Insulation Testing during a condition report/periodic test.

There may be loads connected L-N that mean that you will stress if you raise one side and not the other to 500VDC for the  test. Traditional  loads like filament lamps and DOL motors are unlikely to care, and in effect the insulation meter is seeing  L and N in parallel anyway, but some modern electronics does not appreciate being held up in the air by one leg only.of course if you disconnecto all loads and fittings, then you can do separated L and N to gnd tests, but that really is a lot of work, compared to switching off at the incomer to remove the connection of the NE short at the substation, plugging in an L-N short and then testing from there to ground and in effect testing the whole site at once.

I agree if you have to break it down by sub circuit to find a problem it rapidly gets very tedious.
Mike

I would add to mapj1's comment that you may think that you have disconnected all loads, but have missed some electronics somewhere - an aerial booster in the loft, for example.

So generally a good idea to test on 250 V initially.

That makes sense, thank you for your response I really appreciate it. 

Thanks for the response, very helpful. 

Ben, the idea is to prevent the test voltage appearing across the load, more especially, load with sensitive electronic components that may be damaged by said voltage. Now consider a neutral to earth connection, perhaps a fault of negligible impedance such that the neutral and earth are connected at the load. Applying the test voltage between phase and earth also applies it across phase and neutral.

Proprietary portable appliance testers automatically short phase and neutral when the IR mode is selected to avoid such a circumstance.

Especially necessary if the test Voltage is 500 Volts. It prevents damage to sensitive electrical equipment.

Z.

A pretest at 250V L + N to E . Then if you feel happy with the result you might do the L to E and the N to E.

Depending upon what you`ve got you might then decide which 500V tests to do.

Or you might just do the 500V L + N to E (personally I do the 250V test first anyway unless it`s a circuit I`ve just installed)

Or we might say that if the R.C.D. isn't tripping off then all is good, and no insulation resistance tests are needed at all!

Z.

I'd be wary of that - the RCD might not trip, even with 29mA of leakage (though real ones usually have thresholds in the low to mid 20mAs) That could represent  a touch under 10k ohms  - which if it is concentrated at a single point is probably quite a serious fault and worthy of investigation. A DC test actually checks the insulation, and is not fooled by capacitance. An RCD 'headroom' test alone cannot distinguish between a dead mouse glowing on the wire resistivity  and a problem, or several nano-farads of capacitance L-E from a long and meandering installation in perfect health, or a room of IT equipment with 'Y' c connected filter capacitors on the mains inputs  to tame the worst excesses of the switch mode power supplies. Both capacitive cases are legitimate earth leakage, but the resistive case is not.

That said, a test of L-N leakage with the clamp meter around L and N looking (RCD-like) for the out of balance current is a good confidence check that there will not be RCD false alarms in the future, though I suggest an investigation limit of 10mA on a 30mA RCD

Insulation tests do have merit, especially on wiring that is either brand new (looking for nails in the wire from the new skirting board or whatever) or in poor condition ( post flooding or buildings with suspect rodent issues come to mind as special cases).
Mike.

I am reminded of the sad Emma Shaw case, where an insulation resistance test would not have picked up a serious fault, but an R.C.D. may have saved a life.

https://www.dailymail.co.uk/news/article-2077015/Young-mother-electrocuted-mopped-water-leaky-boiler-died-catalogue-workmens-errors.html

And...........

www2.theiet.org/.../messageview.cfm

Z.