missing lighting cpc

[gts860]

Good Evening,
I have failed an EIC test, C1 lighting circuit has induced voltage on light switch screws.
The lighting circuit is on a split load CU non RCD side only MCB fitted.
All switches and pendants have a connected R2 however the CPC conductor present does not connect to any Earth, its floating.
The break or disconnection is probably under the floorboards.
Is there any reason why one cannot take from a tested radial socket cct, on 20A RCD, which is 1.5M distant from one of the light pendants in this failed cct, run it through the loft and connect it to the pendant ? I am told one cannot do this and the electrician wants to drill holes through the walls, run an extra CPC from the CU, up a wall surface, within trunking and tap it into one of the light switches on this circuit.
He tells me my suggestion cannot be done "the cpc must come from the same cct" Surely the many and various cpc on the other circuits are all connected to the same Earthing point so i do not see the logic in his explanation
The entire house is rewired to 16th Edition

 

[i=p/u]

C3, if plastic fittings. Rcd it , take switch away and add microwave sensor fitting of some sort or replace cpc, but not from the socket

 

[gts860]

Hello i=,
they are plastic fittings, but because the cpc is physically present but not connected to a earth then it is subject to an induced >100v potential on the light switch screws,taking the switch away is not an option all light switches on this circuit have "live" fascia screws. What I would like to know is the reason why one cannot take from another cpc, surely its all the same point.
Which REG is contravened ?

 

[westward10]

There is nothing preventing an earth being shared by multiple circuits providing it is suitably selected. How long did they spend looking for the fault.

 

[i=p/u]

I'd need a look myself haha, if its only going to light switch, disconnect it at light maybe.

 

[Lucien Nunes]

C1 lighting circuit has induced voltage on light switch screws.

Surely that would be a C2 - it's merely a symptom of the missing CPC rather than a separate fault. I would give a C1 if the screws were live from a spurious connection of the CPC to line.

The break or disconnection is probably under the floorboards.

A concealed damaged cable or junction that has an open-circuit CPC could be defective in other ways too. E.g. It might be a cable with a screw through where the CPC has blown itself clear on one side but the screw is still live. I would not be content to supply a replacement CPC, from the board or another circuit, without attempting to locate and rectify, or bypass completely, the cable segment with the open-circuit CPC.

All switches and pendants have a connected R2 however the CPC conductor present does not connect to any Earth, its floating.

You say they have a resistance value that indicates they are connected, but then say they are not connected. Am I missing something?

 

[gts860]

There is nothing preventing an earth being shared by multiple circuits providing it is suitably selected. How long did they spend looking for the fault.

Hello Westward, i do not know i was not there. I suspect i am being given the run around if what i suggested is a viable and according to regs acceptable solution then it probably would not take more than 15 mins and 2M of eth cable. I have already pointed out to them that the ETH arrangements in the house are not what they wrote on the form, that the Ze value is too high, (incoorectly filled in form) a suspected C2 on a workshop with 20 sockets because open grommets have been used on the knock outs and more than 3mm gap is present. This is quite a time consuming job taking all the socket faces off disconnecting all the conductors, fitting closed grommets and putting it all back. When questioned they decided it was not a C2 but a C3 so easy fix for them different tick box.
Anyway, what do you mean by "suitably selected"

 

[gts860]

hello Lucien,
to clarify, all the components of the lighting circuit ETH points, switches and pendants are connected together. This is the R2 test performed, However this conductor is not actually connected to an earthing point, perhaps it has always been missing,

 

[westward10]

I think you need an electrician who can actually fault find to give you a clear opinion as to what is going on.

 

[gts860]

I think you need an electrician who can actually fault find to give you a clear opinion as to what is going on.

 

[gts860]

i thought i had of one those, he has all the certificates but clearly non of the answers

 

[westward10]

I would ask where the fault is through the circuit any competent electrician would be able to tell you after some fundamental testing.

 

[Lucien Nunes]

R2 is specifically the resistance of the CPC from the furthest point back to the CU. But if your CPC is not connected to the CU, you cannot have a valid result for R2.

If the CPCs of all points (bar none) are interconnected, but not connected to the CU, then the break must be in the 'home run' of the cable. If that cable is accessible at a point so that it can be disconnected, my suggestion would be to run not just a CPC but a complete new cable to the CU, and thereby isolate the faulty section completely.

