Skip to content
FUSEPLAN

RCD type B+: when type B isn't enough

From AC to B+: what separates the six RCD types, and when type B stops being enough.

A few months ago I was planning a distribution board for a house with a PV system, battery storage, and a wallbox. The PV circuit was originally going to get a standard type A RCD, the kind that sits in most circuits. But the inverter’s datasheet said, in plain terms: three-phase, transformerless, capable of producing a smooth DC fault current, RCD type B required at minimum. A type A device wouldn’t have noticed a fault current like that at all. Which RCD a circuit needs depends on the fault current shape that can occur at its far end, not on size or price.

Why there are so many types in the first place

An RCD detects a fault current through a summation current transformer: if more current flows out on the line conductor than comes back on the neutral, the difference creates a magnetic field in the transformer core that trips the switch. That works reliably as long as the fault current is a clean sine wave. But modern devices with rectifiers, frequency inverters, or switched-mode power supplies increasingly produce fault currents that are anything but sinusoidal: pulsating, with a DC component, or at frequencies well above the 50 Hz mains frequency. With a pure DC fault current, a plain transformer core can even saturate and go blind to AC fault currents afterward too. The type classification is what tells you which of these fault current shapes a given RCD can actually detect.

How the summation current transformer in an RCD worksTwo diagrams: left, normal operation, where the L and N currents are equal and flow in opposite directions through the toroidal core, so their magnetic fields cancel out. Right, a fault condition, where part of the current returns via earth instead, leaving a residual flux in the core that trips the switch.L→→N←←ΣΦ = 0Trip coilSwitch closedLoadNormal operationI(L) = I(N), fields cancel outL→→N←IΔ to earthΣΦ ≠ 0Trip coilSwitch trippedLoadFault conditionI(N) < I(L), residual flux trips it
L and N run through the same toroidal core, in opposite directions. As long as both currents are equal, their magnetic fields cancel out inside the core. If part of the current returns via a fault path to earth instead of through N, a residual flux remains, the coil induces a voltage, and the trip coil opens the switch.

The types, from the bottom up

TypeAdditionally detectsTypical use
AConly sinusoidal AC fault currentnot permitted in Germany under VDE 0100-530 section 531.3.3
Apulsating DC fault current up to 6 mAstandard socket and lighting circuits, most household appliances
A AP-Rshort time-delayed, less sensitive tripping to avoid nuisance tripsupstream RCDs ahead of several type A circuits, wallboxes
Fmixed-frequency currents up to 1 kHz, pulsed current with DC component up to 10 mAappliances with a single-phase frequency inverter: heat pumps, air conditioners, variable-speed washing machines
Bsmooth DC fault currents, high-frequency currents up to 2 kHz, high system availabilitythree-phase, transformerless PV inverters, charging infrastructure, UPS systems
B+high-frequency currents up to 20 kHz, enhanced preventive fire protectionlarge PV and storage systems, industrial installations with several high-power frequency inverters

Each type fully covers everything the types below it can do. A type B+ device also detects everything type B, F, A AP-R, A, and AC detect, plus the extended frequency range and the added fire protection.

Try it yourself: pick a fault current shape and see which types keep up.

AC
A
A AP-R
F
B
B+
The curve shows the typical signal shape of the selected fault current, schematic, not a real measurement. Whether a type detects it follows from the cumulative hierarchy in the table above, not from a manufacturer trip curve. A AP-R detects the same waveforms as A.

Type F: the special case for air conditioners and heat pumps

A plain type A RCD rarely cuts it under a modern air conditioner or heat pump, and that’s the whole reason type F exists. Both appliances control their compressor speed through a single-phase frequency inverter, which can produce mixed-frequency currents that a type A simply doesn’t see. Three-phase units usually need type B instead, since smooth DC fault currents become possible there too. The appliance’s datasheet normally states the minimum type required, and it’s worth checking before you plan the circuit, not after.

A AP-R: a more selective A

An A AP-R detects the same fault current shapes as a normal type A, so the AP-R suffix doesn’t mean “more protection.” It trips after a short delay, which makes it less sensitive to leakage currents that add up from several loads sharing one upstream RCD. It typically sits ahead of several downstream type A circuits or ahead of wallboxes, not as a replacement for a type A at the final circuit.

Selectivity: when two RCDs sit in series

Take three breakers for lighting and sockets in the house that sit behind one shared 30 mA RCD in the sub-distribution board. One of those breakers feeds the supply line to the garden shed, where a small distribution board has its own 30 mA RCD for the outdoor sockets and lighting. Say a fault current now shows up out in the shed, at a lawnmower socket or in the lighting. The question is: does the RCD in the shed trip, or the one in the house, taking all three unrelated circuits down with it?

There’s no guarantee. Two ordinary, non-delayed 30 mA RCDs in series aren’t selective against each other. Both see essentially the same fault current at the same time, and which one opens first comes down to manufacturing tolerances and each device’s individual reaction time, not how close it sits to the fault. In the worst case, the RCD in the house trips, and the kitchen lights go out because of a fault out in the shed.

There are two ways to avoid this. The simpler one: don’t put an RCD ahead of the supply line at all. The line to the shed gets just an ordinary breaker in the house, no upstream RCD. The RCD in the shed’s own distribution board handles personal protection entirely. The breaker in the house only protects the supply line against overload and short circuit. This is usually the right call anyway, as long as the supply line is fixed wiring and protected against indirect contact.

The more involved way: build in real selectivity, for cases where that circuit in the sub-board already sits behind an existing RCD and you don’t want to rebuild it. That means the RCD in the house needs a rated residual current well above the downstream one, plus a time delay, usually type S (selective), for example 100 or 300 mA selective in the house against 30 mA standard in the shed. Only then will a fault in the shed reliably trip just the RCD there, with the upstream one left as a fallback in case the downstream one fails for some reason.

This is exactly what A AP-R, from the section above, is for: the same idea, just applied between several type A circuits instead of between two RCDs of different rated currents.

When B isn’t enough and B+ becomes necessary

The jump from B to B+ is mainly about the frequency range covered: 2 kHz for type B, 20 kHz for type B+. For a single PV system with one inverter, type B is enough in most cases. Once several high-power frequency inverters come together, in large PV-plus-storage combinations, say, or industrial installations with multiple drives, their high-frequency leakage currents can add up past 2 kHz even when no single device looks problematic on its own. Type B+ covers that extra range and adds a higher level of preventive fire protection on top. In practice, this mostly matters where a standard, a property insurer, or a project-specific requirement calls for it explicitly.

My rule of thumb

For normal household circuits, type A is enough. A single-phase frequency inverter (air conditioner, heat pump, variable-speed appliance) makes it type F. A device that can produce smooth DC fault currents, three-phase PV inverters above all, needs type B, and that decision comes from the datasheet. I only reach for type B+ when the datasheet, a standard, or an insurer calls for it, usually on larger installations with several frequency inverters. A AP-R goes in upstream positions where selectivity matters more than an extra fault-current type. Whenever a small distribution board with its own RCD sits behind another RCD, I check that the two are selective against each other.

How did you like this article?

Comments

Write a comment

Approved feedback is public. Please don't include personal data such as your email address or phone number.Privacy policy

← All posts

Send feedback

Type of post

Approved feedback is public. Please don't include personal data such as your email address or phone number.Privacy policy