Electric Underfloor Heating - Tripping RCD

[Daniel Coakley]

Hi all, wondering if anyone can help me with a recent issue I've had with electric UFH in a small flat. Found that when the UFH starts, the RCD trips immediately - this started last Saturday while I was away. Prior to this, I had heard some buzzing coming from near the UFH unit controller in my kitchen, but didn't realise there was an issue.

Here are some details of the system

• Controller: TH-132 (15A RES Max.)
• Mat area (total) ~ 5m2 (covers a small kitchen and hallway)
• Installation Date ~2005 (it was installed before I moved in, so I don't have any manufacturers details / specs)

I did some checking yesterday and noticed that the UFH controller is fed from a wall socket in the kitchen. When I opened the socket, I noticed the earth had come free of the terminal, and there was some evidence of arcing damage on the live wire at the terminal also. I pared back the line and earth, and refitted them, thinking this might solve the problem (and might have explained the buzzing I'd heard). When I reconnected the UFH, the problem was still there, and the RCD tripped immediately.

So, pulled out the multi-meter and ran a few tests. Results as follows:

• Tested resistance across mat (LOAD L/N) = ~28 ohms
• L-E > 1 Mega Ohm
• N-E = 29 ohms

I am writing this post from memory of the readings I took yesterday so I may be slightly off. As I don't know the spec of the heating mat, I have to assume the resistance reading I got on the mat seems reasonable, as it would equate to ~8.5A current (which I assumed was about right given what I've seen for a system this size (here's the resistance table I was comparing from another manufacturer).

Insulation test on Line to earth seems okay, but Neutral to earth is not? Am I correct in thinking I should be getting the same reading across both of this. When I carried out the test, the RCD was off, but Main switch was on.

Does anyone have any idea what the problem might be, or what further tests I should carry out to diagnose the issue? Was hoping this was a problem with the main wiring or controller, as I can't afford to take up teh floor right now, and it's not going to be a fun Christmas in Scotland with no heating .

Any help is much appreciated

 

[Pete999]

Hi all, wondering if anyone can help me with a recent issue I've had with electric UFH in a small flat. Found that when the UFH starts, the RCD trips immediately - this started last Saturday while I was away. Prior to this, I had heard some buzzing coming from near the UFH unit controller in my kitchen, but didn't realise there was an issue.

Here are some details of the system

• Controller: TH-132 (15A RES Max.)
• Mat area (total) ~ 5m2 (covers a small kitchen and hallway)
• Installation Date ~2005 (it was installed before I moved in, so I don't have any manufacturers details / specs)

I did some checking yesterday and noticed that the UFH controller is fed from a wall socket in the kitchen. When I opened the socket, I noticed the earth had come free of the terminal, and there was some evidence of arcing damage on the live wire at the terminal also. I pared back the line and earth, and refitted them, thinking this might solve the problem (and might have explained the buzzing I'd heard). When I reconnected the UFH, the problem was still there, and the RCD tripped immediately.

So, pulled out the multi-meter and ran a few tests. Results as follows:

• Tested resistance across mat (LOAD L/N) = ~28 ohms
• L-E > 1 Mega Ohm
• N-E = 29 ohms

I am writing this post from memory of the readings I took yesterday so I may be slightly off. As I don't know the spec of the heating mat, I have to assume the resistance reading I got on the mat seems reasonable, as it would equate to ~8.5A current (which I assumed was about right given what I've seen for a system this size (here's the resistance table I was comparing from another manufacturer).

Insulation test on Line to earth seems okay, but Neutral to earth is not? Am I correct in thinking I should be getting the same reading across both of this. When I carried out the test, the RCD was off, but Main switch was on.

Does anyone have any idea what the problem might be, or what further tests I should carry out to diagnose the issue? Was hoping this was a problem with the main wiring or controller, as I can't afford to take up teh floor right now, and it's not going to be a fun Christmas in Scotland with no heating .

Any help is much appreciated

Difficult to diagnose the problem remotely Dan, could be as simple as remaking the connections you have tried to repair, best bet call an Electrician in to find out what is wrong, and recommend a solution.

 

[Strima]

You say the earth was disconnected? This may have been a previous fault and someone disconnected the earth to get around the tripping RCD.

 

[PEG]

Although you are using the wrong device for some of the tests,were all the conductors disconnected from the controller,when you tested them?

 

[Daniel Coakley]

You say the earth was disconnected? This may have been a previous fault and someone disconnected the earth to get around the tripping RCD.

Thanks, Strima - that's possible, although it looked like the earth had just popped loose, and may have just happened when I removed the socket, as it was sitting just off the terminal, and wasn't taped off with insulating tape or anything.

I assume I've got a short somewhere given the Neutral to earth reading? Just not sure if that's in the wiring to the controller, or in the UFH mat. I'm a little new to this and just trying to get as much info as possible before calling out a specialist (if that's what's needed). It's difficult to get an electrician here on short notice, especially in the run-up to Christmas. If there's anything further I can check in the meantime that could help root cause the problem, that would be great.