 

[westward10]

R2 is specifically the resistance of the CPC from the furthest point back to the CU. But if your CPC is not connected to the CU, you cannot have a valid result for R2.

If the CPCs of all points (bar none) are interconnected, but not connected to the CU, then the break must be in the 'home run' of the cable. If that cable is accessible at a point so that it can be disconnected, my suggestion would be to run not just a CPC but a complete new cable to the CU, and thereby isolate the faulty section completely.

Problem is the electricians description is a bit vague. I believe they are saying there is a cpc at all accessories but an earth fault loop test indicates 100v and subsequently will not perform the test but I could be wrong. Nine times out of ten this is an incorrectly connected cpc.

 

[gts860]

R2 is specifically the resistance of the CPC from the furthest point back to the CU. But if your CPC is not connected to the CU, you cannot have a valid result for R2.

If the CPCs of all points (bar none) are interconnected, but not connected to the CU, then the break must be in the 'home run' of the cable. If that cable is accessible at a point so that it can be disconnected, my suggestion would be to run not just a CPC but a complete new cable to the CU, and thereby isolate the faulty section completely.

Sorry but i have mislead you here by using the terminology supplied by the tester. What is more accurate to say is that all of the earthing points of the lighting circuit components are connected to one-another with a 1.0mm eth cable (not an R2) This (at this moment) is a completely isolated run of cable and is not connected to a central earthing point either at or connected to the CU earth block. So I see why you were confused.
This is why i suggested taking a short spur from a known good CPC adjacent to one of the light pendants and effectively connecting a verifiable earth to the failing light circuit. This I am told is not possible however his solution is effectively the same fix, only he takes the earth not from a remote socket box but from the CU earth via a run of trunking up a wall to pick up on the same circuit, in a light switch.

 

[gts860]

Problem is the electricians description is a bit vague. I believe they are saying there is a cpc at all accessories but an earth fault loop test indicates 100v and subsequently will not perform the test but I could be wrong. Nine times out of ten this is an incorrectly connected cpc.

actually the description is simply that 160V is present on the fascia screws. I think i will be looking to get a second opinion and for him to quote which Regulation does not permit taking a cpc from one circuit and using it for another circuit. If it is of any value the L/E insulation test on this cct passes

 

[westward10]

Has he verified the fault is from the consumer unit to the first light.

 

[gts860]

, my suggestion would be to run not just a CPC but a complete new cable to the CU, and thereby isolate the faulty section completely.

this however is exactly what i want to avoid if at all possible. All cables are either under floorboards or buried and i would prefer not to have trunking running over the surfaces.

 

[gts860]

Has he verified the fault is from the consumer unit to the first light.

i personally do not think he has verified anything, i will ask him to detail what fault finding steps he has applied and what the results where, but based on his performance to date i have little confidence in his thoroughness. Incidently the cct in question incorporates a 3 way hall /landing arrangement and a two way landing/loft lighting switch arrangement.

 

[i=p/u]

Yh , let him do.his job and trust him or get someone.else..

 

[Lucien Nunes]

this however is exactly what i want to avoid if at all possible

You would prefer to have a defective cable / joint remain live and in service without knowing the nature or extent of the defect?

 

[gts860]

I have asked the electrician the following questions
Which light pendant is the first in the run (in which room) ?
have you disconnected the cpc from this pendant and at the CU and measured for continuity, what was the result ?
which wiring REG will be contravened if an earth is taken from a different circuit ?

to the first 2 there is no answer, to the 3rd there is no ref to a wiring reg, but the explanation is that if he takes a connection from the socket then somebody could in the future remove those sockets and in doing so you lose the cpc. Which of course is true, but that argument could equally apply to any other cpc if somebody decided to "remove" it.

