Puzzling IR & Continuity results on an old Ring Main

[bf2k]

Hi, I'm relatively new to this forum and have not long since qualified.

As one of my first job, I recently installed a new CU to feed a large shed. I got the clients provider out to install a service switch so I could safely tail up the new CU. Whilst the engineer was out he plugged his tester into one of the house sockets and found there was no earth. After I'd finished the shed feed I looked at this problem and have uncovered a much bigger problem.

The existing CU is the original from the 70's, standard 3 way re-wirable Wylex box with a 5 AMP and 2x 30 AMP fuses. The lighting circuit is working fine. One of the 30 amp circuits is a radial with a 4mm cable, but I don't know where it is running to. But the 30 Amp circuit I have the problem with is for the house socket ring main (which had no earth). After removing the socket (which to begin with I thought it was a fault with the socket as the rest of the circuit was fine) I found that the earth connection was loose. However, I did the standard IR test on the circuit to issue the minor works cert where I found that there was a problem. My meter (fluke 1653) wouldn't even give me a reading. I've tried doing the test on both 500 VDC and 250 VDC, where both gave me a 0.00mohms reading @ 0.8VDC (which has puzzled me). So I did an R1+Rn gave 53.4ohms (R1+R2 and R2+Rn give >2000ohms). So from this I thought they must be a short between Live & Neutral and proceeded to fault find.

I split the circuit, found faults on both sides. So I split each side and have now managed to narrow this down to 4 sockets (the second in the line, one at the top of the stairs and 2 in the kitchen). What I'm puzzled at is no work has been done recently and the sockets which have faults don't seem to be consecutive sockets in the circuit. Plus before I changed the socket in the hall (which was the first in the line) the system worked. I also did a test from earth and neutral to a back box and got continuity, which again has puzzled me. I didn't do any tests other than the plug in tester to confirm what the engineer found before I changed the socket. They seems to be no logic to which socket follow the last and there are a few spurs running in a few places, but they seem to be fine.

This is the first fault finding mission I've had to do on my own since qualifying and I think I've been dealt a real one. Its a good job I know the client. I guess they can't prepare you for things like this at college.

Does anyone have any ideas, or come across something similar before?

Thanks
Andy

 

[bf2k]

Cheers Marvo

I'll give it another go tomorrow, more light and a fresher mind. I'll have to dig out my analogue multimeter from yesteryear as well.

Thanks (to most) for your help and ideas.

 

[somersetsparks]

A non textbook but useful way of fault finding a 'working' Ring, Is to disconnect 1 earth from the fuseboard, power back up, test at every point between L-E and see where you lose the reading, then change the earths over in the board and repeat the process, you can then narrow down between which points your fault(s) is. Obviously you should unplug everything first. You can also use this method for Live and neutral faults as well,
I wouldn't do it on a dead circuit that I knew nothing about, however this ones been in action for 40 years. The 53 Ohms is a certain load still connected (approx 1kW).
Fault finding out in the field is quite a skill and certainly different to classroom situations! Good luck and be logical and methodical!

 

[bf2k]

A little update as i'm on site now. I've narrowed it down to between the socket at the bottom of the stairs and the socket at the top of the landing. No other socket is between these two points. I've got a 0ohms reafing for ir between these two points and 50.2 for continuity. These are between live and neutral. Live to earth and neutral to earth is fine. Before i lift floor boards had anyone come across this before?

Cheers

 

[Brightspark2]

Sounds like a load giving false readings...as said that type of fault would blow the fuse when it was energized...

Is there an RCD socket or neon?...there's light at the end of the tunnel now, keep looking...

Sent from my Xperia S using next doors WIFI.

 

[RogertheBodger]

From your readings, you still have a load connected, dig a little deeper.

 

[snowhead]

No other socket is between these two points.

Cheers

Have you followed the cable, or are you assuming because you can't see a socket?

Spur for a burglar alarm, or old spur for wall mounted heater or socket under the stairs in a cupboard??

