Minimum IR values

The total insulation resistance value  is always lower than that of  the lowest circuit.

Thank for the reply! I’m aware of that I was just wondering whether the minimum value of 1Mohm is applied to individual final circuits or DB/CU with all final circuit connected in parallel together.

1 over Rt = 1 over R1 + 1 over R2 and so on up to ten.

https://youtu.be/UcSv5Xm8BMc

The honest answer is that 1 megohm is chosen as an easy number to remember.  Really any single figure is a bit of a fudge.

Any  PVC / polyethylene type insulation that is dry and in good condition should measure hundreds of megohms over the 50-100m cable length of a typical final circuit. So for that situation the 1Meg limit is set very low.

However for mineral insulated cables like pyrotenax, and loads like heating elements with ceramic substrates that are slightly porous, a happy reading may be a few hundred k ohms per metre, and for these you may end up having to warm things up and re-test to fiddle a pass, even though nothing is really wrong.

In terms of actual damage, it rather depends if all the leakage is in one place, and heating a small volume to incineration, or spread out over hundreds of metres and raising the temperature by fractions of a degree.

In terms of inspections, test the whole lot from the main double pole switch and if that passes, than all sub-circuits also pass.

If it fails sub-divide, if the sub-circuits pass, that is also OK. IF not then divide until you isolate a defective fitting or fixture or section of cable.

If you actually get a reading really close to 1Meg, on either side you could toss a coin to do something or nothing, but more professionally you could look at the circuit and decide if in the circumstances for that installation, cable type, dampness of location and what it is supplying, if it is a believable no fault reading, or it was worth investigating further or not.

Mike

Measuring the insulation resistance of the entire installation at the main switch it the preferred and easiest method, I have not done the calculations for a long time.

Typical at home or in a large factory?

I venture to suggest that 643.3.2 is not worded helpfully. If the whole board test is > 1MΩ, all is well. If not, check each circuit. If any one is < 1MΩ, that will drop the overall resistance below 1 MΩ because the resistance of a number of circuits in parallel (as Sparkingchip points out) is always less than the lowest one. Then it becomes a matter of either look more closely or FI depending upon what has been agreed at the outset.

So if the whole board test is > 1MΩ, all is well; but if it is not, it does not mean that there is a problem.

HTH.

mapj1 said:

Any  PVC / polyethylene type insulation that is dry and in good condition should measure hundreds of megohms over the 50-100m cable length of a typical final circuit. So for that the 1Meg limit is set very low.

Thanks for the replies everyone it’s greatly appreciated

The only thing slightly confusing is that the distribution circuit supplying all of the final circuits should be tested with all the final circuits connected according to 643.3.2, which could result in all the final circuits being satisfactory but the distribution circuit itself failing due to the parallel resistance of all circuits giving a reading <1Mohm. I hope that makes sense.

OCB99, you will note that the minimum stipulation is in chapter 64 which is concerned with initial verification. The test at that stage is concerned with the fixed wiring only and should not include the equipment connected to it. One needs to be particularly opportune to be able to perform such a test during the erection of an installation. For EICR, it is almost impossible to exclude connected equipment in even simple installations. With respect to periodic inspection and testing, one might take the view that if it is felt IR tests are worth a punt, a good result is likely good but a poor result is not necessarily bad. 
I have a TPN board in the training centre. A number of circuits have RCBOs with the loss of neutral white leads. TNCS system. The board serves the workshop area. Global tests conducted by me reveal 0Mohms and nearly every individual circuit indicates very low levels. Strange really because the EICR conducted several months ago indicated +999Mohms for every circuit! Time we acknowledged IR tests are of limited value for EICR rather than have fellas  fabricate results such that the veracity of the whole report is called into question!

It is a somewhat muddled situation. In practice these days often each final circuit is tested individually - partly because RCDs (both RCCBs and more especially RCBOs with an functional earth connection) tend to mess up global tests, and partly because the model forms seem to require per-circuit results to be recorded. But as you say the regulation itself suggests a test should be done globally on a DB (along with it supply circuit).

As Mike says, it's all a bit arbitrary - if you look back over previous regulations there have been all sort of different approaches - with acceptable values varying from one edition to another and the basis of the test varying between and entire installation, through per-DB, to dividing the result by the number of points served. The general gist has been though that large values (tens or hundrends of MegOhms) are good, very small values (kilo Ohms) usually bad.

The other gotcha is that the tables are specified only for initial verification of new work ... for existing installations there's room for more engineering judgement - which can be handy for some situations (like an old damp Church full of old MICC cableing).

   - Andy.