Consumer unit change - How do you do it?

[Dartlec]

See Regulation 643.3.1

Interesting - is that there to increase safety during the test and avoid 500V appearing on random metalwork for any length of time? Or is it to detect and test any parallel paths?

The CPC of the circuit being tested is probably isolated from all the others CPCs and the Earth, so you would not see a leak from N to true Earth (e.g. a screw, etc, that is touching the N copper and in to plasterboard, etc) as it would not be guaranteed to complete the circuit from true Earth to CPC and back to your IR tester.
[automerge]1601722401[/automerge]
Basic idea is this (sorry for poor diagram):
View attachment 61176
[automerge]1601722565[/automerge]
So IR testing with the CPC isolated will pick up on an internal cable fault (say N to E), but not for external faults (N to some other CPC or true Earth).

Interesting, I'd never really appreciated that. Though a screw piercing only N through plasterboard, would not be detected in any case would it? What would the viable parth to earth be for a screw, assuming it wasn't into a metal stud or in very wet plaster at time of test?

I guess it would pick up where the neutral for one circuit was damaged where it passed through the backbox of a different circuit or where rodent damage allowed the exposed neutral of a lighting circuit to touch cpc of a power circuit running alongside it in the loft.

I'll admit I've sometimes done the IR after first fix, mostly to check that plasterers haven't damaged a cable, and then not rechecked once connected.

On EICRS I usually do them with the cpcs connected, if only because it is sometimes hard to be sure which one is which anyway in a poorly installed board. (Though RFCs have to be located anyway of course)
[automerge]1601724182[/automerge]
Just to add, I'm fairly sure on previous NICEIC inspections when they check my testing procedure (Usually a RFC continuity test, R1+R2 and then IR) they have never mentioned replacing the cpc before testing.

Obviously it's more to see the procedure than to actually do the test at that point, but maybe I'll get brownie points for ensuring I do at the next one - or he'll ask what I'm doing that for!

 

[stevethesparks]

R1+R2 cpc disconnected, Insulation testing CPC connected

 

[Dartlec]

R1+R2 cpc disconnected, Insulation testing CPC connected

Yep, but fairly sure I've not reconnected the CPC before doing IR tests on my last few inspections and it's not been mentioned.

 

[westward10]

If it wasn't picked up then the inspector either weren't doing their job or are unaware of the requirements. The test schedule records a Live-earth test not a Live-cpc test.

 

[telectrix]

problem is, with cpc's all connected to MET, I've several times got IR readings showing extremely low values, even when there are no faults.

 

[stevethesparks]

problem is, with cpc's all connected to MET, I've several times got IR readings showing extremely low values, even when there are no faults.

Then surely the fault more than likely is between LN to elsewhere rather than the actual cpc

 

[pc1966]

Though a screw piercing only N through plasterboard, would not be detected in any case would it? What would the viable path to earth be for a screw, assuming it wasn't into a metal stud or in very wet plaster at time of test?

Simply in to clean dry plasterboard you might not see much, if anything, but I am always reminded of the rather sad case that made the news years ago. This was the fatal accident enquiry summery posted on one of the IET forums:

The case relates to the Death of 22 year old Emma Shaw who was the occupant with her 18 child month of a rented 1st floor flat in West Bromwich in December 2007.

The flat was one of new 42 flats that had electrical installation carried out by Anchor Building and Electrical services an NICEIC Approved Contractor in 2006.

The internal walls were constructed of "C" section metal studwork covered with plasterboard. A hall cupboard contained a pressurised water boiler above which was mounted a consumer unit supplied from a distribution circuit from a switch fuse in a an riser cupboard external to the flat. The means of earthing was TN-C-S. None of the circuits in the consumer unit were RCD protected. Circuit No. 3 supplied one of the immersion heaters in the boiler.

The electrical installation had been 1st fixed the plaster board fitted to the metal studding before the installation was 2nd fixed. Circuit 3 supplied a 20A double pole switch at high level which supplied a cable outlet at low level, this in turn supplied the water heater in flex.

