Insualtuion Resistance - individual circuits V bank of MCBs & RCD readings

[Motty]

Tested each of 9 new circuits in usual way of L.O.C. and IR, all in theory ok ... CCU is a Hagar split-load dual RCD. 1-4 on RCD1 / 5 - 9 on RCD2

Aware that no requirement for "collective" IR on EIC, just individual circuits.

When tested from the outgoing L of the RCD2 to MCB busbar (all MCBs on, RCD off) to the respective N bar, got 0.06 M ohm - fail. Turn off all MCBs, get >999. Then turn on each one (5 to 9) until circuit 9 reveals the sudden drop.

Take out #9 conductors and test again individually (Downstairs Lighting) showing 600-odd M ohm rather than >999Mohm on L-N (L-E and N-E >999M) (All PIR lights have been disconnected before test) Extractor fan for WC connected so switch off at the 3-pole isolator (Assume Test current detecting / shorting / connecting across motor coil / winding) reconnect into MCB, and now L-N up to 100 odd M-ohms. Still concerned LN is lacking behind LE and NE in the readings overall, and had to call it a day at that point.

No fused spurs or neons, no lamps / bulbs, and the individual test without circuit being connected to the MCB is satisfactory, Yet fails (under 1 M ohm) when connected to MCB / RCD. I discounted reading interference from the RCD as the other 4 circuits in this bank all check out (>999) connected. None of it is Live yet (no connection to new property yet) but brain dead after a very long couple of days! ... any thoughts or pointers would be appreciated

cheers

 

[Meggerman]

Tested each of 9 new circuits in usual way of L.O.C. and IR, all in theory ok ... CCU is a Hagar split-load dual RCD. 1-4 on RCD1 / 5 - 9 on RCD2

Aware that no requirement for "collective" IR on EIC, just individual circuits.

When tested from the outgoing L of the RCD2 to MCB busbar (all MCBs on, RCD off) to the respective N bar, got 0.06 M ohm - fail. Turn off all MCBs, get >999. Then turn on each one (5 to 9) until circuit 9 reveals the sudden drop.

Take out #9 conductors and test again individually (Downstairs Lighting) showing 600-odd M ohm rather than >999Mohm on L-N (L-E and N-E >999M) (All PIR lights have been disconnected before test) Extractor fan for WC connected so switch off at the 3-pole isolator (Assume Test current detecting / shorting / connecting across motor coil / winding) reconnect into MCB, and now L-N up to 100 odd M-ohms. Still concerned LN is lacking behind LE and NE in the readings overall, and had to call it a day at that point.

No fused spurs or neons, no lamps / bulbs, and the individual test without circuit being connected to the MCB is satisfactory, Yet fails (under 1 M ohm) when connected to MCB / RCD. I discounted reading interference from the RCD as the other 4 circuits in this bank all check out (>999) connected. None of it is Live yet (no connection to new property yet) but brain dead after a very long couple of days! ... any thoughts or pointers would be appreciated

cheers

Think I have understood post correctly. Have you tested between circuit 9 Live and all the individual RCD2 side neutrals?

 

[Pete999]

Bit confusing, have you disconnected the suspect circuit Line, Neutral and CPC and tried testin just that circuit, as you don't seem to have done that, or am I reading your thread the wrong way?

 

[Meggerman]

Bit confusing, have you disconnected the suspect circuit Line, Neutral and CPC and tried testin just that circuit, as you don't seem to have done that, or am I reading your thread the wrong way?

He says he has done that and that it tested clear. It only shows L&N fault when connected to N bar.

 

[Motty]

Bit confusing, have you disconnected the suspect circuit Line, Neutral and CPC and tried testin just that circuit, as you don't seem to have done that, or am I reading your thread the wrong way?

Hi both thanks for replys -

Take out #9 conductors and test again individually (Downstairs Lighting) showing 600-odd M ohm rather than >999Mohm on L-N (L-E and N-E >999M)

yes did disconnect individual circuit and retested - the results were ok but fails when reconnected into the MCB ...

