Social Housing EICR Contract

AJJewsbury:

If we looking at the best you can do in 1.5 hours then....

I'd forget the R1+R2 tests on radials - they're primarily there to ensure the circuit is safe to energise - but your circuits are already energised - so just do Zs tests - at the very least at each end point (of which there could be more than one per circuit if things branch). There's also a safety advantage in not dismantling things in order to test them (and so re-assembling them after the test, which the consequence that what's put into service isn't quite what was tested). Comparing readings with Zdb should give you an idea if anything is out of the ordinary (although the difference shouldn't really be taken to be a R1+R2 value). Don't forget doing a 'long lead' R1 test if that would be a more convenient way of checking c.p.c. continuity and give an indication of likely Zs. Say if you do a Zs loop test at a ceiling rose (to SL rarther than the loop terminal) and a long lead R1 test to the corresponding switch fixing screws (either from the MET or from the PE terminal in the ceiling rose) and Zs plus you R1 results is still lower than the max permitted Zs, they you can be pretty confident that Zs inside the switch is going to be absolutely fine. There's no need to enter a R1+R2 value on the form if you've already proved c.p.c. continuity and overall Zs by other means.

Continuity on rings I probably wouldn't skip - but if the CU is a mess then do it at a convenient double socket instead as that's likely to be a lot quicker and easier. Parallel paths to the c.p.c. often give Zs or R1+R2 values lower than expected from r1, r2 and rn values - you can't really deduce anything from that fact alone. Different cable types (e.g. one section in 4mm2 rather than 2.5mm2) can also upset any assumptions of R1 (or Rn) to R2 ratios - so again don't get too precious about that kind of thing. Fundamantally as long as Zs is within limits, then ADS should work OK and the circuit should be safe from a shock point of view (at least for earth faults).


Insulation tests can be done once per CU rather than on each individual final circuit (despite the model forms, BS 7671 doesn't actually require per-circuit insulation tests). L+N to PE will reveal the vast majoriy of faults, especially when wired in T&E, so that'll usually suffice. Yes, some electronic equipment has components L/N to PE (usually capacitors for EMI filtering purposes) which can mess up insulation test results (including some RCBOs with a white wire of course). Try the test at 250V first (as that shouldn't damage anything) and only if it's low start hunting for any hidden items. Once it's clear, try the test proper at 500V. In terms of form filling, if you know that the entire installation has an insulation resistance of a certain value, then you can be sure that each circuit will have a insulation resistance of at least that - mathematically it's impossible for it to be lower - so you can correctly write "≥ (your whole CU value)" for each circuit.


Don't skip on the Mk1 eyeball (or nose) tests - they'll probably lead you to problems quicker than meter based tests.


Try to cover yourself in the 'limitations' section of the report - explicity agreed with the person ordering the work of course - and make it part of your template form so you don't have to add it to each report manually.


   - Andy.

Thanks Andy. Ive been moving more towards the R2 method but habit keeps taking me back to what ive previously done so much. I think the R2 wander lead is going to be the staple.


Very happy to learn connecting the R2 lead to a screw terminal of wiring accessory - Im actually excited to witness the potential time saving there!


In the time I have im inclined to do the end to end tests on the ring and provided all are ok, do a ZS and thats it. If any of the results are not good then an individual I.R test and possibly split the ring in to two radials.


70% of the "ring" circuits I have tested so far have been C2. New build is the worst.


Going to try making a scratch on the cooker casing to get an r2 all the way through as I agree regarding dismantling - Old equipment often doesn't tolerate beinng dismantled but was wrking perfectly before trying to gain access. New equipment installed by apprentices - Grub screws rounded or broken, socket back boxes cross threaded or destroyed etc etc Failing that a test on the screw terminal at the flex outlet


Thanks for your reply some tips in there to take forward


 

AJJewsbury:

If we looking at the best you can do in 1.5 hours then....

I'd forget the R1+R2 tests on radials - they're primarily there to ensure the circuit is safe to energise - but your circuits are already energised - so just do Zs tests - at the very least at each end point (of which there could be more than one per circuit if things branch). There's also a safety advantage in not dismantling things in order to test them (and so re-assembling them after the test, which the consequence that what's put into service isn't quite what was tested). Comparing readings with Zdb should give you an idea if anything is out of the ordinary (although the difference shouldn't really be taken to be a R1+R2 value). Don't forget doing a 'long lead' R1 test if that would be a more convenient way of checking c.p.c. continuity and give an indication of likely Zs. Say if you do a Zs loop test at a ceiling rose (to SL rarther than the loop terminal) and a long lead R1 test to the corresponding switch fixing screws (either from the MET or from the PE terminal in the ceiling rose) and Zs plus you R1 results is still lower than the max permitted Zs, they you can be pretty confident that Zs inside the switch is going to be absolutely fine. There's no need to enter a R1+R2 value on the form if you've already proved c.p.c. continuity and overall Zs by other means.

Continuity on rings I probably wouldn't skip - but if the CU is a mess then do it at a convenient double socket instead as that's likely to be a lot quicker and easier. Parallel paths to the c.p.c. often give Zs or R1+R2 values lower than expected from r1, r2 and rn values - you can't really deduce anything from that fact alone. Different cable types (e.g. one section in 4mm2 rather than 2.5mm2) can also upset any assumptions of R1 (or Rn) to R2 ratios - so again don't get too precious about that kind of thing. Fundamantally as long as Zs is within limits, then ADS should work OK and the circuit should be safe from a shock point of view (at least for earth faults).


Insulation tests can be done once per CU rather than on each individual final circuit (despite the model forms, BS 7671 doesn't actually require per-circuit insulation tests). L+N to PE will reveal the vast majoriy of faults, especially when wired in T&E, so that'll usually suffice. Yes, some electronic equipment has components L/N to PE (usually capacitors for EMI filtering purposes) which can mess up insulation test results (including some RCBOs with a white wire of course). Try the test at 250V first (as that shouldn't damage anything) and only if it's low start hunting for any hidden items. Once it's clear, try the test proper at 500V. In terms of form filling, if you know that the entire installation has an insulation resistance of a certain value, then you can be sure that each circuit will have a insulation resistance of at least that - mathematically it's impossible for it to be lower - so you can correctly write "≥ (your whole CU value)" for each circuit.


Don't skip on the Mk1 eyeball (or nose) tests - they'll probably lead you to problems quicker than meter based tests.


Try to cover yourself in the 'limitations' section of the report - explicity agreed with the person ordering the work of course - and make it part of your template form so you don't have to add it to each report manually.


   - Andy.

Thanks Andy. Ive been moving more towards the R2 method but habit keeps taking me back to what ive previously done so much. I think the R2 wander lead is going to be the staple.


Very happy to learn connecting the R2 lead to a screw terminal of wiring accessory - Im actually excited to witness the potential time saving there!


In the time I have im inclined to do the end to end tests on the ring and provided all are ok, do a ZS and thats it. If any of the results are not good then an individual I.R test and possibly split the ring in to two radials.


70% of the "ring" circuits I have tested so far have been C2. New build is the worst.


Going to try making a scratch on the cooker casing to get an r2 all the way through as I agree regarding dismantling - Old equipment often doesn't tolerate beinng dismantled but was wrking perfectly before trying to gain access. New equipment installed by apprentices - Grub screws rounded or broken, socket back boxes cross threaded or destroyed etc etc Failing that a test on the screw terminal at the flex outlet


Thanks for your reply some tips in there to take forward