Is a borrowed earth allowed?

[a-z electrics]

Done an EICR yday on an old house and all ok apart from 2 issues..
1) 32A cooker circuit had high resistance pec continuity (over2kohms) ALTHO it did buzz out on my voltage probes. Hence not able to get a Zs reading and no rcd tripping. The dual rcd consumer unit was fitted 10 yrs ago and those test readings all available and fine. Kitchen was refitted 3 yrs ago and no docs available so they extended original cooker cable prob from the switch which is now tiled over somewhere.
.. I've put in an earth wire link from an immediately adjacent rfc socket into the cooker switch and this solves this problem.
Question .. is this ok as a permanent solution if i label up what ive done?
2) lowish IR readings 1.75M L-E and N-E on light circuits and one rfc but possibly in hindsight due to some bulbs, fan and appliances still plugged in. Could this account for it?

 

[johnduffell]

I think

But you are assuming the break/fault is at the original switch, it could be elsewhere. You would have to assume it is not continuous from the CU. Therefore it is not complaint.

I think the confusion is that sparky chick considers that there has to be "a" cpc present throughout the wiring system, whereas my interpretation is that "the" cpc must be present throughout the wiring system.
Although i think everyone agrees it's not ideal!

 

[SparkyChick]

But you are assuming the break/fault is at the original switch, it could be elsewhere. You would have to assume it is not continuous from the CU. Therefore it is not complaint.

But it is continuous in that he can get a resistance reading end to end. It's just not a very good connection. So assuming it's a single break... both ends are earthed and the CPC will do it's job. Where it becomes a concern is if there are mulitple breaks... if there were two breaks, theoretically there could be a segment of the CPC that does not provide a sufficient good Zs to allow the protective devices to operate should a fault occur on that segment.

In essence, this is assumption... some of it can be proved (by inspection and testing) and some cannot (but we accept it because if we didn't some things would be a mare, like an EICR for example. We have to assume the CPC is continuous and running as expected in the same cable because we have no way of actually proving it.

This is just a theoretical discussion about a less than ideal situation. My personal preference would be to find and fix the fault or replace it. End of. But, I don't believe the solution breaks the regs, at least none that have been quoted so far.

Anyhow, blackcurrant jelly you say... Yum A bar for the big kids is even better... enjoy

 

[Intoelectrics]

my main concern is that, although the bypass cpc provides protection to the point of use, the original cable with it's damaged cpc is not, in itself, protected, so I lean towards the find and fix option.

I can see your point but I don't think this is the intended function of a cpc.
"Circuit Protective Conductor (cpc). A protective conductor connecting exposed-conductive-parts of equipment to the main earthing terminal".

So for example based on your point one could argue that a twin core flex has no protection because there is no cpc present.

Though what the OP has done is not "best practice" no reg has been broken and it has improved the safety of the circuit.
By which it could now be down coded from a C1 / C2 to CFI or C3 if the OP felt that further improvement was needed.

 

[buzzlightyear]

we can all give the good points and the bad, it is down to the op what is conscience tells him he is the one work on it .if was doing it in which I'm not .new feed if possible .

 

[johnduffell]

I can see your point but I don't think this is the intended function of a cpc.

Agree. The cpc in a flex or t&e is not an exposed conductive part it is just a conductor. In swa and metallic conduit the two functions do happen to be combined.
Indeed in my interpretation the regs would allow, if you somehow had an insulating joint in some conduit, you could indeed run a single cpc inside or alongside and use it to ensure the whole length was earthed.
The intent of the reg mentioned earlier is surely to make sure the cpc isn't removed by mistake when another circuit is decommissioned or abandoned?

 

[Lucien Nunes]

I agree that the earth borrowed from the RFC is probably quite adequate to provide the necessary protection to the cooker circuit. However I would not accept the practice as an actual solution specifically for two of the reasons already mentioned:

a) The original cooker cable is either damaged to an unknown extent, or connected in a defective manner. It is impossible to say what other danger exists from this problem, but borrowing the CPC conceals the original defect and is likely to prejudice its timely repair. This is a different scenario to borrowing a CPC for a circuit that was installed without one, where nothing is otherwise defective.

b) In a domestic situation, confusion is likely to arise from the non-standard configuration, especially as one of the circuits involved is an RFC, and erroneous future test results and inferences from them can reasonably be anticipated.

