RCB tripping - tracking it down

[Spenny]

Hi

The RCB on my Crabtree distribution board trips on a regular basis, once every other day. Over the last month I have tried to narrow down which circuit it is on by isolating the individual switches for the different circuits, but with no success - it has tripped out regardless of which circuits have been isolated.

I am reluctant to get an electrician in unless there is something they can specifically test on a circuit to pin it down.

It quite often trips overnight (cold, moisture, damaged wiring) but not exclusively so. I have a lot of computer equipment, but most of it tends to have wall warts which are not earthed so they would be unlikely to be the problem.

So:

1) Is it possible that it is a fault of the RCB itself?
2) If I got an electrician in, is there some equipment that is likely to pick up the earth leakage (my suspicion is that there is more than one fault).
3) Are there any devices I can substitute in place of the devices on the circuit board so that the individual circuits trip, not the whole lot

Crabtree Starbreaker from around 1999.
RCCB 63A 30mA
5 61/B32 and 2 61/B16 with a 61/B40 for the cooker.

 

[sparkyspike]

Yes it could be the RCCB at fault, although this is not the most likely. A spark will have the equipment to test both the RCCB and the circuits.

What you are talking about is RCBOs - individual RCCB/MCB combination breakers which allow each circuit to have its own protection. This is easily done, but quite expensive (£££). This could be an option (many sparks will fit a system like this in the first place), but it's best to find any faults before you start altering the CU.

When checking for an earth fault, you need to isolate the neutrals from the bar as well as opening the breakers. Obviously this involves going into the CU and working with (potentially) live neutrals, so best get a spark to do this, unless you're comfortable with doing it yourself.

Other things to check: Fridge, outside/garage sockets which could be damp, anything which is timed to come on at night, mice. These are usually the culprit. And of course, annoyingly, it can be an accumulation of small faults, making it hard to identify. This is where a spark with an installation tester comes in.

 

[rebuke]

An electrician should have an RCD tester which will allow him/her to verify the RCD is behaving correctly (i.e. it isn't tripping when it shouldn't, and it does trip when it should), so they will be able to determine if it's the RCD itself that's at fault...

As for the wiring, the problem is that when you isolate a circuit, you only isolate the live conductor, the neutral is still connected through, and if you link neutral to earth, you will almost always cause an RCD to trip (as some of the return current will go via earth, rather than via the neutral, creating an imbalance). These sort of faults can be a really big pain to track down.

An electrician would have an insulation resistance tester, which can detect if there is low resistance between any conductors - even if it's not actively in the state where it would cause a trip, it might be noticeable on the IR tester, and that would help pin point which circuit it is.

You could replace the RCD + MCB arrangement with RCBOs (not sure if you can get them for your board, but probably can), but this is effectively the same as replacing the CU, which is a fair amount of work - to do it yourself would require notification to your LABC under Part P (and unless you know a large amount about electrical work, it's not a DIY job anyway), and an electrician would charge quite a lot.

Obvious things to look for are any outside lights that might have got water in them, check the condition of any external cabling that you know of, check for e.g. any light fittings / electricals directly underneath your bathroom, in case water's leaking down the side of your bath/shower and getting into things. Did you get any new appliances (especially something that cuts in and out periodically like e.g. a fridge/freezer) around the time the tripping started etc...

 

[Spenny]

Yes it could be the RCCB at fault, although this is not the most likely. A spark will have the equipment to test both the RCCB and the circuits.

What you are talking about is RCBOs - individual RCCB/MCB combination breakers which allow each circuit to have its own protection. This is easily done, but quite expensive (£££). This could be an option (many sparks will fit a system like this in the first place), but it's best to find any faults before you start altering the CU.

When checking for an earth fault, you need to isolate the neutrals from the bar as well as opening the breakers. Obviously this involves going into the CU and working with (potentially) live neutrals, so best get a spark to do this, unless you're comfortable with doing it yourself.

Other things to check: Fridge, outside/garage sockets which could be damp, anything which is timed to come on at night, mice. These are usually the culprit.

Ah, so my month was not well spent on the basis that the trip switches do not isolate both sides of the circuit. Bah! Perhaps I just better get on the phone.

 

[sparkyspike]

P.S. You can get Starbreaker RCBOs. I've got loads. But they cost between £10 -£40 each, depending on where you get them.

 

[sparkyspike]

Perhaps I just better get on the phone.

Ok. To be honest, having someone round with the right equipment (and intelligence!) should locate the fault fairly quickly.

Fixing the fault is a different matter! Good luck.

 

[maltaron]

I had a case where the RCD tripped only when a plug was in a particular socket, any appliance which was OK in another socket on the same circuit tripped the RCD when in this socket. The fault vanished when one of the fixing screws was removed. The screw had nicked the neutral and only tripped when power flowed to the neutral via the appliance.

 

[Spark123]

I have had the same, but the RCBO wouldn't hold in, someone had used the knockout behind the mounting screw on a surface mount plastic back box.

 

[Spenny]

OK, had a qualified electrician round, ripped the consumer unit wiring out and organised it to fault trace and I now know I don't have to rewire for a long time, got rid of two faulty sockets. Took a long time!

Basically the RCD on a ramp test was firing at 15mA. Running 5 PCs and a raft of other stuff just seems to be too much. The dishwasher, washing machine had recent repairs and not been PAT tested, so they now are eliminated. Once everything was disconnected we discovered there was another symptom - the RCD would not reset without the other circuit breakers being opened then closed one by one which when I think about it, did not use to be the case.

Anyhow, though there is not one piece of equipment that is obviously faulty, it is clearly sailing close to the wind and if the clean up of the wiring isn't the answer, then I'm going to get the electrician to split off my home office wiring onto a separate RCBO, though this means that one of my lighting circuits needs to be put over onto the RCB side to make room (as my lighting is nearly all energy saving light bulbs, this is probably a minor issue).

 

[ColJack]

if you're going to do that, make it the lighting circuit with the bathroom on it.. you've killed 2 birds with one bush then...

 

[rebuke]

Was it only going at 15mA on a ramp test, or repeatably at a normal test, as if it trips at 15mA, then assuming it's a 30mA RCD, that's wrong - the regs require them to not trip at half the rated tripping current (i.e. 15mA), trip within 200ms at the rated tripping current, and (with an RCD not exceeding 30mA rating, if providing protection against direct contact) within 40ms at 5 times the rated tripping current, i.e. 150mA.

 

[Spenny]

I watched the test, it was only on the ramp test. 37ms at 30mA, 13ms at 150mA. The electrician thought it was marginal and was expecting to see around 18mA (he reckoned that Crabtree did tend to be sensitive and I think he was a little surprised to see it trip on 15mA rather than after). He was wondering whether to follow up with Crabtree though.

I think if it is the difference between 15mA being technically out of spec and 16mA being in spec, then that's not a big deal - I think I would be wanting something tripping at 20-25mA (assuming that his expected 18mA is not much extra headroom).