Electrical work in the work place

[JohnW2]

This issue was covered in issue 172 of NICEIC's Connections magazine.

Replacement of accessories The replacement of an accessory, such as a socket-outlet, ceiling rose or lighting switch, on a like-for-like basis may be regarded as maintenance rather than installation work.
However, NICEIC expects 'Approved Contractors, Conforming Bodies and Domestic Installers undertaking such maintenance work to carry out essential inspection and testing to ensure that the replacement accessory is safe to use before being put into service.
As a minimum, tests to confirm that shock protection has been provided are essential. These essential tests are:
:- a test to establish the earth fault loop impedance [the value of which needs to be,checked against the characteristics of the protective device)
:- polarity, and
:- where an RCD is used for fault protection or addition protection, the correct operation of the RCD.

Thanks. At first sight that appears to be a very clear answer to the question I was asking. However, when I start thinking a bit more (always dangerous!), I'm not so sure that it all makes complete sense. I obviously have no problem with a desire to 'ensure that an accessory is safe to use before being put into service', but I'm struggling to decide in what sense, for example, a replacement (non-metallic) lighting switch would/could be unsafe to put into service without undertaking those tests (even once one has decided how to undertake them on a light switch!)?

Kind Regards, John.

 

[OwainDIYer]

but I'm struggling to decide in what sense, for example, a replacement (non-metallic) lighting switch would/could be unsafe to put into service without undertaking those tests (even once one has decided how to undertake them on a light switch!)?

Live wire is trapped when pushing switch back against metal back box, box becomes live, box isn't earthed, protective device doesn't operate, faceplate screws remain live. If a child pushes finger against faceplate and makes contact with screw the risk of injury is not just shock but if the child jumps back, falls over and bangs head against desk, suffers brain injury.

 

[JohnW2]

Live wire is trapped when pushing switch back against metal back box, box becomes live, box isn't earthed, protective device doesn't operate, faceplate screws remain live. If a child pushes finger against faceplate and makes contact with screw the risk of injury is not just shock but if the child jumps back, falls over and bangs head against desk, suffers brain injury.

Fair enough - that's not impossible, even though it requires at least two faults/errors. However, which of NICEIC's 3 required tests (EFLI, 'polarity check' & correct operation of RCD (if fitted)) do you think is going to reveal that situation - and exactly where/how in relation to the light switch would you conduct those tests?
(and, of course, if you removed the switch from the box to undertake any of the tests, the faults may well go away!).

Kind Regards, John.

 

[RF Lighting]

Well if you've proved EFLI at the switch, then you can not end up with a two fault situation as automatic disconnection would take place.

 

[JohnW2]

Well if you've proved EFLI at the switch, then you can not end up with a two fault situation as automatic disconnection would take place.

I'm not sure I understand. If you measured Zs using live side of the switch and the CPC, you'll presumably get a 'normal' reading (assuming no third fault on the circuit!) and no disconnection - and the L-box fault will also probably have gone away when you removed the switch from the backbox to do it.

The scenario postulated should really be detected without any testing. The first fault (unearthed back box) should be detected by inspection, not testing - and once that's detected and rectified, a protective device ought to operate as soon as the switch is put back in its box (and the circuit energised), if the second postulated fault (trapped L conductor contacting the box) is still operative.

I can see this discussion is heading in the direction of my being accused of being awkward or silly - but I really just asked a genuine question to which I was interested to learn the answer!

Kind Regards, John.

 

[westie101]

I think part of the issue is going too far into the theoretical, this is where experience and technical knowledge come into play by assessing each situation and finding an appropriate solution.

The reality is that not every situation can be predicted so trying to find a "one size fits all" or a written procedure is impossible.

 

[JohnW2]

I think part of the issue is going too far into the theoretical, this is where experience and technical knowledge come into play by assessing each situation and finding an appropriate solution. The reality is that not every situation can be predicted so trying to find a "one size fits all" or a written procedure is impossible.

