RCDs, RCBOs, and ELCBs — What They Protect Against and What They Don't

A residual current device (RCD) works on an elegantly simple principle: in a healthy circuit, the current flowing out through the live conductor must exactly equal the current returning through the neutral conductor. If some current is "missing" — flowing to earth through a person's body, through damaged insulation, or through a fault — the RCD detects this imbalance and disconnects the supply.

Inside the RCD, both the live and neutral conductors pass through a toroidal (ring-shaped) magnetic core. Under normal conditions, the magnetic field created by the current in the live conductor is exactly cancelled by the equal and opposite current in the neutral conductor. The net magnetic flux in the core is zero, and nothing happens.

When current leaks to earth (the "residual" current), the live and neutral currents are no longer equal. The difference creates a net magnetic flux in the core, which induces a voltage in a sensing winding wrapped around the core. This voltage operates a trip mechanism that opens the contacts and disconnects the circuit.

An analogy: imagine a hotel lobby where every person who enters through the front door must leave through the front door. A doorman counts people in and people out. If the count does not balance at the end of the day, someone is still inside (or has left through a fire exit). The RCD is that doorman — it counts current in and current out. If the count does not balance by more than the rated threshold, it trips.

The beauty of this principle is that the RCD does not need to know why the current is imbalanced. It could be a person receiving a shock, a cable with damaged insulation touching a metal pipe, or moisture inside a junction box. The RCD detects them all because they all result in current leaving the circuit through a path other than the neutral conductor.

RCDs are manufactured with different rated residual current thresholds (IΔn), each designed for a specific protection purpose:

30 mA — Personal Protection (Additional Protection)

The 30 mA threshold is designed to protect people against electric shock from direct contact with live parts. Medical research shows that a current of 30 mA through the human body for up to 40 ms is unlikely to cause ventricular fibrillation (the lethal heart rhythm). A 30 mA RCD typically trips within 30 ms at its rated current and within 200 ms at half its rated current (15 mA), keeping the exposure below the dangerous threshold.

BS 7671 Regulation 415.1.1 requires 30 mA RCD protection for all socket outlets rated up to 32 A, all circuits in bathrooms, and all circuits supplying mobile equipment used outdoors. AS/NZS 3000 Clause 2.6.3 similarly mandates 30 mA protection for socket outlets, lighting circuits in domestic installations, and circuits in wet areas.

100 mA — Fire Protection

A 100 mA RCD is sometimes used where 30 mA sensitivity would cause excessive nuisance tripping (e.g., long cable runs to outdoor equipment) but fire protection is still required. At 100 mA, the RCD will not reliably prevent lethal shock to a person, but it will detect insulation faults and earth leakage currents that could generate enough heat at a fault point to start a fire. An earth leakage of 100 mA at 230 V dissipates 23 W at the fault point — enough to ignite dry timber or plastic over time.

300 mA — Earth Fault Protection

A 300 mA RCD provides protection against sustained earth faults and insulation degradation rather than personal protection. It is commonly used on distribution board incomers, lighting circuits in commercial buildings, and applications where the primary concern is detecting insulation breakdown before it escalates to a major fault. BS 7671 Regulation 422.3.9 recommends 300 mA RCDs for fire protection in locations with combustible materials (e.g., wooden buildings).

10 mA — High-Sensitivity for Medical and Special Applications

Some RCDs are available with 10 mA sensitivity for medical locations and environments where even 30 mA is considered too high a risk. These are used in Group 2 medical locations (operating theatres) as part of the medical IT system, though the primary protection in these areas is typically insulation monitoring rather than RCDs.

These three terms are often used interchangeably, but they refer to different devices with different capabilities:

RCD (Residual Current Device)

An RCD detects earth leakage current only. It does not provide overcurrent protection. If the circuit is overloaded or a short circuit occurs (live to neutral), the RCD will not trip. A separate MCB (miniature circuit breaker) must be installed in series to provide overcurrent protection. In a typical consumer unit, an RCD protects a group of circuits, with individual MCBs on each circuit.