 

[Meggerman]

Have you tested between circuit 9 Live and all the individual RCD2 side neutrals one at a time?

 

[Hawkmoon]

Not quite fully understanding but if you only get this fault when connected to neutral bar could possibly be borrowed neutral? though you said all lamps were out.

Definitely no loads still connected? loft light, aerial amplifier, outside light etc etc

Can you confirm that all is good when you test the cable alone and separate

 

[Meggerman]

Not quite fully understanding but if you only get this fault when connected to neutral bar could possibly be borrowed neutral? though you said all lamps were out.

Definitely no loads still connected? loft light, aerial amplifier, outside light etc etc

Can you confirm that all is good when you test the cable alone and separate

He has confirmed this.

 

[Hawkmoon]

This may be the RCD (I normally disconnect them for IR).

Edit - though if MCB off when testing, not RCD.

 

[Motty]

This may be the RCD (I normally disconnect them for IR).

Hi again

The bank of MCBs 5-8 on RCD2 are fine, its only when i switch mcb9 to on that the IR dropped to 0.06.

Tested across L busbar / RCD terminal to RCD2 neutral bar.

Load of Ext fan is still connected but off at the 3pole isolator, but have'nt physically detached the wires from isolator to fan.

Circuit #9 Live, Neutral and CPC conductors removed and tested individually = LN = 600+MO, LE = >999MO, NE= >99MO

No borrowed neutrals - all new circuits in new build house.

No lamps or neons, no fused spurs, 8 x downlighter connections disconnected at the click-flow connectors.


Meggerman
have you tested between circuit 9 Live and all the individual RCD2 side neutrals one at a time?

no haven't done this - guessing this in case of cross connection with another circuit? ie. borrowed neutral?

 

[Hawkmoon]

Hi again

The bank of MCBs 5-8 on RCD2 are fine, its only when i switch mcb9 to on that the IR dropped to 0.06.

Tested across L busbar / RCD terminal to RCD2 neutral bar.

Load of Ext fan is still connected but off at the 3pole isolator, but have'nt physically detached the wires from isolator to fan.

Circuit #9 Live, Neutral and CPC conductors removed and tested individually = LN = 600+MO, LE = >999MO, NE= >99MO

No borrowed neutrals - all new circuits in new build house.

No lamps or neons, no fused spurs, 8 x downlighter connections disconnected at the click-flow connectors.


Meggerman
have you tested between circuit 9 Live and all the individual RCD2 side neutrals one at a time?

no haven't done this - guessing this in case of cross connection with another circuit? ie. borrowed neutral?

Is this with the other MCBs off? If so then definitely RCD

 

[Motty]

additionally following on from testing circuit 9 Live to all neutrals on RCD 2 there are 2 circuits with fused spurs (immersion heater outlet) but I have removed the fuses from the fascia cartridges

Hawkmoon - didnt't try that either ... think my process of logical elimination went to pot ... however the other four MCBs on that RCD seemed fine switched On and connected to that RCD, hence i discounted the RCD at that point.

 

[Pete999]

Disconnect these two heaters and test again

 

[Hawkmoon]

I think this may be one of those 'you have to be there' kind of faults!

 

[Engineer54]

Obviously (if i'm reading this correctly) the problem lies on the lighting circuit, and not knowing what you actually mean by ''PIR lights disconnected'', because if you are testing with the PIR L+N still connected and the lights not, you will still be reading through the sensor. If that's the case, then you'll be lucky if the PIR is still functioning, if your testing at 500V...

 

[Motty]

I think this may be one of those 'you have to be there' kind of faults!