He should be able to determine which rose is the first in the circuit, but to do this i imagine you must disconnect each rose in turn until you can measure continuity back to the mcb. If he does this and measures the associated cpc on this first run rose and it is open cct then obviously either the conductor is broken somewhere under floorboards or maybe loose at an underfloor junction box. If that is the case then either one lifts the floorboards at several places in the house to check where the cpc is lost or perhaps one can borrow another cpc. I would agree that taking the cpc from elsewhere is not the most elegant solution but lifting carpets and floorboards is a very expensive hammer to crack a nut.

in order to counter his argument about somebody removing those sockets I suggested that the incoming supply to the loft radial sockets could be cut (in between the loft joists),a junction box introduced, preferably under a loft board, the radial cct is reconnected and from the junction box run a separate cpc to either the light pendants in the loft or the light pendants for any of the 1st floor which are accessible from the loft.

 

[i=p/u]

Theres defo a solution , as you say wheres the break??

 

[westward10]

I would be using using the earth from another circuit as a last solution. Someone needs to establish where the problem is as nine times out of ten it can be rectified. I suspect you are wasting your time with this electrician.

 

[i=p/u]

Who knows where it is , il go check my lights too

 

[gts860]

You would prefer to have a defective cable / joint remain live and in service without knowing the nature or extent of the defect?

no that is not what i was suggesting, as a C1 failure it has to be resolved, my preference is for a more elegant solution than his suggestion, tacking in an extra wire via external trunking. If a cpc is derived from another cable then the induced voltage collapses and the circuit is no longer live and is safe, is it not ?
I was hoping that somebody could explain to me the electrical theory behind the assertion that one cannot take a cpc for one circuit from another circuit, all the good tested cpc's return to exactly the same point, i would think the functionality of the cpc protection is retained irrespective of the route it takes.

 

[gts860]

I would be using using the earth from another circuit as a last solution. Someone needs to establish where the problem is as nine times out of ten it can be rectified. I suspect you are wasting your time with this electrician.

the problem for me is how many qualified people do I need to employ until I find one who is capable of faultfinding. If the conductor can be shown through measurement to be o/c in the leg from the consumer unit, and this cable is largely inaccessible then there are only two options open 1a) lift all the carpets, floor boards, buried trunking and run a new cable to the first ceiling rose in the loft 1b) run a new cable in external trunking to the first ceiling rose in the loft 2) take a cpc from somewhere adjacent to the first ceiling rose in the loft. What would you suggest to your customer if this was the 1 time out of 10 where it could not be readily rectified ?

 

[i=p/u]

if it's good enough for pefc it's good enough and label accordingly.. but wouldn't really be done in my eyes. It would be found and fixed. You could disconnect circuit from board and supply from a spur off another circuit.. and see where that leaves you.

 

[Lucien Nunes]

no that is not what i was suggesting, as a C1 failure it has to be resolved, my preference is for a more elegant solution than his suggestion, tacking in an extra wire via external trunking. If a cpc is derived from another cable then the induced voltage collapses and the circuit is no longer live and is safe, is it not ?

No, not necessarily. The CPC of a cable does not suddenly go open-circuit by itself. Something is significantly wrong with that cable run, but you don't know what. You know only the symptom that the EICR has thrown up, i.e. open-circuit CPC, and you propose to alleviate that symptom by providing an alternative CPC. Anything else that is wrong will remain live and in service.

As I said back in post #6, the cable might have been partially severed by a screw or nail which is now live. Or it might contain a thoroughly defective joint made by an unskilled person using sellotape, of which one of the three cores has so far given way. Or it might have been ripped to shreds by catching on a Skil saw when somebody was cutting floorboards, and now the bare length of line conductor is resting against a non-extraneous gas pipe. You don't know whether any of these apply to you, only that the broken CPC has alerted you to a problem with the cable. Do you intend to ignore any other ramifications once your CPC is reconnected to the MET?

 

[gts860]

No, not necessarily. The CPC of a cable does not suddenly go open-circuit by itself. Something is significantly wrong with that cable run, but you don't know what. You know only the symptom that the EICR has thrown up, i.e. open-circuit CPC, and you propose to alleviate that symptom by providing an alternative CPC. Anything else that is wrong will remain live and in service.

