Extending the ring vs spur

[John D v2.0]

... I would say that EFLI (aka Zs) was the one that is "easy" (at least, for someone with the appropriate equipment)

Yes agreed but in this context i think the op is more likely to have an idea of circuit length than have the appropriate equipment to measure zs.
Also you can measure zs directly but in theory you should measure ze and then add on the R1+R2 just in case there's a boiler or similar nearby on the ring, which may later disappear.

 

[ericmark]

On the 17th edition being released I wondered about claims on the person doing an EICR if they missed some thing which would cost to correct, for example with change of owner, that would mean new owner is landed with a bill which had it been picked up, old owner would have paid for. If the volt drop was not measured, could the electrician be expected to correct it, without charge to new owner?

I realised I had never actually calculated volt drop when doing an EICR, for new installation yes, but never on an old one. Since so complex, I set to making a java script program to check it, enter the loop impedance supply and socket outlet select cable sizes etc, and it would show the volt drop. But then I started to look at possible errors. It really does not matter if measuring loop impedance or prospective short circuit current, the meter is not that accurate.

The meter puts a load on the supply and works out the figures based on the volt drop, so any fluctuation in the supply volts can affect the reading, taking the reading three times and discarding any one reading wildly different to out two could help, but then the meter started to get hot, so yet another error.

Old meters used a fair load, but it would trip the RCD, so latter meters had a very small load so it would not trip the RCD. This again resulted in being less accurate as any item switching off/on during the test can upset the reading. So only real way would be to switch off all circuits and use a low ohm meter, the loop impedance or prospective short circuit current may ring alarm bells, but to get an accurate reliable reading it has to be the low ohm meter.

The same is also clearly a problem with earth loop impedance as to if low enough to trip a MCB. The incoming ELI for a TN-S 100 amp single phase supply is considered OK up to 0.8Ω. At that it would be hard to wire up a shower with a 45A MCB and ensure with a short circuit to earth it would trip, so in the main we rely on the RCD to protect us. Having lived in houses with 30 mA RCD protection for years, I realise the short to earth is in real terms covered by the RCD.

Clearly if the ELI is below spec do high light it, but can you fail a circuit because either the tripping of the magnetic part of MCB or volt drop is exceeded where a RCD is used? High light in notes yes, but not really a failure, as in real terms no danger, the old code 4 has gone, so you can't code simply because it does not comply with BS7671. It has to be a danger, and volt drop does not mean there is a danger.

BS7671 contradicts itself, it says we should not rely on the RCD, but allows use of RCD as sole protection with a TT supply. Can't really have it both ways, if reading permitted with a TT supply if it has a RCD then it must still be safe (even if it does not comply) with any other type of supply, so it will not attract a code.

So nitty gritty, if having too much cable will attract a fail code, then clearly it is to be avoided, so be it volt drop or ELI what code can you attach to it if spec is exceeded? Over 500Ω ELI is clearly not safe due to change in earth rod resistance in dry weather, but unless over that how can it attract a code?

 

[JohnW2]

Yes agreed but in this context i think the op is more likely to have an idea of circuit length than have the appropriate equipment to measure zs.

Yes, probably true but, as I said, actually determining the length of an existing ring really requires measurement of R1 or R2 and, even then, that can only be translated into a VD by the use of assumptions.

Also you can measure zs directly but in theory you should measure ze and then add on the R1+R2 just in case there's a boiler or similar nearby on the ring, which may later disappear.

Again, I suppose that's theoretically true, but I wonder how often it's actually done like that? Apart from anything else, other than as a means of measuring Ze (rather than assuming a value 'by enquiry'), it would mean that folk would not require a means of loop impedance measurement!

Put another way, do you think that people would actually 'fail' a circuit if Ze+R1+R2 was above the 'maximum permitted' even though a (current) Zs measurement was below the required maximum? Whilst it is true that "...there could be a boiler or similar nearby on the ring, which may later disappear.", so is it possible that the Ze might subsequently change - so I'm not sure to what extent one should (or needs to) take into account 'future possibilities', rather than just rely on 'current measurements'?

Furthermore, as above, estimating VD also theoretically requires measurement of R1 or R2 (to determine ring length), as well as the assumptions, so there's not that much difference.

