thoughts on non notifiable work

[Tom__40]

Hi. I would like to extend my ring main which I understand is non-notifiable. I also recently had an electrical inspection and know that I need a new consumer unit in the near future. I'm confident I can do the work safely (cables would be clipped direct to joists using the metal cable clips, and run in plastic conduit in vertical chases in the walls), however am I right in thinking it's also a legal requirement to certify the work, i.e. a minor works certificate in this case? Options as I see it:-

1)Do the work myself shortly before having the consumer unit upgraded, possibly even fill in minor works certificate myself (if that's allowed, I have limited testing equipment, I think the only thing I'd need would be a loop impedance tester as all I have for that now is a plug in tester that tells you it's below a certain value). The consumer unit upgrade will then give everything a full test anyway?

2) Do the work myself and try and find one of these 3rd party certifiers to write me a certificate. Not sure what I'd pay for this.

3) Get it done at the time of the consumer unit install at ??how much extra cost.

Be interested on peoples thoughts!

 

[ericmark]

There is no legal requirement to do inspection and testing, but the home must be save for workmen, be it police, fire, ambulance, or carpet fitter at some time some one will need to work in your house so health and safety at work act will apply. Also those considered as not able to look after their own safety children etc. But the 18th edition wiring regulations are not law, but can be used in a court of law.

For DIY it is near impossible to comply with every requirement of HSE, building regulations, and wiring regulations, some rules will be broken, so really what is done is a risk assessment, how likely are you to get it work and produce danger?

The regulations state for a ring final the volt drop at centre should not exceed 11.5 volt, that equates to 106 meters of 2.5 mm² cable, we can work out with maths that is the in coming supply has an impedance of 0.35Ω and the centre socket has an impedance of 0.94Ω line - neutral in both cases then we are right on the limit and there will be 106 meters of cable.

However there is very little today which has a problem with volt drop, fridge/freezer if not inverter type, old audio equipment before they started using switch mode power supplies, I am sure there are more, but as far as danger is concerned it does not really cause a danger as such, breaking the ring final could cause an overload danger and cables could over heat, and for the earth loop impedance going over the limit can result in the magnetic part of the MCB/RCBO not being able to work, but in real terms the RCD will trip with a line - earth fault so may be regulations require it, but is does not really cause danger.

On a new build an electrical firm can use semi-skilled labour who are monitored by an inspector, this has been done throughout time, it is of course how an apprentice learns his trade, however there is one big difference to having a third party tester or LABC tester, no one needs power until tester has visited, in the main once a house is occupied we want power restored within the day.

So in real terms we take a chance, the consumer unit upgrade will mean the ring final it tested to ensure it is a ring, but they will not test volt drop, he may test the earth loop impedance at some random points, but unless there are some glaring faults he/she is unlikely to find them.

As to faults like surface wiring without metal clips very unlikely to find that. Most of the plug in testers with loop test the pass mark is 1.5Ω or higher, and a ring final needs I think 1.36Ω it use to be 1.44Ω and I would have to work it out, but the basic problem is with a 16A supply as used in Europe yes they will show if loop impedance too high for the magnetic part of the MCB/RCBO to work, but in the UK with a 32A supply not really going to show a fault. A 25 amp type B MCB needs 1.75Ω and most of the plug in loop testers will show if above that figure, first light comes on with most between 1.5Ω and 1.7Ω so they will rest a 25 amp supply, but to test if under 1.36Ω you need an expensive loop impedance tester with a actual reading, not simple go/no go lights.

I we want to keep to regulations then there is really no point in DIYNOT as all we would be saying is don't do it. £500 for test gear means it is not financially viable to DIY, my loop impedance tester has failed, I have a insulation tester they are cheap enough around £35 and a low ohm meter, but RCD tester and loop impedance tester stopped working, as to if a RCD tester gives a true result anyway not sure, we were trained to test it with nothing switched on, however we have type AC, A, and if car charger likely type B, and the tester I had would test type AC, but it would not show if it still worked when there was DC on the circuit, so it is like the MCB we don't really test, we just hope it does what it says on the can.