 

[somersetsparks]

Well that would be a load of 4.58A, so in theory would not blow fuses but would be a hell of a drain on finances! However Live to neutral faults are very rare in T+E installations, A connected load would be much more likely, being that its in the hallway, this would be the most likely place for a spur which could be anywhere in the house! Also, since this is a likely load connected I would be very careful meggering it further as you could end up buying some form of domestic appliance! That would be far more of a pain then the fault :wink5:
My domestic experience only extends to testing/repairs and fault finding and I have yet to find a genuine L-N fault!

 

[Richard Burns]

Cables from upstairs to downstairs can be damaged if they pass through the stair treads and are constantly trodden on, also are you sure that there is not something in the cupboard under the stairs as it is an easy place to spur from, tap in to get a light, etc.
1kW appliance sounds unlikely in that location but could be possible, or a failed accessory.

 

[bf2k]

I've located the fault. A cable run just under the floor boards had worn. Replaced cable and all fine now on the live neutral ir front. However I've now lost continuity between the two ends of the ring. Going through everything I've done now but everything is connected correctly.

I've got a reading of 0.31ohms between neutral and earth (evetything is disconnected from cu). With this being an old installation this could be a bare earth touching a back box or neutral leg, is that a correct?

I'm going to connect a fly lead to each leg and go round to see if i can see where the live i as lost.

Pain in the arse these fault finding missions aren't they?

Ps all the meggering has been done at 250v incase of connected appliances ive missed.

 

[Amp David]

If you had ring continuity before doing the repair, then it has to be your repair. Is it non continuous on each conductor?

 

[bf2k]

Again an update for those who are interested.

The faults I found were, loose neutrals in a kitchen socket, a poor (very poor in fact) connection for the boiler - were loose strands were hanging out everywhere (British Gas so called fitters for you ) and a broken cable under the floorboards were a wardrobe had recently been moved. All fixed and working.

But the biggest thing I found, which once realised made a lot of sense to my test readings, was that the sockets which were fed into the same 30A fuse labelled "Socket Ring Main" is actually two radial circuits. So I've de-rated the fuse to 15 Amp and moved one of the circuits to a spare fuse rating that at 15 amp as well as a temporary measure. I've notified my customer that if she uses too many appliances at once she may blow the fuse and that she really needs a re-wire and new CU. As expected she doesn't want the expense and mess of decorating. So I've said she really needs the circuits to be protected by a 20 amp fuse and have suggested replacing the re-wirable fuses with the plugin mcb's (its a wylex board). Because I found a short between earth/neutral I can't put rcbo's in so wondered if putting an rcd before the board would still work, although I think the short is happening between the installation and the building.

I suppose not finding that the sockets were if fact radials is due to a lack of experience and will be learnt from, not to take other peoples notation as gospel is one of them.

Thank to all that helped and allowed me to have a Sunday off for once

 

[Strima]

Glad you sorted it mate, it's a good feeling in the end. An upfront RCD will still trip.

 

[bf2k]

It does, especially when you flick the switch and plug your tester in and you don't hear snap crackle pop and get the results you want

I thought it would do but thought I'd look into the idea. I've told her that she really needs a new CU, which in turn will mean she needs a rewire (mainly because of the N+E fault and the IR between L+N was 8mohms. Must be puzzling for her though because the system has worked for 40 years. I've told her to go and get a second opinion if she needs but at least give me a chance to quote . She has taken me up on the suggestion of recplacing the re-wirables with plugin mcb's.

 

[Strima]

She has taken me up on the suggestion of recplacing the re-wirables with plugin mcb's.

Some people are even reluctant to do that. However I have pointed out the benefits even even just doing the lighting circuits so they're not trying to cram a nail in the carrier with a match between their teeth...

 

[ayjay]

why not find the 2 ends of the radials and link them to form a ring,
then find the e/n fault and fix it
(btw what size cable are the radials ? some are 4.0 mill)