In any event, we are agreed that determination of the maximum circuit length theoretically requires consideration of both Zs and VD. Whilst, numerically, VD may well be the length-limiting-factor (particularly with TN-C-S), I would personally feel that it is by far the 'less important' of the two. Apart from anything else, given that the majority of installations probably have a supply voltage around 10% above the 'minimium permitted', I would find it hard to get excited about a circuit which, if fully loaded (itself very unlikely), would have a VD a bit over 5%!

Kind Regards, John

 

[EFLImpudence]

Put another way, do you think that people would actually 'fail' a circuit if Ze+R1+R2 was above the 'maximum permitted' even though a (current) Zs measurement was below the required maximum?

I would say "Yes" - albeit 'fail' might not be the correct term to use if referring to EICRs.

I would think you, with your installation and bonding (assuming it has not changed) are more qualified to judge.

 

[JohnW2]

... can you fail a circuit because either the tripping of the magnetic part of MCB or volt drop is exceeded where a RCD is used?

I don't think that one could ever 'fail' a circuit on the basis of an excessive VD, since there is no explicit regulatory requirement in relation to VD - other than it should not be such as to "impair the safe functioning" of equipment. Hence, unless one really believed that the potential VD might 'impair that safe functioning' (and I struggle to think of when that could be the case), I can't see how one could 'fail' a circuit for that reason alone.

... and don't forget that the "maximum theoretically possible VD" we estimate (on the basis of circuit length and assumptions about load distribution) will virtually never actually be seen in practice.

BS7671 contradicts itself, it says we should not rely on the RCD, but allows use of RCD as sole protection with a TT supply. Can't really have it both ways, ...

I think that's rather unfair/unreasonable/unrealistic. Unless they decided to 'ban' TT, to allow it to be protected "by the best/only means possible" is surely the only sensible course - even if that method of protection, alone, would not be regarded as adequate for TN installations?

Kind Regards, John

 

[JohnW2]

I would say "Yes" - albeit 'fail' might not be the correct term to use if referring to EICRs.

Well, I suppose that we are essentially talking about EICRs (or EICs, come to that).

However, as I just wrote, if one is going to 'fail' a circuit on the basis that the Zs measured at the time of an inspection might subsequently rise if incidental paths to earth were to disappear in the future, why not also 'fail' it (particularly if, at the time of measurement, Zs is close to the 'maximum permitted' and/or Ze is particularly low) on the basis that Ze might rise at some point in the future?

I would think you, with your installation and bonding (assuming it has not changed) are more qualified to judge.

Nothing has changed here but I think that the goalposts are in a very different place when considering TT installations. Yes, if I undertake measurements with bonding in place, all of my final circuits have Zs less than the 'permitted maximum' (for satisfactory ADS with MCBs). However, the hypothetical future total loss of all the incidental paths to earth would have a dramatic effect on all my circuits' Zs (all of which would then be above 50Ω), hence the need for RCD protection 'just in case' - whereas the consequences of total loss of all incidental paths to earth in a TN installation would be far far less.

Even though extraneous-c-ps in the form of metal utility supply pipes are obviously on the decrease, I'm sure that a very substantial proportions of installations still have incidental paths to earth via bonded extraneous-c-ps. Are you suggesting that there is no point in direct measurement of Zs in circuits in such a property, since they would usually be lower than Ze+R1+R2 (which is the criterion which should be used for assessing the acceptability of the circuit)?

Kind Regards, John

 

[John D v2.0]

I would say "Yes" - albeit 'fail' might not be the correct term to use if referring to EICRs.

I would think you, with your installation and bonding (assuming it has not changed) are more qualified to judge.

Agree, i can't see how an installation that relies on a parallel path could be compliant, the cpc needs to be part of the same wiring system as the live conductors, a copper gas pipe and bonding conductor would not comply.
What the dno does is another matter but the change from a boiler removal could be much greater than any change the dno might do. It is also an unnecessary risk. Dno work also falls outside of bs7671. The changes by the dno would be not picked up until the next inspection.

 

[John D v2.0]

Are you suggesting that there is no point in direct measurement of Zs in circuits in such a property

They're certainly useful as a finger in the air check, but so are those beeping socket testers.
I think the crucial point is that using a gas pipe as part of the cpc is a non compliance in the scope of the installation that may give rise to a danger, whereas the dnos wiring is outside the installation so isn't in scope.
Maybe the dno might change the end that some houses are fed from but that's a risk that only they have the power to mitigate. Internal wiring is in the homeowner's power to mitigate.

