Insulation resistance problems

[newfutile]

During an insulation resistance test IR at an EICR I measured 0.0014 MegOhms from line to earth.
When I disconnected the neutral from the neutral bar it's 116 MegOhms lowest (on the outside bollards).

However measuring from disconnected neutrals to earth all read over 200 MegOhms.

So it does not comply as measured but complies with neutral disconected.

Does anyone have an idea why?

 

[westward10]

You have to disconnect the neutral to do an IR test or it will be flat down.

 

[newfutile]

Why if the rcd and the main switch are both off,I usually conduct global IR line to earth and neutral to earth and I only start to dismantle if high reading are observed

 

[Marvo]

What voltage were you using for the IR test? Maybe the neutral was being pulled down to earth via surge arrestors somewhere on another circuit.

 

[Lucien Nunes]

If I understand you right, every circuit individually achieves over 200MΩ but when the neutrals are commoned, one circuit reads 1.4kΩ L-E? Do we know what the aggregate N-E reading was at that moment? You mention bollards; presumably this means permanently-connected lighting loads (on which we preferably wouldn't be testing L-E and N-E separately.) If there was any kind of N-E fault anywhere on any circuit, commoning the neutrals with it will equally affect the line on any circuit with load. I would be wondering what if anything was disturbed that created an intermittent low IR on the neutral bar, even if it only happened to manifest while testing the line.