Full rewire required?

[TinnyB]

My new house has four floors including a basement and an attic. I have been inspecting the wiring of the sockets circuit. There is a single socket circuit which has a 32amp (6kA type B) MCB. An RCD is installed. There are 26 sockets. I disconnected the live conductor(s) at a socket and then tested each to see which was live before reconnecting them and repeating the process at the next socket with the following results (and inferences):
AA) 1 socket – 3 connected cables, 1 live (radial type connection supplying two radial type connections);
BB) 4 sockets – 3 connected cables, 2 live (ring type connection supplying a radial type connection);
CC) 4 sockets – 2 connected cables, 2 live (ring type connection);
DD) 10 sockets – 2 connected cables, 1 live (radial type connection supplying a radial type connection);
EE) 7 sockets – 1 connected cables, 1 live (radial type connection - end);
FF) 1 fused supply connection to boiler – 1 live connection (radial type connection – end);
GG) 1 fused connection to bathroom fan (not investigated);
HH) 1 fused connection to toilet fan and ceiling light (not investigated).

All the cabling has 2.5mm live (red) and neutral (black) conductors (or thereabouts – based on comparison to known 2.5 sqmm cable). The cables to CC) sockets have insulated earth conductors. The rest are not insulated (except by enclosure in the grey cable sheath and with yellow/green sheathing inside sockets). All the cabling looks to be in good condition and neatly terminated.

I intend to get advice from an electrician. I suspect he will advise a complete rewire and I may well take that advice. However I am wondering if a reasonable course would be to have the 32amp MCB replaced with a 20amp MCB and disconnect the eight kitchen sockets from this circuit and connect as a new ring circuit fed from a new 32 amp MCB. I would ask to have Zs, maximum fault current and continuity of cpc checked. My question is – would this approach be reasonable?

 

[Strima]

A periodic inspection and test would determine if a rewire is needed. It would be best to engage a local electrician to conduct and inspection and test and raise an Electrical Installation Condition Report (EICR) to identify any problems you may have.

 

[TinnyB]

A periodic inspection and test would determine if a rewire is needed. It would be best to engage a local electrician to conduct and inspection and test and raise an Electrical Installation Condition Report (EICR) to identify any problems you may have.

Thanks Strima. Any response to the question I asked?

 

[Deleted account]

A periodic inspection and test would determine if a rewire is needed. It would be best to engage a local electrician to conduct and inspection and test and raise an Electrical Installation Condition Report (EICR) to identify any problems you may have.

Agree.


whats the sort by votes by the right of posts? Maybe @Dan could let us know.

 

[DefyG]

If the cables are pvc insulated/sheathed its unlikely a rewire will be required (maybe a partial rewire depending upon alterations).
Personally I would advise splitting the circuits rather than all 26 on one circuit. How thats best achieved would be down to the electrician doing the work. Of course as part of the work the alterations will be tested and certified (R1+R2, Zs etc.).
As per post #2 a good place to start is with an EICR especially if its a new property to yourself.

 

[pc1966]

As above, get it professionally inspected to see the condition and have the cable IR tested, etc. Generally speaking PVC cable can last many decades if not abused (overheating, UV light outside, rodents, etc) so in most cases you might only be looking at a small amount of work to get everything in good shape. Don't simply go for the cheapest quote for an EICR as they are either (a) not planning on doing much work for it, or (b) will be looking to make a profit on the work the reccomend. You are looking at 1/2 - 1 day of skilled time to do it properly and write a full report, depending on the size of the propery and number of aspects to be investigated.

However, before you proceed think carefully about the room layouts. Are the lights and sockets where you want them? Are there enough sockets in each room for its intended use? If you are having any work done it is a lot cheaper and easier to do it before you decorate, etc, and in the one go.

If you have any doubts about the EICR then post it here with any personal details redacted (your home/address, electrician doing work) along with a couple of photos showing things like your consumer unit (CU = fuse box).