 

[motorbiking]

have a read of this https://electrical.theiet.org/bs-76...england-and-wales/frequently-asked-questions/

I wouldn't clip the cables to joists, its good for them to have a chance to move if a nail or drill bit comes their way. Make sure any non-serviceable junctions boxes use maintenance free boxes with strain protection (i.e. not the old circular junctions).

You don't say why your consumer unit needs replacing. If there is no RCD/MCB protection, I would prioritise this over the extension. You don't want to be in a position that you get some leakage that is not enough to burn the fuse, but will cause problems when you upgrade. Not to mention the safety benefits of having better protection.

Non-Notifiable work - is just that. Unless you are planning to rent the place out.

 

[AndyPRK]

you could put the back boxes in, run cables in, and let the electrician connect them up?

Maybe best to get agrement from sparky where they will be extended from first

 

[JohnW2]

Hi. I would like to extend my ring main which I understand is non-notifiable.
1)Do the work myself shortly before having the consumer unit upgraded, possibly even fill in minor works certificate myself (if that's allowed, I have limited testing equipment, I think the only thing I'd need would be a loop impedance tester as all I have for that now is a plug in tester that tells you it's below a certain value). The consumer unit upgrade will then give everything a full test anyway?
2) Do the work myself and try and find one of these 3rd party certifiers to write me a certificate. Not sure what I'd pay for this.

Almost by definition, virtually all (non-notifiable) DIY electrical work is done per your (1). You can complete a MWC yourself but, as you say, you might well not have all the kit necessary to undertake all the testing that is, strictly speaking, required in order to complete a certificate - but in that respect you are in exactly the same position as 99%+ of people who do DIY electrical work - so 'go figure'

As for your (2), "3rd Party Certififiers" (rare though they are) are really relevant only to notifiable work (I suppose they should really be called "3rd party notifiers"). You could, in theory, go to any electrician and ask him/her to undertake what was effectively just a very limited-scope EICR, on the circuit you had worked on but, even if you could find someone prepared to so that (for a sensible price), I very much doubt that (m)any people would do that!

Kind Regards, John

 

[Tom__40]

Hi. To give some background, the consumer unit needs an upgrade as it's just MCBs with no RCDs (well, one for a shower circuit that's not in use but that's it) - I only got the house a year ago. The wiring itself is all PVC (old colours), done around 37 years ago (judging by newspaper clippings next to old junction boxes, house was built ~1955 and there's still bits of old rubber wiring that's been cut to show that it has been rewired once. Inspection revealed N-E faults on the lighting, which I couldn't reproduce with an insulation resistance tester after changing the junction boxes on the circuit to WAGO's in WAGO boxes (very easy to use, strain relief and rated as maintenance free if I needed it, which I don't). I only installed the WAGO boxes to make it easier to try and narrow down the fault....wasn't counting on not being able to find it at all. A couple of bits of lighting cable could do with replacing where there's single core double insulated stuff used for the two way lighting, beyond that I figure it's fine as the IR readings I was getting were >200 Megohms and it's lighting, very low load anyway. The existing ring main tested fine and is ~32m long currently based on the resistance readings in the report, and if I extended there's no RCD to test at the moment. I've seen/read that come the 18th edition cables need to be adequately supported in the event of premature collapse, so I was thinking the easiest way is metal cable clips clipped ~50mm below the top of the joists. You'd have to be really cack handed to hit it at that depth...I suppose it could be run in metal conduit to provide actual mechanical protection but seems unnecessary to me - if you're nailing down floorboards you can see the joists and you can see how long the nails are - just pick ones short enough or drill and use screws (what I intend to do to keep things accessible).

Your dead right ericmark, pulled out my plug in tester and green is 1.8 Ohms or less. It's tempting to see if I can find a loop impedance tester at a reasonable price, as I said can't test the RCD anyway as there isn't one, and it'd be done on consumer unit change - I plan to have RCBO's as I've not got many circuits.

Presumably I could easily test volt drop with a multimeter on AC setting (being very careful of course as it'd be live)?