The cable to the flex outlet was longer than it should be and when the plasterboard was installed the cable was trapped between the plasterboard and the stud-work. In addition a plasterboard screw passed through the cable clipping the Line and CPC conductors and in to the studwork.

The installation had been energised without any testing and the fault current had blown away part of the screw and CPC. This was verified by forensic examination at the HSE laboratories. This left a high resistance fault from the line conductor to the metal studwork.

18 months later the stat. on the water heater failed and the cylinder over pressurised and the safety valve operated and the safety valve operated and discharged in to a tun dish. The plastic waste pipe from the tun dish had parted because it was not glued and the water spilled on to the floor soaking the carpet. The water soaked under the foot of the wall and in to contact with the live studwork.

Miss Shaw was mopping up the water and had texted her partner to come home telling him the hall was flooded and the "electricity was sparking". he partner tetxed back to say to turn off the water stop cock sited in the same cupboard as the boiler.

Later she was found kneeling in the cupboard slumped forward and apparently lifeless. Subsequent examination and a Post Mortum determined that she was kneeling in the charged water and received a fatal shock when she touched the earthed stop valve.

West Midlands Police, the ambulance service and the fire service attended the scene. The police commenced an investigation for a suspicious death.

The police later arrested both the QS Hoult and Tomkins the electricians mate and they were interviewed under caution. Tomkins had signed an Electrical Installation Certificate as the Inspector. He admitted he was not qualified or competent to do so. He said he had been taken to the flat by the electrical site foreman and told to do ring continuity and loop impedance testing.He said when he got to the flat the installation was already energised. He did no other inspection and testing. He said that in the site hut they had sat around the table with the site foreman and was told the other tests had been done and he was told what to write on the test certificate.

This EIC was submitted to Anchors offices where a type written certificate was produced. The type written version had differences to Tomkins hand written version. The typed version had "P/P C Tomkins" in the single signature box for design. construction and inspection. Tomkins said in court he had not seen the typed form and he would not have consented to his signature being used. Mr Hoult the QS said that he just checked the figures on the form and signed it as the QS. he did not go to site and was not involved in the testing and inspection. Mr Hoult had known Tomkins for many years and it was alleged he knew about Tomkins status as a mate and his lack of qualifications.

Those are the brief facts that will no doubt generate a lot of questions. I have some of the detail from the expert witnesses, the police and others. The HSE produced an excellent DVD of a computer generated simulation of the accident shown in court to explain the circumstances and the consequences of the events in non-technical language.

 

[Dartlec]

Simply in to clean dry plasterboard you might not see much, if anything, but I am always reminded of the rather sad case that made the news years ago. This was the fatal accident enquiry summery posted on one of the IET forums:

The case relates to the Death of 22 year old Emma Shaw who was the occupant with her 18 child month of a rented 1st floor flat in West Bromwich in December 2007.

The flat was one of new 42 flats that had electrical installation carried out by Anchor Building and Electrical services an NICEIC Approved Contractor in 2006.

The internal walls were constructed of "C" section metal studwork covered with plasterboard. A hall cupboard contained a pressurised water boiler above which was mounted a consumer unit supplied from a distribution circuit from a switch fuse in a an riser cupboard external to the flat. The means of earthing was TN-C-S. None of the circuits in the consumer unit were RCD protected. Circuit No. 3 supplied one of the immersion heaters in the boiler.

The electrical installation had been 1st fixed the plaster board fitted to the metal studding before the installation was 2nd fixed. Circuit 3 supplied a 20A double pole switch at high level which supplied a cable outlet at low level, this in turn supplied the water heater in flex.

The cable to the flex outlet was longer than it should be and when the plasterboard was installed the cable was trapped between the plasterboard and the stud-work. In addition a plasterboard screw passed through the cable clipping the Line and CPC conductors and in to the studwork.

The installation had been energised without any testing and the fault current had blown away part of the screw and CPC. This was verified by forensic examination at the HSE laboratories. This left a high resistance fault from the line conductor to the metal studwork.