 

[TinnyB]

If the cables are pvc insulated/sheathed its unlikely a rewire will be required (maybe a partial rewire depending upon alterations).
Personally I would advise splitting the circuits rather than all 26 on one circuit. How thats best achieved would be down to the electrician doing the work. Of course as part of the work the alterations will be tested and certified (R1+R2, Zs etc.).
As per post #2 a good place to start is with an EICR especially if its a new property to yourself.

Thanks for that DefyG. By R1+R2 do you mean a ring main continuity/integrity test? As you will realize the socket circuit is a bit strange. I think it might have started as a ring type circuit (there are two cables to the consumer unit MCB but only 8 sockets now have two live feeds – see table in original post). So many radial type connections have been added that it is now more akin to a radial circuit (without adequate protection of the cabling) as you will see from the table. I do intend to get an electrician to review the system and make recommendations but I would like to be forearmed with some thoughts so that I can discuss possible options and make a judgment on the guy’s competence.
As far as splitting the circuit I was thinking of a separate circuit for the kitchen. I realize my this would not be a perfect solution but I think it might be a big improvement on what is there now. I think it could also be code compliant but I realise there are other factors to consider.
Thanks again.

 

[pc1966]

Thanks for that DefyG. By R1+R2 do you mean a ring main continuity/integrity test?

R1 and R2 are the line and earth (CPC) resistances, used to estimate Zs given the supply Ze.

r1 and r2 (and rn) are the ring final loop resistances, used to verify the ring is complete and the ratios look sane (ro r1 should be close to rn, and r2 typically 67% more due to the smaller CPC in T&E cable).

It sounds like you have a bit of a butchered RFC. If it were me I would probably look at rewiring it as a proper ring (usually per floor, but upper and attic is reasonable, etc). If that is not practical or affordable, then dropping to 20A for the MCB (or better still, lokking at an all RCBO board) is safe but you are risking the odd trip if you use much in the way of high power appliances (so heaters, kitchen appliances with heaters like dishwasher, washing machine).

 

[TinnyB]

R1 and R2 are the line and earth (CPC) resistances, used to estimate Zs given the supply Ze.

r1 and r2 (and rn) are the ring final loop resistances, used to verify the ring is complete and the ratios look sane (ro r1 should be close to rn, and r2 typically 67% more due to the smaller CPC in T&E cable).

It sounds like you have a bit of a butchered RFC. If it were me I would probably look at rewiring it as a proper ring (usually per floor, but upper and attic is reasonable, etc). If that is not practical or affordable, then dropping to 20A for the MCB (or better still, lokking at an all RCBO board) is safe but you are risking the odd trip if you use much in the way of high power appliances (so heaters, kitchen appliances with heaters like dishwasher, washing machine).

Thank you pc1966. My intention is to get an electrician to recommend the most appropriate approach given all the circumstances. You seem to have summed it up neatly - it is a bit of a butchered RFC. A complete rewire would be best but practicality and cost provide motivation to look at alternatives. Maybe a rewire would not cost as much as I fear but I am sure it would disrupt mightily. As I said I was thinking a new circuit to supply the kitchen would be relatively easy to add and most of the items you mention would be supplied from it.

 

[DefyG]

So many radial type connections have been added that it is now more akin to a radial circuit (without adequate protection of the cabling)This is why an EICR is a must, so that the current circuit arrangements can be established before working out what to do next.

This is why an EICR is a must, so that the current circuit arrangements can be established before working out what to do next.

By R1+R2, Zs etc. do you mean a ring main continuity/integrity test?

By 'etc.' yes I meant all other relevant tests.

 

[TinnyB]

This is why an EICR is a must, so that the current circuit arrangements can be established before working out what to do next.


By 'etc.' yes I meant all other relevant tests.

Will an EICR actually establish the circuit arrangement - in the sense of establishing the configuration of the wiring? Presumably if it does it will provide a schematic wiring diagram (of, in this case, the connection of each socket, fan etc to the consumer unit)? I have seen a couple of EICR certificates and I am pretty sure than none of them included, or had appended, wiring schematics.