As I said back in post #6, the cable might have been partially severed by a screw or nail which is now live. Or it might contain a thoroughly defective joint made by an unskilled person using sellotape, of which one of the three cores has so far given way. Or it might have been ripped to shreds by catching on a Skil saw when somebody was cutting floorboards, and now the bare length of line conductor is resting against a non-extraneous gas pipe. You don't know whether any of these apply to you, only that the broken CPC has alerted you to a problem with the cable. Do you intend to ignore any other ramifications once your CPC is reconnected to the MET?

the proposal to use another cpc is not mine it is the suggestion of the author of the fail certificate, I am simply suggesting a less obtrusive version of his suggestion. If the conductor has been partially severed and is now live would an insulation test L/E indicate >500Meg ? The electrician has confirmed that the voltage present on the fascia screws is induced, if it really was line voltage then would not the connection of an cpc from elsewhere to this potentially partially severed cpc cause the MCB to trip ?
The gas and water/CH bonding has passed

 

[loz2754]

On the question of which regulation states that you cannot use a CPC from one circuit to add a CPC to a different circuit, you will not find any such regulation. On the contrary 543.1.2 and onward shows the requirements for where two or more circuits share the same CPC. So in some circumstances it might be perfectly acceptable.

However, as Lucien and others have pointed out, this would not be an appropriate solution to the "missing" CPC in your situation. You would only be masking the symptoms, not correcting the fault.

A bit like taking painkillers to treat a bleeding artery. You need to find the problem and repair it.

 

[Lucien Nunes]

If the conductor has been partially severed and is now live would an insulation test L/E indicate >500Meg

The CPC is completely severed. I'm not so worried what that might or might not be in contact with, as one side is already earthed and the other side is going to get earthed. It's the other bits that might be involved. What was the insulation reading from line to shelf bracket?How about line to shower-curtain rail? (we had one of those some years back, the rail was solidly live from where a fixing screw had penetrated a cable.)

I won't bother trying to convince you any further.

 

[gts860]

On the question of which regulation states that you cannot use a CPC from one circuit to add a CPC to a different circuit, you will not find any such regulation. On the contrary 543.1.2 and onward shows the requirements for where two or more circuits share the same CPC. So in some circumstances it might be perfectly acceptable.

However, as Lucien and others have pointed out, this would not be an appropriate solution to the "missing" CPC in your situation. You would only be masking the symptoms, not correcting the fault.

A bit like taking painkillers to treat a bleeding artery. You need to find the problem and repair it.

Thanks for that reference

 

[gts860]

The CPC is completely severed. I'm not so worried what that might or might not be in contact with, as one side is already earthed and the other side is going to get earthed. It's the other bits that might be involved. What was the insulation reading from line to shelf bracket?How about line to shower-curtain rail? (we had one of those some years back, the rail was solidly live from where a fixing screw had penetrated a cable.)

I won't bother trying to convince you any further.

I appreciate your engagement. Thanks.

 

[Lucien Nunes]

Having pressed the point about not leaving a damaged joint / cable in service, perhaps I should give my approach to locating one.

1. Walk the course. Is there any evidence of work carried out since the installation in areas where the cable is likely to be? Sometimes it's a no-brainer.

2. Inject tone into cable and search with tracer along likely routes. If it can be found at least at a few waypoints, estimate total length along that route.

3. Ping cable from both ends (CU and first lighting point) with a TDR or capacitance meter to discover distance to the break. Compare sum of lengths with measured route. Take average, measure along assumed route, look carefully around epicentre for likely causes of damage.

4. If possible, make access hole behind some other fitting or fixture that will conceal it later. Insert borescope. If nothing interesting visible, return to step 2.

etc.

 

[Mike Johnson]

2. Inject tone into cable and search with tracer along likely routes. If it can be found at least at a few waypoints, estimate total length along that route.

Lucian what do you use to Inject a tone into the cable?

 

[Lucien Nunes]

Three choices:

The normal AF warbler, ideal for close-up work, finding a core or pair in a cable, following it around a panel etc, but range limited to about 12".

801K | Tempo 200FP Tone Generator, 2 Tone, 1100Hz Tone Frequency | RS

uk.rs-online.com

Then there's one of these around the place somewhere . Ironically, when it's lost, you can't locate it by waving a piece of cable around.

RS PRO Cable Tracer Kit, Cable Detection Depth 2m CAT III 300 V, Maximum Safe Working Voltage 400V | RS

uk.rs-online.com

Many moons ago I made one myself out of the guts of a couple of transistor radios. Depending on how one were to do that, one might end up with a device that radiates on bands where one is supposed to keep quiet. One would need to use discretion as it might have significant range.