18 months later the stat. on the water heater failed and the cylinder over pressurised and the safety valve operated and the safety valve operated and discharged in to a tun dish. The plastic waste pipe from the tun dish had parted because it was not glued and the water spilled on to the floor soaking the carpet. The water soaked under the foot of the wall and in to contact with the live studwork.

Miss Shaw was mopping up the water and had texted her partner to come home telling him the hall was flooded and the "electricity was sparking". he partner tetxed back to say to turn off the water stop cock sited in the same cupboard as the boiler.

Later she was found kneeling in the cupboard slumped forward and apparently lifeless. Subsequent examination and a Post Mortum determined that she was kneeling in the charged water and received a fatal shock when she touched the earthed stop valve.

West Midlands Police, the ambulance service and the fire service attended the scene. The police commenced an investigation for a suspicious death.

The police later arrested both the QS Hoult and Tomkins the electricians mate and they were interviewed under caution. Tomkins had signed an Electrical Installation Certificate as the Inspector. He admitted he was not qualified or competent to do so. He said he had been taken to the flat by the electrical site foreman and told to do ring continuity and loop impedance testing.He said when he got to the flat the installation was already energised. He did no other inspection and testing. He said that in the site hut they had sat around the table with the site foreman and was told the other tests had been done and he was told what to write on the test certificate.

This EIC was submitted to Anchors offices where a type written certificate was produced. The type written version had differences to Tomkins hand written version. The typed version had "P/P C Tomkins" in the single signature box for design. construction and inspection. Tomkins said in court he had not seen the typed form and he would not have consented to his signature being used. Mr Hoult the QS said that he just checked the figures on the form and signed it as the QS. he did not go to site and was not involved in the testing and inspection. Mr Hoult had known Tomkins for many years and it was alleged he knew about Tomkins status as a mate and his lack of qualifications.

Those are the brief facts that will no doubt generate a lot of questions. I have some of the detail from the expert witnesses, the police and others. The HSE produced an excellent DVD of a computer generated simulation of the accident shown in court to explain the circumstances and the consequences of the events in non-technical language.

Definitely a tragic case and should make anyone think, though seems to be a whole series of incompetencies that caused it from plumbing to install. And if I recall they did no real testing at all before signing the certificate which was the real crime

I've worked (post build) in several flats like this so can well believe the sequence of events though.

I dislike metal studs personally, and cost seems to be the only benefit. (Not that it results in cheaper properties, just more profit for the builder).

Were they able to determine if the high resistance fault to metal stud would have been picked up by an IR test to MET? Metal studwork isn't normally bonded is it? Though it may be bonded to some extent by contact with the concrete floor...

(working) RCDs on each circuit will hopefully minimise the risk of it happening again.

 

[pc1966]

Definitely a tragic case and should make anyone think, though seems to be a whole series of incompetencies that caused it from plumbing to install.

I suspect most serious accidents are a result of many failings.

And if I recall they did no real testing at all before signing the certificate which was the real crime

That is why there was a criminal prosecution, for failing in their duty to test to ensure safety and falsifying the records saying it was tested.

I've worked in several flats like this so can well believe the sequence of events though.

I dislike metal studs personally, and cost seems to be the only benefit. (Not that it results in cheaper properties, just more profit for the builder).

Thankfully my job spares me most of that, but it is a worry generally to see how shoddy many new properties are done, even at high prices.

Were they able to determine if the high resistance fault to metal stud would have been picked up by an IR test to MET? Metal studwork isn't normally bonded is it? Though it may be bonded to some extent by contact with the concrete floor...

(working) RCDs on each circuit will hopefully minimise the risk of it happening again.

I don't know.

In this case the screw shorted L-E as well but CPC was blown by the fault "bang test", but more generally I suspect an IR test would have shown 10s of M ohm which is a "pass" but always worries me as good clean accessories, even with decades old PVC T&E, always show 1G ohm or more when I have tested them.

I think this, and possible a few other similar incidents, was the push to have the regs to specify RCD protection on most domestic circuits as a functioned RCD here would have saved her.