RCDs come in two physical forms:

• RCCB (Residual Current Circuit Breaker): A standalone device that provides earth leakage protection only. This is what most people mean when they say "RCD."
• RCD as an add-on block: A module that clips onto an existing MCB to add earth leakage detection.

RCBO (Residual Current Breaker with Overcurrent Protection)

An RCBO combines earth leakage detection and overcurrent protection in a single device. It replaces both the RCD and the MCB. Each circuit gets its own RCBO, which means a fault on one circuit does not trip other circuits — unlike a shared RCD arrangement where an earth fault on any circuit in the group trips the RCD and disconnects all circuits in that group.

RCBOs are increasingly popular in modern installations because they provide individual circuit protection. The downside is cost: an RCBO costs significantly more than an MCB alone, and a full consumer unit fitted with RCBOs costs more than one using shared RCDs. However, the improved selectivity (only the faulted circuit is disconnected) is usually worth the investment.

ELCB (Earth Leakage Circuit Breaker)

An ELCB is an older technology that has been largely replaced by RCDs. There are two types:

• Voltage-operated ELCB: Detects voltage on the earth conductor relative to a separate earth electrode. If a fault raises the earth conductor voltage above a threshold (typically 40–50 V), the ELCB trips. This technology has significant limitations: it only works if the fault current returns through the monitored earth conductor. If the fault finds an alternative path to earth (through pipework, for example), the ELCB does not detect it.
• Current-operated ELCB: This is functionally identical to a modern RCD (detects current imbalance). The term "current-operated ELCB" has been replaced by "RCD" in modern standards, but the old terminology persists in some regions.

The key distinction: if someone asks about an "ELCB," they usually mean a voltage-operated device (the old type). If they mean a current-operated device, the correct modern term is "RCD" or "RCCB."

This is the most important section of this article. RCDs are sometimes treated as a universal safety device, but they have significant blind spots:

1. Live-to-Neutral Faults (No Earth Involvement)

If a fault occurs between the live and neutral conductors — for example, a short circuit inside an appliance — the current flows from live to neutral in the normal path. The current in and current out are balanced. The RCD sees zero residual current and does not trip. Overcurrent protection (MCB or fuse) must handle this type of fault.

2. Live-to-Live Faults (Phase-to-Phase)

In a three-phase system, a fault between two phase conductors does not create any current imbalance that an RCD can detect (the current flows between phases, not to earth). Again, overcurrent protection is required.

3. Overload

An RCD does not detect overload current. A circuit carrying 50 A through a 20 A cable will overheat and potentially cause a fire, but the RCD sees balanced live and neutral currents and remains closed. This is why an RCD must always be paired with overcurrent protection (MCB), or an RCBO must be used.

4. Shock Between Live and Neutral

If a person simultaneously touches the live and neutral conductors (e.g., inserting both fingers into a socket), the current flows through their body from live to neutral. From the RCD's perspective, this is normal current flow — current out through live equals current back through neutral. The RCD does not trip. This scenario is rare in practice because it requires simultaneous contact with both conductors, but it is a real limitation.

5. DC Earth Leakage

Standard AC-type RCDs (Type AC) only detect sinusoidal AC residual currents. They cannot detect smooth DC leakage or pulsating DC residual currents. This is increasingly relevant with the proliferation of electronic equipment containing rectifiers (EV chargers, solar inverters, VFDs). BS 7671 Regulation 531.3.3 requires Type A RCDs (which detect pulsating DC) or Type B RCDs (which detect smooth DC) for circuits supplying equipment with rectifiers. Type A is now the minimum requirement for most circuits in modern installations.

One of the most frustrating practical problems with RCDs is nuisance tripping — the RCD trips when there is no genuine fault. The most common cause is capacitive leakage current on long cable runs.

Every cable has a small capacitance between its conductors and between conductors and earth. This capacitance allows a tiny AC current to flow to earth even in a perfectly healthy cable — the insulation is not a perfect insulator at AC frequencies. The leakage current is typically in the range of 0.5–1.5 mA per 100 metres of cable, depending on cable type and voltage.

