Funny RCD trip

[Highresdis]

Hi guys.
I wonder if I can inquire of your collective knowledge on this one. A colleague passed me a job where the customer had a couple of faults. One was that kitchen lights would trip the RCD when switched on.
The installation is on a 16th edition split board, and none of the lighting circuits are on the RCD. Interestingly, the trip occurs only first thing in the morning, after which the lights can be switched on and off with no tripping.
I'm scratching my head. I tested the whole house, and found no other clues.
What do you think?

 

[SimpleSimon]

Anything on a timer? Not neccessary lighting?

Sent from my GT-I9100 using Tapatalk

 

[Highresdis]

Hmmm, they do have storage heaters. But they're on their own individual radial circuits...

 

[SimpleSimon]

Could be a link. Its strange that the lights arent connected to the RCD side but are tripping. Would go back, switch the heaters off at the board, turn the lights on and see.
another thought, immersion heater on timer? A long shot but if its downstairs light tripping, is there a leak from the tank? Possibly a leak from tank is causing a path between earths of lightong circuits and immersion circuits?causing an imbalance and RCD trip

Sent from my GT-I9100 using Tapatalk

 

[Highresdis]

Good suggestions. The difficulty is I can't reproduce the fault, cos it doesn't trip most of the time. I tried a continuity test between each non-RCD neutral and the RCD neutral bar, but no joy.

 

[SimpleSimon]

Which may show something intermittent. Something causing an imbalance.

Sent from my GT-I9100 using Tapatalk

 

[WDMDL]

Could be the high inrush current on the lighting is causing an induced current in adjacent wiring - possibly even inside the consumer unit. Is the MCB for this lighting circuit right next to the RCD?
Maybe try re-routing the cables for the lighting circuit inside the CU to minimise coupling between circuits.

Simon.

 

[Highresdis]

Thanks for that Simon. Actually, it's a double decker Hager CU (I know that's unlikely to be its official designation), and the lighting circuits are not near the RCD, so any induction must be in the cable routing I guess.
I was forgetting the RCD tripped on the 1/2 current test, so it may be it's just oversensitive. I just worry that there's a background induced current, and even if I supply a new RCD, it'll also trip on the 1/2 current test. I guess I need to retest with all other circuits isolated first.

 

[Des 56]

When your testing between the neutral bars and getting no continuity,it could be that the load causing the problem is not on demand so it wont show itself

 

[Highresdis]

Indeed.
Actually, testing between neutral bars always tripped the RCD, but I'm not sure if it would always do that on a PME?

 

[telectrix]

might be an idea to ramp test the RCD, then, if the trip current is below 22mA, replace RCD and ramp again. just thinking that it's only about 30mins work to do this, before you start delving into the circuits.

 

[Highresdis]

Mmmmm. I'm pretty sure my Megger doesn't have a ramp test facility. Trial and error I guess.

 

[Highresdis]

Ah, I find my Megger does have the cheese options! It's just that I've never used it... I guess a bit of clamp current measuring might help too.

 

[Richard Burns]

Does sound like an over sensitive RCD when a high current flows (when cold) on the lights if there is a slight neutral connection on adjacent circuits then may be enough to trip. have you tried an insulation resistance test between the neutral bars this may highlight a slight leak, but would probably be to small to show on the range.

 

[Highresdis]

No, but it's a good idea.
Thanks for these inputs guys, it gives me some leads to try..