Tragic IR fail on EICR

putalightuponce · Apr 25, 2015

Hello all,
I could use some help here:
I was recently asked to do an EICR at a small warehouse/office unit.
The installation was, to put it mildly, less than satisfactory; with multiple C1 and C2 conditions. However, of greatest concern to me were the insulation resistance readings I was getting. I left the job, trying to explain the severity of the condition to the client, but still feel under the circumstances, he is unlikely to take any further action, despite his report been top to bottom in 'further investigation's.
Whereas I'm sure everybody on here has had this at some point (if not daily), it is the specifically the IR fails I am still scratching my head over.
2xsockets in floor boxes, 3xsockets in dado trunking, 3xlighting circuits, and 2xhand dryer spurs were all returning 0Mohms IR L-E, yet none of the associated breakers were tripping or had been problematic in any way. Breakers and board were old (maybe late 70s early 80s), no markings on breakers except obvious ratings, but I'm assuming wiring can't have been of same period unless floor boxes and dado trunking wre popular back then?? The decor and fittings were certainly more recent.
I was pleasently suprised to find a single A4 schedule of circuits reinforced with several layers of tipex and the board had a Ze and PFC reading written in it from 2002, but that was it for records.

Any thoughts???

davesparks · Apr 26, 2015

Found an bit from the 10th edition too

[ElectriciansForums.net] Tragic IR fail on EICR

Knobhead · Apr 26, 2015

The 1/10000[SUP]th[/SUP] FLC is still relevant today on large power circuits. They would fail if the 1MΩ recommendation was applied to many installations.

Almost all of the UK’s LV distribution system would be a miserable fail but its still working.

There’s also the minor problem of just pressing the button and taking the first reading that is shown. A long cable needs a 30 second test at least, if the needle keeps rising keep your finger on the button. Every cable has a capacitive element, it needs to charge before a stable reading can be taken.

We tested a 1.7KV cable on a regular basis. At first the reports were coming back at around 5MΩ which would be in the words of the OP a tragic fail. A 5 minute test came back with 50GΩ (50,000,000,000Ω). The meter could go up to 10TΩ (10,000,000,000,000Ω).
The rule I set was if the test reached 1GΩ in 1 minute then the test was a pass, to extend the test leads to stress on the cable.

The one value for all doesn’t take different voltages in to account. 1MΩ, doesn’t fit them all.