Insulation Resistance??

[deb152000]

HI All,

I am really confused!
When checking the insulation resistance do the MCB's need to be open or closed?

I have had a look in my books but cant find the answer.

Thanks in advance

Debbie

 

[ezzzekiel]

you are testing the wiring from the outgoing side of mcb so technically it would not matter, however if the busbar is live then you gonna hit a problem - so in answer off.

if testing all from main tails then you would need the mcb on so test is applied over cables

 

[Adam W]

The way I learned it is you have the main switch off, turn on the MCBs for the ccts you want to test and put the test clip on the busbar.

 

[ezzzekiel]

there are various methods to achieve the same thing, but to follow Gn3, main switch and breakers off test between outgoing side of mcb to l/e l/n n/e

 

[Risteard]

Closed, because an insulation resistance test should be applied to the entire installation/distribution board. Obviously main switch locked off as it is a dead test.

 

[deb152000]

OK thank you all, but clearer now

 

[SEC-Tom]

Took my AM2 the other day and did it with the MCBs closed but I wasn't getting >200ohms on a new install was around 149 ohms. I did ask the assessor if he wanted me to test each individual circuit to find the problem and he said I wouldn't bother as it still complies I think I just proved that I was competent and he couldn't be bothered to watch me do it all.

Would you all personally test each circuit to find the problem?

 

[Risteard]

149 million ohms for an entire installation isn't really a "problem".

 

[SEC-Tom]

149 million ohms for an entire installation isn't really a "problem".

I did mean to say 149 M ohms which obviously is above the minimum of 1 M ohm and the 2 M ohm that should be investigated but surely it suggests that there is something not quite right on one of the circuits?

 

[Risteard]

I did mean to say 149 M ohms which obviously is above the minimum of 1 M ohm and the 2 M ohm that should be investigated but surely it suggests that there is something not quite right on one of the circuits?

Not necessarily. What fault does it suggest?

 

[SEC-Tom]

Not necessarily. What fault does it suggest?

As I haven't done much testing in the past iv only ever seen >200 M ohms so this threw me a little when in my exam it showed up at 149, I assumed it meant that there was some kind of minute leakage or trapped cable somewhere. I am not experianced enough so if I sound silly its unintentional.

 

[ezzzekiel]

149 Mohms is still a very high reading and nothing to be concerned about, a lot of people when just starting out testing make the same mistake and expect to see a reading of >299 every time

readings will always vary and depending on age of installation etc will depend on how you react to it,

for example an old instalation showing 20- 30 Mohms would be recorded but not concerned about, however if this was a new build or rewire etc then id be wondering why the values are low

 

[rocker]

Do you mean 1.49 megohms or 149 megohms. I would investigate 1.49, but 149 is probably one of the best readings you're going to get on a non-new build. Nothing to worry about.

Going back to the first question, if you're doing it properly the line should be D/C'd from the MCB altogether. That's the way I was shown, anyway. We all put a proble on the line screw and a clip on the relevant bar, but to get accurate readings you'd want to be clipping to the cables themselves. And you want the MCBs open i.e. turned off.

 

[Risteard]

Going back to the first question, if you're doing it properly the line should be D/C'd from the MCB altogether. That's the way I was shown, anyway. We all put a proble on the line screw and a clip on the relevant bar, but to get accurate readings you'd want to be clipping to the cables themselves. And you want the MCBs open i.e. turned off.

Not correct. The insulation resistance of the entire installation or distribution board should be tested at the one time.

If you were to test circuits individually then you would have to calculate the cumulative effect as the limits in the Regulations state that they are for the installation as opposed to individual circuits.

 

[robskiing]

Im with rocker on this one doesnt take much more time to megger each circuit individually on the dead testing procedure before reconnecting .

 

[Risteard]

Im with rocker on this one doesnt take much more time to megger each circuit individually on the dead testing procedure before reconnecting .

The problem is it doesn't address what the test is designed to confirm, which is the cumulative effect.

1/Rt = 1/R1 + 1/R2 +1/R3 etc.

Rt = 1/(1/R1 + 1/R2 + 1/R3 etc.)

 

[xraytek]

Here is a extract from Guidance Note 3 if that helps.

2.7.7 Insulation Resistance:
"Simple installations that contain no distribution circuits should be tested as a whole"

 

[andekoch]

149 Mohms is still a very high reading and nothing to be concerned about, a lot of people when just starting out testing make the same mistake and expect to see a reading of >299 every time

readings will always vary and depending on age of installation etc will depend on how you react to it,

for example an old instalation showing 20- 30 Mohms would be recorded but not concerned about, however if this was a new build or rewire etc then id be wondering why the values are low

Agreed. did an insul resistance test on a new install and got a 40M ohm on the downstairs ring main. Investigated and found rodents had munched their way though the cable. Told the customer his new wiring needed to be replaced and the rodents 'evicted'.

 

[rocker]

And if we aren't testing the full installation, for a minor works cert for example? It would be unreasonable to power down a whole house for the sake of changing a lighting pendant, but if we measure at the bars for L2 and LN with a live installation then we are going to get inaccurate readings, no? Also. if we are supposed to be testing the whole installation, then once again again when we are changing that lighting pendant we will be siging our names to a cert that says the whole installation is safe.

