2 circuits done. 2 to go.

[JohnW2]

I thought I had corrected that before posting?

Clearly not, as you can see from my quote

However, returning to what I said in my post, if it were 88 metres, that would mean that the furthest socket could be no more than 44 metres from the origin - pretty similar to the corresponding radial.

Kind Regards, John

 

[JohnW2]

Not if you take note of the full stop. And the need to have an earth run even if not used.

I disagree, and don't understand your obsession with the full stop.

The sentence before the full stop is irrelevant to this discussion, since it merely says that G/Y cannot be used to 'identify' (e.g by oversleeving) anything other than a protective conductor - which is not the situation if a G/Y-insulated conductor is 'identified'(e.g. by oversleeving), and used, as something other than a protective conductor. It does not say that a G/Y-insulated conductor cannot be 'identified' (e.g. with brown or red over-sleeving), and used, as something other than a protective conductor.

However, the sentence after the full stop goes on to say that, despite that previous sentence, in the specific case of a single-core cable, a G/Y -insulated conductor may not be identified (by over-sleeving) and used as anything other than a protective conductor.

That all seems clear and straightforward to me - do you read/interpret it differently?

Kind Regards, John

 

[ebee]

I seem to remember 88 meters limit for ring final with 2.5 mm² cable, and 4% volt drop, when we went to 5% it went up to 106 meters.

It seems a lot, but by time you take in the 1/3 rule on drilling beams, very easy to exceed. What seems odd is we don't seem record the lenght of cable used, or volt drop.

I at one point did worry about exceeding the volt drop, but no one ever seems to test it, may be just as well, if it was an EICR failure to have too much volt drop, I think many would fail.

88 metres 106 metres etc etc etc.
In fact I quoted one version around 2007 or thereabouts based on 91 metres, depends on which OSG is current and which method of fusing and earthing type. In this instance it was around 91 metres at the time (I think it was 2007) .

The reason I used it was to break a "Myth" - I had heard several, including College Tutors, stating the double helix R1 + R2 test resistance method giving exactly the true R1 + R2 at any point. I worked it out and it varied approx 6% I think, so I stated it should not read "exactly" but be substituted by the word "substantially" because using 91 metres as a max ring length the difference in theory to reality was not actually great and might well be dwarfed by field errors and meter reading accuracy anyhow. Some might have thought (quite rightly probably) that I was being a pedantic git, however I objected to the dogmatic stating of "exactly" in some quarters.

I probably quoted my reasoning about this on the DIYNOT Forum at the time I think

 

[ericmark]

I heard about the maximum length being increased at an IET lecture, I tried to calculate this, and failed, so at the next lecture I asked the question, and it seems it uses the assumption that the ring final will be loaded 20 amp in centre and 12 amp even spread.

So it is calculated at 26 amp not 32 amp. It also uses the formula given in BS 7671 to correct volt drop per meter when the circuit is only part loaded.

Using this I did get 106 meters well since saying exact actually 106.7 meter. Using Appendix 12 Low voltage installations supplied directly from a public low voltage distribution system other uses 5%, with lighting being 3%.

So with a line - neutral loop impedance of 0.35Ω at the DNO head, we should get 0.94Ω at the mid point of the ring final. It is clear this can be easy measured, we are looking at 0.59Ω greater loop impedance at mid point, however when it comes down to actually doing the measurement the loop impedance meter will often give readings with 0.02Ω variations, both at the DNO head and mid point so the measurement can be 0.04 out. So until we get over 1Ω we can't really say the ring is too long.

And this also assumes we are using Design current for circuit Ib as 26 amps, for all we know when the ring was installed it was to supply a floor cleaner only throughout the building, I actually had this with a massive fridge for cheese, the ring final with sockets was only there to use the cleaning equipment.

But of course we also have to keep the loop impedance low enough so the MCB will trip with a short circuit on the magnetic part of the trip, we are told to allow for 5% volt drop, so around 1.37Ω for a B type, so in real terms, if the loop impedance is within the allowance for tripping the MCB, then we have to ignore the volt drop limits.

I have been an electrician some 50 odd years, and only twice have I seen real problems with volt drop, once in Algeria where the Dutch electrician thought two phases of a 110 three phase supply would be 220 volt, in fact since out of phase, only 190 volt, and this caused the air conditioning unit to regularly trip, and it would burn out the built in overload devices in the motors.

