Old wife's tail, myths, and other stories we were told as an apprentice, before the days when BS 7671 started.

[ebee]

PS, none of us should undertake actually doing a Zs test without bonds attached (or we would never admit it )

 

[EFLImpudence]

When measuring Ze it is the earthing conductor that should be disconnected.

 

[JohnW2]

Yup, basically , in an ideal world, Ze should be measured without bonds etc.

Quite. Hence my question as to why 'we' don't do that.

If you are wiring a complete installation from scratch or a complete rewire of an existing installation then, hopefully, you should enough skill and knowledge to minimise the risk for a fairly limited time. .... ... When adding a socket, lighting point etc etc then the risk of disconnect all para paths including bonding is difficult to justify, both for safety and the additional costs/upheaval that might bring.

That seems to be the only answer that anyone has been able to produce to my "why not" question above, but is is really particularly valid? There are intrinsic (albeit tiny) 'risks' of various aspects of testing of an electrical installation and a competent person should act in order to mitigate/minimise those (already tiny) potential risks.

Moving from 'safety' to "...additional costs/upheaval" (of disconnecting bonding to measure Zs), that would surely be offset by the "additional cost/upheaval" associated with having to measure Ze if one took the alternative approach (and 'used the equation'), wouldn't it?

No problem. ..... So new/rewired or major works ( and EICRs/PIRs) justifies this risk and the equation holds good.

I think there is some potential confusion in the way you talk about this 'equation' [ which is essentially just saying 'the obvious', namely that (A+B) = A + B ]. Zs = Ze + (R1+R2) is always totally correct (in terms of what Zs would be if parallel paths were to disappear).. A direct measurement of Zs will also be 'totally correct' (within limitations of measurement error) in terms of the situation at the time of measurement, but may under-estimate what Zs would be were parallel paths to disappear (and one can't rely on them not disappearing at some point in time).

However, I remain confused by the apparent contradiction in what you keep writing. You repeatedly say that one should not disconnect bonding to measure Zs (which I agree is what guidelines and common practice), but the only alternative, if one wants a Zs figure that one could rely on if parallel paths were to disappear is by use of 'the equation' [based on Ze and (R1+R2\) ] - yet you seem to also repeatedly say 'negative' things about use of 'the equation'. Can you clarify what you are actually advocating?

Any use of the formula by using Zs then subtracting R1 and R2 to give Ze would be unreliable therefore no use (not much) .

I've never suggested doing that - although, if Zs were measured with bonds disconnected it would give an exactly correct answer (give or take the 'measurement errors').

That`s why we don`t use it (we say it doesn`t work - well actually as an equation does work, after all it is an equation just like any other but we say "doesn`t work" to bring to attention it is unreliable so not much use).

See above. It would 'work' if Zs were measured with bonds disconnected.

As any electrician if they actually remove bonds to confirm Ze then tests Zs before reconnecting bonds on a small addition to an existing circuit (like an extra socket or an extra lighting point) an there will be extremely few.

Quite, but, as above, they could (and I might suggest "should")

Metered tests are always dependant upon the actual accuracy of the meter at best, plus additional field errors on the day, huma errors plus other additional sources of error (such as how steady your source at origin is on that particular minute/hour/day/week/month and say ambient temperature. Any tests you make at the time of Ze, Zs, R1,R2 etc etc are only so good as plus/minus 1000 percent or 1000000 percent or something else, ...

Yes, we know that there will be measurement errors and dependence upon conditions at the time of the measurement, but that has to be accepted and we have no choice but to assume that the measurements are as 'correct' as we can get.

we minimise this to a reasonable extent but do not take extra step of using derived equations.

I really don't think that adding together two resistances/impedances that are in series to get a total really qualifies as "using a derived equation", does it?

It is considered allowable to measure Ze without bonds with an instalation isolated if required and undertaken using robust procedures but is never considered reasonable to increase risks by energising a circuit(s) or parts thereof with bonding not connected because of risks.

I realise that is your view, but it's what I have been questioning. You seem to be assuming that, because of that perceived issue, electricians routinely use 'the other approach' [measuring Ze and (R1+R2) and 'using the equation] to obtain a Zs figure which would reflect the situation if parallel paths were to disappear - but my suspicion/concern/fear is that I suspect that some simply use their MFT to measure Zs with bonding connected - hence producing answers which are 'potentially dangerous'.

