Adding to a ring final.

[ericmark]

Prospective short circuit current and loop impedance are related and it does not really matter which is measured but what we look for is the worst reading which will be as near to centre of ring that we can measure.

There are two limiting factors one is the volt drop and the other is the current required to ensure the automatic disconnection device will disconnect within the prescribed time.

Before the advent of the RCD we would be more interested in the line – earth reading but today it’s the line – neutral reading which is more likely to be the limiting factor.

The gives us the first problem as although top of the range meters may measure PSCC line – neutral and ELI line – earth not all meters do this so often it means making leads to allow this to be done without having to resort to probes.

But the real question is what is the line – neutral limit? To trip a B32 MCB in 0.01 seconds then 1.44 ohms but volt drop is not so easy. We are given complex formula and rules like 20 amp in centre are 12 amp even spread so for volt drop we use 26 amp not 32 amp. But the loop impedance will work out at somewhere near 0.79 Ω assuming the supply at 0.35 Ω. The ELI likely will be around the 0.94 Ω mark.

Simple maths without correction factors 230/11.5 = 0.44 + supply resistance of 0.35 likely with a TN-C-S supply = 0.79 Ω. At 0.018 Ω per meter = 44 meters double it up and 88 meters use the correction factors 106 meters of cable which is fair enough.

But question is measure say 0.70 Ω and your not sure that’s dead centre and customer wants ring extending to 4 extra sockets. Up and down walls 3 meters each and between sockets another 10 meters. So you need to add 22 meters. This will likely add 0.198 Ω to the reading so 0.90 Ω total although you have measured hot and meters not really that accurate.

I have seen ELI measured with an EICR, but line – neutral I would think is rarely measured. So getting caught is remote. Will the extra volt drop cause a problem? unlikely. Running 4 fused spurs is a real pain and does not look professional.

Exceeding the 1.44 Ω could cause danger, but with a RCD this is removed, and I personally wonder what danger is there with excessive volt drop? To me the danger is some one some time in the future will find out, and use ones own paperwork to prove you knew it did not comply, and want you to correct FOC. So should it be written on the departures? At least that way you can say it was agreed at the time.

I tell the DIY guy before extending a ring final he should measure the loop impedance and work out if once his alterations are completed will it comply. But I really question how many professional electricians actually sit there and work out volt drop before quoting for the work? Once quote is done to then say can’t do it for that price because what ever is really too late. Seeing RCD missing is easy but how many even measure the loop impedance before starting the job?

So does it really matter? And if some one did measure the volt drop and ask for the problem to be solved what would be your answer? That’s assuming there is over 5% volt drop. Would you fix it? I have heard the answers so many times, I told your wife “xxx” and she said go ahead anyway. “Or I did tell you that before I started.” My answer would be of course then why is it not on the list of items not complying.

When I bought my house new in 1980 I was given no installation certificate and so there is no way I could prove what it was like when installed. Today there are reams of paper work so even 20 years on some smart owner could try it on. If the installation certificate shows it did not comply when installed we can hardly call it fair wear and tear. We are seeing where sex offenders are being taken to court some 40 years after the event. Could we be taken to court in 20 years time for work we have done today?
Personally I would say 10 year limit as after 10 years there should have been an EICR, but seems there is no such limit with sex offenders no one says why didn’t you say something when you became adult at 21.

So what is the opinion on volt drop limits? Does anyone really worry? And if one retires would insurance valid at time of the work still cover you? With my Vauxhall car sold to my daughter some 10 years latter she still had a new fuel tank fitted under a recall where Vauxhall found there was a design fault. We are here talking about a design fault so would there be any point where a customer could not ask for it to be fixed?

 

[JohnW2]

There are two limiting factors one is the volt drop and the other is the current required to ensure the automatic disconnection device will disconnect within the prescribed time. ... Before the advent of the RCD we would be more interested in the line – earth reading but today it’s the line – neutral reading which is more likely to be the limiting factor.

That’s surely not really true, given that one is not normally meant (allowed?) to rely on RCDs to achieve the required disconnection times in response to L-E faults. Given that, and since Ze will invariably be higher than R1+Rn+Re (since CPCs have smaller CSA than live conductors), it will surely always be the Ze (i.e. L-E loop impedance, EFLI) which is the “limiting factor”?

But the real question is what is the line – neutral limit? To trip a B32 MCB in 0.01 seconds then 1.44 ohms but volt drop is not so easy. ... Exceeding the 1.44 Ω could cause danger, but with a RCD this is removed, and I personally wonder what danger is there with excessive volt drop? ...

