Testing on renovation head scratching !!

[Soulsurfer]

Hi all, I am undertaking a partial rewire of a house that a client is needing to do quite a bit of surgery on as tenant just left and what a mess they left behind !

TNC-S BS88 main fuse hoping at least 100A as client wants to add storage htrs all over the place ! So far looks like a no go as would be 90.6A just for heaters without the gen purpose CU or added Garage CU !! :crazy:

Issue I have is weird readings on prelim tests while breaking up downstairs ring final circuits.

Outline: entire downstairs all on 1 ring from 3036 re-wireable CU and old red/black 2.5mm old nasty cable, think it's pirelli before my apprenticeship 3 silver strands for the cpc.

I have tried to rewire kitchen on own ring, and separate other ground floor ring for other rooms but owner wanted to try and re-use some legs in the old cable too to save cash !

r1, rn, r2 on both rings are fine,
Kitchen
r1 0.30 ohms
rn 0.30
r2 0.58

Other ground floor ring approx twice the floor area
r1 0.55
rn 0.56
r2 0.95 Ohms

now the fun part:

Insulation R
Kitchen ring all linked at points not chased in as yet with terminal block or in existing sockets:

L-N 56 MOhms
N-E 47 MOhms
L-E 15.54 MOhms ?? all seem lower than anticipated but L-E seems bizarre !

Other D/S ring
L-N 49.3 M @500V
N-E 49.6 M @ 500V
L-E 0.11 M ??!! WTF ? ( on this test my Fluke MFT could not reach 500V test voltage and indicated just 166V !!

I tried again at 250V IR and got 0.10 M ohms ! and that time tester wouldn't shift to 250V test V and stopped at 140V test V.

Tried my MFT at home tonight and not any issues at all.

Also tried Continuity again to see if some odd L-CPC break but doubt that would find anything if IR wasn't good ? but anyway I don't get that reading it's baffling. I found a 240V jigsaw plugged in on d/s ring but dropped into floor void next to some boards that have been lifted but that didn't change any readings. can't see anything else plugged in or connected on these 2 rings. Also MFT was acting strangely by showing lightning symbol and locking out from further operation of test button if tried after first odd result at 166V when set at 500V & gave off a kind of NO buzzer sound. Had to disconnect crocs and turn off totally then back on to try test again!


Advice: Would anyone break each leg to socket all way around ring and IR test one socket to socket at a time and check good value or just save time and rewire rest of rings x 2 in new cable entirely ? saving labour costs messing around ? Very aged house and as a friend pointed out, may well be loads of long forgotten dodgy wiring or similar items connected somewhere else. I found lots of 30A jb's in voids under kitchen floorboards from possibly avoiding using a longer new length of 2.5mm at time so jointed in little bits of cable. ( now replaced ) and some just fabric taped under floors in small bakelite connector strip. Any ideas would be great as I was expecting altogether much higher than even 50 MOhms too. Even though above 1M is o.k. I like them way up high. Sorry to rattle on a long post.

Using as new Fluke 1653B MFT.

 

[Richard Burns]

Kitchen ring I would test repeatedly and see if the reading rose as it may just be damp. Other than that it is a good pass for an old circuit.

I would expect something still connected for the 0.11MΩ reading on the other ring, pretty good value for an old style neon to be in place across to earth (found a few handy indicator lights dropped to earth for function before an RCD is fitted). However the easy option is to replace the old wiring, the cost will be comparable to fault finding the old wiring.

 

[Soulsurfer]

Thank you Richard again ! I was under the impression with my more clued up friend who runs a testing business that there may be something stashed away causing a problem with the 0.11M but as is on 30 fuse wire and not modern CU with RCBO isn't going to trip. Also he was agreeing there that exactly that fact a few lengths of new cable as opposed to labour poking about would be cheaper or same cost and also future proof over much older twin that will only get worse.

Just couldn't get why tester wouldn't fire the full test V of 500V across and stopped at 166V or crept up a little after ?

 

[Richard Burns]

If the tester were putting the full voltage across the faulty section then the test current would be too high to be safe and so the tester limits the test current by lowering the test voltage.

 

[Soulsurfer]

Hey all, just got all accessible legs of offending rings out and drew up a diagram, now if I was to short out say from cu to first socket outlet L-E with a jumper & put crocs of IR tester on other L&E could I effectively test insulation of one leg loop at a time to find duff one or would tester not entertain that ? Just trying to avoid rewire of all old legs

 

[Richard Burns]

To test for the fault and narrow down where it is you need to break the circuit up.

Connecting any sort of jumper between points will simply ensure that there is zero (in MΩ) resistance between those points and along the rest of the circuit. So you will be in a worse situation than before.

I have drawn a diagram using only line and earth of the effective circuit for a ring with both legs separated out at the cu and a fault at the second socket.
It is not in ring form but this is what you have in principal.
Then with a jumper across L-E at the first socket (not exactly what you said but I could not follow that statement)

Then with the legs of the ring separated out at another socket as well and testing each end of the ring.

 

[Soulsurfer]

More testing during 1st fix today on renovation / upgrade and found 3 x lots of lighting cable was black rubber stuff and breaking down on internal insulation quite badly so replacing those tomorrow ! also had 0 Meg on L-N IR tests on both lighting circuit ways. is it likely still a lamp in somewhere or dodgy fitting ? Subsequently tested L+N together to E and had 38.2Meg and other circuit had 15.16Meg @500V so I guess thats o.k. as not new install as is normally a Limitation is it not ?

 

[Richard Burns]

If you have 0.00MΩ on a circuit when IR testing line to neutral it is often worth testing on continuity range to see if you get a reading. If you get 0-2000Ω then it is likely to be a load in the circuit (or a bad fault).
If the resistance reading is off scale then it tends more toward a faulty circuit.
However if you test L+N to E and get OK readings (as you did) then this would be acceptable.
It is an acceptable method of measuring IR for a periodic inspection according to GN3 so would not be a limitation, just standard practice.

 

[Soulsurfer]

Thanks Richard, I was under the impression it was a lim on a new install as there would be no excuse to not test fully as wiring along during first fix etc.., While you're here helping me yet again ;-) do you recall if it's anywhere in osg or 7671 to fit an above work surface d.p. switch or sfcu to isolate say a socket in kitchen units for larder fridge or dishwasher etc.. I usually fit them for ease of isolation but client would rather avoid extra switches on wall. Sure was a requirement but ?

 

[Richard Burns]

There is no requirement to fit any isolator or switch or fuse for sockets in a kitchen.
However it can be useful and also covers ease of access to switchgear when the socket is inaccessible behind equipment or at the back of cupboards.

 

[Richard Burns]

On a new installation there is no possibility of putting LIM on the records as the new installation must be tested fully.
I know various people may not agree with this, but it is the (unspecified) requirement.
On an existing installation L+N to E is acceptable.