RCD both tripping and refusing to trip!?

[Greg:]

I have been away from a refurb job while other (inc. the owner) have been in and on my return I found that the RCD (for a cooker circuit and three power circuits) is tripping. It triped even when a fairly basic piece of equipment such as my makita drill charger of the owners kettle was plugged in. Thankfully my kettle was fine. It tripped when items were put onto each of the 4 circuits with the other circuits switched off. I put in a new RCD and the same happened.

Then, at least for me, it gets stranger. I did an RCD auto test on the new RCD in connection with the 4 circuits mentioned and it refused to trip. I consistently got a reading of >30mA and I was perplexed and began to question my tester. I went out of the premises, knocked on some neighbours doors and presented a request which roughly said "this may sound strange but it would really help me if I could test your electrical equipment". With good grace someone let me and my tester tripped the RCD in an expected fashion.

I took the new RCD out and wired it in as an external item to the board. sequence of equipment was Main switch> breaker> RCD> socket> tester and the results that I got for my auto test were 34ms, 34ms, 24ms and 10.9ms.

I did some brief IR tests (I was expect elsewhere and missed NE on three of the circuits) and got results typically in the 300 Mohm range. Ze measured 0.25 ohms.

Nothing was plugged in while tests were conducted and the isolator for the boiler was switched off.

This whole thing doesnt make any sense to me and advice would be appreciated.

 

[trev]

Yes and it seemed to be going well until today.
I just cant see how all four circuits on the one RCD (cooker and three power) can be affected when the only other circuits are lighting!? I've not seen evidence of a circuit crossover.

Think about what the RCD is doing and how it works mate

 

[Greg:]

??

You also note issues with the kitchen circuit - so best you start there!

gulp OK

I had done Zs measurements at initial testing but only once recently at a socket near the DB

I also just realised that when I said I had tested one circuit at a time I had only switched off the other three breakers but not disconnected the related neutrals. I'm thinking that this is a good way to go.

 

[Deleted member 9648]

Sounds to me like you have only half tested it. Disconnect each circuit one at a time and carry out IR tests on it,also worth testing between the disconnected neutrals and the rest of the connected neutrals for 'borrowed' neutrals. Always keep in mind that if circuits are good,and RCD's are not faulty then they will hold.It always surprises me the number of threads coming up in which the OP states all tests are fine and the RCD replaced yet it still trips.....clearly all tests are NOT fine then!

 

[Greg:]

Sounds to me like you have only half tested it.

yep i didn't have time for three of the NE IR tests and hadn't thought through the rest. Am really grateful for the advice on testing between neutrals etc.

What's also on my mind is that the house seems to have been working fine. It was bought with no certification, I conducted initial testing and from then it didn't trip once.

its a basement and ground floor flat and yesterday I worked on the ground (and kitchen inclusive) floor in other rooms than the kitchen converting pendants to downlighters. I have been really careful to drill around the conductors and core out separate holes into which the conductors will be redirected. In all this I only saw one instance in which a bit of sheathing was scuffed but I guess I have to face the posibility that I missed something.

On the topic of the ceiling there was another thing that really worried me. It is plaster/fire proof board city plus ... three layers on the basis of lard and plaster. the thing that worried me, and counting down from the lard and plaster, was that there was a one cm gap between the first and second layers.

what advice or comments would you make to the owners?

 

[trev]

Lard and plaster. Ew that'll be horrible and sticky.
Best tell the homeowners to scrape it out before they get mice mate

 

[Octopus]

yep i didn't have time for three of the NE IR tests and hadn't thought through the rest. Am really grateful for the advice on testing between neutrals etc.

What's also on my mind is that the house seems to have been working fine. It was bought with no certification, I conducted initial testing and from then it didn't trip once.

its a basement and ground floor flat and yesterday I worked on the ground (and kitchen inclusive) floor in other rooms than the kitchen converting pendants to downlighters. I have been really careful to drill around the conductors and core out separate holes into which the conductors will be redirected. In all this I only saw one instance in which a bit of sheathing was scuffed but I guess I have to face the posibility that I missed something.

On the topic of the ceiling there was another thing that really worried me. It is plaster/fire proof board city plus ... three layers on the basis of lard and plaster. the thing that worried me, and counting down from the lard and plaster, was that there was a one cm gap between the first and second layers.

what advice or comments would you make to the owners?

So there is your starting point!

 

[ipf]

Lard......Sounds like it must have previously been a commercial kitchen or takeaway.

 

[Sintra]

Yes and it seemed to be going well until today.

I just cant see how all four circuits on the one RCD (cooker and three power) can be affected when the only other circuits are lighting!? I've not seen evidence of a circuit crossover.

I also don't know how it is possible for an RCD not to trip when tested in one circumstance but to trip as expected in another. This is a burning question for me.

I will also fabricate a "circuit" from the RCD in situ test it in another way.

Is this a split board? Are there lighting circuits on both sides? If so turn off all lighting mcb's and try your rcd.

 

[snowhead]

The only way forward it to assume every single cable on every circuit has been vandalised by persons unknown.
Not forgetting outside as well, e.g any excavations that may have damaged the incoming supply cable.

Start at the begining with all circuits disconnected, test the obvious ones first where you know there has been a problem, cut cables, exploding radios etc.

 

[jaypp]

When I first returned to the job I tinkered with a lighting circuit

its a basement and ground floor flat and yesterday I worked on the ground (and kitchen inclusive) floor in other rooms than the kitchen converting pendants to downlighters

if you used jb to replace pendant and wire d/lights, check connections there (particularly if pendant was loop in/out) and perhaps check 2way from basement re borrowed neutral as stated in other replies (if feed @ switch)
just a thought if n-e tests don't help. also re-check ring continuity and figure of 8 and maybe socket that didn't like your choice of music.