 

[brianmoooore]

An EICR which has been done thoroughly will need to establish the circuit arrangements in order to check that nothing contravenes the regs.
As an example, you would need to confirm not only that a RFC is complete, but that any spurs are single spurs, and not a spur fed from a spur.
This won't produce any schematics as such, but it would be a fairly small extra step to do so.

 

[DefyG]

Schematic? No not normally. It will for example, provide the number of sockets on each circuit and what areas covered for compliance with BS7671 so shouldn't be too difficult to draw your own schematic from the EICR.
'Circuit Arrangements/Configuration' as far as BS7671 goes will include circuit type, cable type/size, protective device type/rating, floor area (in the case of radial circuits for 13a outlets) etc.

 

[TinnyB]

An EICR which has been done thoroughly will need to establish the circuit arrangements in order to check that nothing contravenes the regs.
As an example, you would need to confirm not only that a RFC is complete, but that any spurs are single spurs, and not a spur fed from a spur.
This won't produce any schematics as such, but it would be a fairly small extra step to do so.

From the table in my original posting you will see that there are 11 sockets which are spurs and feed spurs - these are sockets listed under AA) and DD). With the current MCB this does contravene the regs... hence the tentative proposal in my original posting. So would the EICR establish the wiring configuration or simply establish the non compliances of the existing system? It seems to me that it is quite a big step to produce the schematics in question - not particularly difficult but time consuming (and perhaps unnecessary).

 

[brianmoooore]

Every socket which is part of a RFC can feed a spur*. It used to be quite common to find a RFC feeding upstairs sockets only, with spurs being taken from them to each feed a socket in the room below.
What cannot be done is to have an unfused spur in 2.5mm2 from a RFC, which then loops on to another socket.

* Wouldn't be considered good practice now, but still within regs.

 

[DefyG]

So would the EICR establish the wiring configuration or simply establish the non compliances of the existing system?

It would establish what was compliant or non-compliant about any given circuit. Non-compliances are coded. No code = compliant circuit.

not particularly difficult but time consuming (and perhaps unnecessary).

Yes which not necessary part of EICR as customers do not want to pay the extra time.

 

[TinnyB]

Schematic? No not normally. It will for example, provide the number of sockets on each circuit and what areas covered for compliance with BS7671 so shouldn't be too difficult to draw your own schematic from the EICR.
'Circuit Arrangements/Configuration' as far as BS7671 goes will include circuit type, cable type/size, protective device type/rating, floor area (in the case of radial circuits for 13a outlets) etc.

How simple would it really be to draw the schematic? In this case there are 8 sockets (sockets BB) and CC)) - connected in the ring main. There are 18 sockets connected as spurs with 11 of these being 'intermediate' spur sockets (sockets AA0 and DD)) and 7 being 'end' spur sockets. I find it difficult to see how the actual configuration could be determined without testing specifically for this purpose

 

[TinnyB]

Every socket which is part of a RFC can feed a spur*. It used to be quite common to find a RFC feeding upstairs sockets only, with spurs being taken from them to each feed a socket in the room below.
What cannot be done is to have an unfused spur in 2.5mm2 from a RFC, which then loops on to another socket.

* Wouldn't be considered good practice now, but still within regs.

Are you sure that you cannot have an unfused spur which loops on to another socket? What if the consumer unit MCB is selected to protect the radial type connections. As far as I can determine this would be compliant - hence the tentative proposal floated in my original post. In post #8 pc1966 at least seems to concur.

 

[brianmoooore]

A radial circuit can have sockets looped from one to the next, but a conventional RFC, fused at 32A and made up entirely with 2.5mm2 cable can only have one socket (single or double) on the end of an unfused spur.

 

[TinnyB]

A radial circuit can have sockets looped from one to the next, but a conventional RFC, fused at 32A and made up entirely with 2.5mm2 cable can only have one socket (single or double) on the end of an unfused spur.