 

[Lucien Nunes]

What would the viable parth to earth be for a screw, assuming it wasn't into a metal stud or in very wet plaster at time of test?

The most hazardous situation is, like the above, where the conductor with breached insulation is in contact with some other conductive item or structure (e.g. studwork, non-earthed pipework, exterior masonry) that could present a shock risk. The bigger and more widely spread that conductive material the more likely that it will have a low enough resistance to true earth and the MET, somehow, to show up as less-than-ideal IR. It might not at that point reveal the extent of the hazard, but just trigger a further investigation into something that doesn't seem right.

Looking at it another way, of all the things that a live conductor could possibly be faulted to, everything but the CPC is excluded from the test if the CPC is disconnected from the MET (unless there is another connection somewhere that leaves the CPC earthed anyway). Something in our sphere of work that drives home the significance of this 'conductor-to-rest-of-world' test was the plug-patch for theatre lighting, which works a bit like a manual telephone exchange with perhaps 48 plugs choosing from 24 sockets. The circuits running from the plug cords to the remote sockets are like installed extension leads, leading to the unusual situation that any plug that is pulled out and sat in its rack, has exposed pins connected to de-energised conductors that share a wiring system with energised conductors. If each circuit were tested separately, a fault between say circuit 30 line and 31 line would not show up, but the exposed pin of plug 31 would be fully energised when circuit 30 was in use. Therefore line and neutral pins of all plugs had to be earthed at once using a box with 48 sockets with all pins earthed, then one at a time each plug would be withdrawn from the earthing box and tested against all the other plugs L, N & CPC (& true earth) simultaneously.

In a domestic installation the equivalent would be a special box with hundreds of flyleads ending in magnets, that could be stuck onto every screw in the kitchen to provide a return path for the IR test. You saw it here first!

 

[ChrisElectrical88]

That story’s awful. I take it that case had a massive baring on use of RCD protection.

 

[Dartlec]

In a domestic installation the equivalent would be a special box with hundreds of flyleads ending in magnets, that could be stuck onto every screw in the kitchen to provide a return path for the IR test. You saw it here first!

Don't give Megger ideas for the 1900 series!

Thinking aloud, would the use of a wander lead to do IR to any dodgy looking screw that caused concern show up all possible faults like this? I've never seen the use of a wander lead for IR testing, but only for R2 testing

It sounds like in the Emma Shaw case, the screw was probably plastered over and not accessible , and neither was the metal studwork - it was the pool of water that allowed the connection - so there were no potentially 'extraneous conductive' parts to test to.

 

[Pete999]

Dont change many consumer units, usually go ok but had one today that I did cheap for next door neighbour (Why is it always the ones you do cheap for friends and family that bite you). Anyway changed the board etc, all ok. Then came testing and obviously have various alterations in the last 40yrs, still had original wylex rewireable fuses.

I found Neutral ring missing - A socket had a neutral out.
Conservatory had two sockets, a light and an automatic window opener all spured off a spur in the kitchen with no grommets in the back box (Apparently done by an electrician)
One socket of two gang no neutral - just change face plate
One socket no neutral (Spur) - Just had to wago and put blank plate as obviously a junction box under the floor
There was some others as well.

Anyway in my haste to speed up testing I did insulation resistance with a couple of sockets off (I know stupid), when powered on RCD kept tripping if I removed socket neutrals it was fine. Finally tracked it down to a really tight socket in the kitchen that had a damaged neutral cable, managed to sort everything.

Started at 9.30, was running around in the dark finished about 8pm with everything ok, im totally shattered.

I am thinking of insisting on a full EICR before I do a consumer unit change again, at least you know what you are getting into mostly and you dont spend an entire evening trying to chase down issues as once you have changed the board and especially adding RCD's your kinda committed.

Just wondering do others do EICR beforehand or just wing it?

If you just wing it how often do you have issues?

