Rcd tripping strangely

[ejmoore]

Hi

Don’t often post on here but I’ve dreamt about this issue last night. Need advice please

Yesterday I attended a property to complete an EICR, home owner had moved in the previous week.
Customer has not had any issues with the install, all circuits powerd on a 16th edition cu split load sockets on RCD1 3 ring final circuits 2.5 1 mm T/E 2 of which were on B32 1 on a B40 which is none compliant. Would swap this for a B32 if test proved good
So I come to end of my EICR with just a few C3 no earth sleaving labels, recomend rcd for circuit 1-6 on none rcd side. Then couple c2 lack of supplementary bonding in bathroom and ip rated fitting needed in bathroom.

So end of my test I checked the trip times of the RCD 126mA 1x 36mA 5x trip button operated.
So all good!

On trying to reset the mcb on the none rcd side while the rcd was set it would trip the RCD when I tried to set any of the b6 lights. Then also when I set any of the sockets circuits.
Thought borrowed neutral, so checked all wiring on no rcd all good. Found a N-E fault on upstairs sockets, the circuit which is was on the B40. This did not present itself on my dead test.

So I split the ring located the leg and split again found my 2 points of fault.
It was getting late on in the afternoon, I had another job to attend. Explained to the customer I would return. Left the circuit dead, to carry on on my return, 1.5hrs later.
When I returned, continued with my test, no fault present.
Scratching my head, for a while.
I put everything back together reconnected to the cu, changing the B32 now for aB16 Powered up, all good, normal operation. Left the customer with a working circuit, to attend after the weekend for FI

I’m lost on this, need an explanation if anyone could shed any light I would very much appreciate it. Something I’m not seeing, missed?

Many thanks
EJ

 

[Cid]

Too much going on there. Unable to decipher your question.

 

[ejmoore]

Sorry, just trying the tell you of the happenings. I’m not overly good at getting my point across through literature

Probably why I’m not understanding the issue, waffling in my head. I’ll try taking one step at a time

1st why would an rcd trip when setting mcbs on the non rcd side. No N-E fault on that side. Rcd side mcbs off. Seems there is an imbalance.

 

[Wilko]

Hi - there is some chance Owner has changed something while you were away. Disconnected a lamp or appliance that had a fault perhaps. Another possibility is the fault was at the backbox (?) where you've taken it apart and it's no longer in contact.

 

[Cid]

Possibly one of the non-protected circuit was/is connected to the protected neatral bar?

 

[ejmoore]

Hi - there is some chance Owner has changed something while you were away. Disconnected a lamp or appliance that had a fault perhaps. Another possibility is the fault was at the backbox (?) where you've taken it apart and it's no longer in contact.

Plastic back box, although this particular back box had a telephone cable extension cable running through.
I know I may of disturbed something during inspection and played with the movement of the cables, still didn’t re-present the fault in doing so.
I suppose the owner could of done something while I was away though.

 

[ejmoore]

Possibly one of the non-protected circuit was/is connected to the protected neatral bar?

Nope, I checked followed all T/E at cu all good.

 

[snowhead]

Something along the lines of loose Neutral at the C.U Isoator or meter or cutout.

 

[DC-backfrom the past]

could the cables be nicked at the consumer unit? when you disconnect and move them to test it could be allowing contact. then separating them again its taking the fault away again . I would check at consumer unit end. years ago I saw sparks cutting the insulation with stanley blades and sometimes nick the insulation.

 

[Toneyz]

I have read your post a couple of questions that are not clear.
I assume that you changed the 40A TO 32A?
Why did you change a 32A to 16A?
Did you do a global insulation test to earth? (there could be a nail or something between different circuits)
Were all the loads disconnected when testing?