Jaymack said:

I suggest that the problem is with limp wristed installers or the switch location being covered up with excess, thermal insulation.

The general observation being it's "usually the Neutral" that shows signs of overheating, sort of leads us down the path that the problem is not only due to "limp wristed installers" as you put it (although overtightening could also be a problem ... having said that, now more manufacturers are specifying torque settings with most of their range, that ought to be addressed over time).

Just wondering, though, whether installers check the shower end of the cable too - to check whether it's overheating at that end as well? I've certainly seen Neutrals at the pullcord and shower overheat at the same time, with cables arguably right at the limit to comply with BS 7671, but not under-rated, if you ignore a very short - < 300 mm -  through thermal insulation.

I am erring towards the N not checked on testing scenario as being the leading factor. I think the only case where L might be longer than N is in lighting switch circuits but not so much the reverse and therefore N being longer. Yes I`d be rather inclined to suspect the N not checked as much for being the main deciding factor if most other things are equal.

I checker the shower-end and all was well. The cable looks to be 10.00mm but I couldn't find a KW rating label on the shower unit itself. The old isolator was a Contactium Inspire range one. I don't see many problems with Contactum gear as a rule, but can no longer get it.

We must also take into account the gradual dilution of product quality which has taken place over decades due to value engineering exercises mandated by bean counters. Smaller steel cage terminals with poorly plated steel screws which have a different thermal coefficient of expansion/contraction to that of copper which doesn't  help.

MK, the former Rolls Royce of electrical accessories is a prime example of this. I won't fit MK any longer as a consequence. Even their Logic range of light switches seem to fail with monotonous regularity.

whjohnson said:

We must also take into account the gradual dilution of product quality which has taken place over decades due to value engineering exercises mandated by bean counters. Smaller steel cage terminals with poorly plated steel screws which have a different thermal coefficient of expansion/contraction to that of copper which doesn't  help.

Whilst I don't want to discount this observation, I'm interested in how "dilution of product quality" explains why "it's the neutral that burns" ... well at least most of the time, if the general observation (without statistics) on this observed phenomena is true?

If product quality were the issue, then surely we'd expect to see as many Lines burning as Neutrals?

gkenyon said:

If product quality were the issue, then surely we'd expect to see as many Lines burning as Neutrals?

I suspect that there may be some confirmation bias going on.

And of course - like the washing machine repair man, firmly convinced that all modern machines are rubbish as he only sees the broken ones, we do not count all the ones we go past that are working just fine. Especially for pull switches that get damp from below and cold air from the loft coming in with the cables from behind, the conditions are good for a condensation problem, and that may well finish otherwise perfect contacts.

I am not sure if any of the switches where whcih pole is "N" has been  defined break neutral after live to allow the neutral contacts to be undersized - it is not true in DP sockets, where the L and N rockers are the same,  but in 3 phase switch gear having an odd-legs neutral is not uncommon. NExt time someone is staring at a hot neutral shower switch, if they could take a hammer to it, and compare the weight of the metal work on the N and L sides it would be very interesting.
Mike

How about a contactor located at the DB, with a control circuit only going to the shower isolator? It could even be an ELV control circuit.......

I know that the contactor cannot be relied on as actual, proper,  isolation, but it could eliminate burn out in the shower and cooker isolators, which I too find extraordinarily common and almost always in the neutral conductor........this is probably a greater danger than the isolation - which should be carried out by a professional anyway. 

Might be a way forward for the future 19th edition.......

My contact in the social housing provider's maintenance team has reported back that there is indeed a sizeable number of shower pullcord switches with thermal damage with a disproportionate number where the neutral is burned back. He tells me that often bubbling is evident on the surface face of the pullcord.

I am sure that if someone in IET wanted to establish more precise facts, he would be willing to engage.

I have replaced a few of these in my own home (around one every couple of years), such is the paranoia I actually keep a new one in stock in my garage. When I change them the cores are not found loose. When the new one goes in I take time in manipulating the cores in the box for the least stress, push the switch into the box then lower it just enough to get a screwdriver on to recheck the terminal screw tightness. I have taken old switches apart and generally the switch contact faces are blackened, mostly on the neutral.

Interestingly if you consider the operation of the shower; current breaking at the ceiling switch should not be a problem because everyone turns the shower off at the shower then at the ceiling switch. Logically at the time the ceiling switch is operated you are only switching off the neon lamp within the shower that indicates the power is on.

Not sure if anyone else has taken an old shower apart for fun, sadly I have. When you switch it on a solenoid valve operates which admits the water to the heater and this has a pressure switch on it that switches the heater elements in. This is done for safety so that the heaters cannot be on unless the water flow will be adequate to prevent overheating.

From memory inside the shower unit only the live is switched on the heater and solenoid valve and the wiring in there is not particularly large, certainly smaller than the 10mm^2 supply cable. I am wondering if there is something else going on, perhaps due to the inductance/capacitance of the heater elements, is the damage to the neutral contacts occurring when the switch is opened?

If the neutrals are being disproportionately affected, why? The current is the same. The terminals are not obviously smaller. The conductors are the same. If any current were being diverted down the CPC the RCD would trip, but if not fitted, the current in the neutral would be reduced compared with the line. So let's have an explanation please!