imersion heater thermostat fault finding

[Andrew B]

Hi all, hoping someone can offer some knowledge/experience of immersion heater fault finding.

My parents-in-law's imersion heater stopped working, on inspection I found the time switch had failed open. Also the isolation switch neon indicator was dead. The heater itself had a good resistance of 20ohms, good insulation resistance and good continuity back to the time switch via the thermostat control.

I replaced the time switch and isolation switch, tested and called it a job done.

A month later and the fuse has blown in the CU, re-wireable 15A type. Testing of the circuit didn't reveal a fault. I put it down to a very old fuse and replaced it. It's a Memera CU so unfortunately can't get a CB pulg in for it so had to stick with fuse wire.

A week or so later and the circuit has gone again. Not a blown fuse but the live wire at the CU to the immersion heater had melted, and unfortunately the wire terminal in the fuse slot had fused making that slot unusable.
Tests of the heating element & insulation resistance were all fine. So now I'd put this down to a damaged wire after the initial fuse blow and then a failure of the now weakened wire.
At this point I've left that circuit disconnected, no hot water for the in-laws, as it needs further work and investigation.

I should add that checking the timer (mechanical type) I can tell that the circuit has failed after several hours each time I've had to investigate. So it's not switching into a dead short.

Questions:
1) I have to revisit my assumption of an old fuse causing the blown fuse and damaged wire causing the second failure. A genuine overcurrent is more likely. So why can't I find an insulation failure or shorted heating element?
2) I've not been able to test the thermostat turns off as I've a cold tank and a minimum temperature of 40 on the thermostat. Could the fault be that the thermostat is stuck on and the fuse (and the wire) failed after hours of full load current, where as it would usually only be an hour max until it reaches temperature.
3) the damaged slot in the CU, and the age of the CU make reinstating the circuit as it was impossible. The preferred option is to completely replace the CU and bring it up to regs. However they are in the middle of selling the house and the new owners (if they have any common sense) will rewire the lot. So in the immediate term my suggestion is to put the immersion heater on a FCU from the ring main. Do you think that's sensible?

Thanks for reading and for your help.

 

[Midwest]

Hi Andrew, welcome to the forum. I can't read your profile, so we have no idea of your competency. I think despite your bravest efforts, you have thus far failed to resolve this fault, and it seems you now have some damaged cabling & terminals/connections that will need replacing.

If you are not a competent electrician, I would recommend your parents-in-law employ one, no offence meant to yourself. This forum does not allow step by step guides to DIY'ers, and as it seems there might be several things wrong with this install, it would be wrong to suggests fault finding from afar, which may leave other faults in situ without rectification.

 

[telectrix]

my first thought would be the element corroded and although it reads OK when cold, it's failing when hot. expansion causing the fault to become apparent.

 

[Andrew B]

No offense taken. I've done my electrical inspection and test qualification, 2391-52, so am considered competent. Although my day job is as an electrical design engineer for aerospace so I do have limited experience on domestic wiring.
When I say I've tested the circuit, I mean I've tested to the procedure set out in 2391 with a multi function tester. As I say I've not been able to see any fault with the immersion.
When it was just a time switch that looked faulty it was an easy fix but for changing the CU or changing the circuit to a FCU I would get an electrician in, mainly as I'm not Part P registered.

Thanks for the input telectrix, any idea how to diagnose that type of intermittent fault. Or is it just a case of change the heater and see if the problem goes.

Just to say again, I will get an electrician in now but just wanted to learn if you would have approached this differently. And if you think a FCU will be ok? As the in-laws are selling the house I'd prefer not to install a new CU at this stage if possible.

 

[Wilko]

Hi - it certainly seems that an overload current has been carried in the circuit and this has now damaged some components. I agree with the previous posts, you'll need help to fix this one.

 

[Octopus]

Did you IR the circuit?

 

[telectrix]

Did you IR the circuit?

according to OP, 2nd para.

The heater itself had a good resistance of 20ohms, good insulation resistance

specsavers are open all day..

 

[Midwest]

Is it me, do I need to go to specsavers, but I have OP #4 (as an alert)coming on-line after tel's #7, yet the order of posting is shown as above? Very strange, seen it before on other threads.

Back on topic, I would agree as tel's suggested, is the original fault. What are the IR readings when testing the element (disconnected), live to earth?

 

[Andrew B]

Is it me, do I need to go to specsavers, but I have OP #4 (as an alert)coming on-line after tel's #7, yet the order of posting is shown as above?

This could be because as a new member my posts have to be approved by a moderator first before they appear publicly.

IR readings @500V, element (isolated at DP isolation switch) to tank & PE >5GOhms. isolation switch supply L-N >5GOhms, isolation switch supply L-PE >5GOhms.

So sounds like i need to replace the immersion element as well as fix the connection to the CU.

Would a normal electrician be able to fit a new immersion heater or do i need a plumber as well now?

Can any one see an issue with using a FCU to connected the immersion circuit to the ring main, thus saving a new CU?

 

[Octopus]

^^ what do you mean by normal electrician?

Are you not a spark?

 

[Midwest]

Are you sure of your Gohm unit? Not farmliar with that myself.

If the element is pants, I would use a plumber to replace it myself. Some others on here would do themselves. Putting immersion on an RFC is not a great idea, but as a short term measure it might be ok

 

[Andrew B]

^^ what do you mean by normal electrician?

Are you not a spark?

I've passed my electrical inspection and test BS2391-52. But i don't work as an electrician, and i'm not Part P registered. In my day job i design power electronics for the aerospace sector - I've done the BS2391 qualification so we have someone on site that can inspect and test our test equipment installations.

