Newish immersion heater element has just croaked

[Dave Baker]

Went to run a bath this morning and found only cold water coming out of the hot tap. A quick look at the circuit breaker box, the immersion has its own terminal, showed that switch still in the on position so it hadn't tripped overnight.

Took the plastic cover off the immersion element and one of the wires has just melted, more like exploded actually, away from its terminal. The wiring inside it is the positive mains wire goes to the thermostat, a black wire comes out of the other side of that to one immersion terminal and the neutral mains wire goes to the other immersion terminal.

It's the black wire out of the stat that has melted from the immersion terminal. On the wire is a little sticker saying only replace with a similar link wire or summat and do not remove this sticker. The immersion is only 18 months old so is it just a crap make or should I be looking for something else that could have blown the wire apart but not tripped the circuit breaker? The wiring from the wall socket to the immersion is pretty old now (25 years maybe) but it's the pukka heavy duty stuff and anyway it's not that that has failed. The broken link wire is part of the immersion kit.

It would be a soldering job to replace this bit of wire anyway which I guess I could do rather than drain the whole thing down and fit a new immersion but I'd rather not bother if the failure is indicative of a bigger problem that's only going to recur anyway.

[Roger Mills]

If the link wire is like the one on an immersion heater I bought recently, it will have been spot welded onto the heater terminal. If that joint got hot and burnt through, it must have been a duff weld. You may be able to get a good joint using some sort of clamp-on connection. Failing that, you'll have to replace the immersion heater.

You should complain to the supplier of the immersion heater. I would be inclined to ask them to supply *and* fit a replacement - but I'm not sure how far you would get.

[Dave Plowman (News)]

Sounds like that link wire had a poor (high resistance) connection causing it to run hot and melt - you also see this in sockets where the screws haven't been tightened properly.

I'm not quite sure what you mean about it being a solder job to fix - surely the link wire is removable to allow the fitting of the immersion and or thermostat? If things are so melted that the terminals no longer exist then replacement is the only real option - they're not that expensive. The thermostat can be replaced easily - it should just pull out after disconnecting, the immersion requires draining down.

I'd normally say contact the makers about a free replacement when something fails this early - but this has all the signs of faulty installation. Even if the link wire came attached to one or the other that terminal should still have been checked for tightness on installation.

[Andy Wade]

How was said wire originally connected to the element, or is that unclear because of the damage? I think this connection is usually a

0.25 in. push-on blade type, so the wire end would have had an uninsulated crimp receptacle. If this was loose or the crimped connection wasn't properly made it could easily lead to the symptoms described.

The wire's insulation will have a high temperature rating - 105 or 180 deg., probably.

Who installed it? Was the tightness of _all_ connections checked at the time?

Check the resistance of the element (should be around 18-20 ohms for a 3 kW heater) also check the insulation resistance to earth at 500 V DC (several megohms at least). After repair (if repaired) check with a clip-on ammeter that the current drawn is around the 12-13 A mark.

I wouldn't recommend soft soldering because of the high temperature these leadouts reach. I suppose you could try silver soldering a bare wire and applying heatshrink insulation after, but then you risk cracking the ceramic bushing. Also the other end of the wire, if it goes into a screw terminal on the 'stat will probably need a crimped bootlace ferrule.

I suspect you will end up replacing the heater. At least if it's only been in for 18 months it shouldn't be too difficult to get out...

[Dave Plowman (News)]

Just thinking on, are these spade terminals? Rather like Lucar ones? If so Huge is having the same problem.

[Huge]

What's the emoticon for crossed fingers? After some very careful squeezing of the crimp terminals with needle nose pliers, it's been OK since I enquired. of course, it'll fail again now...

[Dave Baker]

The element terminal has a crimp connector, basically an open ended box made of about 0.75mm thick metal welded or soldered onto the end of it and the 2 inch bit of wire was crimped into this at the factory. Whether there was also solder on the joint there is no longer clear. Nothing is replaceable or pull-offable. The other end of the wire goes to the stat with a screw connection tightened by yours truly when I fitted it all 18 months ago and still completely tight and unmelted.

Maybe I can lever open the box with a small screwdriver to get another bit of wire into it but it'll be tricky in situ.

Resistance of the element is 14 ohms. That's about as much as I can check easily.

[Dave Baker]

The link wire doesn't need to be removeable to fit it to the stat. It just needs the stat to be in place so the link wire doesn't have to be 2 feet long. The element terminal end was crimped up and non moveable from the factory and that's the end that's blown. Maybe a bad crimp connection.

Cheeky scrote.

Anyway I doubt I kept a receipt for something costing 20 quid so I'll probably just have to drain the bugger down yet again and fit another one.

[Dave Plowman (News)]

But at least one end has to be removable - or is the immersion and thermostat one unit now?

Sounds like it. I've not come across this type.

Wasn't implying you fitted it. ;-)

Even without a receipt a crimped connection shouldn't fail, so might be worth having a go if you remember where it came from. They're *meant* to be more reliable than a screw terminal.

[John Stumbles]

Drain down? Drain a little bit off more likely (unless the immersion's at the bottom of the cylinder).

If you drain it all down the cylinder tends to move around when you're unscrewing the immersion, rather than the weight of water holding it in place.

You *did* make sure the gate valve feeding the cylinder was OK/replace it with a lever valve, last time, didn't you? Or I bet you wish you had done. ;-)

[Dave Plowman (News)]

Ah - that might explain it.

[Andy Wade]

Unless it's a 4 kW heater (v. unusual) that's suspiciously low and suggests that the element might be en-route to self-destruction...

[cynic]

Or the meter was one of those =A35 from the market stall ones

[Roger Mills]

Are you sure? According to my calculations, it needs to have a resistance of about 19 ohms when *hot*. But presumably the resistance when cold will be less than this - but I don't know by how much.

[Andy Wade]

Fairly, unless his meter is telling fibs.

Nichrome has quite a low temperature coefficient of resistivity[*]. The increase in resistance when hot will be around 3% which is about an order of magnitude less than the discrepancy in Dave's measured figure. Try measuring the resistance of a kettle element (a) cold and (b) immediately after boiling.

[*] Some data here:

[Roger Mills]

of, and that it runs at about 300 degC, which seems reasonable.

In that case, yes, the resistance of the OP's element is a bit low!

[I was probably thinking of things like light bulbs - which run *much* hotter, and where there *is* a significant increase in resistance when they're hot.]