This latter point does not in my mind stand comparison to a metallic containment providing the CPC for multiple circuits, which is a normal and expected configuration, nor the deliberate design of a wiring system with one CPC cable serving multiple circuits. This occurs quite commonly in entertainment lighting, where cabling often runs in blocks of 6, 12, or 24 identical circuits to a single point of use. There's no point running 24 CPCs in parallel to the same box, one cable will do, but it is plainly obvious which CPC serves which circuits even without documentation and the circuits can be tested without unexpected side-effects from cross-connections.

By contrast, this ad-hoc arrangement is anything but obvious. Part of the cooker circuit cable is protected by the usual CPC from the CU, the other part terminating in an unknown location, and neither properly connected to, nor insulated from, the first part, is back-fed along with the cooker itself, from a CPC consisting of two legs of a functionally unrelated RFC.

All that said, I might be tempted to do this as a fix but I'd make it visible and ugly, so that the problem wouldn't get swept under a rug!

 

[johnduffell]

Part of the cooker circuit cable is protected by the usual CPC from the CU, the other part terminating in an unknown location

I agree with the general thrust of your post, but I'm intrigued about this need to protect t&e cable with a cpc. As i understand it the regs only require the earth at every termination, and cpcs in the same wiring system would not provide any protection unless the cpc was also an exposed conductive part eg SWA or metallic conduit.
I appreciate that nailing t&e has a chance of hitting the cpc but what protection is actually given by that?

 

[BruvDunk]

Im with the replace cable side !
Too many unknowns and guesses...

 

[Lucien Nunes]

Johnduffell you're right, I didn't word that very well. What I meant to say was that any accessory inserted into the cable, such as the putative bypassed isolator which may have a metal back box, would be protected by... rather than the T+E cable itself.

 

[a-z electrics]

I disagree, his post suggests the cable was extended at the original switch meaning there are two cable segments, one from the CU to the switch and another from the switch to the cooker. Both segments will be wired in T+E (valid assumption) meaning the CPC is contained within the wiring system (the cable, again... valid assumption).
.

Yes the cable from DB is red and black. At switch it's brown and blue.. I understand the original switch position is in the wall less than 3M from the DB. The new switch is about 2m as crow flies from DB, albeit thru a thick wall with lots of cables and pipes along it. I'd say it was a good chance the cpc join at the original switch isn't sound .. there is ecery infication the live cable joins are sound. The test readings match almost exactly those done 10 yrs ago after I did the cpc repair..

 

[a-z electrics]

Funnily enough, I made a similar mistake a couple months ago.. I put in a socket spur from a socket on the rfc which was being covered over.. after plasterer had been I retested and found there wad no earth at new socket! I had to cut out an inspection hole and found the new cpc had actually snapped off inside the terminal for some reason (possibly a hurried stripping operation).. so it DOES happen!

 

[Midwest]

You not inferring you plastered over a joint?

 

[a-z electrics]

You not inferring you plastered over a joint?

Not me.. that was the builder

 

[a-z electrics]

But it was MF..

 

[SparkyChick]

But it was MF..

Just out of interest... what sort of joint was it?

 

[Midwest]

Hope it wasn't one of these;

 

[Pete999]

But it was MF..

What type of connector was used?

 

[mrloy99]

There are other concerns too..maybe there could be another radial in the kitchen with a 'lost ' earth and the previous spark has already borrowed the cooker earth......So new cooker circuit in /old cooker circuit out might leave another circuit unprotected...Full test might identify that..I suppise...I dont really see a problem with just installing a new separate earth..Why does it need to be in the same cable.

 

[Midwest]

Why does it need to be in the same cable.

We were arguing that last weekend. You'll have to read the posts yourself.

 

[Deleted member 9648]

Funnily enough, I made a similar mistake a couple months ago.. I put in a socket spur from a socket on the rfc which was being covered over.. after plasterer had been I retested and found there wad no earth at new socket! I had to cut out an inspection hole and found the new cpc had actually snapped off inside the terminal for some reason (possibly a hurried stripping operation).. so it DOES happen!