If I'm testing the whole installation, I will do each circ seperately, then test bar to bar CPC to neutral, and test from bars to lines on each circ to test for faults between different circs, but on most minor jobs, I wouldn't.

 

[ezzzekiel]

its also advisable to disconnect the circuits neutral or earth

 

[rocker]

Yea, that's what I'm saying. But if, as is being suggested here, we should not be testing individual circuits and should be testing at the bars, then doing this will mean that we are testing every other circuit, and not the one we have worked on. If I am testing an individual circuit, I:
open breaker
D/C line
Find N and CPC for circuit and D/C those
Test L1 and L2
Test L2 and LN
Test LN and L1
Look for missed neons
Test Ln and L1
Reconnect and enliven.

If we left all the CPCs in the earth bar, and all the neutrals in the neutral bar, and the rest of the installation is turned on, then we are going to get very duff readings from where the Neutral and Earth are connected somewhere on the other side of the main isolator (PME, anyone?).

These are my thoughts, anyway.

 

[Adam W]

Surely the main isolator (DP) would be off though, otherwise you'd be trying to do IR live?

 

[rocker]

No because the breaker would be open (off) and all the wires would be disconnected from the board altogether. All the other circs will be live, but the one I'm working on will be deader than dead.

 

[keniff]

its also advisable to disconnect the circuits neutral or earth

i have always been told to leave the earth connected as this would pick up any problems related to structural earthing/pipes etc

 

[Risteard]

i have always been told to leave the earth connected as this would pick up any problems related to structural earthing/pipes etc

It should be connected to the means of earthing.

 

[Jimmy Boy]

Are RCD's/RCBO's able to with stand a 500V test ? I refer you to GN3 test 2 on page 38 which states that simple installations should be tested as a whole, and IF a reading of less than 2M ohm is recorded then the circuits should be tested individually

 

[Adam W]

No because the breaker would be open (off) and all the wires would be disconnected from the board altogether. All the other circs will be live, but the one I'm working on will be deader than dead.

I'd always been told not to work on a board live, even taking the cover off, because it exposes live parts.
Any parallel paths should be taken into consideration because they will be present when the circuit is used.

 

[Benny_Boy]

IR is a dead test, surely??? I wouldn't fancy scratching around in a consumer unit with the busbar exposed if the thing wasn't proved dead!

 

[rocker]

The busbar wouldn't be exposed. The cables for the circuit would be disconnected as I keep saying.
Surely if this is the wrong way to test, there is no need to have access to the screw terminals in the top of an MCB, so why aren't they put behind the shield that covers the busbar.

Are you guys seriously telling me that if you replace a lighting pendant that you power down the whole installation? Or do you just not bother to test/certify?

And as someone else has said, I would be very dubious about putting 500V through an RCBO. I understand testing in many circumstances from the neutral and earth bars, but I have never seen or heard of testing from the line busbar in a consumer unit. It just seems like madness to me. What do you do if you find a fault, just say 'oh, there's a fault somewhere on the system. Oh well.'?

 

[rocker]

It should be connected to the means of earthing.

I'm not being funny, but are you confusing an insulation resistance test with an earth fault loop impedance test?

 

[Des 56]

I but surely it suggests that there is something not quite right on one of the circuits?

Best to ask yourself why we do the insulation test

eg
What are the effects of heat ,age, mechanical damage etc and what they mean to the results of the insulation test that is being done on your installaton

 

[Jimmy Boy]

Does anyone/Has anyone used test method 2 in GN3 as stated in regs book 612.3, where to join the 2 Line conductors together and test to earth ? this removes the need to disconnect sensitive equipment and must be quite useful if you have a circuit full of fluorescent lights or similar.

 

[semaphore]

Steve,
"Does anyone/Has anyone used test method 2 in GN3 as stated in regs book 612.3, where to join the 2 Line conductors together and test to earth ? this removes the need to disconnect sensitive equipment and must be quite useful if you have a circuit full of fluorescent lights or similar."

We use this method virtually all the time. I'ts very rare we can disconnect equipment.
I think you mean join the line(s) and neutral together and IR test between the live conductors (joined together) and Earth.

 

[Jimmy Boy]

Thats what I meant yes I was calling the Neutral a Line conductor I should have said 'Live' conductors

 

[Risteard]

I'm not being funny, but are you confusing an insulation resistance test with an earth fault loop impedance test?

No, I'm not.

Insulation Resistance should be performed with the system de-energised and the means of earthing attached.

 

[Risteard]

What do you do if you find a fault, just say 'oh, there's a fault somewhere on the system. Oh well.'?

You open circuit breakers or remove neutrals until you find the faulty circuit. That is why it says that below 2 Megohms requires further investigation. This is the further investigation.

 

[rocker]

You open circuit breakers or remove neutrals until you find the faulty circuit. That is why it says that below 2 Megohms requires further investigation. This is the further investigation.

Which, for the billionth time, is what I do to start with, because, for the billionth time, there is no need to test the full installation if you are only working on one circuit.

 

[rocker]

Just realised my last post comes off as really chippy, which was not what it meant to sound like. Just find it odd that we've all been taught in different ways.

 

[Risteard]

Which, for the billionth time, is what I do to start with, because, for the billionth time, there is no need to test the full installation if you are only working on one circuit.

I'm not talking about working on an individual circuit.

I am pointing out that insulation resistance is cumulative.

And no need to apologise.