Garage RCD conundrum.

[Bobble23]

Can anyone help with this one I've been scratching my head over for a week?

Called to a customer whose garage RCD keeps tripping. TNCS supply, three core SWA submain to RCD garage unit, two radial circuits (6A lights, 16A sockets). IR test from main board on whole garage installation 0.01 mega Ohms so thought great, find the problem, sort it, job done. Problem was, when I tested the individual garage circuits there was no problem, >200 mega Ohms. So I wondered if there was a problem with the sub main and tested it separately, again >200. So next I checked if there was a problem with the board, and, by accident, found out that the low IR reading only happened when the (metal) board was screwed into the (concrete garage).

Given that, I could mount the board on a piece of timber screwed into the garage, but I can't get my head round why having the garage presumably acting as a big earth rod would be causing the tripping.

Any ideas?

Thanks.

 

[telectrix]

so the low IR is when the DB is connected to earth via the garage concrete?

 

[Dave OCD]

I wonder whether the garage is precast with steel reinforcing rods maybe with one of the DB fixing screws in contact with one and the SWA has damage in the ground ? It's an odd one ! Have you tested between the conductors and armour with both ends disconnected ?

 

[Bobble23]

Yes, telectrix, low IR when DB screwed into concrete garage.

And I wondered the same thing Dave, but made some new holes in the garage wall and the same thing happened. To be honest, I haven't tested the armour, just between the conductors. I'll do that tomorrow. But assuming I find nothing, I suppose I'm trying to work out if there's any potential danger from fixing the board onto timber so the problem goes away. Although I'd really like to find out what's happening first.

 

[Dave OCD]

Keep us updated please mate, it's an interesting one for sure !

 

[goasis]

Could there be a screw/nail clip etc damaging a live or n conductor somewhere in the garage, which itself is screwed into the fabric of the garage? No damage between conductors, hence the individual circuit IR 's being ok but a connection through the building fabric to the earth terminal of the cu. Assuming your low reading is on a l/n - cpc test?

 

[Bobble23]

Thanks for the replies all. I also thought about something going through the conductor to the wall without hitting another conductor.

The customer wants a couple more sockets putting in so I'm going back tomorrow and will check over every inch of cabling/accessories (fortunately not too much of it to look at, just a couple of sockets and two lights) but if nothing shows up I think I'll mount the DB on timber. Not ideal but if I do all of the test results will come up perfect which I suppose is all I can do.

I'll post again if I find anything, but thanks all for the help anyway.

 

[ackbarthestar]

Being a TNC-S and supplying an external structure that may have extraneous metalwork, it might be logical to assume that there are two methods of earthing that is interfering with the correct operation of the RCD. The low IR value could well be the garage RCD being energized when 500V DC is put across it.

 

[Bobble23]

Thanks Ackbar. So does that mean I could test that by bypassing the RCD on the garage unit for my IR test? If the reading is then acceptable your theory is right, timber mounting gives the practical solution and conundrum solved?

 

[Wilko]

Hi - just my thoughts, for what it’s worth .
If the metal CU in the garage was earthed it should not matter whether it was on the wall or floating in the air. So I suspect 2 faults - board not earthed (confirm continuity from garage earth bar back to MET with long lead) and a NE fault that’s causing the trip.

 

[davesparks]

What has a low IR when the board is screwed to the wall, how low and between which conductors?

I'm curious as to what made you unscrew the whole board from the wall as part of the fault finding process?

 

[Kidsolo]

I suppose I'm trying to work out if there's any potential danger from fixing the board onto timber so the problem goes away.

You will fix the problem on the board... but you are still leaving the low reading on the walls, thus making a potentially dangerous situation between the wiring and the garage walls. If the board is properly earthed your problem will return on the board.

 

[Bobble23]

Thanks for all the help everyone. Just as an update, if I'm honest I still can't quite understand how this makes sense, but I went round last week and replaced the SWA sub-main from the house to the garage and that solved the problem, all readings acceptable, even with the garage unit screwed direct into the concrete.

I inspected the existing SWA as much as I could and couldn't see any problems, but couldn't get to approx 5 mtrs of it under some decking at the house end (20 mtrs away from the garage!).

So, one of those ones where it's solved but I can't work out why. Thanks again anyway.

 

[DPG]

Thanks for letting us know.

 

[Timbo]

Cable damaged when decking installed?