Nuisance tripping on a compressor circuit: bigger breaker is not the answer
A 20 A type B MCB in a domestic switchboard supplies a 2.2 kW single-phase compressor in a garage. The circuit is 4 mm² TPS, about 25 m long, and the MCB sits in a row of twelve in the same enclosure. The compressor starts three or four times an hour, and the MCB trips perhaps once a week, always within the first second of starting. It never trips while running. Measured Ze is 0.45 Ω, and the circuit is protected by a 30 mA RCD that does not trip. The homeowner has been told by a friend that 'you just put in a bigger breaker'.
1. What must be ruled out before any protective device is changed?
Read the worked answer
Rule out a genuine fault first - a trip in the first second can be a motor fault, not a discrimination problem.
- Insulation resistance of the compressor and the circuit (motor windings, capacitor, supply lead, and the fixed wiring).
- Start/run capacitor condition - a failing capacitor raises starting current and lengthens start time.
- Loose or heat-damaged terminations at the MCB, the socket and the compressor - resistance here raises current and can cause nuisance trips and fire risk.
- Actual starting current, measured with a clamp meter on min/max, not estimated from nameplate. Record the value and the duration.
- Supply voltage during start - a significant dip lengthens acceleration and raises current.
- The MCB's thermal environment - twelve devices in a row derate each other, and a warm enclosure shifts the thermal trip point down.
- Other loads on the same circuit or final circuit sharing the compressor's supply point - a fridge or welder starting at the same time changes the picture.
Judgement point: the friend's advice ('bigger breaker') would raise the cable's protection threshold without confirming the cable can carry the higher current, and would conceal a genuine fault. The measured data decides the answer, not the guess.
2. Calculate the earth fault loop impedance for the circuit and determine whether a 20 A type C MCB would be acceptable on that basis. State what else must be checked.
Read the worked answer
Loop impedance
Using approximate operating-temperature conductor resistances (AS/NZS 3008 values):
- 4 mm² Cu ≈ 5.5 mΩ/m; 2.5 mm² Cu ≈ 8.9 mΩ/m
R1 + R2 = (5.5 + 8.9) × 25 / 1000 = 0.36 Ω
Zs = Ze + R1 + R2 = 0.45 + 0.36 = 0.81 Ω
Type C 20 A check
Magnetic threshold: 10 × In = 200 A
Maximum permissible Zs = (0.95 × 230) / 200 = 1.09 Ω
0.81 Ω < 1.09 Ω - a 20 A type C is acceptable on loop impedance at the far end of this circuit.
Still to check
- Cable capacity with derating - 4 mm² must have a derated Iz ≥ 20 A for the installation method, grouping and thermal insulation present.
- Prospective short-circuit current at the MCB must be below its breaking capacity, and above its magnetic threshold so a short circuit trips instantly.
- Coordination with the upstream device so a fault at the compressor does not take out the main switch.
- Overload protection for the compressor - a 20 A device on a motor with a nameplate FLC around 14 A gives only about 1.4× FLC, which is fine for the cable but may be marginal for the motor's own thermal protection; the compressor's integral overload is what protects the motor.
3. If the measured starting current is 95 A, does the type C solve the problem, and why is a type D not the right answer here?
Read the worked answer
Type B (3–5 × In): magnetic range 60–100 A. A 95 A start is inside that band, so whether it trips depends on the individual device's tolerance - which explains the 'once a week' pattern rather than a consistent trip.
Type C (5–10 × In): magnetic range 100–200 A. A 95 A start will not operate the magnetic element, so yes, the C-curve addresses the symptom, and the loop impedance check above confirms it is permissible.
Type D (10–20 × In): magnetic range 200–400 A. Maximum Zs = 218.5 / 400 = 0.55 Ω, but the circuit's Zs is 0.81 Ω. The device would not disconnect within the required time on an earth fault. On loop impedance alone, a type D is not acceptable here.
A type D would only be arguable if the 30 mA RCD were the means of fault protection (which changes the disconnection analysis rather than the Zs figure), and even then it gives much poorer protection against low-level faults and should not be used simply to mask a high starting current.
Better engineering answers
- Fit an unloader valve or a soft starter to reduce the start current - this addresses the cause.
- Check whether the compressor's own start/run capacitor or unloader is faulty.
- Consider a 25 A type C only if the 4 mm² cable's derated capacity still supports 25 A protection.
Judgement point: the device must protect the cable and clear faults within the required time. A device selected because it stops tripping, but which cannot clear an earth fault, has moved the problem rather than solved it.