UK RCD tripping explanation

[KeenPensioner]

As a keen adult learner any help would be appreciated

I understand that an RCD trips when there is an imbalance detected between L and N.

If a 30mA RCD is monitoring several circuits (in a split board for example) and one of the circuits is protected by a 32A Type B MCB. I'm trying to understand what would happen in the following situations and would appreciate some help.

In a L-N short circuit the RCD would not trip but the MCB would when 160A passes through the MCB?

If a L-E or N-E fault then the RCD would trip before the MCB?

Any advice on fault finding a tripping RCD would also be appreciated

 

[timhoward]

In a L-N short circuit the RCD would not trip but the MCB would when 160A passes through the MCB?

Correct

If a L-E or N-E fault then the RCD would trip before the MCB?

L-E, either could win.
N-E, MCB wouldn't trip.

Any advice on fault finding a tripping RCD would also be appreciated

Insulation Resistance testing is the best way. Also an earth leakage clamp meter can help.
The common suspects are water ingress into outside lights, and anything with a heating element.

 

[brianmoooore]

In a L-N short circuit the RCD would not trip but the MCB would when 160A passes through the MCB?

It often happens that there is a fault between neutral and earth of a few thousand ohms, remote from the CU, that has been present for some time, but has not caused the RCD to trip because the voltage between neutral and earth, (caused by voltage drop from the current flowing in the neutral), is insufficient to imbalance the live and neutral anything like enough to trip it.
When a L to N fault occurs, the voltage drop in the neutral rises momentarily to a much higher level before the MCB trips, proportionally increasing the current flowing through the existing N -E fault, which may then unbalance the neutral and live currents enough to trip the RCD as well.
Note that the N-E fault might not be any way near the site of the L-N or even on the same circuit, as long as it's fed by the same RCD.

 

[KungFuTurkeyLeg]

Reference to finding the fault.

If you had nothing but a screwdriver to fault find, first job on a split board would be to try and identify the circuit to blame if the RCD has tripped. For a bog standard fault the below would be an indicator. By bog standard I mean a piece of load equipment is faulty or something, not a wiring issue.

Knock all the MCBs to open and close the RCD. Close the MCBs one at a time until the RCD trips.

Identify the outlets for that circuit, such as lights or plug sockets, and disconnect the loads. Be it lights bulbs or appliances.

Attempt to close the RCD and MCB for that circuit again. If it trips again, double check you have unplugged everything.

If it doesn't trip again, plug each load back in one item at a time. Switch on the lights or TV or something on that circuit that you can see or hear. That way you will lose light or sound once you plug the faulty load back in. Make sure to turn each load on after plugging it in, might only be faulty when powered up.

If it is still tripping, then it's likely a wiring fault. I'd call a spark at that point. As you're then into the realms of doing IR testing, working out resistances, splitting rings, etc. to track down the location of the fault. Could be water ingress as mentioned above, or vermin chewing through wiring, or a new picture on the wall had a cable strike.

Hope that helps.

 

[KeenPensioner]

Thank you very much for taking the time to reply - appreciated.

 

[timhoward]

Knock all the MCBs to open and close the RCD. Close the MCBs one at a time until the RCD t

Certainly worth a try, and I would do this too.
A L-E fault can often be detected this way.

If it’s a N-E fault then they only tend to show up when large currents are flowing as Brian explained above. At that point it becomes impossible to work out what’s actually going on without doing some testing.

 

[pc1966]

If a L-E or N-E fault then the RCD would trip before the MCB?

It depends on how "hard" the fault is and how much current flow. In your 32A B-curve socket example:

• 0 to 0.015A nothing happens
• 0.015 to 0.030A possibly the RCD trips
• 0.03A to 96-160A, RCD trips, no MCB trip as too slow on thermal side
• 160A or more MCB trips first on magnetic side, but RCD is already committed to tripping, so also goes

MCB are energy-limiting devices when you hit the "instant" magnetic trip point and generally they fire in under 10-20ms at that point and as you get well in to the magnetic trip region that drops down to 5ms or maybe less.

The usual 30mA RCD are intended for shock protection and so are only specified to trip in under 300ms (at 30mA threshold) down to 40ms (at 5*30 = 150mA or more). However, many will trip in region 8-20ms when tested normally but they are not designed for energy-limiting, and often can only break around 1kA themselves so they depend on being cascaded with a fuse or MCB that is faster opening at high currents to protect the RCD contacts from a kA-level fault.

Just to add, for N-E you wont trip the MCB (in UK only on L side), and fault currents are only likely in the tens of amps region at most.

 

[KeenPensioner]

It depends on how "hard" the fault is and how much current flow. In your 32A B-curve socket example:

• 0 to 0.015A nothing happens
• 0.015 to 0.030A possibly the RCD trips
• 0.03A to 96-160A, RCD trips, no MCB trip as too slow on thermal side
• 160A or more MCB trips first on magnetic side, but RCD is already committed to tripping, so also goes

MCB are energy-limiting devices when you hit the "instant" magnetic trip point and generally they fire in under 10-20ms at that point and as you get well in to the magnetic trip region that drops down to 5ms or maybe less.

The usual 30mA RCD are intended for shock protection and so are only specified to trip in under 300ms (at 30mA threshold) down to 40ms (at 5*30 = 150mA or more). However, many will trip in region 8-20ms when tested normally but they are not designed for energy-limiting, and often can only break around 1kA themselves so they depend on being cascaded with a fuse or MCB that is faster opening at high currents to protect the RCD contacts from a kA-level fault.

Just to add, for N-E you wont trip the MCB (in UK only on L side), and fault currents are only likely in the tens of amps region at most.

Thank you for taking the time to help explain this - much appreciated