Magic gel compound, sealing galv conduit boxes with swa cabling.

[jktelectrical]

Hi.

I have recently completed a garden lighting installation comprising of the following.

i have installed a 3 core 1.5 swa from the inside of the house from a db board backed up by a 6amp rcbo 30ma the to a one gang surface mounted switch, this then runs outside to a galv tee box swa terminated using galv 20 mm boxes cw20s glands and a small 20mm stuffing gland to a gu10 spike light 5a connectors have been used in side the box all terminated fine, box was then sealed up using magic gel then a gasket and lid have been installed, I have installed five lights in total and the las box was through box,

the problem I have is that there was mess made to the grass so after the grass was put down 24 hours after sealing in sunny weather conditions, the owner was then watering his grass, rcd triped out.

i have used magic gel in the past and its been fine, so I removed the galv boxes from ground to inspect them magic gel was quite sticky, I then threw a bucket of water of the conduit boxes and it triped out.

before connection of spike lights I carried out an ir test and this was clear.

sorry for the essay but I have explained what I can.

any help is grateful, I think it could be a faulty batch of gel.

 

[Plonker 3]

Have you disconnected the light from the SWAs?

 

[jktelectrical]

I would have to take all the gel out, can't see it being the lights.

 

[Plonker 3]

I would have to take all the gel out, can't see it being the lights.

So you visually see the lights are perfect and wouldn't trip the RCD, I think you need to revisit and actually prove it with a proper tester rather than just guessing.

 

[diditrain]

i think he means that he is unable to disconnect his lights as his joints are enclosed within the magic gel.

 

[Plonker 3]

Well unfortunately he is going to have to to find the fault.

 

[diditrain]

He will

 

[jktelectrical]

Cabling was installed and lights were connected without lamps in and dead and live testing was carried out this was all fine before boxes were gelled up.

 

[Plonker 3]

You need to disconnect the lights, regardless of what the test results were previously, it is evident there is a fault somewhere. Until you find it it will continue to trip unless it is water damage and eventually dries out.

 

[diditrain]

I think that your crude water test has answered your query

 

[Taylortwocities]

I am going to make a guess here that, although the magic gel has been poured in, there may be an area at the bottom, possibly under the terminal block where the gel has not penetrated.
This may create an air space where any moisture will gather. Its a dig out and redo issue. I have never used galv boxes under ground. Prefer the use WISKA with magic gel, etc.

Agree with the above sentiments. You can't fix problems by trail and error, testing is needed. You should be able to locate an (earth) fault by connecting L&N together and doing an IR test to earth.

 

[not clever]

Get some of the new magic POWER gel, you can mix it let it set then push it into your boxes after removing the old stuff.
As said you bucket of water has gone some way to proving you have a probable air gap in the gel.

 

[flyingsausage]

You have a 4mm threaded hole in the back of these boxes for earthing,it would make sense to put a 4mm brass bolt in there to stop water ingress.

 

[Engineer54]

Standard conduit boxes were never made to be installed in underground conditions be they filled with gel or not!! There are propriety connection/JB's specifically made for this purpose!!

 

[markythesparky]

This must at be a first, I've never come across someone testing a circuit by throwing a bucket of water over its joints! ;-)

 

[diditrain]

It`s not an approved method yet. Still under review