Lights Tripping RCD when they turn off........Sometimes

Talk Electrician Forum

Help Support Talk Electrician Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Joined
Mar 9, 2008
Messages
2,022
Reaction score
56
Location
Doncaster
See the attached image.

I have been sent to a building that is about 15 years old. It has recent had its Communal consumer unit changed and since then the communal Lights have been having issues with tripping the RCD. It is a shared RCD but for testing I have disconnected all conductrs of other circuits. So it is the only circuit on the RCD. I have ramp tested the RCD and it trips at 27mA. It's a 30mA so all good there.

I am assuming the lights weren't RCD protected before.

There are 24 of these fittings on the circuit and they are switched with a Staircase timer setup, though I have removed the timer to test and the results are the same. 12 of the fittings are in the communal areas of the 4 storey block and each of the 12 appartments in the block have 1 fitting in it's own respsective entrance hall. I have only managed access to view 4 of the appartments fitting and all those are non-emergency. So it looks like there are 8 emergency and the rest are non-emergency.

The issue is this. If the circuit has been dead for a while, when you first turn the lights on, the RCD trips. Flick the RCD back on and the lights will work. If the lights are only on for say, 30 seconds, there doesn't seem to be an issue. If they are on for over a few minutes, when they turn off the RCD will intermitently trip. Turn it all back on and same again. Works fine if on for a short time, can trip if been on a while. or will trip if been off for a while.

I have inspected all the fittings I can get access to and other than the original installer using black for neutral on new colours, they seem fine. I have checked the CU end an all matches up so it isn't the conductor use that causes the issue.

It has been completely bending my melon today. With it being intermitent, you can think you are making progress whilst breaking the circuit down and seeing how much you can energise and then a random trip.

In a final ditch attempt to make any sort of progress, I have disconnected all the conductors from the emergency style fittings, parked up the emergency feed in a connector block and then put the switched feed and the neutrals into a pendant. I cycled the timer 6 times before leaving for home and the circuit held the hole time. If the circuit holds throught the night with tenants using it through the evening I will have hopefully narrowed it down to either an emergency fitting being faulty or a cummulative issue with all the emergency fittings. I don't know enough about these fittings to know if they have leakage and if they do, would they be enough to trip a 30mA RCD???

The mobile reception at the building is terrible though I did manage to get the data sheet up for the emergency controller and it states they have 0.5mA leakage. I assume this is the type of leakage an RCD could detect???

I have fitted loads of these fittings in the past, but none protected by an RCD (all environments that reg didn't dictate RCD required at the time) so don't know if they have a inherent characteristical issue.

Anyone got a clue what the issue could be as my head is hurting???

Thanks

20191112_130010.jpg

 
that'll be 0.5mA when new. Personally I would stick it back on an MCB without RCD protection . This is a life critical system and needs to be totally reliable.

 
that'll be 0.5mA when new. Personally I would stick it back on an MCB without RCD protection . This is a life critical system and needs to be totally reliable.


Could be tricky as the OP says they are on a shared RCD

Why do people insist on fitting dual boards - this situation seems a highly inappropriate installation IMHO

 
Could be tricky as the OP says they are on a shared RCD

Why do people insist on fitting dual boards - this situation seems a highly inappropriate installation IMHO


depends if it has suitable 3rd neutral bar, but totally concurr twin RCD boards are ****e in my opinion, especially in situations like this.

 
Done any IR testing on the circuit or any of the fittings?

But agree with others, most split load boards can be reconfigured to give at least one circuit not protected by an RCD. You will need a spare bit of busbar and I have even added a third neutral bar to facilitate that.  If you have not modified the circuit no need for it to retrospectively be rcd protected.

 
I do wonder if commercial lighting needs RCD  protection  ,  a bit OTT  perhaps .      

You get much bigger circuits of , say,  fluorescent fittings  winding up the residual currents flowing  ...personally I think its well over the top . Creating problems that wern't there before. 

 
Done any IR testing on the circuit or any of the fittings?

But agree with others, most split load boards can be reconfigured to give at least one circuit not protected by an RCD. You will need a spare bit of busbar and I have even added a third neutral bar to facilitate that.  If you have not modified the circuit no need for it to retrospectively be rcd protected.


Yes. at 250V with L&N together to earth. I get 0.00 M Ohm. But I'm unsure if that is to be expected or not with the type of fitting. I simply Don't know. Hence I'm asking. I cannot access the first 2 fittings on the circuit (going to try today) but if I split the circuit after them and test after, the reading is the same. If I connect just one fitting, the reading is the same. I've tried multiple. Which is what makes me think it just might be normal, I don't know.

With the issue only arrising with the lights have been off or on for a while, it's like a capacitor is charging or something and causing the issue.

Could be tricky as the OP says they are on a shared RCD

Why do people insist on fitting dual boards - this situation seems a highly inappropriate installation IMHO


Tell me about it. The board does have 2 unprotected ways, and the lights were temped onto one of those prior to me visiting. but powers above me say they need to go on RCD 🤦‍♂️

If I don't get anywhere by dinner time today, I'm going to suggest tthat the light switches are RCD protected and the Lights themselves, not. See if that will satisfy them. The switces come back to board  for staircase timer. So, thinking switches activate staircase timer, Timer operates a new contactor. Lights circuit runs through contactor. I can only imagine they want the lights RCD protecting for the switch drops that are <50mm, so the above should satisfy that.

 
Tell me about it. The board does have 2 unprotected ways, and the lights were temped onto one of those prior to me visiting. but powers above me say they need to go on RCD 🤦‍♂️


I would try them on their own RCBO .........

what reg number is being used to state they should be on a RCD?

 
I would IR test just 1 fitting on it's own.  It that shows 0.0M then there is your problem.

I too would like to know the reg number that says an existing circuit has to retrospectively be RCD protected.

It "powers higher up" insist they must be RCD protected tell them all the light fittings need to be changed for something suitable.

 
Bit of an update.

So today I spent a good bit of time on the job. I split the circuit down as much as I could access and tested each leg and each fitting. Sure as damn it I went up the floors disconnecting and back down the floors testing. Found an insulation resistance fault with the leg between the board and a light in at one end of the ground floor. 0.08M between switch live (Grey) and CPC. There is an appartment in between the 2 but couldn't get in to it. No idea if the leg goes Board > Appartment > Hall light or Board > Hall Light > Appartment. I set up a temp from the Board to the Hall Light, leaving out the faulty leg. The full circuit then tested out fine. With the IR L&N to CPC you could see the figure go up as the caps in the emg fittings charged, then went >200M within about 10 Seconds. Ran the circuit all afternoon whilst doing other jobs and it was fine. Spoke to the client / contractor and explained will need to get in said appartment when possible but everything else was hunky dory. Happy days.

 
Last edited by a moderator:
It is the height of frustration when you cannot get access to the whole circuit. You have done all you can and found where the fault is.

 
Top