Fusebox/RCD fault

[John D v2.0]

Agreed with the two previous posts, actually the electrician should have tested the wiring with his calibrated test equipment to make sure there is no leakage or faults. Then he should have tested the two RCDs to make sure they do and don't trip at the right levels. If he's tested the wiring and the rcd he will know which are within specification and will not be using this binary trial and error technique.

If the rcd is not tripping with nothing connected, then with only the circuit connected and no appliances, it shouldn't trip even if it's well out of spec. As the wiring should be >1Mohm so an rcd would struggle to even notice that current.

 

[stillp]

the amp rating of the new RCD is 100 while the old one was 63 - is that a problem?

No. It's the tripping currect (mA) that is critical, the 63A or 100A is just the maximum current it can carry.

 

[ban-all-sheds]

Agreed with the two previous posts, actually the electrician should have tested the wiring ...

You forgot either the quote marks around the word 'electrician', or the phrase 'soi-disant' before it.

 

[garviel]

I didn't watch everything he did. He was there nearly 3 hours. Did a lot of remedial work on most sockets where the Earth was connected to the backbox rather than the socket itself. Separated some old wires that had been touching.

After replacing the the RCD he carried on testing, including via the sockets to check the system tripped (using the same large diagnostic box).

I'm not sure what more I am supposed to have done!

 

[Iggifer]

You're not supposed to have done more, changing RCDs until one holds isn't really the way to fix it.

Insulation resistance testing should have quite quickly found where the fault was/is.

As others have said, RCDs are rated to trip at 30mA of leakage, but in reality they're more likely to actually trip in the mid-low 20s. This isn't an issue in a working installation as there should be little to no leakage. What may well have happened here is he's come across an RCD that's actually closer to 30mA and you might still have 25mA of leakage

 

[JohnW2]

As others have said, RCDs are rated to trip at 30mA of leakage, but in reality they're more likely to actually trip in the mid-low 20s. This isn't an issue in a working installation as there should be little to no leakage. What may well have happened here is he's come across an RCD that's actually closer to 30mA and you might still have 25mA of leakage

Indeed - or, if he didn't test it properly, even maybe more than 30mA.

It seems to me that the most odd thing is this ... the 'electrician' seems to have believed that the problem was a 'faulty' RCD, so he replaced it. However, he also got trips with the other RCD in the CU - so, by his reasoning, that one was also 'faulty', so it seems irrational that he didn't also do that.

As you say, the most probable explanation would seem to be that there is a 'leakage' somewhere (which needs to be found and rectified) and that both of the original RCDs had low enough trip thresholds to respond to it, but the new 'replacement' one has a higher threshold, and therefore doesn't quite trip.

Kind Regards, John