Proper testing of Ze without and real world testing of Zs with. Renders the equation not useable - like comparing apples with oranges.

Yes, as I've said, if one measures Zs with bonding connected, one will get a correct answer in relation that that point in time, but one which will probably (perhaps 'dangerously' underestimate what Zs would be if the parallel paths were to disappear - and, as above, I suspect/fear that such is what is commonly done.

Kind Regards, John

 

[JohnW2]

When measuring Ze it is the earthing conductor that should be disconnected.

It is, but the primary purpose of doing that is to disconnect any bonded parallel- paths to earth, which would result in an underestimate of true Ze.

 

[securespark]

We are all told 1966 the earth rules changed, and thanks to our @flameport I have now seen what was written, this was a surprise, as it seems wall lights had needed earths well before that, but they did not fit them to parents house built 1954, but likely the electrician like me in my early years had never seen a copy of the regulations. It was hard enough getting his head around the new ring final circuits.

I have been through the 13th Ed., and all I can find so far is that it says metal conduit must be earthed.

So, if you have a lighting circuit wired using conduit, that conduit must be connected to earth.

I have come across 3 1950s LA houses in my time that had lighting circuits in conduit and the Zs was a very respectable reading.

I will continue to read through the 13th Ed., and will then read through the 12th Ed. to see if I can find the rules on lighting. Sadly, I don't have anything earlier than the 12th.

 

[EFLImpudence]

It is, but the primary purpose of doing that is to disconnect any bonded parallel- paths to earth, which would result in an underestimate of true Ze.

Yes, that is the reason but all anyone has said so far is disconnect the bonding but that is not correct.

It might not be known where all the bonding is connected; not to mention any appliance CPCs that are connected to pipes.


The real purpose is to isolate the earthing conductor which can be done by disconnecting one conductor.

 

[JohnW2]

Yes, that is the reason but all anyone has said so far is disconnect the bonding but that is not correct.

I don't know about "anyone', but I've only been talking about 'disconnecting bonding' when measuring Zs, for which it is correct. One obviously couldn't measure Zs with the earthing conductor disconnected !

It might not be known where all the bonding is connected; not to mention any appliance CPCs that are connected to pipes.

One would obviously have to know "where all bonding was connected" if one wanted to disconnect it ('all') to measure Zs - so, if one knew that, could theoretically do the same for Ze measurement. However, for that it is simpler/foolproff just to disconnect the earthing conductor - something which, as above, one can't do for a Zs measurement.

The real purpose is to isolate the earthing conductor which can be done by disconnecting one conductor.

Quite - as above, the simplest/foolproof approach (but only for Ze measurement)

 

[ebee]

When measuring Ze it is the earthing conductor that should be disconnected.

Can I clarify,
I usually use an MET outside the consumer unit rather than using the CU earth bar as the MET.
But whichever one I disconnect the installation earthing conductor at the CU end so all you have is a bare end of the earth you will be using to make the test.
If a separate MET is used then the bonds are removed from it and all you have are the suppliers E coming to it and my E going from it and the MET is purely a joint to those two pieces for the duration of the test. Nothing else helps it or hinders it.
So we now have the E end as it will be connected to the CU later on.
So L, N, E as it will be once connected/switched on.
That is all you can do to measure what comes in on the supply.

Moving from 'safety' to "...additional costs/upheaval" (of disconnecting bonding to measure Zs), that would surely be offset by the "additional cost/upheaval" associated with having to measure Ze if one took the alternative approach (and 'used the equation'), wouldn't it?

Making sure everybody is away from the property or from any part of the installtion and will not touch anything/pipework for which you the installer would be responsible if injury was suffered is non mean quick task and the more people that atre onsite or likely to be is , sometimes, impossible to control adequately. From experience, no matter how much you tell and explain to everyone, they will immediately go against what you`ve just said, getting everyone out of harms way can be unbelievably impossible, you would not believe it.