All

Thanks for the replies and interpreting my waffle. its given me some fresh thoughts and stuff to eliminate on retesting. Thankfully not live yet, and not even DNO connected for another month.

bit of a nightmare and another lesson learned for me to do my testing with a clear head and not at the end of a 14hr day, with a 50 mile drive to go on top!

cheers

Engineer54> I did remove these conductors and taped the ends before the IR (500v) test, on all PIR fittings. Certain nothing other than the Ext fan was connected and this was switched off at the 3pole isolator

 

[Octopus]

I've got a bit lost on this one BUT if there is a very low IR as you suggest the only thing to do is closely inspect the specific circuit in question.

It sounds like a new build so as you installed it then fault finding should be a breeze!

 

[topquark]

If you're sure there's no obvious short (from inspection) ... sounds like may be a tranny/dimmer/relay or lamp still in circuit somewhere, something's obv current hungry somewhere.

Time to start successive dividing of the circuit to close in on the fault; if reinspecting reveals nothing. Start by disc at CU end, then move to approx middle of circuit and try both legs there and repeat.

 

[Badged01]

For those of you who understood this thread and how to solve the problem I tip my hat to you ........ I couldn't make head nor tail of it!

 

[Risteard]

Tested each of 9 new circuits in usual way of L.O.C. and IR, all in theory ok ... CCU is a Hagar split-load dual RCD. 1-4 on RCD1 / 5 - 9 on RCD2

Aware that no requirement for "collective" IR on EIC, just individual circuits.

No idea what you mean by "L.O.C.", but insulation resistance should be tested with all circuits in parallel (and with the means of Earthing connected) rather than each circuit. If you test circuits individually then you would need to calculate the parallel resistance to see if it complied with the minimum acceptable for BS 7671.

 

[Engineer54]

No idea what you mean by "L.O.C.", but insulation resistance should be tested with all circuits in parallel (and with the means of Earthing connected) rather than each circuit. If you test circuits individually then you would need to calculate the parallel resistance to see if it complied with the minimum acceptable for BS 7671.

Unless i'm misreading this, Pass that by me Again please!!! lol!!

 

[Risteard]

No probs. I was just saying that the minimum insulation resistance quoted in BS 7671 is for the installation/dis board, and as such if individual readings were taken for different circuits then the effect of the parallel resistances on the overall insulation resistance would need to be considered.

 

[Meggerman]

No probs. I was just saying that the minimum insulation resistance quoted in BS 7671 is for the installation/dis board, and as such if individual readings were taken for different circuits then the effect of the parallel resistances on the overall insulation resistance would need to be considered.

That is incorrect. The insulation resistance values quoted are for individual circuits and that is why you record it as such. If you carry out a global IR then all the circuit resistances are in parallel so will be lower than any individual circuit. If global IR gives you an acceptable level then you need do no more and record all as that value.

 

[davesparks]

No probs. I was just saying that the minimum insulation resistance quoted in BS 7671 is for the installation/dis board, and as such if individual readings were taken for different circuits then the effect of the parallel resistances on the overall insulation resistance would need to be considered.

Think you've hot the wrong end of the stick there!

Insulation resistance is tested for each circuit, if you tried doing one global test on a large installation it will come out with a seriously low value.

 

[Pete999]

What is L.O.C.? Please

 

[Engineer54]

That is incorrect. The insulation resistance values quoted are for individual circuits and that is why you record it as such. If you carry out a global IR then all the circuit resistances are in parallel so will be lower than any individual circuit. If global IR gives you an acceptable level then you need do no more and record all as that value.

Global IR's are fine for initial checking, but are totally meaningless as far as official recording of circuit IR values in say EICR etc.... If anyone accepts a global IR value of an installation on any official recording medium, something somewhere is very Wrong!!

 

[Hawkmoon]

What is L.O.C.? Please

Yes that's that new approved test......Licking Of Conductors

 

[Octopus]

Yes that's that new approved test......Licking Of Conductors

Best done when you've mastered safe isolation!

 

[Badged01]

Global IR's are fine for initial checking, but are totally meaningless as far as official recording of circuit IR values in say EICR etc....

Agreed, but it seems to be the cheap and cheerful way of doing it for your £50 drive-by landlords special. Worst still is they only do a L/N to E test so they don't need to disconnect anything!. Guess if you put down what you're doing in the limitations box ..........

