Blown fuse on underfloor heating.

[redharry123]

Hi,
I have an underfloor heating mat that has been down for 8 years laid on a insulated concrete floor and under tiles in the hall. It was checked with a multi meter before and after installation.
It has worked perfectly all that time but yesterday it blew the 13a fuse on the switched fused spur.
I replaced the fuse and it switched on and started working then after about half an hour it blew the fuse again. Because the fuse holder cracked when getting the blown fuse out I have put a new fused spur on and again the fuse has blown after about 20 minutes or so.
The floor hasn't been damaged or the electrics disturbed in any way, no water has got into the electrics or under the floor.
I have checked the wiring into the fused spur and from there to the thermostat and everything is nice and tight and no signs of overheating or anything.
I can get access to where the tails go into the floor but don't want to rip the skirting board off just yet.
Has anyone any ideas where to go from here ? any help would be appreciated.
thanks.

 

[Taylortwocities]

Is there a programmer that controls this? If so, one thing you could try, is to connect the heating element direct to the supply, thus eliminating the programmer, it may be faulty. Run it like that for an hour and see if the floor warms up and the fuse stays intact.

Otherwise...You are going to need some rather out-of the-ordinary test equipment to investigate this. Probably a low-ohm test meter and an insulation resistance tester. If you do not have these then you are going to need an electrician to check this out.

 

[ericmark]

What type of heating mat, there are three basic types, the fixed length type, you can't cut it, you have to order length required, then there is the twin wire with element between there is a necking every so many inches and you can cut it after each necking point, and third is chemical an impregnated material between feed and return warms up and the higher the temperature the more the resistance so is to an extent self regulating. Some mat has an earthed braid around it so suitable for bathrooms, some does not.

You can measure the resistance both line - neutral and line - earth, most need RCD protection so a line - earth fault should take out the RCD not the fuse, with the chemical type crushing can cause damage, wrong tile cement can fail causing the cable to be crushed.

You say it was tested with multi-meter on laying have to still got the readings and have these changed? Really speaking a multi-meter is not good enough, need an insulation tester which uses 500 volt and a continuity meter which uses at least 250 mA, but the multi-meter may still help. I would be looking for physical damage, loose tile etc. In the main UFH stops working rather than starting to draw too much current.

So big question is it RCD protected?

 

[Taylortwocities]

I wondered about an RCD too. He hasn't mentioned that one had tripped, but may help to know if we are looking for a true short cct or a potential earth fsult.

 

[redharry123]

Thanks to both of you for quick replies.
To answer your questions in order.
There is a programmer and thermostat combined - I have just tried bypassing it as suggested and the result is the same, 13a fuse blown after a few minutes.
The heating mat is a fixed length type.
There is an RCD, the UFH fused spur is on the ground floor sockets circuit. The first time the fuse blew it also tripped the RCD, subsequent times it has not tripped it.
I don't have the original multi meter readings unfortunately.

thanks for your help so far.

 

[ban-all-sheds]

If you have to lift the tiles, replace the mat, and re-lay the tiles, would you do that, or would you decide you no longer had UFH?

Because if it would be the latter, then you should consider going straight there, instead of via paying an electrician.

 

[redharry123]

Yeah, I will have to consider dumping the UFH as it is a costly nightmare ripping it up and doing it all over again.
I am hoping that there may be some reasonably cheap way to determine the problem so I can make a sensible decision. For example if the mat is faulty and the problem spot could be pinpointed I could take a tile up repair the spot and put a new tile down as I have some spares.

 

[endecotp]

Is there a hot spot on the floor?
(Spray some water and see where it dries first?)

 

[ban-all-sheds]

The UFH is drawing 20-25A - I wouldn't suggest running it in order to see if you can find where it's too hot.

 

[Agile]

This is a very interesting scenario because the mat is drawing about 20 amps whereas the normal current, depending on the size which we are not told, might perhaps be 4-8amps .

Does anyone know how the resistance elements are arranged? Live and neutral down each side and resistance in between?

I would want to know if there is any resistance to earth.

If it can be run for a shortish time then an IR thermometer might give some indication of where the fault is by scanning the surface temperatures.

If it was mine then I might cut the live side before the fault point to enable me to still run the remaining area if that was much. I would also be considering controlling the power input if the whole mat was apparently taking an over current.

Has anyone encountered a fault like this where it takes a constant but over current?

Tony

 

[redharry123]

Thanks for your thoughts on this.
Re some of the questions.
the mat is approx. 14 square metres, it runs through the hall into the kitchen.
I have not found any hot spots when it was running, and I now can't get it running long enough for it to really start warming up much. I believe (but not certain) the earth in the mat is between the live and neutral, although where the cable tail comes into the programmer the earth is braided around the live and neutral.
I have obviously got a few people scratching their heads. Any more thoughts would be welcome.

 

[Taylortwocities]

In a UFH element, the earth braid is on the outside. Like TV coax cable. The element is a single conductor that starts at L and goes all the way round and returns to N.
SO you need to be looking for fault between line (or neutral) and earth. And a big fault like that may be seen on a regular multimeter.

 

[ban-all-sheds]

A regular digital camera might serve as an IR detector. If the room was cold and dark (try leaving the window open overnight and testing before daybreak), and the UFH turned on, a hotspot might show when looking at the screen.

 

[endecotp]

It's doesn't seem to be tripping the RCD (except the first time) so it sounds like it's not a live-to-earth fault.

A miltimeter might reveal that the L-N resistance is much less than it should be.

 

[ban-all-sheds]

This is a very interesting scenario because the mat is drawing about 20 amps whereas the normal current, depending on the size which we are not told, might perhaps be 4-8amps .

Maybe even a bit more..