The second was a shrink wrap machine, this would fail to work when the voltage dropped.

I have tried to consider what danger a volt drop could be, for example the fan failing to start on a fan heater, but this would be very unlikely.

So on an EICR we may plug in the loop impedance meter, and we may think this reading makes me think the ring final is too long, but one is appendix 12 part of the regulations, we are tested on the appendix in the C&G 2391 exam, but I would say the appendixes are outside the regulations, and we are looking for potential danger not complying with regulations.

However when the LABC have to decide if they can issue a completion certificate, they can't ask the inspector to complete an installation certificate only the person doing the work can do that, so he completes an EICR, which is very similar to the installation certificate, and he is instructed to test as if it's an installation certificate, so all deviations from the current edition of BS 7671 are noted.

If he feels the limit has been exceeded he could note this, as to if the LABC would then refuse to issue the completion certificate I don't know.

I have only once worked with the LABC issuing the completion certificate, and they insisted I fitted an extractor fan, even when the bathroom had opening windows, as they said visitors could walk past the window.

Clearly the do it again and do it properly is the last thing we want to hear from the LABC inspector, and he is within his rights to say this if the installation does not comply, so the last thing I would want is to give him an excuse to fail the installation.

 

[JohnW2]

.... The reason I used it was to break a "Myth" - I had heard several, including College Tutors, stating the double helix R1 + R2 test resistance method giving exactly the true R1 + R2 at any point. I worked it out and it varied approx 6% I think, so I stated it should not read "exactly" but be substituted by the word "substantially" because using 91 metres as a max ring length the difference in theory to reality was not actually great and might well be dwarfed by field errors and meter reading accuracy anyhow. Some might have thought (quite rightly probably) that I was being a pedantic git, however I objected to the dogmatic stating of "exactly" in some quarters.

It's not for me to say whether you are a pedantic git! However, whilst I do not doubt what you say about some College Tutors (and others), for what it's worth, my (albeit fairly ancient) copy of GN3 does say that the 'double helix' method gives substantially the same readings at all sockets on a ring (which is very true, any differences being very small- see below), but it then goes on to saying, seemingly incorrectly, that this is equal to approximately one quarter of r1+r2 (whereas it should presumably say 4 times, not 'one quarter').

It then goes on to say that the maximum value measured (which would be at the furthest point from origin of the ring) represents the RI+R2 which should be recorded on the Schedule of Results - but, again, it presumably should be one quarter of the measured value which is recorded as the (maximum) R1+R2.

I find this all a bit confused/confusing. Quite apart from that apparent "quarter vs. 4 times" error, they do not initially make it clear that, although the 'double helix' measurement is 'substantially the same' at all sockets, R1+R2 most certainly won't be - in fact it will approach zero as one approaches an end of the ring. Hence, as they go on to say/imply, if one wants the 'maximum R1+R2' one has to find the socket with the highest 'double helix' measurement (and divide, not multiply, it by 4 to get 'max R1+R2').

The graph and tabulation below show 'the truth'. They illustrate the very small reduction in 'double helix' measurement as one moves from centre of ring towards an end of the ring, and also how actual R1+R2 varies across the length of the ring.

I don't understand what is said to be 'approximate' about a 'one quarter of double helix' measurement at the middle of the ring as a measure of R1+R2 at that point - as can be seen it is exactly equal to true R1+R2.

What I don't really understand is why, if one wants to determine the true R1+R2 at the most distant socket ('exactly', within the limits of accuracy of the measuring equipment), one doesn't just leave the ring intact, join L and CPC at the origin of the circuit and then measure the actual R1+R2 at the socket of interest.

Kind Regards, John

 

[flameport]

I had heard several, including College Tutors, stating the double helix R1 + R2 test resistance method giving exactly the true R1 + R2 at any point.

Education starts with mostly lies.
As the students get older, the number and magnitude of the lies gets smaller.
However there are always some lies left.

 

[JohnW2]

Education starts with mostly lies. As the students get older, the number and magnitude of the lies gets smaller. However there are always some lies left.

All true - however, as I've just written, GN3 appears to include some 'lies' (as well as some truths), and is all rather confusing (at least for me) about this matter

Kind Regards, John

 

[ebee]

Education starts with mostly lies.
As the students get older, the number and magnitude of the lies gets smaller.
However there are always some lies left.

Brill Answer that one Flameport