Just to remind people, that 1.44 Ω figure has only got two or three weeks life left (or, at least until 1st July) and, as I said above, it is not really acceptable to remove a ‘danger’ due to inadequate OPD-based fault protection by relying on an RCD, However, as for your question, I would personally say that, in a domestic situation, ‘excessive voltage drop’ would virtually never represent a significant ‘danger’.

In any event, the whole concept of stating limits for voltage drop from a non-fixed starting point is surely just plain daft? Considering the case of a house whose normal supply voltage is about 240V (very common), to suggest that it was in some way ‘dangerous’ for a final circuit to have a VD of, say 6%, in premises, when it would have been regarded as perfectly acceptable for the supply voltage to have been about 10.3% less than it normally was (with a permitted further 5% drop below that) would make absolutely no sense, would it? ... it would mean that a delivered voltage (at socket) of 226.2V (240V-13.8V) would not be acceptable in one house (because there was a “6% VD”), whereas a delivered voltage of 204.7V (216.2-11.5V) would be acceptable in another house (whose normal supply voltage was 216.2V). The difference would obviously be even more dramatic if the house’s normal supply voltage was approaching the maximum of 253V.

I tell the DIY guy before extending a ring final he should measure the loop impedance and work out if once his alterations are completed will it comply. But I really question how many professional electricians actually sit there and work out volt drop before quoting for the work? ... So does it really matter? ... So what is the opinion on volt drop limits? Does anyone really worry?

I think you have more-or-less answered your own question. Although it is fair enough to estimate (or even measure) VD, I would think that it would be very unusual for it to be an issue (even in terms of the regs’ guidelines) in domestic situations. I live in a very large house, but none of my ring finals even approach the lengths that would produce VD ‘problems’. As above, I don’t see that VD is ever going to represent a ‘significant danger’ in domestic situation - and that’s even without considering the fact, as above, that a ‘delivered voltage’ (at socket) which was 'unacceptable' in one house might be more than 20V higher than the voltage that would be acceptable in another! ... and, of course, our estimated VD figures are based on maximal loading of the ring final, which is very rarely going to arise, and very very rarely for significant periods of time.

I realise that some electricians are constrained to work to what they regard as the letter of the regs, but is this not a situation in which, at least ideally, common sense should prevail?? As for that “letter of the regs”, don’t forget that the infamous 5% and 3% figures only exist as a guideline in an ‘informative’ appendix of the regs. All that the regs themselves require is that equipment be supplied with a voltage which does not impair their ‘safe operation’.

Kind Regards, John

 

[WabbitPoo]

thought this was a diy forum

 

[JohnW2]

thought this was a diy forum

Well, as far as electrical work is concerned, I presume I count as a "DIYer" - I certainly am not an electrician or electrical engineer

Kind Regards, John

 

[ericmark]

I thank JohnW2 for his answer. I was thinking myself the volt drop seemed a little silly and that was underlying the question.

Not sure what you mean about the 1.44Ω going as I thought it was simple calculation for a B rated 5 times stated thermal trip current so for a 32A MCB it will trip at 160 amp so 230/160 = approx 1.44Ω.

As to not relying on an ELCB we do no really have an option with a TT system I am not sure when we started using an ELCB it must have been after 1954 when my parents house was built but by the time I started the ELCB-v was used for TT systems. I left to work abroad and on return the ELCB-v was banned and the ELCB-c replaced it.

To even get an earth rod to 7Ω one would be hard pushed other wise we have to rely on the ELCB/RCD/RBCO or what other name you want to give it to disconnect the supply with a line - earth fault.

So unless TT is banned then we have to rely on the RCD.

I would argue that the 1.44 figure line - neutral is not set in stone as the disconnection times are for line - earth not line - neutral. To my mind the idea is to disconnect the supply before it can cause damage where the problem arises is where a cold replacement can be inserted. Be it a fuse or plug in MCB if after tripping due to over current a cold replacement can be used then the already hot wires could damage insulation. But other than that even with a short circuit the fuse/MCB will always interrupt supply before damage to cables.

Clearly if we are trying to stop damage to a motor or semi-conductor then we need better than standard fuse or MCB.

To me the most dangerous system we have at the moment is the split phase 110 volt but this is because the standard brick does not comply with BS7671:2008 rather than the regulations being at fault. With a 10A overload trip on incomer one can draw nearly 42 A line to earth without opening the trip.

 

[ericmark]

thought this was a diy forum

If you read my post it is all about the advice given to a DIY person. Without test gear the DIY guy can't test his work and so he has no idea if the volt drop is within the permitted limits.

I have said many times for an electrician to extend a ring final is likely the best way to provide extra sockets but for a DIY guy or girl it is likely the FCU is a better option as far less likely to exceed the figures without realising.

What I am doing is questing my own advice. If the volt drop does not matter and if we have RCD protection then assuming the DIY guy or girl does check continuity between lines, neutrals and earths then no reason why he should not extend the ring final.

