Insulation resistance results weird on actuator circuits

[alexcarnes]

Interested in people's thoughts/experience here. I've just been testing out the insulation resistance on a 20 year old surface water pollution control facility. It has a large number of penstocks with Auma actuators.

Here's the puzzle: on a few of them, the resistance - although passable - is conspicuously low at 3-4MR in one case, between phases, but absolutely fine phase to earth. The cables are SWA four core, with CPC and the armour connected to CPC both ends.

I don't see how the insulation resistance can measure 3-15 MR between phases, but several hundred MR or even GR phase-earth in the same cable? Any thoughts? The only thing I can think is if an overcurrent occurred at some point and heated up the phase conductors and left the CPC insulation relatively unscathed? All the actuators are healthy.

The actuators were complete disconnected during the test, of course, so this is purely cable insulation, although there is a plug-type connector on one end that would be onerous to remove.

Interested if people with more experience than me see this all the time or is it odd?

 

[KungFuTurkeyLeg]

My guess would be moisture at the plug end. Not necessarily wet, but vapour inside the plug. Or if it's a really old cable, could be starting to weather and crack?

 

[Marvo]

It is a bit odd. It's not an essential service that's got a high impedance IT earthing arrangement for first fault tolerance is it? I'd suggest just accurately record all the IR test results and monitor their trends over time.

 

[alexcarnes]

My guess would be moisture at the plug end. Not necessarily wet, but vapour inside the plug. Or if it's a really old cable, could be starting to weather and crack?

On this particular site, when insulation resistance has tested a bit lower that expected - say 50MR - it's always turned out to be due to a rotary isolator in the field, enclosed in a cheap plastic box. As you say, suspect moisture ingress. These are decent quality metal plugs though and everything appears to be watertight. I'm going to have to dismantle a couple of the suspect ones and test literally cable only with nothing whatsoever attached to it except fresh air!

 

[alexcarnes]

It is a bit odd. It's not an essential service that's got a high impedance IT earthing arrangement for first fault tolerance is it? I'd suggest just accurately record all the IR test results and monitor their trends over time.

Thanks, I was hoping someone could confirm whether it was odd or not. I've done a fair bit of insulation testing but only on a small number of sites of similar vintage. No nothing weird and wonderful about the earthing, just a few with weird readings. Unfortunately this is (as far as I know) their first IR test in 20 years so no historical data. I presume they were tested at commissioning but I haven't seen the records.