Quite so. No disagreement with this. Hence the final para of my original post. And for example post #18. I apologise if somehow I have suggested that what you say in post #19 was not the case. I did not mean to do so. I agree with what you say in post #19

 

[DefyG]

How simple would it really be to draw the schematic?

Simple but time consuming! Time costs!

I find it difficult to see how the actual configuration could be determined without testing specifically for this purpose

Depends what you mean by ‘actual’?
Without ripping a place apart the actual cable routes cannot be determined during an EICR for establishing the circuit arrangement.

 

[brianmoooore]

The two requirements are that no cable and no MCB is likely to be overloaded. Non standard circuits can be used, as long as these conditions are likely to be met, as determined by a 'competent person'.
There are also things like voltage drop and loop impedance to be considered in the design.

 

[TinnyB]

In this case there are connections to

Simple but time consuming! Time costs!

Depends what you mean by ‘actual’?
Without ripping a place apart the actual cable routes cannot be determined during an EICR for establishing the circuit arrangement.

I was thinking of what each cable connects to at each end - as opposed to the physical layout of the cables. On reflection perhaps it doesn't matter - if the idea floated in my original posting is adopted. Voltage drop could be an issue but presumably that could be assessed based on system impedances which presumably would not be difficult to measure.

 

[brianmoooore]

You have to determine where the 'far end' of radial circuits are, whether they are lighting or power circuits, in order to measure the loop impedance.
Repeat EICRs are much easier than the first one you undertake at a particular installation, as long as your records are good enough to be able to establish if any changes have been made.

 

[TinnyB]

You have to determine where the 'far end' of radial circuits are, whether they are lighting or power circuits, in order to measure the loop impedance.
Repeat EICRs are much easier than the first one you undertake at a particular installation, as long as your records are good enough to be able to establish if any changes have been made.

Presumably you could measure the loop impedance at each socket which would tell you where the 'effective far end' of the radial circuit is?

 

[TinnyB]

Thanks for all the info and ideas. I have a follow up question.

I have now determined that connections GG) and HH) are both what I have termed ‘end of radial’ type connections making a total of 10 (connections EE), FF), GG) and HH) in my original post). There are 4 sockets connected as ‘origin of radial’ connections (BB) in original post). Hence there must be a total of 6 branches in the radials. Socket AA) is one branch and I have sighted one branch which is at an underfloor junction box and can easily be eliminated. That leaves four branches at one or more (four seems most likely) junction boxes.

Presumably non intrusive testing could provide at least an indication of the integrity of these junction boxes. My question is - how reliable would such testing be?

Alternatively I could identify and eliminate the branched cables, or upgrade the junction boxes (to Wago boxes perhaps). This could be a major task and of course it is possible that there are other junction boxes in the system which would not be identified by my testing – although this seems fairly unlikely.

Thanks again everyone for all the help so far.

 

[Deleted account]

Thanks for all the info and ideas. I have a follow up question.

I have now determined that connections GG) and HH) are both what I have termed ‘end of radial’ type connections making a total of 10 (connections EE), FF), GG) and HH) in my original post). There are 4 sockets connected as ‘origin of radial’ connections (BB) in original post). Hence there must be a total of 6 branches in the radials. Socket AA) is one branch and I have sighted one branch which is at an underfloor junction box and can easily be eliminated. That leaves four branches at one or more (four seems most likely) junction boxes.

Presumably non intrusive testing could provide at least an indication of the integrity of these junction boxes. My question is - how reliable would such testing be?

Alternatively I could identify and eliminate the branched cables, or upgrade the junction boxes (to Wago boxes perhaps). This could be a major task and of course it is possible that there are other junction boxes in the system which would not be identified by my testing – although this seems fairly unlikely.

Thanks again everyone for all the help so far.

I thought I understood, but I’m totally confused now. Sounds like you;understand your system, so with the help of a sparkies yo7 should crack it good luck.