Recurring 30 mA RCCB Trips in a Wet Workshop
A polytechnic workshop is fed from an 8-way sub-distribution board whose incomer is a 4-pole, 63 A, 30 mA residual current circuit breaker. Downstream are eight 6 A switched socket outlets in the machine shop, a 1.5 kW wash-area pump, and an inverter welding set. The RCCB trips two or three times a week, mostly when the pump or the welding set starts, and once overnight when no one was in the shop. The maintenance electrician proposes swapping it for a 100 mA unit so that the nuisance tripping stops, and separately proposes disconnecting the earth conductor of the welding set because its casing gives a tingle. Insulation resistance measured at 500 V DC: wash-area pump circuit 0.35 MΩ, welding set 2.2 MΩ, all other circuits above 50 MΩ. The earth electrode resistance is 18 Ω and no circuit is individually RCD-protected.
1. Evaluate the proposal to replace the 30 mA RCCB with a 100 mA unit.
Read the worked answer
Reject the proposal.
- 30 mA is chosen because it is the residual current at which a person's grip is normally released and below the level generally accepted as causing ventricular fibrillation in a healthy adult for short exposures. A 100 mA device does not provide this protection.
- Socket outlets in wet and outdoor locations, and circuits feeding equipment used by students, are required to have 30 mA residual current protection. Downgrading to 100 mA removes a code requirement, not just a convenience.
- A 100 mA device is a fire-protection device, not a personnel-protection device. It would still trip on the real fault, only later and at a higher current.
- The tripping is a symptom. Suppressing it hides a deteriorating insulation and a high-resistance earth path, and the next shock will be the one that injures someone.
Correct use of the number: the 0.35 MΩ insulation resistance on the pump circuit is below the 1 MΩ minimum expected at 500 V DC for a 230 V circuit. The measured leakage at that value is only 230/0.35 MΩ ≈ 0.66 mA when dry, so the trip is not explained by the dry reading alone - it is explained by that resistance collapsing under moisture and voltage stress at each start. Repair the pump circuit; do not weaken the protection.
2. Explain the trips that occurred when no one was working, and give a testing and remedial sequence that would identify the cause.
Read the worked answer
Why an overnight trip is significant
An RCCB trips on the vector sum of all leakage currents through it. The likely contributors here are:
- Moisture ingress in the wash-area pump circuit, whose 0.35 MΩ insulation falls further as the area cools and condensation forms - consistent with a trip with no staff present.
- Cumulative standing leakage from several healthy but capacitive loads; good practice is to keep total standing leakage below half the rated residual current, i.e. below 15 mA for a 30 mA device.
- High-frequency leakage and transient charging currents from the inverter welder and from long circuit capacitance at each start.
Sequence
- Isolate, lock off, and megger each circuit separately at 500 V DC; record the pump circuit, which fails.
- Split the board and test the RCCB itself with an RCD tester: it should trip within 300 ms at 30 mA, 150 ms at 60 mA and 40 ms at 150 mA. If it trips at less than half its rating with no load, it is faulty; if it trips correctly on load, the load is at fault.
- Megger the pump motor windings and cable, inspect the terminal box and glands, and dry or replace the damaged cable and motor.
- Clamp each circuit with a leakage clamp to find standing leakage and check the welder's filter and its earth connection.
- Give the pump and the wet-area sockets their own 30 mA RCBOs, so that one faulty circuit does not take out the whole workshop and the source of future trips is obvious.
3. Respond to the proposal to disconnect the earth conductor of the welding set, using the measured 18 Ω electrode resistance.
Read the worked answer
Refuse, and record the refusal.
A protective conductor is not optional. Disconnecting it means that a fault from live parts to the metal casing has no low-resistance path back to the source; the casing stays live until someone touches it, and the RCCB becomes the only protection. It also defeats the welding set's own leakage path to earth.
What the 18 Ω means
- 18 Ω is high for a workshop electrode. With a 30 mA RCD protecting the installation, disconnection is still achieved (30 mA × 18 Ω ≈ 0.54 V of touch voltage is trivial), which is why the RCD has been masking the earthing problem rather than the earth masking the RCD.
- Without the RCD, a 230 V fault through 18 Ω would pass about 12.8 A, far below the 63 A breaker's trip level - the casing would sit at most of the supply voltage indefinitely. That is exactly the situation the proposal would create.
The tingle is a fault indication. De-energise, megger the welder (2.2 MΩ is passable but should be watched as it is below the 50 MΩ seen elsewhere and the set is inverter type), inspect the input cable, plug and filter, check the earth continuity from the plug to the casing and to the electrode, and improve the electrode (additional rods, bentonite backfill) to bring the resistance well below 18 Ω. Only then re-energise and re-test the RCCB.