Nuisance Tripping of an Origin RCD on a Kitchen Board with VFD Loads
A four-star hotel reports that the 300 mA Type AC residual current device at the origin of the TN-S kitchen distribution board trips two or three times a week, always within about ten minutes of the breakfast shift starting. The board feeds twelve final circuits: four 30 mA RCBOs for socket outlets, direct-on-line combi ovens, a 15 kW dishwasher, and an extract hood fan driven by a 7.5 kW variable-speed drive on a 45 m motor cable. An earth-leakage clamp meter applied to each outgoing circuit at rest reads between 0.6 and 3.8 mA, summing to about 11 mA. Insulation resistance on every circuit is above 50 MΩ at 500 V DC with loads disconnected. The drive manual states earth leakage of up to 10 mA per drive, and the drive has no dedicated residual current device. The maintenance team proposes replacing the 300 mA device with a 500 mA unit to stop the nuisance trips.
1. Identify the most probable causes of the recurring trips and explain why the 11 mA standing-leakage reading does not prove the installation is healthy.
Read the worked answer
Why the reading understates the real leakage
- The clamp was applied with the loads at rest. The oven elements absorb moisture overnight and leak most during the first warm-up; the dishwasher's heater and solenoid valves behave the same way. The 11 mA figure is a cold, idle snapshot, not the operating value.
- VFD earth leakage is dominated by high-frequency common-mode current from the DC link charging the capacitance of a 45 m motor cable. That current rises with cable length, switching frequency and dv/dt. A standard leakage clamp is a power-frequency instrument and reads a fraction of it.
- The 10 mA figure in the manual is typical, not a maximum, and it is quoted for a reference cable length.
Why the trip happens at shift start
- All leakage paths on the board are additive at the origin device: idle leakage plus warmed element leakage plus drive leakage plus start-up transients all coincide within a few minutes.
- A Type AC device is designed to respond to sinusoidal alternating residual current. A mixture of pulsating DC and high-frequency leakage can desensitise it or make its behaviour unpredictable.
- A genuine N–PE fault or a shared neutral between circuits should not be dismissed until excluded by test.
Conclusion: the most likely cause is cumulative leakage from damp heating elements plus VFD cable leakage acting on an origin device of the wrong type and with no discrimination from the downstream 30 mA devices.
2. Set out the sequence of checks and tests you would complete before recommending any change to the protective device.
Read the worked answer
- Safe isolation of the board, prove dead, lock off, and record the state before any disconnection.
- Insulation resistance per circuit at 500 V DC, line and neutral to protective earth, with appliances disconnected. Values above 50 MΩ at rest are normal and reveal nothing about a load-dependent path, so this is a baseline, not a conclusion.
- RCD testing on every device - ramp test to find the actual residual operating current, trip times at 1× and 5× IΔn, at 0° and 180°. A device tripping at 40–60% of rating is a symptom in itself.
- Load-sequenced leakage survey: clamp each outgoing circuit and the main earthing conductor while loads are switched on one at a time, recording the increment attributable to each appliance, especially the drive. This is where the real numbers appear.
- Repeat the drive reading with a wide-band instrument or a power quality analyser with a current probe on the motor feeder, and compare against the clamp reading; the difference quantifies what the clamp missed.
- Exclude a neutral-to-earth fault: insulation test neutral to earth with the main switch off, and check the drive's internal EMC filter link configuration, which some manufacturers require to be changed for certain supply arrangements.
- Inspect thermally and physically: terminal tightness, discoloured neutral bars, water ingress at element glands and in the extract hood.
Only after steps 2–6 is the picture complete enough to design a remedy.
3. Propose the remedial scheme, and state why fitting a 500 mA device is not an acceptable solution.
Read the worked answer
Why 500 mA is wrong
- The device is there for fire protection and, on socket circuits, for additional protection. Raising the threshold removes protection from the whole board without removing the leakage current, so the same conditions will eventually trip the larger device too.
- It changes nothing about waveform sensitivity, so a Type AC device will still behave unpredictably with pulsating DC and high-frequency residual current.
- It also worsens discrimination with the downstream 30 mA devices rather than improving it.
Recommended scheme
- Split the board so that the drive and the cooking appliances are not grouped with socket circuits on a single residual current device. Keep the standing leakage on any one device at or below roughly 25–30% of its rating.
- Match the device type to the load: Type A for the general circuits and for a three-phase drive where the manufacturer permits it; Type B where the drive has a single-phase rectifier, DC injection braking or the manual requires it. Follow the drive manufacturer's instruction.
- Reduce the leakage at source: screened motor cable, shorten the 45 m run where possible, fit an output dv/dt or sine filter, and verify the EMC filter earthing and screen bonding.
- Retain 30 mA protection on all socket outlets, as individual RCBOs rather than a shared block.
- Verify after the change: ramp and trip-time test every device, repeat the load-sequenced leakage survey, confirm discrimination with the upstream device, and record actual measured leakage per device in the handover file.