 

[HappyHippyDad]

It looks like you have a N-E fault. You have 28ohms N-E which means they are touching.
I think this is just the mat you are testing which I have assumed means you have completely disconnected from the mains supply (hence the results indicate a fault). If not then you may still get a similar reading (N-E) if your supply is Tncs or TNS as the earth is connected to the neutral.

 

[Daniel Coakley]

Although you are using the wrong device for some of the tests,were all the conductors disconnected from the controller,when you tested them?

The conductors were terminated in the back of the TH132 control module, but the front panel was removed, so they would have been isolated from each other (I think). See here for the user guide for my controller: http://www.electricity.gg/media/55479/Aube-TH132-User-Guide.pdf

 

[davesparks]

The conductors were terminated in the back of the TH132 control module, but the front panel was removed, so they would have been isolated from each other (I think).

In which case the N to E resistance you have measured could well be the neutral to earth connection of the incoming supply to the property rather than a fault.

 

[Daniel Coakley]

In which case the N to E resistance you have measured could well be the neutral to earth connection of the incoming supply to the property rather than a fault.

I think I am just checking the mat, as the L/N connection to the UFH controller spur from the socket, and I just checked the resistance across the Neutral from the mat, and the common earth. I attach a drawing of the circuit, the way I think it's currently wired (may be incorrect, as I've not had a chance to pull out everything to check. Therefore, I think when the faceplate of the controller is disconnected, everything should be isolated (there are just 6 terminals for the conductors in the backplate on the wall). Regardless, the mains is not fed directly to the UFH controller, but it seems to be spurred from the back of a kitchen wall socket. Hope this makes sense

 

[Daniel Coakley]

I've updated the diagram as I didn't show the earth to the mat, which is what I was testing. I am just going off what I can see at the moment without pulling out off some of my wall panels. However, I think this is the way it's been wired up.

 

[Intoelectrics]

You need to disconnect the mat and test it separately to see if the fault is with the mat or the controller. Since you say it only trips when the mat is energized, I assume anything plugged into the socket its fed from does not trip the RCD?
Is the controller fed directly from the socket outlet not via a FCU? Very poorly installed if so.

 

[HappyHippyDad]

Just keep it simple.
Turn off the main switch in the consumer unit. This will guarantee that you have isolated the N from the earth. Then if you still get the same reading of 28ohms (or similar) N-E then you know you have a fault.

 

[Daniel Coakley]

You need to disconnect the mat and test it separately to see if the fault is with the mat or the controller. Since you say it only trips when the mat is energized, I assume anything plugged into the socket its fed from does not trip the RCD?
Is the controller fed directly from the socket outlet not via a FCU? Very poorly installed if so.

Yeah, I will do a bit more testing when I get home later. As you suggested, I will disconnect the mat conductors from the terminal and test those individually to avoid interference, and also switch off the mains.

The controller is fed directly from the socket as far as I can see, no FCU in sight :/. This isn't the first bit of poor wiring we've found in the flat - there was a separate issue on the lighting circuit last week which was resolved by an electrician that called out yesterday (I don't know the details of what was wrong, as I wasn't home when they fixed it). I'm not sure if the issues could be interrelated in any way.

Anyway, I think I know what I need to do next - I'll isolate the mat (switch off the mains etc.) and check the insulation across L/N, N/E and L/E. If there is a fault in the wiring to the mat or controller, I'll get that fixed, and also look to installing a 13A FCU between the socket spur and the controller (controller is rated to 15A max switching current). Would that be correct? Perhaps it's the controller itself that's the issue.

 

[DPG]

I'm surprised there isn't an FCU feeding the controller.

Keep us posted with what you find.

 

[Daniel Coakley]

Forgot I had a photo of the socket on my phone. Bit of a mess - you can see the three conductors feeding out from the terminals to the UFH. The Live is a little charred as it seems to have been arcing at the socket, neutral seemed intact, and the earth was disconnected. I'm going to replace the socket itself also, as it's quite old, and is possibly a source of the issue. Looking at it again, it's quite likely the neutral feeding out the back of the socket may be slightly damaged, and is now shorting to the casing possibly. Will update with findings later

 

[Silly Sausage]

View attachment 46545

Rough as ...!

 

[Markyd]

grey cable to live?........is that yellow cable definitely an earth?
sorry C.P.C

 

[Strima]

Looking at the state of those conductors I would completely disconnect and get someone round to test it fully.

 

[DPG]

Agree with above. It needs a spark in to a) test it, b) tidy it up, and c) ensure it has proper fuse protection.

 

[telectrix]

as hhd said, dis. the mat leads completely from the contoller and IR test them. L-E and N-E.... L-N should read on continuity, anything from 20 ohms upwards depending on the wattage rating. if those readings are good, then look back towards the socket.