Commercial EICR strategy?

There is no 'hard and fast' limit for what leakage is OK - it may be made higher if there are motors with inverter controls or lots of PCs with switch mode power supplies - normally it is the interference filters on electronic equipment that have some capacitance L-E (and N-E) that gives rise to the largest leakage component. Next after that are metal clad heating elements in cookers and kettles and so on.

As first stab at what is reasonable, expecting no more than a mA per amp or two of circuit rating  is a good start - so a 30mA RCD covering a 30A circuit might be leaking up to 10mA or so and not troubled. For a 100A sub-main, perhaps 30mA of leakage, and so a 100mA RCD if it was a TT supply.

However, the type of loads connected could affect this a lot - a single PC of the metal cased kind may add 3mA of leakage on its own, as may the electronics in an induction hob. Indeed sites wite many PCs (or many induction  hobs I suppose -  but less likely) benefit from being wired as a larger number of under loaded circuits - like small area rings and radials, so that the pre-load on any RCD is not too great, and the losses if it trips are minimised.

Mike.

Check tightness of terminals in db's and sampled accessories

If the suggestion is to do that live, I'd be very wary. Insulated tools might be fine if all goes well, but it only needs an unexpected failure - like the heat of a screw shearing off - to put you in a very sticky situation (remove the screwdriver and the broken part might fall onto live busbars, but you can't sort it out without moving...)


I'd ask why they want an EICR - likely it's not just because they like collecting pretty sheets of paper - but do actually need to prove properly to someone else that their system is safe, and something full of ifs and buts and gaps and limitations won't really do. I suspect that in the end the answer will be to arrange an organized shutdown (likely out of hours) and do a proper job.


   - Andy.