Are you sure of your Gohm unit? Not farmliar with that myself.

result was 5000 Mega Ohm, which is 5 Giga Ohms (GOhm).

If the element is pants, I would use a plumber to replace it myself.

Thanks, i thought that would be the answer. I'd prefer to be sure the element is faulty first but so far no one has been able to suggest a method to test the thermostat.

 

[Andrew B]

Ah looking into it, it seems you can replace the thermostat without replacing the whole heater assembly.

So now i need to be sure its the thermostat that is faulty (not turning off once temp has been reached) as i suspect.
Or if the element has an intermittent fault that only appears when its been heated.

Considering the thermal damage to the circuit wires in the CU, and the fact that the circuit has failed only after several hours of being active. ( time switch shows the element had been on for hours before the circuit failed). I'm leaning toward the thermostat being stuck on and so the circuit is conducting 11A continuously for hours which gives a longer term thermal overload.

 

[buzzlightyear]

need a plumber

IMO probably. the best lager in the word !
having the copper cylinder that element is inside needs taking out and replaced .

 

[westward10]

I most definitely would not connect it from the ring final until you have resolved why it ruptures a BS3036 15A fuse first.

 

[Midwest]

result was 5000 Mega Ohm, which is 5 Giga Ohms (GOhm).

Thanks, i thought that would be the answer. I'd prefer to be sure the element is faulty first but so far no one has been able to suggest a method to test the thermostat.

A reading of 5Mohms, would be quite an acceptable reading IMO, but as tel says heating of the element can affect the insulation of the element.

A modern thermostat will have an overheat stat, but that's only to prevent the water overheating and circulating into the cold water tank, its not going to cause the fuse to go. As Westward said, and as I did before, you need to rectify these faults. Not suggesting domestic circuits are more complicated than aerospace circuitry, but its horses for courses; I still would recommend getting a competent sparks to investigate.

 

[Wilko]

Hi - just wondering what instrument you've got with 5,000 meg Ohms fsd ? (Mine only goes to 999).

 

[telectrix]

Hi - just wondering what instrument you've got with 5,000 meg Ohms fsd ? (Mine only goes to 999).

my megger Ir tester reads up to 999 Gig.so ner, ner.

 

[davesparks]

There seems to be a bit of a plumbers approach to fault finding going on here, just replacing parts until it starts working again without a real logical process of fault finding.

If the thermostat has failed closed and the circuit remains live for a couple of hours then someone would have noticed the water boiling or the overheat cutout (if fitted) will have operated.
If the thermostat has failed open a simple continuity test will identify it.

As far as I can see only the supply cableing has been IR tested and not the actual heater element so far?

The cable melting at the terminal in the CU suggests a possible poor connection there.

 

[Andrew B]

If the thermostat has failed closed and the circuit remains live for a couple of hours then someone would have noticed the water boiling or the overheat cutout (if fitted) will have operated.
If the thermostat has failed open a simple continuity test will identify it.

Continuity of thermostat has been checked and is closed as expected with a cold tank so i know it hasn't failed open.

Your might be right that someone would have noticed the water boiling if it had failed closed- although my parents-in-law are old and have medical issues so i wouldn't bet on it. (hence them currently selling the house to move into sheltered accommodation). This thermostat doesn't have a cutout

As far as I can see only the supply cableing has been IR tested and not the actual heater element so far?

I tested between heating element line connections and PE, this gave a high insulation resistance of 5 GOhms. I did actually state this in post #9

just replacing parts until it starts working again

I think that is a bit unfair.
On the first time round I identified a failed mechanical timer and no other faults: continuity, IR and element resistance all good. The mechanical timer is at least 30 years old. So i think it was fair to consider it a failed part and simply replace it.

The second time around a fuse had blown - again all tests came back OK, no obvious cause of the blown fuse. All tests, continuity, IR & element checked out. No visible damage other than the blown fuse. Again with nothing to go on I think it was fair to replace a fuse, consider it was perhaps due to a 30 year old fuse being disturbed when the circuit was first examined.

On the third time round, no blown fuse and no power to immersion heater. clearly continuity issues which on investigation was found to be broken wires due to thermal overload in the CU. Clearly there is an unidentified fault and that's how I've left it.

I'd like to hear what you would have done differently?

 

[telectrix]

suggest clamping feed to heater and waiting for it to heat up. see what happens.

 

[Midwest]

I think that is a bit unfair.

I'd like to hear what you would have done differently?

Don't mind davesparks, he's like that with everyone.

I think you need to go back to basics, back to the beginning.

If your CU has been damaged by this issue, that needs repairing or replacing.

Then get your immersion final radial circuit repaired, tested and inspected, so its functioning correctly, without any load attached to it.

After you have done all that, then you can move onto the immersion heater itself, and its ancillary controls.

Still think, you should admit defeat, and employ a competent electrician, which if your CU needs replacing, you'll have to do anyway.

 

[PEG]

Hi,i think you have done more than expected,to help the in-laws,it may be time for help.
As for if a "normal" spark could tackle the tank side...well,a normal one may have a go,but "special" ones,always have an 86mm spanner and some towels...

 

[DefyG]

What is happening to the fuse wire as that can tell you what sort of fault you are looking for.
Have you re-tested since replacing the timer as its possible that the element was good then but faulty now?
Is the flex between the timer and the element good, not squashed, burnt or damaged in any way?
Are you sure the time switch is connected correctly and you haven't created a dead short as soon as the timer operates, maybe contacts welded together causing permanent dead short?