 

[AndyPRK]

If it was rewired in say 1965 the lighting circuits wouldn't have CPC unless you have metal flourscent tubes. and even then most of the lighting points would not have CPC.
As you have mentioned single core double insulated used. Nothing wrong with this as long as switches and lights are insulated (plastic) or double insulated metal
So I am rather surprised its failed N-E on the light circuit lol !

 

[plugwash]

It's 2020, so "37 years ago" would be 1983, well into the era when earths were mandatory.

 

[oldbutnotdead]

Big problem you've got is your proposed cables chased into the wall. Unless you use proper steel conduit (not capping) your new cables will be non-compliant cos no RCD on the circuit.
Cheapest option is of course to fire them in (in safe zones), part of CU replacement should be tests of insulation resistance, ring integrity & ELI so that'll cover the testing bit.
Better option is get the self-certifying spark who's doing the CU to look at your plan and do whatever they're happy with you doing prior to the swap (chases for definite, if you're lucky they'll be happy with you pulling cable through as well.
If you go for Plan A try and get one of those 'This installation has wiring colours to 2 versions of BS7671' stickers on the existing CU

 

[winston1]

I've seen/read that come the 18th edition cables need to be adequately supported in the event of premature collapse, so I was thinking the easiest way is metal cable clips clipped ~50mm below the top of the joists.

That does not apply to cables under the floorboards.

 

[ericmark]

RCD protection which works when the test button is pressed must be better than no RCD protection, maybe it will not trip in 40 mS, but in the main that does not matter, when my roof leaked, the RCD tripped, not really worried if it took one second, as long as it tripped.

Last house I fitted two RCD's and swapped the fuses for MCB's main consumer unit tested the loop impedance, second consumer unit never tested as it always tripped the RCD so would have needed to short out RCD to test it. And to be frank I was not worried that earth loop impedance not tested, at end of the day it didn't really matter.

You must accept you can't follow the BS7671 to the letter, and assess the risks.

I know excessive volt drop will cause fridge or freezer to stall and do multi re-tries which will burn out the over load, so if I hear the unit trying to restart, I would test, but unless some thing alerts me to a volt drop, I don't go around with a meter testing.

We learn from experience, last house I would get bouts of RCD tripping, never found a fault had 4 MCB's on each RCD, this house not making same mistake twice, so all RCBO consumer unit. I assessed the risk of sockets causing lights to trip was too high with just two RCD's so I have 14. But others may assess the risk different, you need to consider the risks and use BS7671 as a guide rather than a bible.

 

[JohnW2]

I know excessive volt drop will cause fridge or freezer to stall and do multi re-tries which will burn out the over load, so if I hear the unit trying to restart, I would test ...

You often say that, but I find it very difficult to understand unless you're unlucky enough to have a very low supply voltage.

In my house (which has a fairly 'typical' supply voltage, a bit over 240V), there would have to be a VD of 30V or more on the circuit before the voltage reaching the appliance fell to what could, theoretically, be the usual supply voltage in some other property - and I would have thought that any appliance sold in the UK would be expected to perform satisfactorily with any permitted UK supply voltage (and would not be considered 'fit for purpose' if it didn't).

Kind Regards, John

 

[JohnW2]

Cheapest option is of course to fire them in (in safe zones) ...

I've tried pretty hard, but can't really work out what this was intended to say

Kind Regards, John

 

[ericmark]

You often say that, but I find it very difficult to understand unless you're unlucky enough to have a very low supply voltage.

In my house (which has a fairly 'typical' supply voltage, a bit over 240V), there would have to be a VD of 30V or more on the circuit before the voltage reaching the appliance fell to what could, theoretically, be the usual supply voltage in some other property - and I would have thought that any appliance sold in the UK would be expected to perform satisfactorily with any permitted UK supply voltage (and would not be considered 'fit for purpose' if it didn't).

Kind Regards, John

I would agree, where I had the problem was Algeria, they had a 110 to neutral 3 phase supply, the Dutch electrician had connected the AC units phase to phase, and had assumed two phases on 110 volt = 220 volt however that is not the case, it is actually 190 volt. We has a succession of Dutch refrigeration items fail, at that time we had not harmonised so Dutch stuff was still 220 volt. So if the DNO volt drop and the premises volt drop were both on the lower limit, it could cause problems, and you will note nearly every fridge/freezer says don't use an extension lead.