 

[Motty]

Yes that's that new approved test......Licking Of Conductors

Brilliant :clap::clap:

Low Ohm Continuity

All circuits, bar circuit 9, individually, came out as >999 M ohm across L-N, L-CPC, and N-CPC. it was only when I connected them all up to the RCDs, and doing a collective test, that the L-N came out as 0.06 M ohm on RCD 2 / Circuits 5-9.

With circuit 9 disconnected, circuits 5-8 all read >999 m ohm, and Circuit 9 caused the drop. testing circuit 9 individually across its respective conductors gave the following:

L-N = 600 M ohm
L - CPC = >999 M ohm
N - CPC = >999 M ohm

I got in a spin when connecting it up and got that 0.06 reading across L-N as a "collective".

Additionally I can't connect the CCU to the main earth as the supply cable hasn't been installed by the DNO yet.

I did leave an extractor fan connected on circuit 9, although off at the 3-pole isolator. As its a load, I should have really pulled the wires out. It had been a long day and I was panicking a bit when I started this topic ...

Cheers for all the replys, you have all given me fresh hope!

 

[Pete999]

Yes that's that new approved test......Licking Of Conductors

Nice one

 

[Risteard]

That is incorrect. The insulation resistance values quoted are for individual circuits and that is why you record it as such.

I disagree. My understanding is that the reason for the box on the forms next to every circuit was simply to make the form look tidier.

However the Regulations require the minimum insulation resistance of 1 Megohm for the installation/part thereof and not for an individual circuit, therefore the effect of parallel resistances need be considered. In fact, Guidance Note 3 even suggests that insulation resistance should be tested with all circuits in parallel and not individually (except to find the cause of low readings of course).

 

[Hawkmoon]

I disagree. My understanding is that the reason for the box on the forms next to every circuit was simply to make the form look tidier.

However the Regulations require the minimum insulation resistance of 1 Megohm for the installation/part thereof and not for an individual circuit, therefore the effect of parallel resistances need be considered. In fact, Guidance Note 3 even suggests that insulation resistance should be tested with all circuits in parallel and not individually (except to find the cause of low readings of course).

If you get a lowish reading on one circuit, say 40 Meg (low but still acceptable) wouldn't it be appropriate to record it individually on the cert?

 

[Meggerman]

I do more than consider it, I test it!! I don't understand your post, you are saying you disagree but then also saying that GN3 suggests IR tests are done with all circuits in parallel.

 

[Meggerman]

I would not call 40M/Ohms low IR

 

[Hawkmoon]

I would not call 40M/Ohms low IR

It is if new install and all other circs are >200 - I was just making a point that if a circuit is unexpectedly low with no fault then it is right to record it - even if just for future monitoring of insulation deterioration.

 

[Octopus]

I use a Metrel and if I do all 3 cores automatically, I get >30

Low IR is less than 2!

 

[Hawkmoon]

I use a Metrel and if I do all 3 cores automatically, I get >30

Low IR is less than 2!

Yes but my point was mainly about individual circuit IR readings and the merits of separately recording them on a cert...

Your 'low' IR being < 2 Meg is ok but if you had a reading of say 4 Meg would you not consider that at least a code 3 to be investigated? At least record on the cert?

Or am I misunderstanding the point of your post?

 

[Octopus]

Yes but my point was mainly about individual circuit IR readings and the merits of separately recording them on a cert...

Your 'low' IR being < 2 Meg is ok but if you had a reading of say 4 Meg would you not consider that at least a code 3 to be investigated? At least record on the cert?

Or am I misunderstanding the point of your post?

Not a code 3, if I was testing as part of a board change I would comment on it and on an EICR I would recommend a retest in 3-5 years. The BIG issue with IR tests is that so many now seem to be done as L&N to E so the "real picture" isn't disclosed.