I'm sure you're right. I innocently asked, out of interest, what tests an 'average electrician' would usually undertake after just replacing a mechanically damaged socket - and we seem to have (as so often) been led off into discussions of rules/requirements/guidelines and hypothetical (unlikely but not impossible) scenarios! ... and I guess that I'll be the one who'll get blamed for the 'contrived and pedantic' discussion

Kind Regards, John

 

[westie101]

This sort of discussion can be useful and often occurs in the office!

But as with a lot of the situations on here it cannot be easily answered without being there.
To be honest I would be more concerned with a replacement without some thought as to the cause of the damage and an assessment of the fitness for purpose of the replacement. (or even the original for that matter)

 

[JohnW2]

To be honest I would be more concerned with a replacement without some thought as to the cause of the damage and an assessment of the fitness for purpose of the replacement. (or even the original for that matter)

Sure, particularly in a school, or other public place I agree - the location might be putting it at risk of damage, and maybe it ought to be of more sturdy construction etc.(or not there at all). However, this (tangential) discussion has been entirely about what testing should be done after it is replaced (decisions about whether/how/where//withwhat it should be replaced presumably already having been made).

Kind Regards, John.

 

[westie101]

Though not decrying the need and value of testing, it is in a lot of ways like an MOT only valid until you drive out of the garage!
We had a large number of failures at on time where we were correctly testing 11kV cables only for them to fail within minutes of energising them

 

[Adam_151]

Well if you've proved EFLI at the switch, then you can not end up with a two fault situation as automatic disconnection would take place.

I'm not sure I understand. If you measured Zs using live side of the switch and the CPC, you'll presumably get a 'normal' reading (assuming no third fault on the circuit!) and no disconnection - and the L-box fault will also probably have gone away when you removed the switch from the backbox to do it.

Your ELFI test will prove that the box is earthed, if your phase-> earth fault does occur when scewing the switch back after testing then it'll safely disconnect.

The scenario postulated should really be detected without any testing. The first fault (unearthed back box) should be detected by inspection, not testing

And if its not that a CPC is not there, but that its not connected to earth the other end... how would inspection pick that up?

 

[securespark]

When changing a light switch I check EFLI at the fitting it controls, then do a quick flying lead check to ensure the switch/box has an acceptable reading too.

 

[JohnW2]

Your ELFI test will prove that the box is earthed, if your phase-> earth fault does occur when scewing the switch back after testing then it'll safely disconnect.

True, but we are talking abou OwainDIYers hypothetical scenario in which he said that the box was not earthed. If you used the box for the EFLI test, then that would obvioulsy pick up the fault. If you used the CPC (which was OK, but not connected, adequately or at all, to the box), then you would not pick up any fault - and the accessory (screws) would remain dangerous, despite the tests, if the L-box contact then recurred when the plate was fitted back to the box.

And if its not that a CPC is not there, but that its not connected to earth the other end... how would inspection pick that up?

I confess that when he said that the box was 'not earthed', I took that to mean the absence of a connection (or a satisfactory connection) to the box, not a situation in which the local wiring was all OK, but the CPC was not providing a satisfactory path to earth. I agree that could not be picked up by inspection, but would be detected by an EFLI test.

However, I'm still interested in knowing whether electricians normally would understake such testing after replacing a light switch, say in a domestic setting? Some clearly would - but would most of them?

Kind Regards, John.

 

[JohnW2]

When changing a light switch I check EFLI at the fitting it controls, then do a quick flying lead check to ensure the switch/box has an acceptable reading too.

Fair enough. Thanks. That's pretty thorough.

Kind Regards, John.

 

[Adam_151]

If you used the box for the EFLI test, then that would obvioulsy pick up the fault. but the CPC was not providing a satisfactory path to earth. I agree that could not be picked up by inspection, but would be detected by an EFLI test.

You obviously visual the termination of the cpc at the same time as you do the loop test, you don't test to a cpc terminted in a connector block and then totally ignore the fact its not to the box!

However, I'm still interested in knowing whether electricians normally would understake such testing after replacing a light switch, say in a domestic setting? Some clearly would - but would most of them?

I'd do a quick loop test at switch and visual at the DB, and fill out a minor works cert with the zs value and tick for polairty entered and make it clear it refers to worked upon point only