For a single short circuit (say, a 10-metre cable to a nearby socket), this leakage is negligible. But when multiple long cable runs share a single RCD, the cumulative leakage can approach the RCD's trip threshold:

• 6 circuits × 50 metres average × 1 mA/100 m = 3 mA steady-state leakage
• Add transient leakage from electronic equipment (LED drivers, switch-mode power supplies) starting up: +5–10 mA spikes
• Add slight residual leakage from older appliances: +2–5 mA
• Total: 10–18 mA on a 30 mA RCD — very close to the trip threshold

The RCD may trip during transient events (motor starting, LED driver inrush, lightning-induced surges on long cable runs) even though there is no genuine fault.

Solutions for Nuisance Tripping

• Use individual RCBOs per circuit instead of a shared RCD for multiple circuits. Each RCBO only sees the leakage from its own circuit.
• Use time-delayed RCDs (Type S or selective) on upstream devices to prevent sympathetic tripping.
• Limit the number of circuits per RCD — a practical maximum is 6–8 circuits, depending on cable lengths.
• Check and fix actual leakage sources — old appliances, damp junction boxes, and failed cable insulation can contribute hidden leakage that pushes the total over the threshold.
• Use surge protection devices (SPDs) at the consumer unit to absorb transient surges that cause momentary leakage spikes.

RCDs are classified by the type of residual current they can detect, defined in IEC 61008 and IEC 61009:

• Type AC: Detects sinusoidal AC residual currents only. The most basic type. Increasingly inadequate for modern installations with electronic loads.
• Type A: Detects sinusoidal AC and pulsating DC residual currents. Required by BS 7671 and AS/NZS 3000 as the minimum for most circuits. Necessary for any circuit supplying equipment with single-phase rectifiers (washing machines, dishwashers, EV chargers in Mode 2).
• Type F: Detects everything Type A detects, plus composite residual currents from frequency-controlled equipment (variable frequency drives). Required for circuits supplying VFD-controlled appliances like inverter air conditioners and washing machines with variable-speed motors.
• Type B: Detects all forms of residual current including smooth DC. Required for three-phase rectifier loads, EV chargers (Mode 3 with three-phase supply), solar inverters, and medical equipment. Type B RCDs are significantly more expensive than Type A.
• Type B+: Type B with enhanced sensitivity at frequencies up to 20 kHz. Used for specialised applications with high-frequency leakage currents.

The general rule: if the equipment contains a rectifier or power electronic converter, a Type AC RCD is not sufficient. The smooth DC component of the fault current can saturate the RCD's magnetic core, rendering it completely blind to both DC and AC residual currents. This is not a theoretical concern — Type AC RCDs have failed to trip in documented incidents involving equipment with rectified DC output.

Every RCD has a test button that creates an artificial leakage current through an internal resistor. Pressing this button verifies that the trip mechanism operates correctly.

BS 7671 and AS/NZS 3000 both require the user to test RCDs regularly — the recommended interval is every three months (quarterly) using the test button. During periodic inspection and testing by a qualified electrician, the RCD is tested with a calibrated instrument that verifies:

• Trip current: The RCD trips at or below its rated residual current (e.g., ≤ 30 mA for a 30 mA device).
• Trip time at rated current: ≤ 300 ms for standard RCDs, ≤ 40 ms for Type S (selective/time-delayed) at 5× rated current.
• Trip time at 5× rated current: ≤ 40 ms for standard RCDs (to verify rapid disconnection at high fault currents).
• No-trip at 50% of rated current: The RCD must not trip at half its rated residual current (e.g., 15 mA for a 30 mA device), verifying it is not overly sensitive.

An RCD that fails these tests must be replaced immediately. RCD trip mechanisms can degrade over time, particularly in humid or dusty environments. A failed RCD looks identical to a working one from the outside — only testing reveals the failure. This is why the quarterly test button press is genuinely important, not just a regulatory formality.