But today even freezers and fridge/freezers are inverter powered with three phase motors at least mine are, so only cheap ones are likely to have problems with volt drop. But in this case I was actually considering the reverse, and was saying that measuring volt drop and any of the other limits as long as some common sense is used, and if you see some thing untoward you investigate, then in the main we can get away without it.

I do find the idea of not relying on a RCD seems odd, when we have to rely on the RCD with a TT supply, why should it be different with a TN supply? We know a type B MCB should trip between 3 and 5 times its rated valve with the magnetic part of the trip, so a B32 MCB needs between 96 to 160 amp to trip, simple ohms law 230/96 = 2.39 and 230/160 = 1.44 and allow 5% for volt drop so 1.36Ω is the recommended limit so the MCB will trip on the magnetic part of the device, as I said most the cheap plug in testers first light is over that limit, however they are below 2.39Ω so chances are the MCB will trip on magnetic part if the plug in tester with loop test passes the installation, and with RCD protection as well it is unlikely it would not trip with a fault to earth.

Yes it may not comply with regulations and with a rented property it would need the proper tests, but I assess the risk is low, and to say you should not DIY then may as well get rid of this whole site, it is right to tell people there is a risk, but they must be given the opportunity to assess that risk and decide if it is within acceptable limits.

 

[JohnW2]

I would agree, where I had the problem was Algeria, they had a 110 to neutral 3 phase supply, the Dutch electrician had connected the AC units phase to phase, and had assumed two phases on 110 volt = 220 volt however that is not the case, it is actually 190 volt.

That's obviously nothing to do with voltage drop - it's simply a matter of connecting equipment to the incorrect supply voltage.

So if the DNO volt drop and the premises volt drop were both on the lower limit, it could cause problems ...

Yes, of course - but I was saying that (a) that would be incredibly uncommon/unlikely and (b) I would not expect an appliance to be considered 'fit for purpose' for sale in the UK if it could not function satisfactorily with the lowest permissible UK supply voltage (probably minus the maximum recommended in-premises VD)

I do find the idea of not relying on a RCD seems odd, when we have to rely on the RCD with a TT supply, why should it be different with a TN supply?

That's an interesting question, and I suspect at least some of the answer is a matter of 'history' - in as much as when RCDs first appeared, they were (and, in most cases continue to be) regarded as providing 'secondary' protection (secondary to over-current devices) in TN installations.

In a TN installation, we do, of course, have to "rely on" the OPD (MCB or whatever) to provide adequate overload protection, and the only reason why that does not necessarily automatically also provide adequate fault protection is that CPCs commonly have a smaller CSA than the L & N conductors. If a circuit is wired in singles with the CPC having the same CSA as the other conductors, then adequate overload protection (which is always required) automatically means that fault protection provided by the MCB will also be adequate

We know a type B MCB should trip between 3 and 5 times its rated valve with the magnetic part of the trip, so a B32 MCB needs between 96 to 160 amp to trip, simple ohms law 230/96 = 2.39 and 230/160 = 1.44 and allow 5% for volt drop so 1.36Ω is the recommended limit ...

The "3 times" is obviously irrelevant, since one has to design on the basis of the worst case scenario - i.e. that it could take a current of 5In to result in a magnetic trip. What you describe as "allowing 5% for voltage drop" is what BS7671 now does with "Cmin", but that only relates to 'voltage drop' upstream of the installation (i.e. in the distribution network) causing the supply voltage at the origin of an installation being <230V - and, in any event, really ought to be 6%, rather than 5% (the UK supply voltage is allowed to be as low as 230V - 6%).

Yes it may not comply with regulations and with a rented property it would need the proper tests, but I assess the risk is low, and to say you should not DIY then may as well get rid of this whole site, it is right to tell people there is a risk, but they must be given the opportunity to assess that risk and decide if it is within acceptable limits.

I totally agree with that.

Kind Regards, John