 

[ebee]

if Zs were measured with bonds disconnected it would give an exactly correct

Yes, if you actually did that, but could you do it safely and could you guarantee you have actually eliminated all parallel paths besides bonding? No

 

[ebee]

but my suspicion/concern/fear is that I suspect that some simply use their MFT to measure Zs with bonding connected - hence producing answers which are 'potentially dangerous

200% with you on that

 

[JohnW2]

Can I clarify,

That would be helpful

I usually use an MET outside the consumer unit rather than using the CU earth bar as the MET.

So do I.

But whichever one I disconnect the installation earthing conductor at the CU end so all you have is a bare end of the earth you will be using to make the test.

Fair enough.

If a separate MET is used then the bonds are removed from it and all you have are the suppliers E coming to it and my E going from it and the MET is purely a joint to those two pieces for the duration of the test. Nothing else helps it or hinders it.

Again, fair enough - but why don't you do the same as when the MET is within the CU, in which case you again only have to remove one conductor?

So we now have the E end as it will be connected to the CU later on.
So L, N, E as it will be once connected/switched on.
That is all you can do to measure what comes in on the supply.

I'm not fully sure what you are saying.

However, I presume you are talking about 'for Ze measurement' - in which case your plan would presumably be to determine Zs from that [and the (R1+R2) of the circuit ] by 'using the equation' - about which you appear to have negative things to say. I therefore feel the need for a bit more 'clarification'!

Making sure everybody is away from the property or from any part of the installtion and will not touch anything/pipework for which you the installer would be responsible if injury was suffered is non mean quick task and the more people that atre onsite or likely to be is , sometimes, impossible to control adequately. From experience, no matter how much you tell and explain to everyone, they will immediately go against what you`ve just said, getting everyone out of harms way can be unbelievably impossible, you would not believe it.

I can't argue with your theoretical concern but, as I've said, I do wonder whether it is a 'significant risk' If one undertakes a Zs measurement with bonding disconnected, the theoretical hazard does not arise simply because 'someone touches something' but, rather, only if they touch two relevant 'somethings' and (other than in very exceptional circumstances) do that during the interval of a second or three whilst the test is being undertaken.

You might argue that any risk should be avoided - but if (as I believe) it is very common for the alternative to be to measure Zs with bonded connected, then the price one pays for eliminating the (I would say tiny) risk during the very brief testing could be to create a long-lasting potential danger because one had under-estimated Zs of the circuit concerned. I therefore find it fairly hard to believe that a risk-benefit analysis would come down on the side of leaving the bonding connected.

Kind Regards, John

 

[JohnW2]

Yes, if you actually did that, but could you do it safely ....

See my recent post

.... and could you guarantee you have actually eliminated all parallel paths besides bonding? No

Very littlein this world is certain. As you imply, it'spossible that there are some unbonded extranepus-cps which are somehow in continuity with the installation's earthing system.

However, in terms of (the potential danger of) underestimating what Zs would be if parallel paths disappeared, it will always be better/'safer' to measure it with at least all the 'known' bonds' disconnected than to measure with them all still connected.

Kind Regards, John

 

[JohnW2]

200% with you on that

That doesn't surprise me, but do you disagree with my suspicion thst such is probably what the majority of electricians usually do do?

I think I feel a poll coming on

Kind Regards, John7

 

[securespark]

I will continue to read through the 13th Ed., and will then read through the 12th Ed. to see if I can find the rules on lighting. Sadly, I don't have anything earlier than the 12th.

I have now got a copy of the ASEE guide to the 11th Ed. While not the regs, it's a good publication.

The only guidance I can see about earthing WRT lighting seems the same as before.

That if the circuit is run in conduit, that conduit must be earthed.

So, if you run a lighting circuit in twin VIR, it does not require a CPC.

 

[ericmark]

It is not only the earth, RCD testing also seems to be flawed. We were told we test as even a new RCD can fail, and I have had this where the strain of the cables wrapped the case slightly, so clearly must be tested with all cables attached, or it may not work. And the idea was allowed 30% background leakage, so most 30 mA will trip at around 24 mA or more, so if it does not trip at 15 mA likely the background leakage is within limits.

OK, today we have clamp on meters to see background leakage, so we know within the 9 mA limit, but to test a RCBO one needs to know the background leakage to know if tripping within limits, it is rather pointless disconnecting it as the reconnection can affect how they work.

I think it would be interesting to go back to collage and see how the teaching has changed.