Immersion heater wire overheating

kavelot

Hello


I've been fighting with my 3 kWh immersion heater for a couple of months.
The neutral wire is connected directly to the heating element, while the live wire goes to the rod thermostat, then to a resettable thermal cut-off attached to the tank, then to the heating element.


The thermal cut-off (85°C) began tripping often (every other day or so) and I tried everything. I replaced all wires and connectors for ones rated 20A+ more; and I also replaced the thermal cut-off a couple of times.


The behavior is always the same:
In the first 3-4 days, everything works well. The temperature of the wires and the thermal cut-off is about 40°C when it starts running (cold tank) and increases slowly up to 50-60°C.
After a few days, even with the cold tank the temperature of the wires goes to 70°C or more after 10 minutes running and can reach 90°C until tripping the cut-off (even while the tank is <50°C)


If I bypass the thermal cut-off, then it works as expected and doesn't heat.


My current hypothesis is that even though those cut-off are rated for 16A, it deteriorates over time with the high (70-75°C) temperatures of the water tank; but I don't think this makes much sense since they're rated for high temperatures..


So I'm pretty much out of ideas.
I thought about replacing the heating element, but it works fine if I bypass the thermal cut-off


Has anyone seen this before or have any other ideas?


Thanks!

MOD NOTE: I think electrical is best for this post.

2011

Most likely the element needs to be replaced.

2011

kavelot wrote: »

If I bypass the thermal cut-off, then it works as expected and doesn't heat.

Does it heat or not?
Because you go on to say:

I thought about replacing the heating element, but it works fine if I bypass the thermal cut-off

kavelot

2011 wrote: »

Does it heat or not?

Sorry, it's a bit confusing
The heating element always work/heat as expected, with or without the thermal cut-off (if it's resetted/closed).



With the thermal cut-off, the wires overheat.
With the thermal cut-off bypassed, the wires do not overheat.



But if I replace the thermal cut-off (as I did 3 times), it always work as expected for 3-4 days, then it starts overheating.

2011

It’s overheating because the element needs replacing.

kramer1

Change the rod stat?

meztli

@kavelot I think I have a similar problem, may I ask you how you measured the temperatures?

Chomper.

kramer1 wrote: »

Change the rod stat?

Hmm ya, he hasn't mentioned checking it?
That would be obvious thing to check


Heater caked with limescale ? Likely

John.G

If you are talking about a dual immersion then the neutral always does the switching. (picture courtesy of Bruthal)

Chomper.

John.G wrote: »

If you are talking about a dual immersion then the neutral always does the switching. (picture courtesy of Bruthal)

Yes internal(immersion) switches the neutral so control switch(immersion switch) can switch phase

kavelot

meztli wrote: »

@kavelot I think I have a similar problem, may I ask you how you measured the temperatures?

I used those laser thermometers

kavelot

John.G wrote: »

If you are talking about a dual immersion then the neutral always does the switching. (picture courtesy of Bruthal)

Mine has those timer switches that only triggers one of the heating elements


Here's a picture of my bottom one (it's a mess because of my tests)

kavelot

As for the problem being the rod or limescale: Those make sense, but I can't understand why they'd stop when I bypass the thermal cut-off switch

John.G

kavelot wrote: »

As for the problem being the rod or limescale: Those make sense, but I can't understand why they'd stop when I bypass the thermal cut-off switch

Is the live wire also attached to something at arrow1?.
What is arrow3 attached to, if anything?
Is arrow 4 the neutral attached to one of element terminals?

Can you identify on the photo the two element terminals?

If you leave the thermal cut out out of the circuit does the control (rod) stat switch the immersion on/off as per normal without any wiring o/heating?

kavelot

John.G wrote: »

Is the live wire also attached to something at arrow1?.
What is arrow3 attached to, if anything?
Is arrow 4 the neutral attached to one of element terminals?

Can you identify on the photo the two element terminals?

If you leave the thermal cut out out of the circuit does the control (rod) stat switch the immersion on/off as per normal without any wiring o/heating?

the 2 element terminals are arrow1 and arrow3
arrow1 is connected to neutral (you can see the blue wire coming from below it, pressed on the screw)
the live wire doesn't touch arrow1, it just goes behind it and connect to the rod


yes, without thermal cut-out the circuit works as expected (no heating and it switches off after reaching ~70C)

kavelot

oh, and the ground is connected just to the cap screw, not to the heating element


sorry about the weird angle, here's a better one

Chomper.

I'd say you guys are overthinking all this, carrying out repairs at the heater terminals etc

As first suggested it probably needs replacing assuming control stat is working properly

John.G

kavelot wrote: »

the 2 element terminals are arrow1 and arrow3
arrow1 is connected to neutral (you can see the blue wire coming from below it, pressed on the screw)
the live wire doesn't touch arrow1, it just goes behind it and connect to the rod


yes, without thermal cut-out the circuit works as expected (no heating and it switches off after reaching ~70C)

TCO = Thermal Cut Out

1. If the TCO is still in the circuit but not affixed to the immersion boss, is the
problem still there

2. If the TCO is still in the circuit but IS affixed to the immersion boss, is this
when the the problem appears?.

If the problem is still there in 1. then it can only be bad connections or the TCO
isn't rated for the current and is overheating, if the TCO is over heating then
this will be very apparent as it will be very very hot to the touch.

If the problem only appears in 2 then a bit strange because the control stat should switch off the immersion at 70C and its difficult to see the water temperature rising another 15C to trip the TCO.
You could say that the element is so covered in limescale that the whole immersion is getting very hot before the water temperature reaches the control rod temperature but I think the element would burn out before this happens.

Your answer to 1. will reveal a lot, if no problem there but problem at 2. then the element should be renewed IMO.
If the problem is there at 1 then I would buy a combined control and overtemp (rod) stat and throw away that TCO.

Curious_Case

Maybe the heat is being generated by a bad connection rather than a component drawing too much current.

Possibly a contact that's not quite open enough when it should be open or not quite closed enough when it should be closed.

Any sizzling sound or that characteristic smell in the air when sparks cross gaps?

Any sign of flashes around element connections at night time when room lights are off?

whizbang

Thats a lot of corrosion for a device thats supposed to be dry and warm. Is that a concrete floor.?
Maybe something galvanic going on..
Is there a RCD on the supply?

I suspect every connection is too corroded for usefull current capacity.