A merry Christmas to WHJ and everybody else too.

Clearly there has to be some form of functional switching, but why isolation?

For maintenance purposes? If say, a plumber needs to work on any fittings withing the shower enclosure?

Or let us look at the problem another way - what if any reg would I be contravening if I just straight-thru crimped the former isolator connections instead of simply replacing the isolator?

Click do a decent shower pull switch with 2 screws per terminal like the decent ones used to be. The problem i have with them is the cable entry at the back should be a bit bigger so you can slide some slack up and down into the ceiling so not forcing it up and putting stress on the terminal because the cable wont bend (usually 10mm). I have started making it a bit bigger with a multi tool and it works fine so long as the cable doesn't get disturbed by somebody above if it's a roof space. 

Gary

whjohnson said:

Given that on Xmas Eve was called out to yet another example of an electric shower isolator burning off it's neutral at the terminal, is it time to cease fitting isolators for the likes of ovens and showers?

To answer that question, we need to get to the bottom of what's happening in that particular case(s).

And I think the answer could be, at least in some cases, that the Neutral conductor gets slightly hotter than the line conductor for shorter lengths of run from shower to isolator. Copper is a good conductor of heat as well as electricity, and, depending on installation method, if there's not a lot of opportunity for some heat to be lost through the sheath, if the cable is close, or over, its 70 deg C operating temp for its current carrying capacity, the isolator switch (and possibly shower terminations) are overheating.

Why more often Neutral and not Line under these conditions? On a number of electric showers, the Neutral comes straight off the element via a very short wire to the fixed wiring terminal, whereas there's a loop through the thermostat/over-temp cutout, and switching, internal to the shower.

So, it's possible in some cases it's cables near, or above their conductor operating temp for one reason or another, rather than a particular problem with isolators?

Other cases ... well, cable access vs enclosure size etc. has often been stated.

But I think more evidence/statistics of the real causes might be needed to determine what ought to be advised?

Why more often Neutral and not Line under these conditions?

I've always had a theory that bad L connections get spotted by routine L-PE loop tests (or R1+R2) so when the reading is a bit on the high side everything gets tightened up - whereas bad N connections could go completely unnoticed (until they overheat at least).

+1 for the design of shower isolator that has twin screw terminals - and terminals fixed to the back part - so no re-arrangement of the conductors to put the cover on. (It might be better if the design aligned the screw directions (say all E-W, rather than star fashion) so that there wasn't always one that was just too tight to the wall to get a decent sized screwdriver in straight.)

   - Andy.

gkenyon said:

But I think more evidence/statistics of the real causes might be needed to determine what ought to be advised?

I will have a go at getting that, at least in terms of how common it is.

I took the electrical team of one of the biggest social landlords in the U.K. through the 2391. They told me that one of the top maintenance issues found during inspections was what they called “neutral burn-back” at the shower pullcord. 

I will seek to get further info from the section manager. 

I note that these isolators are now being marketed as being rated at 50A rather than the traditional 45A, although I do wonder if it is simply a rebranding exercise and that the internals remain exactly the same.

There may be some substance in what you say about the cable length of the neutral when it comes to the internal wiring Graham, although this is pretty much out of our control. I do not know what the installation method details are for this latest job. The isolator is a 2 gang type mounted in a 2 gang cavity box in a stud wall. Cable looks to be 10.00mm T&E, MCB is 40A rated and sited in a dual RCD hager board which looks quite recent, although it is not labelled up in any way!

Returning to my question - is it a crime against BS7671 to omit the isolator altogether?

Get your tin hat on right now Whj , you are starting a fight here

Returning to my question - is it a crime against BS7671 to omit the isolator altogether?

Not a crime, but maybe takes a bit more to achieve a similar level of safety and convenience using the CU. Single pole MCB for isolation? Certainly not on a TT system and probably not on a TN system if its downstream of an RCCB (or accidental touching of N-PE risks tripping the RCD and taking out other circuits), The CU is probably remote from the equipment to be isolated - so the means of isolation will need to be able to be secured - how many plumbers or DIYers are likely to carry lock out kits for all the various brands of MCBs/CUs?  And it's simpler and easier to pick the correct isolating device if it's right next to the equipment.  Maybe more of a design challenge that simple BS 7671 compliance - but if a compliant solution is in some ways less conveniently safe than 'the norm' - is it really acceptable?

Maybe it's just that the accessory standards need  bit of improvement....

   - Andy.

gkenyon said:

if the cable is close, or over, its 70 deg C operating temp for its current carrying capacity, the isolator switch (and possibly shower terminations) are overheating

I can see that the CSA of a slightly loose connector could be less than that of the cable, but otherwise wouldn't the whole cable show signs of distress? (Of course, it might be distressed where it is buried in the wall, etc.)

I am going along with WHJ - from the sound of it, these switches add danger.

If we really do need 2 pole isolation, what about FCUs?

How often are electric showers replaced? If it comes to it, the main switch could be used for isolation. Granted the wifi would be off, but would that really matter for a few minutes?