GASP!

 

[a-z electrics]

Just out of interest... what sort of joint was it?

Push in Wagos .. but I have a feeling the cpc wire was nicked whilst stripping, hence snapped off whilst being pushed into the box..

 

[alasdairp]

Sorry this is late, I only usually look once a week when the summary comes around. I have carried quite a few 'serious' Grounding / Earthing related works in the past. In one case, in a very large shopping and business complex in the upper Midlands, with its own substation in the basement, I determined (and they eventually admitted after legal pressure) that they had failed !!!!! to reconnect the Earth to the whole building after they had tested and commissioned the substation! So much for the "professionals" in the DNO's.

Another case was a multiply-supplied (with originally 6 separate meters) industrial/scientific starter-unit complex (with a different DNO) that eventually had one pharmaceutical owner and had really high EMF magnetic fields in their labs (like 50 microteslas) from stray currents which were affecting the biological research results. I determined what the problem was. It was in the local distribution network and not their own supply cables. The only way to sort it was to remove all but one of the network protective Earths and rely on that and the buildings steel structures into the ground (like super Earth Rods). That sorted the problem and the DNO agreed with me that it was the only practical solution. So we did that BUT also affixed signed warning notices at all the incoming supply points to say what we had done in some detail and why.

I agree with SparkyChic that ideally the CPC should be of adequate size, but, if as you say, it is a dual RCD protected installation then all any CPC needs to carry will be 30 mA (or possibly 100 mA). This is not yet reflected in BS7571 - even in the latest 18th Edition.

I think, if you leave it that way, you should stick a notice near the CU saying that is what you have done. It should be safe, but it should not be un-noted. I would have done what you did.

 

[alasdairp]

Sorry this is late, I only usually look once a week when the summary comes around. I have carried quite a few 'serious' Grounding / Earthing related works in the past. In one case, in a very large shopping and business complex in the upper Midlands, with its own substation in the basement, I determined (and they eventually admitted after legal pressure) that they had failed !!!!! to reconnect the Earth to the whole building after they had tested and commissioned the substation! So much for the "professionals" in the DNO's.

Another case was a multiply-supplied (with originally 6 separate meters) industrial/scientific starter-unit complex (with a different DNO) that eventually had one pharmaceutical owner and had really high EMF magnetic fields in their labs (like 50 microteslas) from stray currents which were affecting the biological research results. I determined what the problem was. It was in the local distribution network and not their own supply cables. The only way to sort it was to remove all but one of the network protective Earths and rely on that and the buildings steel structures into the ground (like super Earth Rods). That sorted the problem and the DNO agreed with me that it was the only practical solution. So we did that BUT also affixed signed warning notices at all the incoming supply points to say what we had done in some detail and why.

I agree with SparkyChic that ideally the CPC should be of adequate size, but, if as you say, it is a dual RCD protected installation then all any CPC needs to carry will be 30 mA (or possibly 100 mA). This is not yet reflected in BS7571 - even in the latest 18th Edition.

I think, if you leave it that way, you should stick a notice near the CU saying that is what you have done. It should be safe, but it should not be un-noted. I would have done what you did.

The secod one was was due to PME problems in the area.

 

[SparkyChick]

Push in Wagos .. but I have a feeling the cpc wire was nicked whilst stripping, hence snapped off whilst being pushed into the box..

Were the Wago's in a Wagobox by any chance?

 

[buzzlightyear]

Just out of interest... what sort of joint was it?

a smoking one why

 

[gerry whitwham]

I would say that borrowed earths are not good practice ……..one alternative may have been to run a separate 4 0r 6mm cpc back to the consumer unit.

 

[Intoelectrics]

I would say that borrowed earths are not good practice ……..one alternative may have been to run a separate 4 0r 6mm cpc back to the consumer unit.

I don't think anybody is arguing that it is good practice. The op was about whether it was an acceptable alternative in the circumstances he/she faced. If you are going to the trouble of running a new single cpc back to the CU, I suspect in this case you may as well run in a new twin & earth.