The first picture has four wires, red & black, brown &Blue, on a 16A RCBO, presumably a centre fed radial. i might try splitting them and test for leakage separately. 

There is another 16A RCBO next to it. I might try to swap the circuits between those two RCBOs to check if results the same.

With something that did trip (and should trip) which doesn't, suggests something wrong in the RCBO

it is certainly a head scratcher

This has been going on since before Christmas, there was an old Crabtree consumer unit with an upfront 30 mA RCD that started tripping frequently, on occasions several times a day.

The new RCBO consumer unit has an occasional trip on a different socket circuit, this one supplies a few sockets in a garage conversion, but there had not been a trip for over a month until a few days ago.

So I was again looking for reasons why there have been trips, but then it appears there should be a trip when there isn’t and this circuit tested out okay.

I came home feeling like I had fried my brain. the two fan heaters I used to apply additional load are both double insulated, the kettle isn’t as good as it could be, all the RCBOs are double pole so the rest of the installation is isolated when they are off, I check for shared neutrals and so on and so forth.

Oh wow.  I feel your pain - these can be a real !"$£ er  It is the sort of thing that can give RCDs a bad name ;-)

OK if the fan heaters and kettle are altering the clamp meter reading, and 2/3 of them are 2 wire loads (I guess we  could lift the green wire in the kettle plug to get 3/3 if you like living on the edge. Probably no need) then there has to be something getting off the L-N loop. Normally an NE confusion but the IR tests do not find it. So perhaps to the N of another circuit, or to something that is not really earthed but is neutralled in some way, so would pass the IR test. Or a cheeky backfeed from the neighbours ?

I presume your ultimate IR test would be  (L+N)  relative to (CPC+MET+ bonding+ other circuits neutral bar)  with the test loads plugged in.  (I know, ECSQR, but it is only a 10 minute diagnostic, and most other countries permit)

I presume there is no time of day, rainy days or heating on or off pattern to the RCD trips and that N and E are not many volts from the fork in the flower bed or screwdriver in the lawn depending on the type of garden.

Otherwise as you describe it, I am as stumped as you are.

Mike.

It is TT and the supply may be looped through next door, so adding 6kW of load with the kettle and heaters may bump up the neutral voltage significantly, but there wasn’t anyone at home next door so I don’t know if it is looped and I cannot imagine there were any significant loads next door any way.

Four doors up there’s an EV charger, but that’s the only one I could see and I could not see any PV on neighbours houses, this house does not have either.

Hi Andy, your testers seem to be set to dc measurement… Why is that? 

www2.theiet.org/.../messageview.cfm

The Megger is measuring AC, but the Di-log is set to DC because by then I was looking for DC.

The Di-log does not zero out as I would like hence the picture of it lay on the shelf, but it definitely seems to indicate there’s DC floating about in the installation, though that picture of the DC measurement with the meter clamped around both meter tails probably raises more questions than answers.

Thanks for the reply Andy. Why were you our looking for DC?

I was looking for DC because the RCBO doesn’t trip when apparently it should.