Zs taken by live test and r1 r2 calculated by deducting Ze or Zdb

Ro said:

This regulation is in Chapter 64 Initial Verification.

Yes, Regulation 651.2 refers to the 'appropriate tests and measurements from Chapter 64'.

I understand there are circumstances where it's difficult to carry out EICR without removing power, but there are dangers involved of carrying out live tests on circuits where you don't know there's continuity to the cpc ... so this should be an absolute last resort and backed up by some very good, documented, reasons, and a documented risk assessment. EAWR Regulation 8 makes earthing (where it's needed for safety) an absolute requirement, so if you or anyone else gets a shock because of lack of earthing (such as broken cpc or loose/ineffective cpc connection) whilst carrying out loop tests or RCD tests, it's something that's quite difficult to defend against if the HSE get involved.

It's also really difficult to carry out a comprehensive inspection with the installation 'live' ... Regulation 14 of EAWR is relevant also.

Ro said:

if on an EICR you are doing live testing and the cert is asking for one or the other r1r2 or r2, what is the difference?

Then the industry practice would be to record 'LIM' in the R2 and R1+R2 columns because you didn't do those (continuity) tests ... and explain the reason for the limitations in the relevant section of the EICR form.

Ro said:

I'm asking why, if we can do ZDB and r1r2 to calculate Zs why not back calculate. I know doing two tests would give you different values but I'm asking if you done ZS-ZDB to put r1r2 in, why is this frowned upon if you are keeping your circuits connected to the earth bar when doing r1r2? On an EICR it asks for reading in the box either r1r2 or r2.

See above. If you didn't do those continuity tests, there isn't a result.

Is the only reason you do not calculate r1 r2 by deducting Ze/Zdb from Zs is that it is down to parallel paths. The reason I am asking is, when the apprentices are doing their trade test, they are encouraged to short the circuit they are working on to the earth bar via a crocodile clip to get r1 r2. Surely this is picking up parallel paths also? Or is there is another reason?

So no one can give me a definitive answer to above. If you are not meant to live test why do you have a meter that does. As asked why is it ok to do live ZE but not Live Zs? Surely if you do a live test at end of line you are proving continuity on the cpc? Or am I missing something?

Ro said:

Is the only reason you do not calculate r1 r2 by deducting Ze/Zdb from Zs is that it is down to parallel paths.

No, I think that GK has explained it earlier today.

You are checking for CPC continuity, i.e. R₂ is let's say, around 0.5 Ω (obviously, depending upon the length of the circuit). You can do this one of two ways. First, by connecting your meter to both ends of the CPC using a 'wander lead', you measure R₂ directly. The second method is easier: join line and CPC in the DB and measure between them at the far point. This gives you R₁ + R₂. You put one or the other on your generic schedule.

At this stage, you have yet to energise the installation, so you are working safely. So that is the main reason for obtaining R₁ + R₂ that way.

If there are parallel paths, they may be included in your R₂ value. You can reduce the chance of that happening by joining line and CPC before you put the CPC in the earth bar.

Z_e is one of the first live tests, which you do at the intake. Having measured that, and knowing R₁ + R₂ for all of your circuits, you can calculate all the values of Z_s and check that against observed values.

It is all about working as safely as possible.

Surely this is picking up parallel paths also?

Yes and no. Parallel paths can occur in a number of different ways - yes ideally you'd measure R1+R2 (or R2) without any parallel paths at all - but that very difficult in practice. Not having the c.p.c. connected to the earth might eliminate some of the parallel paths, but other may well remain. On the other hand, testing just the final circuit bit of the loop separately from the supply side can reduce the effect of some parallel paths, even if the c.p.c.s remain connected.

A while back people were trained to disconnect both L and c.p.c. and connect them together to do a R1+R2 test ... but that leaves the single most important joint in the circuit untested ... so whatever's suggested there's always a downside. The more you get into the regs the more you realise they're based less on perfect logic and more on compromise and weighing one disadvantage against another.

   - Andy.

BTW - I think you mean R1 and R2 - rather than r1 and r2 ... upper/lower case is significant - the latter mean something rather different (end-to-end continuity of a ring).

Ro said:

Surely this is picking up parallel paths also?

So, for an EICR, are we saying that, as it includes parallel paths, it's a more effective assessment of the real-world conditions than the original certification? Is that an issue? Certainly for initial verification removal or parallel paths gives the best result ... but as you've said in another way, the real-world conditions are far more favourable to our protection than the worst-case envisaged by Chapter 41 of BS 7671 ?

This thread has gone a way I never thought possible, I'm now getting pulled up for grammar......... LOL..... Can I refer back to the original question...... and take Initial Verification out..... If we can do live testing ZE/ZDB and ZS, why is it frowned upon to back calculate to put in R1 R2 (hope I got it correct this time)? This is all I'm asking nothing more? As I have already stated that the FICA asks them to use a crocodile clip from the circuit they are testing on the earth bar so you are using the earth for the whole installation. I AM ASKING WHAT IS THE DIFFERENCE FROM DOING IT THAT WAY OR LIVE TESTING AND BACK CALCULATING. NOTHING MORE I AM ASKING WHY???????????????????

Ro, have we not answered your question? If something is unclear, please say so in order that we may try again.

AS ABOVE

I do feel that the whole thread has ended up as being 'at cross purposes'.

On the one side we have technical explanations about the quality of the different approaches.

On the other side we have (or want) administrative explanations needed for the apprentice training.

We had two approaches (to taking a measurement) that appear to give 'different' results, with a 'correction' allowed one way, but not, apparently, the other way. 

It's the "administrative" 'excuse' for the 'not the other way' that's required.

If I read the threads correctly, it is because we shouldn't make things look 'better' than measured, plus other factors based on the different ways the tests are set-up (short at the board vs at the far end, etc).

And we are looking for a solid apprentice "yes/no" level clarity. No wishy washy qualitative senior engineer stuff .  Is that a fair assessment?

Well done Philip, that is exactly what I'm looking for, an idiot proof answer to the original question which, going by this thread is very hard to get.