Unstable IR reading

Steedy · Oct 15, 2013

Hello,
New to the forum.
Advice please.
Just run a new circuit of 6mm cable from a 32A MCB, on an additional new CU, which is to run a SPA in a back garden. There is a shut off switch installed for the Spa and an additional fused spur with sockets in an IP rated enclosure.
All readings fine, including IR on L-cpc and N-cpc, which were >299.
However IR on L-N is @ 52M ohms and unstable, i.e. moving up and down, which I believe is indicative of water ingestion into a cable somewhere. It has been very wet during the installation, so this is highly possible. I'm aware that 52M ohms is acceptable, although not ideal on a new circuit.
My question is, the SPA is due to be installed in two days time. Would it be DANGEROUS for any reason to install it before getting a perfect IR reading?

Grateful for any advice.

Marvo · Oct 15, 2013

Welcome to the forum Steedy.

What test instrument and at what voltage did you do your IR tests?

not clever · Oct 15, 2013

If you are worried about it, break the circuit down & test in sections.

Marvo · Oct 15, 2013

52Megs is well within any safe limits, it's the 'unstable' bit you've got to decide about. Are you testing just the new cable or is it already terminated at both ends? Maybe you're getting a false reading from a neon in the isolator switch? Low IR between L&N could be a small load or even a surge arrestor connected somewhere. If it was moisture I'd expect to see a similar reading from live and neutral to the CPC.

Steedy · Oct 16, 2013

Thanks Gents,
I'm using a Meggar 1553 at 500v for the IR test.
It would be really difficult to break the circuit down as the complicated termination end I put together in my shed and transported it as a whole to site.
If it helps, the circuit has 8m of 6mm internal cable onto another 12m of 6mm SA, when it goes outside. This runs into the cut out switch for the Spa, with 5m of 6mm SA tails running off it loose. From the switch it goes to an IP65 13A switch, with a quadruple socket running off it, housed in an IP rated box. The socket was disconnected for the test as it has a power indication light on it. All materials brand new.
So, it's the fact that the reading is unstable that's potentially dangerous?