2+ faults on ring final? suggestions please

[GarethH]

Hi all

I'm renovating a room in our house and came across a badly installed socket with a blackened wire running next to a CH pipe in a notch between floorboards. This socket was spurred off the ring final using bare terminal blocks. We have two 32 A RF circuits - one for kitchen and one for the rest of the house. The socket isn't used so I decided to get rid of it.

I was going to remove the wire to the unused socket and replace the blocks with crimps. I tested continuity / resistance between the conductors of the two remaining cables first. All like conductors (L to L, N to N and CPC to CPC) were open circuit which suggests the ring was broken. With the wires separated (and everything unplugged), 5 double sockets that previously worked don't receive power.

All wires are 2.5mm2 and running off a 32A circuit.

I removed the face from a different socket, on the same circuit, in the next room and tested continuity between the conductors at that point and got a low resistance reading - 0.4 ohm.

In summary, I'm 100% confident this is the same RF circuit and:
o At one point like conductors are open circuit
o At another point like conductors give a low resistance.

Having thought about this it seems I have at least two faults to rectify - the ring is broken somewhere and there is a bridge in somewhere else. Is there another explanation or something specific I should be looking for?

For now I've replaced the 32A breaker with a 16A one from the immersion circuit (which we never use).

 

[JohnW2]

In summary, I'm 100% confident this is the same RF circuit and:
o At one point like conductors are open circuit
o At another point like conductors give a low resistance.
Having thought about this it seems I have at least two faults to rectify - the ring is broken somewhere and there is a bridge in somewhere else. Is there another explanation or something specific I should be looking for?

As you say, it's a mess, and may take a bit of unravelling. The ring final needs to be systematically tested/investigated and, unless you are very confident (and suitably equipped), that's probably a job for an electrician. One problem in trying to do it yourself is that you would need a means of accurately measuring low resistances, which requires kit that most non-electricians won't have.

Kind Regards, John

 

[333rocky333]

could it be the two cables are part of a radial circuit or a branch of the ring, hence parting them broke the ongoing supply to the other sockets

 

[Taylortwocities]

I removed the face from a different socket, on the same circuit, in the next room and tested continuity between the conductors at that point and got a low resistance reading - 0.4 ohm.

If you are doing any testing, like this, you must remove all the connections at the consumer unit.
Remember that neutral and earth are connected together, either at the supply head (TNC-S) or at the substation

So you'll always get a neutral- earth connection unless you isolate the circuit completely.

 

[JohnW2]

If you are doing any testing, like this, you must remove all the connections at the consumer unit.

Indeed - but, obviously, not just remove them from the CU, but also join the corresponding conductors from the two legs of the ring.

Remember that neutral and earth are connected together, either at the supply head (TNC-S) or at the substation So you'll always get a neutral- earth connection unless you isolate the circuit completely.

Indeed, and (if one did not disconnect from CU) that could result in failure to detect a lack of ring continuity in N or E (since the other would result in apparent continuity, due to the connection between them). However, such considerations obviously could not result in erroneously finding no continuity (L-L, N-N or CPC-CPC), and nor do I think it could result in erroneously finding L-L continuity on a circuit which was at least partially working (hence no L-N or L-E fault at the CU) and which had no connected loads.

Kind Regards, John

 

[GarethH]

There's a double pole isolator which I'd presumed to be sufficient although I guess it does leave CPC connected back to neutral at the substation or service head. I'm aware the MCBs are single pole so only isolate L.

I've been testing between like conductors (i.e. L to L, N to N etc.). It's interesting that there's a complete break between all pairs in places around the circuit as this indicates something more fundamental than a single loose connection.

John - I agree RE getting an electrician in and I've contacted a friend in the trade who is dropping round in a couple of weeks time. In the meantime, I'd like to try and narrow down some potential locations for the fault(s) as I suspect we may end up emptying a room to lift the carpet and floorboards. I have a good DMM (Tektronix).

There's one room (next to the room I'm renovating) which has three sockets, two of these don't have L-L continuity and one does so that seems a good location to start investigating further.

 

[JohnW2]

There's a double pole isolator which I'd presumed to be sufficient although I guess it does leave CPC connected back to neutral at the substation or service head. I'm aware the MCBs are single pole so only isolate L.

Are you talking about an isolator between your meter and CU? If so, when that is 'off', it would remove the problem which TTC mentioned - since, although the CPC would be connected to the supplier's neutral at some point, your installation would be isolated from that neutral.

I've been testing between like conductors (i.e. L to L, N to N etc.). It's interesting that there's a complete break between all pairs in places around the circuit as this indicates something more fundamental than a single loose connection.

Indeed - and, as you've said, your findings seem to indicate that [in addition to fault(s)] what you are dealing with seems to be anything but a simple ring - which is why I think careful systematic tracing of the arrangement of the circuit is required.

John - I agree RE getting an electrician in and I've contacted a friend in the trade who is dropping round in a couple of weeks time. In the meantime, I'd like to try and narrow down some potential locations for the fault(s) as I suspect we may end up emptying a room to lift the carpet and floorboards. I have a good DMM (Tektronix).

Yes, I understand what you want to achieve. Can your DMM accurately measure resistances below 1 ohm?

Kind Regards, John

 

[GarethH]

The isolator is located in the cutout before the meter so some good news at last. My original intention was only to paint a room and replace the skirting.

The DMM is a tektronix TX-3 performance for the 50 ohm range, 5,000 counts: resolution 0.01 ohm, accuracy ±(0.1% + 10 counts). Sounds reasonable although I'm not sure how to interpret "counts".

