Overheating immersion heater

[Huge]

As summer approaches and heating oil heads ever upwards in price, I guess I better do something about my immersion heater...

Last year, it overheated, melted the plastic cap and eventually the insulation on the cables and tripped the RCD. I replaced the thermostat with one with an overheat trip, replaced the supply cabling, and "bodged" a new cap, but it has overheated again a couple of times since then. Since it was winter and the central heating was on, I switched the damn thing off and forgot about it.

So, any suggestions as to why it's overheating? I suspect the connections to the heating element are not that good, so I guess I need to get a proper crimper and make decent connections to the element. Presumably using some kind of heat resistant cable? Also, where can I get a new cap? The one I made from 4" soil pipe and PCB is not exactly a thing of beauty.

[John]

Do you mean that you didn't use special heat resistant cable? Bear in mind that the terminal is almost directly onto the heating element so a lot of heat gets conducted into the cable.

[David Hansen]

On 29 Apr 2008 10:00:11 GMT someone who may be Huge wrote this:-

Not using the right cable may have been the cause of what appeared to be overheating.

, the 1.5mm 3 Core - Buytl Heat Resistant Cable, is the one to use. If that doesn't sort things out then it is probably time for an new heater.

[Harry Bloomfield]

Once cable and/or connections have suffered overheating, the oxide formed on the outer surfaces will present a high resistance. The relatively high resistance will cause over heating and further damage.

It does need to be heat resistant, but all connections need to be clean, undamaged and tight. That likely means replacement of everything which has suffered any damage.

[js.b1]

As stated, you *must* use high temperature cable for 3 reasons

- 3kW immersion, 13Amps, will heat cable conductors

- immersion element directly heats cable conductors

- ambient temperature is typically 35oC+

The cable should be 1.5mm minimum really

- 1.5mm is ok for 16A, 1.25mm for 13A

- 1.5mm is the standard size for immersions

- 2.5mm will result in lower cable temperatures again

Anything affected by heat needs replacing & terminals need cleaning. Think yourself lucky that the melted cable tripped an rcd (0.03A RCD), rather than a shower of sparks from the mcb-side tripping (100A for

20A type b).

Butyl (85oC) is favoured over high temp PVC (85oC) but does discolour (yellow) with age.

The cost difference between 2.5mm & 1.5mm is often small, but provides a little bit of safety where someone stuffs clothing on top of the cable (cable can overheat from that). Pick up a new safety cutout thermostat if you have not already (I need to do same too).

[David Hansen]

On Tue, 29 Apr 2008 18:23:03 +0100 someone who may be Harry Bloomfield wrote this:-

The last time I checked the IET disagreed. Cables must be suitable for a number of things, including the temperatures they will be exposed to when installed. Is this something which is changing with the 17th Edition?

[dennis]

Just replaced a bit of flex.. toolstation do a 1m length of 2.5 mm2 heat resistant flex for £2.70 if you just want a bit. You need 2.5 mm2 if you have a jacket that covers the wire BTW.

[David Hansen]

On Tue, 29 Apr 2008 10:57:15 -0700 (PDT) someone who may be [e-mail] wrote this:-

MCBs operating produce showers of sparks? Fascinating.

[Grimly Curmudgeon]

We were somewhere around Barstow, on the edge of the desert, when the drugs began to take hold. I remember Huge saying something like:

Any jobbing plumber will have an embarrassingly large stock of old caps he's not got round to tossing out. Usually connected to s**te old immersion heaters he's not tossed out yet.

[js.b1]

No, not from the mcb, from the melted cable :-)

An immersion cable melts & shorts... a) L-N short -- 100A to trip a type-b 20A mcb (5*In) b) L-E short -- 0.03A to trip a rcd/rcbo

If the immersion cpd is a 20A rcbo...

- L-N short will trip out the mcb-side (100A flows)

- L-E short will trip out the rcd-side (0.03A flows)

The former, 100A across 1.5mm flex, will produce a shower of sparks.

A L-N short will not trip out the rcd-side of an RCBO.

[David Hansen]

On Tue, 29 Apr 2008 13:22:03 -0700 (PDT) someone who may be [e-mail] wrote this:-

Cables melt when MCBs operate? Fascinating.

I am actually making a serious point. Provided that the protective device and cable have been designed properly there will be no sparks, no melted cable and no damaged insulation. That is because the energy let through by the protective device depends on the current that flows and the time for which it flows.

In a dead short the current that will flow is generally going to be more than 100 amps, probably several thousand amps in most circumstances. The cable will tolerate this for a short period.

I wouldn't recommend that anyone tries this out, but I can confirm that meter tails will withstand currents of over 10,000 amps.

[js.b1]

Read what the original poster wrote.

"melted the plastic cap and eventually the insulation on the cables and tripped the RCD"

Insulation melted sufficiently to cause a L-E short thereby tripping the RCD as opposed to L-N short tripping an MCB.

Probably. Instant trip for a type-b cpd is typically 5*In with the arc tube fragmenting the arc before full pf current can be realised. Whilst pfc at the CU may be several kA (tn-c-s/tn-s) at the end of 10m of ft&e 2.5mm it is more like 500A (0.5ohm).

What the cable will tolerate depends on size & length - adiabatic calculation. Short a 10mm cable close enough to the CU and it will weld the isolator shut, blow a BS88 fuse and reveal the path of the cable ready for its replacement.

When insulation has melted off cables and conductors short there will be brief spray of sparks if bare flex strands short L-N. A small amount of copper will be vapourised. Like a recent YouTube video where someone shorted bare live wire of a 110V cord to the metal receptacle in the USA - sparks fly.

They have to. TN-C-S PSC can be up to 16kA or 24kA (within london). BS1361 fuse will tend to crap out at about 800-1100A for 60-100A.

The poster had an overheating immersion which melted the plastic cap and eventually the insulation on the cables. When insulation melts off cables any L-N short can produce sparks and a fire.

[David Hansen]

On Tue, 29 Apr 2008 14:18:40 -0700 (PDT) someone who may be [e-mail] wrote this:-

I did.