rewiring a light circuit

Hello,
I didn't look close enough at the electrics before I bought this house. I have found that none of the sockets are on a ring: they are all wired as radials, some with spurs. I know you can have a radial with a 20A MCB but they are not wired like that (32a MCB in CU making it look like rings)! I found at least one socket that did not have an earth connection!
The lights are wired curiously. Some of it is red and black T&E. Some of it is T&E that has stranded conductors. And some of it is (if such a thing exists) "single and earth": grey outer insulation, with a red insulated wire inside and an un insulated earth wire. Though earth wires may be present in the cable, they are not always connected; sometimes they are cut level with the grey insulation.
What is frustrating is that a professional firm of electricians fitted a new consumer unit a few months before I bought the house. How did they let the vendor get away with this? Shouldn't they have tested for earths at the sockets, etc? It seems to me fishy that someone should buy a new CU just before moving house. I wonder whether someone else made an offer on the house, had a survey, and withdrew their offer because of the electrics, so the vendor fitted a new CU to give the appearance that everything was ok?
I thought I would begin by rewiring the lights. My house is like this:
ground floor: lounge and kitchen first floor: front bedroom, rear bedroom, bathroom second floor (loft conversion): bedroom and en suite
I thought I would start on the first floor as I am already doing some decorating on that level.
The loft conversion above is making it very difficult to follow wires. What do the professionals do when rewiring houses: remove carpet and cut holes in the floor above or cut holes in the ceiling below?
Rather than move the bed and roll back the carpet in the loft conversion, I went with cut a hole in the ceiling on the first floor. It seems that the builders have run chicken wire between the old ceiling joists and placed rock wool insulation slabs on top of this. This has interfered with me using rods to push cables along. Why would they put insulation between the first and second floors? Surely it would be better to put it above the loft conversion, not under it?
It is very filthy up there, lots of black dirt has fallen onto me.
To make matters worse, the loft conversion is on new joists that run between the old ones (which are still used to hold up the ceilings of the rooms below).
The light at the top of the stairs is on the ground floor light circuit, yet the light in the rear bedroom is connected to that lamp for its live and neutral. It seems the loop connections on the light at the top of the stairs are used for the upstairs light circuit but the switched live for that light comes from the ground floor circuit! I think this is dangerous as you would not expect one fitting to have conductors from two circuits. Do you agree?
What is the best way for me to rectify this? Should I put the light at the top of the stairs onto the upstairs circuit? Or should I leave that light on the ground floor circuit and just remove the first floor live connection to that fitting? What is the common practice with stair lights; what circuit do you put them on, or do you give the stairs a circuit all of their own (perhaps shared with fire alarms?)
Thanks, Stephen.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Friday, 22 January 2016 20:01:43 UTC, Stephen wrote:

Radials are permitted & common. 32A with 4mm^2 is OK. If it's 2.5mm^2 you s hould use 20A MCBs. However these figures change if wire is buried in insul ation. See Wiki 'cable' for full detail.

easily fixed. I'd expect the occasional fault in a building with oldish wir ing.

presumably historic cable. Nothing wrong with that as long as its condition is ok, which with pvc it normally is.

Standard stuff

not good, but easily fixed hopefully

It doesn't sound like the work was legal

probably. It's normal.

why? From what you say it sounds like you only need connect the earths & me asure to check things are ok.

both

as Phil said that would be from before the loft conversion

normal

I'm not clear how that makes things worse

It's normal practice, millions of houses are wired that way.

Does it need rectifying for some reason? I can't think of one offhand. I ca n recommend getting a bit more familiar with house electrics before setting off on a good deal of work that by the sound of it is 99% pointless. Meanw hile I'd go round reconnecting all the earths - just check they are earths first, not a wire repurposed as a switched live.
And you might want to consider meditation.
NT
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Fri, 22 Jan 2016 12:58:40 -0800 (PST), snipped-for-privacy@gmail.com wrote:

It is 2.5mm^2 cable

It wasn't that the earth wire had snapped, it didn't run to the socket at all! There was a junction box beneath the floorboards where the radial split in two and they forgot to take the earth from the junction box.

Only that it means there is less space to poke a rod before you hit one set of joists. The bigger problem is that the rod gets caught in the chicken wire that is supporting the insulation.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Monday, 25 January 2016 09:58:03 UTC, Stephen wrote:

Then if you change the MCB to 20A it should be compliant

normally easily fixed

I can't help thinking you're making an awful lot of unneecssary work for yourself by considering a rewire. It sounds like a fairly easy repair job. Testing circuit resistances would also be best practice.
NT
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 22/01/16 20:00, Stephen wrote:

Technically you can have a 32A radial, but it is impractical due to trying to get 2 large conductors into standard terminals.
Sounds more like someone forgot to "join the ends".