 

[Engineer54]

In fact, Guidance Note 3 even suggests that insulation resistance should be tested with all circuits in parallel and not individually (except to find the cause of low readings of course).

Are you sure you are reading correctly?? If that is actually what is being suggested in GN/3 then get a black marker and black it out because it's WRONG!!

Nothing wrong with conducting a global IR, but you DON'T record the global IR value, on official test record sheets, as the value for every circuit on a DB/CU, that is just plain bonkers!!

Do you know the reason for recording circuit IR values on official test reporting sheets??

 

[Risteard]

I don't understand your post, you are saying you disagree but then also saying that GN3 suggests IR tests are done with all circuits in parallel.

Seems like crossed wires then. I agree that circuits should be tested in parallel.

 

[davesparks]

Nonsense!

IR is tested per circuit, not just once over the whole installation.

On a large installation the parallel resistances of many perfectly healthy circuits could result in a fail on an overall test.

 

[Risteard]

IR is tested per circuit, not just once over the whole installation.

On a large installation the parallel resistances of many perfectly healthy circuits could result in a fail on an overall test.

I suggest you read Regulation 612.3.2

 

[davesparks]

1, that regulation says all distribution ccts with final circuits seperately, not the whole installation.

2, that regulation is nonsense!

 

[Risteard]

1, that regulation says all distribution ccts with final circuits seperately, not the whole installation.

As did I (I said the whole "installation or part thereof"). Simple installations with no distribution circuits should be tested as a whole.

And I don't agree that it is nonsense.

 

[Engineer54]

As did I (I said the whole "installation or part thereof"). Simple installations with no distribution circuits should be tested as a whole.

And I don't agree that it is nonsense.

Recording the Global IR for each circuit ''IS'' 110% Nonsense!! As i have stated earlier above, Global IR tests are fine for a general check(s), but meaningless as far as recording purposes go.

A major reason for circuit IR values being recorded is to determine if the IR value or values over the working life time of the circuit is noticeably degrading, not just to test if it's compliant at each time of testing.
It was drummed into us as apprentices, that you are looking for any ongoing downward tendencies, in order that they maybe investigated and rectified before a categoric failure occurs...

But then if you don't come from a maintenance/industrial based background you wouldn't know that. It seems they don't teach you much about the reasoning behind periodical testing these days, even within the dedicated testing and inspection courses such as 2391 or the latest 2394/95...

 

[Risteard]

Of course I understand about degradation, but if all of the circuits in parallel read say 50 Megohms (50 million Ohms), then we know that every circuit is in excess of this and there is nothing to prohibit us from recording this. Lower insulation resistances are often indicative of long runs etc. and not of insulation failure.

 

[davesparks]

But that doesn't indicate anything useful in the records.
If the global test shows 500meg on year one, 450 on year two, 300 on year three then the conclusion is that the whole installation is going to be knackered in a few years.
However if it is done properly and each circuit recorded, the trend on the one circuit can be identified and dealt with.

 

[Engineer54]

Of course I understand about degradation, but if all of the circuits in parallel read say 50 Megohms (50 million Ohms), then we know that every circuit is in excess of this and there is nothing to prohibit us from recording this. Lower insulation resistances are often indicative of long runs etc. and not of insulation failure.

Why did i just know you were going to come back with this kind of, load of old codswallop...

But unless you conducted individual circuit IR tests, you wouldn't know that!! One circuit could, or should i say, will pull down the global IR value, but you would have no idea of this fact. You could/would be ''recording'' an inaccurate value without even knowing about it.

Please tell me this is not your normal method of conducting an installations IR testing??

As a matter of interest, Is there anyone else here, that conducts a global IR test and uses that value, for recording individual circuit IR values on official test record sheets???

 

[Badged01]

....Is there anyone else here, that conducts a global IR test and uses that value, for recording individual circuit IR values on official test record sheets???

Don't know about on here, but I stumble across lots of re-EICR's where that is exactly what has been done before. And usually its just a test between L&N to E!. It's the quick fly-by-night solution ......