Many thanks for your continued help John

 

[GarethH]

Well, I decided to try and narrow things down a little by lifting some accessible floorboards in the hallway which is currently uncarpeted. This gave me access to the cables of the RF circuit as it goes between rooms. I checked for voltage on the cables with a megger pen before and after breaking the circuit in bed 4.

This diagram shows what I found.

Green wires and sockets are always live (irrespective of whether the circuit is broken.
Red wires and sockets are only live when the circuit isn't broken in bed 4.

I've only drawn in what I can actually see which includes all of the wires which cross the hallway.

View media item 68489
I think this fits with what I've seen from testing:
o There is a complete ring as evidenced from the LR reading from L-L at the green socket in bed2.
o There is also a radial type arrangement that comes off the ring, powers the red sockets but does not appear to return to the ring (as evidenced by the open circuit readings on all three conductors.

It seems I have a P-shaped circuit and It looks certain the fault(s) are under bed2 (which unfortunately, has just been carpeted and filled with all the stuff from bed4. I expect that, when the extension was built, the builder intended to extend the ring but got it wrong. I'll rectify the faults and test the circuit in a few weeks after moving all the stuff out.

 

[Taylortwocities]

Even if you have an isolator, you still should discomnnect the circuit under test.. With the isolator off and the MCB off you will still have all of the neutrals and all of the earths connected together so when you do your insulation resistance testing you'll also be picking up any nasties from the other circuits.

Please do it properly or you can be chasing red herrings for hours.

 

[JohnW2]

Even if you have an isolator, you still should discomnnect the circuit under test.. With the isolator off and the MCB off you will still have all of the neutrals and all of the earths connected together so when you do your insulation resistance testing you'll also be picking up any nasties from the other circuits.

As you will realise, my comments related specifically to continuity testing, which is all the OP reported having done. When it comes to IR testing (if the OP is equipped to do it), what one does obviously depends on what one wants to test - and if one wants to test just one final circuit then, of course, it goes without saying that the cable(s) has/have to be disconnected from the CU!

Kind Regards, John

 

[GarethH]

Will do and thanks for the advice. We'll be doing IR testing when my electrician friend comes over as I don't own the gear.

 

[JohnW2]

I think this fits with what I've seen from testing:
o There is a complete ring as evidenced from the LR reading from L-L at the green socket in bed2.

Yes, there must be a 'ring' of some sort (in the literal sense) but it doesn't have to be a conventional ring final circuit which originates at the CU. I don't think anyone has actually asked you - do I take it that the circuit in question does have the expected (for a ring final) two cables at the CU?

... o There is also a radial type arrangement that comes off the ring, powers the red sockets but does not appear to return to the ring (as evidenced by the open circuit readings on all three conductors.
It seems I have a P-shaped circuit and It looks certain the fault(s) are under bed2

All this really just reinforces the fact that you have some unconventional, probably non-compliant and possibly dangerous circuit arrangement, not to mention a possible fault, and that careful systematic testing/tracing (or tearing up all the floorboards and physically tracing all all cables!) will be necessary to get to the bottom of (a) what the arrangement of the circuit actually is and (b) what fault(s) are present.

Kind Regards, John

 

[GarethH]

Yes - there are two cables (both 2.5mm2 T&E) which come from the CU and go under bed 2. The complete ring seems to supply bedrooms 1, 3 and one socket in bed 2.

The remainder of the sockets in bed 2 and those in bed 4 appear to be fed from the radial type lashup off the ring.

As you say, I expect this is dangerous and non-compliant because it's possible to load the red sockets so as to exceed the CCC of the radial-like part of the circuit without the circuit protection activating (as well as some other considerations such as even distribution of loads around the ring).

I've disconnected this sub-circuit completely for now and am confident that the problem will be easy to rectify when I can get under the floorboards of bed 2.

Lifting the floorboards wasn't much of an problem in this case which was fortunate. I've also now got a clear mental picture of where the cabling for much of the circuit is which could prove to be useful in the future.

 

[JohnW2]

Yes - there are two cables (both 2.5mm2 T&E) which come from the CU and go under bed 2. The complete ring seems to supply bedrooms 1, 3 and one socket in bed 2. The remainder of the sockets in bed 2 and those in bed 4 appear to be fed from the radial type lashup off the ring.

Indeed, that would seem to fit with what you've described. To be sure, you ought to find that one of those latter sockets (the last one in the line) will have only one cable going to it. The 'radial type lashup' would normally be described as an (unfused) spur - perfectly acceptable, safe and and compliant with regs if it supplied only one socket (which clearly in not the case!).

As you say, I expect this is dangerous and non-compliant because it's possible to load the red sockets so as to exceed the CCC of the radial-like part of the circuit without the circuit protection activating (as well as some other considerations such as even distribution of loads around the ring).

Quite so. As above, an unfused spur is only allowed to supply one socket, for the reasons you give. It could be made compliant by inserting a 13A FCU beween the ring, making it into a 'fused spur', but that would,of course, mean that the total load available for all of your 'red' sockets would be just 13A.

I've disconnected this sub-circuit completely for now and am confident that the problem will be easy to rectify when I can get under the floorboards of bed 2.

That sounds sensible. The most obvious solution would be to simply modify the wiring so that those 'errant' sockets were all part of the main ring. As an alternative, if you have a spare way available at your CU, you could run those sockets as a separate radial (i.e. connect them directly to an MCB of their own at the CU), but with 2.5mm² cable, you'd be limited to a 20A MCB, hence total load of 20A for those circuits.

Kind Regards, John