A new CU would have required full circuit testing afterwards, so none of those faults should exist. So much for Part P!!
Have you got any documentation on the CU fitting - because I would refer it straight to the firm's professional body if there was one!

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
wrote:

Yes I have but I suppose I should have taken photos or called them in before trying to rectify the faults.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 22/01/2016 20:00, Stephen wrote:

It's recommended that the 20A radial (usually 2.5 mm^2 cable) should not serve more than 50 m^2 floor area, and the 32 A radial (4 mm^2) no more than 75 m^2. Spurs and branches are allowed, provided the requirements for voltage drop and earth-fault-loop-impedance are met at all points. This is subject to consideration of the likely loading (after diversity)- e.g. the 20A circuit will often not be appropriate for a kitchen.

That was commonly used in the late 60s - early 70s period, especially on low-cost housing. The stranded wire is probably 3/0.029 (pre-metrication ~1969-70).

See http://www.electricalsafetyfirst.org.uk/mediafile/100423440/Best-Practice-Guide-1-Issue-3-.pdf
...

If you're saying that the supply to the lamp is between the line ('live') conductor of one circuit and the neutral of a different circuit then, yes, that's potentially dangerous - although it can still be found in the form of the 'borrowed neutral' in older installations.
If there are only two lighting circuits in the house they should be on seperate RCDs in the new consumer unit, in which case one or both RCDs should trip whenever that light is switched on.

Golden rule: The wiring of the line & neutral of each final circuit must be electrically seperate from that of any other circuit, in order to permit safe isolation.
--
Andy

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
ion, with a red

n.

There was also black and earth.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 22/01/2016 20:00, Stephen wrote:

Much depends on what size cable has been used... also how its been wired. So for example, 4mm^2 T&E radial (with branches in 4mm^2 and / or spurs from it in 2.5mm^2) would be fine. Even multiple spurs in 2.5mm^2 taken from the "origin" of the circuit (i.e. the CU).

Yup, that's never good...

That's fine

Some older imperial sizes of T&E (and modern large sizes) are stranded. There are even some rare cases of modern 2.5mm^2 T&E that's stranded.

It may be that they actually wanted just a double insulated single... (or its a bodge - again the devil is in the detail)

Not necessarily. They would have tested for ring continuity, and probably insulation resistance, but not necessarily tested every socket.

Could be...
I have even see a place that was "rewired" before sale, where all that was changed was the CU, the sockets and switches and short lengths of visible cable, that were then grafted back on to the existing perished rubber insulated wiring!

Fairy snuff...

Depends on what is easiest and requires least making good. Normally you would lift a floor if its floor boards, but in cases where it say has a real wood floor laid over it, it may be less destructive to get through the ceiling from below.

Yup that is a common building control requirement when adding a second storey, to ensure you have 30 minute fire protection between the floors. I had to do it when I built mine. They get very worried about escaping from second or above storeys since you need to get through lots of existing house!

Par for the course alas... all over disposable bunny suit and a good respirator!

Again, good practice for several reasons... it saves having to pull down the existing ceilings and keeps all the new loft work self contained during the build. When done it also means the new floor is decoupled from the existing ceiling so you get less noise transmission.
e.g. http://www.internode.co.uk/loft/floor.htm

Its probably two way switched, so you may have live wires from both circuits present at the switches.
http://wiki.diyfaq.org.uk/index.php/2_Way_Switching

Borrowed neutrals are never a good idea - especially when the circuits are split across multiple RCDs. However it is an unavoidable fact that a lamp switched from more than one storey is at some point going to have live wires belonging to the "other" circuit present. Its just one of those things you need to be aware of.

I would say its best practice to do two way switching in *both* directions - two switches at the bottom and two at the top. So you can switch both lamps from both locations.
http://wiki.diyfaq.org.uk/index.php/File:HalLandingSwitchingSml.gif
You may find some useful background in:
http://wiki.diyfaq.org.uk/index.php/House_Wiring_for_Beginners http://wiki.diyfaq.org.uk/index.php/Electrical_Installation http://wiki.diyfaq.org.uk/index.php/Rewiring_Tips
Plus plenty more electrical articles:
http://wiki.diyfaq.org.uk/index.php/Category:Electrical
--
Cheers,

John.
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 22/01/2016 23:22, John Rumm wrote:

Are you saying you can have a 32A MCB feeding 4mm^2 cable, then run a 2.5mm^2 spur off it with no additional protection?
Andy
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 24/01/2016 21:26, Vir Campestris wrote:

Yes, exactly as with a 32A ring circuit. The same limitation applies - one single or one double socket only if the spur is unfused.
(the logic being, as with a ring, the limitation on the number of sockets will limit the maximum load so you can't overload the cable, and the 32A MCB will still provide adequate fault protection for the 2.5mm^2 T&E is the shit hits the fan (or the screw hits the cable!))
--
Cheers,