 

[Mike Johnson]

Been looking for one of these for a while: Socket and See FFCB100UKA Fuse Finder | ISSWWW.CO.UK - https://isswww.co.uk/socket-and-see-ffcb100uka-fuse-finder?gclid=Cj0KCQjwutaCBhDfARIsAJHWnHtzziofqm8WHz1L4ITtnx4ylS8TlwtITySmtjAXp1RnOctjGdMYPWgaAnNTEALw_wcB ideal for me as it has a removable cable so can be configured for UK and EU via what looks like a standard appliance plug.

 

[timhoward]

All depends on the accuracy required ! I once used a bit of wire wrapped around a lawnmowers HT lead and and attached to a cable conductor , and an AM radio to track a cable run underground. (Not that I’m suggesting that for a domestic lighting circuit of course!)

 

[magnoliafan89]

Personally I think the OP was only looking for someone to say it was OK to take a CPC from somewhere else , not wanting to entertain any other sound advice and now we won't hear from him now someone has said it is OK.

 

[Mike Johnson]

Thread drift does not annoy me, but obviously does others.

 

[gts860]

Personally I think the OP was only looking for someone to say it was OK to take a CPC from somewhere else , not wanting to entertain any other sound advice and now we won't hear from him now someone has said it is OK.

Sorry but I am back to haunt you !
No I did not want to just get an OK, I want to really know the theory behind why such an arrangement is against the regulations. This topic has also appeared on other forums like the IET and it appears to be an issue that some are absolutely against on principle and others see no problem, with the caveat that CSA is considered and appropriately selected and that the work is documented in the system.
As a non electrician i cannot see why the route a fail current takes to get to earth is important the main thing is that it gets there. I appreciate that someone inspecting the system at a later date might assume that the cpc bound to a particular cable does indeed go to the same place as the line and neutral.
At the end of the day the electrician tasked to resolve this issue has to put his name to it and if he is not prepared to do that then I respect this. As another responder here noted it seems that my guy has not actually done any troubleshooting, and that is the main problem here. If you cannot identify where that cable is physically connected then the status of its associated cpc is simply unknown, it could just be loose in the first ceiling rose. If it is identified and the cpc is then measured and found to be open cct then you know what the problem is. If that cable is routed under floors in trunking through walls etc then the question is whether it is reasonable to raise those boards, remove buried trunking out of walls in order to run a new cable or to use an existing cpc

 

[gts860]

Having pressed the point about not leaving a damaged joint / cable in service, perhaps I should give my approach to locating one.

1. Walk the course. Is there any evidence of work carried out since the installation in areas where the cable is likely to be? Sometimes it's a no-brainer.

2. Inject tone into cable and search with tracer along likely routes. If it can be found at least at a few waypoints, estimate total length along that route.

3. Ping cable from both ends (CU and first lighting point) with a TDR or capacitance meter to discover distance to the break. Compare sum of lengths with measured route. Take average, measure along assumed route, look carefully around epicentre for likely causes of damage.

4. If possible, make access hole behind some other fitting or fixture that will conceal it later. Insert borescope. If nothing interesting visible, return to step 2.

etc.

Thanks for the suggestions. That tone generator looks like its a useful bit of kit, what kind of range does it have ?
With only six lighting points in the upstairs circuit it surely cannot be so onerous to unscrew the pendants ,access each ceiling rose, disconnect all conductors und with your wanderlead (or tone generator)measure from the CU to each rose in turn until you can ping it. Once you have located the L/N you have by definition also located the cpc.

 

[Lucien Nunes]

If you cannot identify where that cable is physically connected then the status of its associated cpc is simply unknown, it could just be loose in the first ceiling rose.

Perhaps I had jumped to this conclusion but I thought the two ends of the supply cable had been identified at the board and first point, and that inspection/testing of the ends confirmed that the CPC was O/C somewhere unknown and inaccessible between them.

If, as you imply here, the first point has not been inspected and tested then this is simply daft and he is wasting your time. As you say, there would be a high probability of the disconnect being within the rose.

 

[westward10]

Nine times out of ten it is.