket outlets and mobile equipment used outdoors); 100 mA and 300 mA devices are used for fire protection and for selectivity between RCDs, not for personal shock protection.
For the exact requirements, see BS 7671 Regulation 531.3.3 and the manufacturer's instructions for the connected equipment.