 

[JohnW2]

Agree, i can't see how an installation that relies on a parallel path could be compliant, the cpc needs to be part of the same wiring system as the live conductors, a copper gas pipe and bonding conductor would not comply. ... What the dno does is another matter but the change from a boiler removal could be much greater than any change the dno might do. It is also an unnecessary risk. Dno work also falls outside of bs7671. The changes by the dno would be not picked up until the next inspection.

Whilst I understand what you and EFLI are saying, if one takes that view why on earth do people even bother to (or 'are required to') measure circuits' Zs in a property which has bonded extraneous-cps?

Kind Regards, John

 

[JohnW2]

They're certainly useful as a finger in the air check...

In the presence of bonded extraneous-c-ps, it's presumably only 'useful' if it indicates an unacceptably high Zs. If it indicates an 'acceptable' Zs, that is meaningless (and could result in a sense of 'false confidence') if one is going to say that it can be over-ridden by a Ze+R1+R2 which is too high.

Kind Regards, John

 

[EFLImpudence]

Whilst I understand what you and EFLI are saying, if one takes that view why on earth do people even bother to (or 'are required to') measure circuits' Zs in a property which has bonded extraneous-cps?

"Measure" also applies to determining R1+R2 by resistance; not just a loop test.

 

[JohnW2]

"Measure" also applies to determining R1+R2 by resistance; not just a loop test.

I realise that, but (as you understand) I was talking about 'loop tests' - and I don't recall having ever seen someone not directly measure (and record) loop impedance (as "Zs"), even when they have also measured R1+R2.

Indeed, the standard Schedule of Test Results appears to require, for each circuit, both (R1+R2) and Zs to be recorded, together with 'Maximum Permitted Zs', and one imagines that most people probably compare figures in the "Zs" and "Maximum Permitted Zs" columns in order to determine whether a circuit is compliant. Although Ze does appear elsewhere on the page, there is no column (for each circuit) for Ze+R1+R2 which they could use directly for an assessment of compliance.

I would imagine that most people would record in the "Zs" column the figure they got by direct measurement of the loop impedance. Do you believe that most of them also manually (or 'mentally') add together Ze and R1+R2 (although there is nowhere to record that figures on the Schedule) in order to confirm that it, as well as the recorded "Zs", is less than the recorded "Maximum Permissible Zs"?

Kind Regards, John

 

[EFLImpudence]

The forms do vary but I consider:

R1+R2 to be the conductor resistances,
Zs to be the loop test from the farthest point, and
Maximum Zs to be the figure in BS7671 for the OPD.

Some people, I believe from reading forums, wrongly calculate R1+R2 by deducting Ze from Zs.

 

[JohnW2]

The forms do vary but I consider:
R1+R2 to be the conductor resistances,
Zs to be the loop test from the farthest point, and
Maximum Zs to be the figure in BS7671 for the OPD.

I would think that the vast majority of people (certainly myself) would agree with that.

Some people, I believe from reading forums, wrongly calculate R1+R2 by deducting Ze from Zs.

Maybe they do. I suppose that such a practice is, in some senses 'no worse' than calculating loop resistance by adding Ze to (measured) R1+R2.

However, do you agree with me that, using the forms, most people will look at the recorded "Zs" figure (measured loop impedance at farthest point) and the recorded "Maximum Permitted Zs" figure (from BS7671 tabulations), and compare the two to decide whether or not the circuit is 'compliant' - without usually undertaking any 'confirmatory calculation/thinking' based on (measured) R1+R2?

Kind Regards, John

 

[EFLImpudence]

I would think that the vast majority of people (certainly myself) would agree with that.
Maybe they do. I suppose that such a practice is, in some senses 'no worse' than calculating loop resistance by adding Ze to (measured) R1+R2.

I do not agree with that - nor really know what you mean.
Calculating loop resistance by adding Ze to (measured) R1+R2 IS the circuit impedance - true Zs.
Deducting Ze from tester measured Zs is no indication of the actual circuit impedance.
All this is, of course, assuming the tails are relatively short.

However, do you agree with me that, using the forms, most people will look at the recorded "Zs" figure (measured loop impedance at farthest point) and the recorded "Maximum Permitted Zs" figure (from BS7671 tabulations), and compare the two to decide whether or not the circuit is 'compliant' - without usually undertaking any 'confirmatory calculation/thinking' based on (measured) R1+R2?

Probably.

I repeat. How would that work on your installation if the water pipe is replaced with plastic?