 

[JohnW2]

I thank JohnW2 for his answer. I was thinking myself the volt drop seemed a little silly and that was underlying the question.

Indeed. I personally think that talking about 'voltage drop' within the installation (with 'limits' of 5% or 3%) is just plain crazy when it is theoretically possible for one installation to have a supply voltage which is "16%" less than another (253V-216.2V, expressed as a % of 230V)!

Not sure what you mean about the 1.44Ω going as I thought it was simple calculation for a B rated 5 times stated thermal trip current so for a 32A MCB it will trip at 160 amp so 230/160 = approx 1.44Ω.

Things are about to change with Amd 3 of BS7671:2008 (published 5th Jan 2015, 'effective' 1st July 2015). Your 'simple calculation' will remain the same, except that, instead of using 230V, you will have to use either 218.5V (if they stick to the figure in the draft) or 216.2V (if they take notice of representations from myself and others). The idea is very sound. With current regs and figures, a Zs of just 1.44 Ω would fail to give the required disconnection time if supply voltage were less than 230V. In the draft of Amd 3, they used a voltage of nominal (230V) minus 5% - but (as I and others pointed out to them) it would seem far more rational to use 230V-6%, since that is the lowest permitted supply voltage. Not long to wait to see what they ended up with!

As to not relying on an ELCB we do no really have an option with a TT system ...

Of course. I was (implicitly, not the least because you were talking about a Ze of 0.35Ω !) talking about TN systems. As you say, with TT one has no choice but to rely on an RCD. However, with TN one should not.

Kind Regards, John

 

[JohnW2]

I would argue that the 1.44 figure line - neutral is not set in stone as the disconnection times are for line - earth not line - neutral.

Yes, but that's irrelevant in a TN installation, since the Zs has to be low enough (1.44 Ω or whatever) to achieve the required disconnection times with L-E faults - and, as I said (since CPCs are smaller than live conductors), that will also inevitably mean that the the disconnection times for L-N faults will be at least as short (whether they have to be or not).

Kind Regards, John

 

[JohnW2]

If you read my post it is all about the advice given to a DIY person. Without test gear the DIY guy can't test his work and so he has no idea if the volt drop is within the permitted limits. ... What I am doing is questing my own advice. If the volt drop does not matter ...

As I said, I think this is the least of the issues one needs to worry about. As I said, the chances of a domestic ring final being anything like long enough to result in even a 5% VD (and that only at full loading of the circuit) is, IMO, extremely small - and, even if it did slip a little beyond that 'magic' 5% (at full circuit loading), I have expressed my view that it is an essentially 'silly' criterion which does not relate to any significant safety issues.

Kind Regards, John

 

[ericmark]

With 1.35Ω it is better than with 30A fuse but maybe we will need a type I which was X3 for those installations which would no longer comply?

I have said many times with a fuse getting it slightly out means slightly longer trip times but with a MCB it means difference between 0.01 seconds and 100 seconds and 100 seconds is really too long of a time with a dead short.

But in the main with a domestic dead shorts are only found with lights. Either some one has missed out the sleeve on the blue wire and so we have a line - neutral short when power is turned on. Or ionisation as lamp fails which is near to a direct short as one is likely to get. Again line - neutral.

What I can't understand is using type C RCBO's on the lighting supply that means any electronic switch is likely to fail when bulb blows with ionisation and means more likely hood of the pins welding in the bulb holder. To my mind for lights should always be 6A type B.

 

[JohnW2]

I have said many times with a fuse getting it slightly out means slightly longer trip times but with a MCB it means difference between 0.01 seconds and 100 seconds and 100 seconds is really too long of a time with a dead short....

I'm not sure what you mean by 'slightly out', but, as dar as I can make out, the get the trip time of a B32 to rise from 0.01 secs to 100 secs would require the current to fall by more than half (i.e. the Zs would have to be 'out' by a factor of more than 2) - from 160A to about 70A.

Kind REgards, John

 

[ericmark]

I did not study graph before answering but was referring to the jump between when the magnetic part fails and thermal part takes over.

 

[JohnW2]

I did not study graph before answering but was referring to the jump between when the magnetic part fails and thermal part takes over.

It's not really much of a 'jump', more a gradual progression. Per the curves, the minimum current required remains the same throughout the magnetic part (i.e. as the disconnection time rises from 0.01 secs to about 12 secs) then, when the thermal part takes over, the disconnection time gradually increases (slowly at first, then more rapidly) as the current decreases below the 'magnetic tripping' level. The main difference with a fuse that, not having a magnetic component, the 'gradual progression' continues into the region of very high currents (hence very rapid disconnection times).

Kind Regards, John