John.
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 25/01/2016 01:19, John Rumm wrote:

That seems odd. Either 2.5mm cable is happy with 32A (a double can supply 26A, even without one of the things being faulty) - in which case the ring's 4mm is grossly overspecced - or it can't, in which case obscure circumstances could overload the cable.
A quick google suggests 2.5mm^2 is only good for 20A!
Andy
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

2.5mm when not installed in insulation is good for 27A.
A double socket will soon fail if it is used for supplying 26A for a prolonged period.
--
Adam


Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 26/01/2016 21:05, Vir Campestris wrote:

It depends on the installation method.
http://wiki.diyfaq.org.uk/index.php/Cables#Installation_Methods
To be used in a ring circuit then it needs to be 21A "as installed" or higher. If you can't meet that, with the actual installation (say too much insulation or circuit grouping) then you will need to use a larger CSA cable.
It is also important to clearly differentiate between overload currents and fault currents[1]. With a general purpose circuit you need to provide protection against both types of over current. Normally this protection is provided by the fuse or MCB at the origin of the circuit.
In the case of fault currents, this is the only place you can do it.
With overloads however you can split the responsibility and provide the protection via some other means. This is the design principle that a spur uses for overload protection. The fuse/MCB can't do it, since the rating is higher than the maximum capacity of the 2.5mm^2 cable even in the "clipped direct" mode giving the full 27A, but by controlling the load that can be applied to the end of it[2], you can achieve the same result. Normally a double socket is rated at 20A continuous maximum load (not 26A), and that will be less than the installed capacity of the single piece of cable in the spur.
(I am sure there will be someone along shortly to come up with a scheme involving a pair of 4 way extension leads, 6kW of static load, a length of rubber tubing, and a yak, to highlight why this is not true, and hence why we so commonly see properties burst into flames daily!)
[1] from http://wiki.diyfaq.org.uk/index.php/Electrical_Circuit_Faults#Fault_and_overload_currents
Fault Current
This is the current that flows when a short circuit fault occurs in a circuit between either the live and neutral conductors, or the live and earth conductors. The magnitude of fault currents can be huge (100s or 1000s of Amps) since they are limited only by the resistance of the wires in the circuit between the consumer unit and the fault, and the impedance of your power supply and earth connection as delivered to the property.
Overload Current
This occurs when the total current demand made by the appliances connected to the circuit exceed its design capacity. This significance of an overload will depend on its magnitude and its duration. Small overloads may be tolerable for long durations, while big ones can only be tolerated for short durations, before any damage to the circuit wiring occurs.
[2] Hence the rule about no more than one single or double socket on an unfused spur.
--
Cheers,

John.
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 27/01/16 00:05, John Rumm wrote:

Rod Speed to a T?
--
Future generations will wonder in bemused amazement that the early
twenty-first century’s developed world went into hysterical panic over a
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 27/01/2016 00:06, The Natural Philosopher wrote:

Dunno, I keep him well filtered ;-)
--
Cheers,

John.
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 27/01/2016 00:05, John Rumm wrote: <snip>

So there's no danger to the 2.5mm cable as the socket will burst into flames first?
Though I suppose that's equally true of any double socket anywhere on the ring.
Incidentally I've installed sockets (pre-part P honest!) and noticed the limit. I'll give odds most people have no idea.
Andy
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 27/01/2016 21:52, Vir Campestris wrote:

In many cases there will be no danger to the cable at all (since it will often have a greater installed capacity than the 26A theoretical load of the socket. Even if if you do exceed that, then the cable does not immediately fail - it just runs over the 70 degree temperature limit that allows the cable a virtually unlimited life. If you routinely overheat it, it will have a shorter life before the insulation becomes brittle and starts to crack etc.
There is perhaps more risk to the socket, but again it does not seem to be a problem in real life since its not actually that easy to get 26A (i.e. over 6kW) of load on a socket for a long duration. Most people would probably shy of doing it too often once they realise that it would get 'kin hot running at max capacity for an extended period.

Indeed. Some good quality sockets will carry more current indefinitely anyway, the 20A is the notional design current.

Part P does not prevent you installing sockets anyway... ;-)
--
Cheers,

John.
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Wednesday, 27 January 2016 22:26:15 UTC, John Rumm wrote:

It's hard to come up with a realistic scenario where a double socket would supply 26A for long. A heater & a kettle are normally less than 26A, but even if they were it wouldn't run for long.

has it achieved anything in practice?
NT
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Like hell it is.

But two heaters would.

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

HomeOwnersHub.com is a website for homeowners and building and maintenance pros. It is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.