Intermittent RCBO tripping in a kitchen: device, fault or accumulated leakage?
Call-back to a 1990s flat. The 30 mA Type A RCBO feeding the kitchen radial (six double sockets plus one outside socket on an FCU) trips after 20–40 minutes of use, more often in wet weather, and resets every time. With every appliance unplugged, insulation resistance is >299 MΩ L-N, L-E and N-E. The RCBO ramp-tests at 21 mA; no trip at 15 mA; 24 ms at 1×; 14 ms at 5×. Zs and polarity are normal. With all appliances reconnected and on standby, a leakage clamp on the circuit live reads 18 mA, rising to 23 mA when the fridge compressor starts. Individually, seven items each show 1–4 mA; the outside socket with a damaged gasket shows 9 mA when damp. A brand-new identical RCBO produces the same tripping pattern.
1. Is the RCBO faulty? Justify your conclusion from the test results.
Read the worked answer
No.
BS EN 61009 requires a 30 mA device to hold at 0.5 In (15 mA) and to trip before In (30 mA). This one holds at 15 mA and ramps at 21 mA - inside the band, though at the low end. The operating times at 1× (24 ms, limit 300 ms) and 5× (14 ms, limit 40 ms) are both compliant.
The decisive evidence is the substitution: a new identical device behaves the same way. If the device were the cause, the fault would not follow it.
What the other results rule out
299 MΩ with appliances disconnected rules out a hard insulation fault on the fixed wiring.
- Normal Zs and polarity rule out a N-E reversal at a socket.
What the results do not rule out
Insulation resistance is a DC test with the appliances removed. It tells you nothing about standing leakage from electronic equipment or about moisture that is only present when the circuit is energised and the weather is wet. A ramp test at 21 mA is not a fault - but it does mean this device operates at the bottom of its permitted band, which matters when the circuit's standing leakage is high.
Judgement
Passing the RCD tests is not the same as the RCD being the right answer to the symptom. A device that is compliant can still be tripping legitimately.
2. Explain the mechanism of the tripping, including why it is intermittent and why wet weather matters.
Read the worked answer
The mechanism is accumulated leakage, not a single fault.
An RCD does not measure faults; it measures imbalance. Every item with a filter or a heating element puts a small current into the protective conductor, and on a 30 mA device those currents add up:
- seven items at 1–4 mA ≈ 14 mA of standing leakage;
- the outside socket's damaged gasket adds about 9 mA when damp;
- total 23 mA when the fridge compressor starts and its filter leakage rises.
The device ramps at 21 mA. So the circuit sits just below the operating threshold and crosses it whenever a motor starts or the weather adds moisture. That is why it resets and runs again, and why it takes 20–40 minutes of a wash cycle rather than tripping at switch-on.
Why wet weather is a genuine clue, not a coincidence
The outside socket's gasket is damaged. Damp raises its leakage from near zero to 9 mA. The insulation resistance test missed it because it was done dry and at DC.
Why this matters for the client
The pattern looks like a faulty RCD and is often 'fixed' by fitting a bigger one. The real information in the case is the split between the damp-dependent component (9 mA) and the equipment baseline (14 mA) - which points at two separate actions: repair the outside socket, and manage the standing leakage.
3. Set out your rectification, and state what you must not do and what you record.
Read the worked answer
Rectification, in order
- Replace the outside socket, or its enclosure and gasket, and re-test. This removes the weather-dependent 9 mA and the future water-in-the-box fault.
- Re-measure standing leakage with everything connected. If it is now around 14 mA against a device ramping at 21 mA, the margin is still thin.
- Reduce the equipment baseline or split it: move the fridge or the IT load onto a separate RCBO-protected circuit, or fit an RCBO per circuit so a single device protects one circuit's leakage rather than the whole kitchen's.
- Where an item is genuinely leaky, tell the client which one - replacing a single appliance is cheaper than re-engineering the board.
- Verify with a ramp test and a 1× test, and record the leakage figures.
What I must not do
- Fit a 100 mA or time-delay device on a socket circuit. Socket outlets ≤ 32 A require 30 mA additional protection; uprating removes the protection the installation depends on.
- Leave the outside socket in service because 'it only trips when it rains'.
- Record 'RCD faulty, replaced' when the evidence says otherwise - that writes a false trail for the next person.
Recording
A minor works or EIC depending on the work done, with the outside socket change, the leakage measurements, the ramp test result and the advice given. If the standing leakage remains high, say so on the certificate rather than leaving a circuit that will trip again in the next wet spell.
Judgement
Cumulative leakage is a design and maintenance problem. The assessor wants to see you fix the damp, measure the rest, and refuse to trade protection for convenience.