I think it's safe to say an EICR should be recommended prior to a CU change, at the least testing should be carried out to prove tripping of the RECD does no occur on completion of the CU change, this will or should avoid any confrontation when the CU change is finished and the RCD trips, and you have to explain why everything has not gone as expected.
[automerge]1601735974[/automerge]

9.00 am. arrive on site.
9.30am. quick smoke and get tools off van.
10.00am. start stripping old CU.
12.00. break for 2 hour lunch at nearest pub.
14.30. screw CU on wall and start bunging in cables.
16.00. break for smoke and swig of JD.
17.00.bung lid on and throw main switch.
17.30. OK it works, pass go and collect £400.
18.00.park up at pub to spend some of the ill gotton gains.

you're right. a full day.

C

 

[telectrix]

I think it's safe to say an EICR should be recommended prior to a CU change, at the least testing should be carried out to prove tripping of the RECD does no occur on completion of the CU change, this will or should avoid any confrontation when the CU change is finished and the RCD trips, and you have to explain why everything has not gone as expected.
[automerge]1601735974[/automerge]

C

agree 100%. pete, you seen my post on the coronaviruskeeping it light. john bishop.

 

[Pete999]

9.00 am. arrive on site.
9.30am. quick smoke and get tools off van.
10.00am. start stripping old CU.
12.00. break for 2 hour lunch at nearest pub.
14.30. screw CU on wall and start bunging in cables.
16.00. break for smoke and swig of JD.
17.00.bung lid on and throw main switch.
17.30. OK it works, pass go and collect £400.
18.00.park up at pub to spend some of the ill gotton gains.

you're right. a full day.

Couldn't work like that, I would need to ensure everything is OK before going on the wee wee at the local, nice idea though.

 

[pc1966]

Thinking aloud, would the use of a wander lead to do IR to any dodgy looking screw that caused concern show up all possible faults like this? I've never seen the use of a wander lead for IR testing, but only for R2 testing

Doing a global IR test first (supply isolator off, L+N to E) would show most faults but not crossed circuits.

However, if an all-RCBO panel is fitted you will soon find out if they are!

But more seriously, if you want to check for all fault modes and you can have everything turned off / appliances disconnected then you could do this:

• Switch off input isolator, temporarily connect both L & N to E (using 10k resistors if you worry about no fusing and some muppet switching the supply back on).
• Disconnect each live (L, N) at a time and IR to the above shorted set. Say you test L1 then it shows any fault from L1 to {L2,L3...,MET,CPC1,CPC2,...,N1,N2...}
• Second wire, say L2, then shows L2 to {L3...,MET,CPC1,CPC2,...,N1,N2...} and, as you have already eliminated L1-L2, that is everything it could fault to.
• Repeat until all L & N are checked.

 

[Midwest]

Don't think I would of got many CU change jobs, if I insisted on my customer paying for an EICR first, at what a average cost of £150-300?

You can still factor in some pre testing before replacing the CU, you have to factor a reasonable amount into you quote.

 

[Pete999]

Don't think I would of got many CU change jobs, if I insisted on my customer paying for an EICR first, at what a average cost of £150-300?

You can still factor in some pre testing before replacing the CU, you have to factor a reasonable amount into you quote.

Then I think the ruling should be changed, to avoid the avoidable cases of unscrupulous, people quoting for CU changes without any testing beforehand.

 

[Midwest]

Then I think the ruling should be changed, to avoid the avoidable cases of unscrupulous, people quoting for CU changes without any testing beforehand.

What ruling?

I always tested first, before the old one came off. It doesn't make any business sense to say I'll charge a customer £200+ for an EICR, then £600-1200 (dependant on size of install etc) to actually replace the CU.

If its a very large property, then yes. But a two bed semi, you'd get shown the door.

 

[Pete999]

What ruling?

I always tested first, before the old one came off. It doesn't make any business sense to say I'll charge a customer £200+ for an EICR, then £600-1200 (dependant on size of install etc) to actually replace the CU.

If its a very large property, then yes. But a two bed semi, you'd get shown the door.

Then opinions need to change .don't they?