Insulation Resistance - Head Scratcher

[daveb]

Hi folks,

Was sent out to a job today to add in a hot water pump underneath a kitchen sink in a kitchen that had recently been done up. Took with me just a socket thinking that the electrician doing the kitchen work would have RCD'd it - but no. The woman had her own joiner do the work so I'm assuming he also got someone for the electrics or heaven forbid - done it himself. So, ended up just hard wiring the pump into an RCD spur. Anyway, I digress....

Before I fitted the spur I went to test the circuit - got ring continuity on all but noticed the N-N was 0.73 ohms and lower than the live which was 0.90 ohms. Earth - Earth was 0.61 ohm. R1r2 was 0.50 ohms.

Then, went to IR test between live and Earth, and neutral to Earth to find that it was showing 0.00?? I had everything unplugged bar a fridge and dishwasher which was built in and inaccessible but if anything would expect this to cause a live to neutral fault.

Has anyone experienced this before? I've seen it with RCDs on the circuit that have gone undiscovered but again it's usually L-N. Earth loop was OK as it was 1.01 ohms. Was definitely the correct neutrals and earths disconnected at board too.

I will jot this down on the MW sheet but would like to back it up with possibilities as to why. Obviously there is not a dead dead short across all or the circuit would simply not be working.

Thanks and loon forward to hearing the responses.

Ps - does anyone else really not like the new tapatalk layout for this? It's horrible.

 

[SJD]

When you say you measured IR as L to E of 0.00, and N to E of 0.00, I presume you mean M Ohm. What are the readings in Ohms (they won't be 0.00).

 

[GLENNSPARK]

Did you take the neutrals out of the neutral bar before IRing?

TN-C-S?

 

[Risteard]

Before I fitted the spur I went to test the circuit - got ring continuity on all but noticed the N-N was 0.73 ohms and lower than the live (sic) which was 0.90 ohms. Earth - Earth was 0.61 ohm.

A couple of points:

If the cpc continuity was that low (assuming 2.5/1.5 T&E) then there are loose connections on everything except the cpc (and possibly the cpc also but to a lesser degree) or there are parallel paths. Regardless of this there is very definitely a problem with the phase conductor being substantially higher than the neutral (i.e. a loose connection).

Really this should have been rectified before adding to the circuit. There really isn't any way of writing on a MEIWC that you have altered a ring final circuit but the circuit isn't actually sound! (Suitable for your additions and all that...)

 

[GLENNSPARK]

Always best to IR at 250V first...

As if it is getting through the PME block in the cutout via connected loads then all you've done is tested through said loads with 500V DC....which they may not like

Think about how a portable appliance tester carries out its IR test

Both phase/s AND neutral are tested simultaneously together at 500V to earth...keeping potentials the same on all live conductors prevents damage to sensitive components...

 

[westward10]

Sounds like N/E fault to me, as for continuity readings can't second guess that.

 

[R G]

When you say you tested , did you test where you replaced the socket or did you test at the consumer unit and as already mentioned were all appliances etc disconnected ?

just noticed ends disconnected at board , still if appliances aren't disconnected then all sorts of readings can be obtained , not to mention the damage to them , tip no 1, remove all equipment , tip no 2 if unsure as to wether you have removed all equipment use 250v setting on the IR first before 500v ,

 

[daveb]

Hi guys,

Wow - thanks all for your responses. Much appreciated.

You are all correct with what you say re the 500v IR. This is why I always test on 250v first (unless the circuit is new) for this exact reason, just incase there is anything plugged in that has been missed.

Risteard - for my own learning can you please let me know how you figured out that the live and neutral might be loose? Is this assuming that because the earth was 0.61 ohms, then the lives and neutrals should be about a third of this figure? But because they are both higher and unbalanced then there might be loose connections?

So to sum it up guys, could this potentially be an appliance I am picking up showing the faults and that the unbalanced readings could be loose connections? Hard to say, I understand , but am just trying to figure out possibilities in my head.

Have any of you had a missed appliance still plugged in that would give you - what appears like - a short?

I'm not sure what the reading in ohms would be as I don't know how to get the fluke to show this when set on IR testing. When I have picked up stray RCDs in the past I've usually has a reading of 0.01 mega ohms bit this time it was 0.00

 

[R G]

Fridges are one thing that can give all sorts of readings if not isolated.....

 

[Lucien Nunes]

Items with surge protection VDRs can give low readings L+N>E on 500V but usually fine or O/R at 250V and definitely O/R with a DMM. Did you get low readings at 250V too?

 

[westward10]

If the appliances are not faulty your N/E and L/E will not give false readings. You need to remove them from the circuit to clarify.
Is there a gas boiler connected to the circuit because this can give false end to end readings of the cpcs due to parallel paths.

 

[daveb]

There may indeed be a gas boiler to be honest. Would need to have another look

What would you fill out on the minor works without ripping the house to pieces? Put down your honest readings and a note under the comments on existing installation box?

 

[westward10]

There may indeed be a gas boiler to be honest. Would need to have another look

What would you fill out on the minor works without ripping the house to pieces? Put down your honest readings and a note under the comments on existing installation box?

Unless you are going to investigate/rectify you need to record your findings. I have said this before test before you make alterations.

 

[GLENNSPARK]

If the appliances are not faulty your N/E and L/E will not give false readings. You need to remove them from the circuit to clarify.
Is there a gas boiler connected to the circuit because this can give false end to end readings of the cpcs due to parallel paths.

you need to remove the bloody neutrals from the neutral bar as well....like i`v already told him.

as..(apart from getting false values via the PME block..assuming its a TN-C-S)... theres also the issue of parallel paths on the neutrals

 

[GLENNSPARK]

Items with surge protection VDRs can give low readings L+N>E on 500V but usually fine or O/R at 250V and definitely O/R with a DMM. Did you get low readings at 250V too?

...yep

typically in the 0.220 MEG kind of range for SPDs

 

[westward10]

you need to remove the bloody neutrals from the neutral bar as well....like i`v already told him.

as..(apart from getting false values via the PME block..assuming its a TN-C-S)... theres also the issue of parallel paths on the neutrals

The OP says he had the neutrals disconnected the way I read it.

 

[daveb]

Correct westward - as you read on my earlier post the neutrals were indeed connected. So Glennspark, no need to go all shouty shouty, haha!

 

[Martin burton]

Your L-L should be the same reading as your N-N as they are the same size 2.5. If you want to work out what your Cpc-cpc then x your L-L reading by 1.67 if its 1.5mm. I also agree with the other all appliance need to be disconnected and its sounds like its tn-cs at the cut-out

 

[Flanders]

Yes as others as said it will be the appliances giving you faulse reading they shou,d shou,d be fitted with accessible isolation points, if yhey are not then i would carry out Ir with live and neutral staped together enter n/v or lim in the line to line box and tgen write an explination in the "comments on existing installation " that there is no local isolation for appliances so limited testing of line to line conducters carried out , i know you shou,d not write lim on a minor works , but without ripping out the kitchen it looks like you have no chose.