Question about IR testing related to one on main forum

[newtothis]

http://www.electriciansforums.co.uk...stupid-question-re-insulation-resistance.html

I’ve been taking a great interest in the above post, out in the big pond, rather than take the topic off piste I have 5 stupid questions I‘d like to ask.

I’ve got no real experience, as you all know, I know enough to know there’s no one size fits all approach. I know about factors like age of cabling and extremes of ambient temps having an effect over time.

Q1 Is there a table somewhere which set out characteristics/variables with acceptable IR readings. ?
or is that an experience thing.

All I know >1Mohm is compliant but anything like that low should be investigated.
I know if it was a new installation I’d like it to be to the meter’s max.
Q2 Does that sound right ?

I can’t figure out why I wouldn’t test the circuits individually - If I got 50Mohm globally.
Q3 Hasn’t all the time consuming work been done removing all the vulnerables, lamps etc ?

On the Schedule of test results, if you’ve tested the board as a whole how is that recorded.
Q 4 On a EIC my feeling is it wouldn’t be allowed to put 50 M ohms for each test of each circuit ?
Q5 On a EICR you’d note it was a global result ?

 

[Richard Burns]

I do not remember ever having seen a table of expected insulation resistance results based on characteristics or variables.
The possibilities are very great for minor changes to have a large effect.
For example I was called out to a socket ring tripping an RCD, the insulation resistance measured was 0.006 Mohms.
Once repaired the insulation resistance was >1000MΩ
A 30mA RCD will trip at about 7700Ω or 0.007 MΩ, so the result indicated a slow deterioration with the IR dropping slowly over time until the RCD tripped. The fault was a damp connection allowing line to connect with earth through the water, it would not have tripped until there was enough water to drop the IR down to 7700Ω.
Measuring the circuit at any time before the trip may have given a result between 1000MΩ and 0.008 MΩ depending on how wet the connection was.

Old circuits tend to be below 100MΩ if they are starting to slowly deteriorate, but can still be OK for many years. I have often got low tens of Mohms.
Neons will give about 510kΩ, LED neons about 100kΩ, lower if there are two in the circuit (parallel resistance).
Generally results below say 10MΩ may well be due to connected loads and direct damage to cables, even to earth from filters and so forth.

A new install will generally be near the meter maximum I remember getting >1000 Mohms on all circuits bar one connection that was 996 Mohms, damn:smiley2:!
However I also installed a circuit where it was raining and I got 8 Mohms, I passed it but checked again the following week and it was up over 1000 MΩ as it had dried out.

Even for initial installations it is allowed to do global testing, but it is not something I would recommend, it is best to be sure of each individual circuit.
For initial inspection it is much the best practice to test the installation as fully as possible, so that you can identify any, even slight, faults and give the next person to work on the installation as much information as possible.
For periodic inspections loads are often not removed and the test is global L+N to E; if you are short on time and this passes you can say that it is OK, but gives little information for later inspections. Although in my experience in domestic I have rarely had previous paperwork and even when I have had that paperwork the results do not match with my measurements at all!

For initial inspection the minimum standard is the global test and can be recorded as such, but always try and go one better if you can. Though for installations where one installation person installs everything and then the testing person in the company comes in to check the installation they may not have time to remove all the newly installed kit. This is reality but not quality.

It is always best to indicate the method of measuring IR results on the form; although one value, the same for every circuit and with no Line to Neutral or Earth to Neutral values does give you a clue as to what has happened!

 

[newtothis]

Cheers for the super comprehensive answer, very helpful. I logged in and skim read it this morn at 6ish before work.
Read it again just now,and will give it a further mulling over.

There was a lull at work today and the guvnor gave me the oppotunity to help troubleshoot a problem with the exterior lighting in his garden. Win win He knows I'm interested and he'd get a spare pair of hands.

Lights kept tripping so he disconnected the earth at the Sub CU as a quick solve :whatchutalkingabout
He asked me what test would be helpful and I figured it would be IR.

After checking all the lights/switches we began to break the circuit down. starting at the JB I was almost dissapointed that and we used the continuity function on my fluke voltatge tester to find the prob.
Looking for continuity as something you didn't want, indicating wires that shouldn't be in contact were in contact.
Could I have done the same thing with the IR tester looking for low resistance, that's what I'm asking myself.

Buzzing to have learned from your reply, and the tricks I got shown today.

 

[newtothis]

duplicate post

 

[Richard Burns]

The type of test you would do would depend on the resistance of the fault, if there are wires in contact a low resistance ohmmeter is better as it is more precise and the fault will be in the tenths of ohms range.
If there is a very slight fault in the 1000s of ohms range then an IR tester is easier.
You could also use a multi-meter with a wide range, but this is less conclusive as the test current is so low, even lower than a low resistance ohmmeter, that it may not show a good reading.