On-Site Guide BS 7671:2018 - 7.2.2 Socket-outlet circuits

My guess would be your second option:

 

Or, does it mean from where the spur connects to the circuit to the point which is the furthest away from the protection device in the consumer unit i.e. a point half way around the circuit beginning from the consumer unit? If so, the maximum length of the spur cable would be longest nearer to the consumer unit and shorter as the furthest point away is reached... Surely this cannot be correct, because at the furthest point the maximum length would be zero - !

And zero length at the furthest point does make sense in a way. A spur will always add some resistance - so increasing both voltage drop and Zs above that at the point where it's connected to the ring. If the ring is already as long as allowed - then yes the maximum spur you could add at the furthest point would indeed be zero - otherwise you'd be exceeding permitted loop impedances or voltage drop.


I agree the wording you've quoted isn't that clear. Really the limit should be based on the maximum permitted ring length rather than the length of the ring you've actually got - and indeed they could be clearer about meaning the point on the ring where the spur is connected.


(I'm still puzzling where they've got 1/8th from - I'll have to think about that).


   - Andy.

This statement is without foundation for practical rings, which are very rarely anywhere near maximum length. As RCD protection is mandatory the Zs question is fairly irrelevant, and voltage drop is just inconvenient and again unlikely to be a problem. Strangely what regulation does it reference? I suppose none. I think that the OSG should probably be dumped, it has a number of "old wives tale" comments which are not supported, and any electrician should have a copy of the real regulations, which they should know fairly well (in my experience they don't), and I had a few people turn up for an 18th edition course with just an OSG under their arm! They were really surprised when given a copy of the book to use as to how much was not in the OSG.

I think that if it is to have any real meaning it should be a fraction of the max permitted length of ring less the point it is to be actually  installed at. Or to put it more simply, to not put it any more onerous (Zs and Volt drop)  than any point on max permitted ring length would be.

You`d need to invent a formula to do it

As written it seems to preclude short rings and long spurs - which is a perfectly safe and economical method if for example you have a high ceilings and wish to drop the downstairs sockets as a spur each - the voltage drop is better than if you went up and down the wall each time.

I'm incined to say it is indeed so much cobblers, and a simple total voltage drop requirement (no socket more than 50m from the origin by whichever route is shortest) would be a better approximate 'rule of thumb' for being sure of that.

Clearly there is nothing in BS7671 itself about this.

Thinking of other threads on here, I'm reminded this is just the sort of confusing and unscientific advice that causes false alarms and bitter wrangling when  EICRs are performed by folk who follow a recipe rather than design for a living I can almost see it  "what code do I give an over-long spur ?".

davezawadi:

This statement is without foundation for practical rings, which are very rarely anywhere near maximum length. As RCD protection is mandatory the Zs question is fairly irrelevant, and voltage drop is just inconvenient and again unlikely to be a problem.

The statement that RCD protection is mandatory only really applies in dwellings where cables are buried in walls without earthed metallic containment, where the ring final circuit serves socket-outlets (assuming that BS 7288 RCDs are not permitted, although they would have been considered acceptable to BS 7671 prior to BS 7671:2018). Hence, there is a need to consider


It certainly doesn't cover every installation, and the OSG may be useful for older installations.


 

Strangely what regulation does it reference? I suppose none.

It says it's a rule of thumb - usually, calculation rules of thumb err on the side of caution - so you may have a point - but for many circuits in dwellings, a quick lookup is better than doing lengthy calculations (and making mistakes).

 

I think that the OSG should probably be dumped, it has a number of "old wives tale" comments which are not supported, and any electrician should have a copy of the real regulations, which they should know fairly well (in my experience they don't), and I had a few people turn up for an 18th edition course with just an OSG under their arm! They were really surprised when given a copy of the book to use as to how much was not in the OSG.

I really don't think providing lookup tables that err on the side of caution is responsible for many people being unaware of all of the issues that pertain to the 500 or so pages in BS 7671 (plus all of the attendant standards it references).


I think that's a separate debate all together.

Thank you to one and all for your kind help.


The sentence in 7.2.2 makes no direct reference to the Wiring Regulations, and it is ambiguous.


To quote from the On-Site Guide, ‘Table 7.1(i) has been designed to enable a ... ring final circuit to be installed without calculation’.


My belief is that when faced with any problem it is dangerous to assume anything but, presumably, the IET have included the sentence in 7.2.2 to be helpful to anyone learning about the installation of electrical wiring.


In this instance, the IET appear to have failed to give ‘clear guidance’, and I am not qualified enough to do the electrical calculations myself.


Perhaps my suspicion is wrong but, if I use my first interpretation of the sentence, taking the maximum cable length for a ring final circuit (32A, RCBO Type B, 2.5/1.5 mm2, TN-C-S), as 106m (Table 7.1 (i)); the maximum length of a spur would be 6.62m (106/2 => 53/8 => 6.25).


Similar calculations for two ring final circuits with maximum lengths of 40m and 80m would result in maximum spur lengths of 2.5m and 5m. Both of these maximums would be of useful working lengths.


Taking the meaning of 7.2.2 in the other way would mean that if a ring final circuit were to be extended (for example, a loft conversion), all the new socket-outlets would need to be on that ring final circuit, rather than on spurs taken off the ring final circuit. Which would not be very helpful or practicable.


Presumably, the IET understand the reasoning for the sentence in 7.2.2 and are suggesting to novices and qualified electrical installers alike that, the larger the domestic dwelling (i.e. the longer any ring final circuit is likely to be), the longer the practical and safe working length of any spur taken from that ring final circuit. Hence my summation being, that the maximum spur length for any given ring final circuit (shorter in length than the maximum given in Table 7.1(i)), is 1/8 of 1/2 of its full length (effectively, 1/16).


I am not assuming that I am correct, but I am trying to understand.


Thank you once again to everyone who has given my question their considered time and thought.

The sentence in 7.2.2 makes no direct reference to the Wiring Regulations, and it is ambiguous.

:

I am not assuming that I am correct, but I am trying to understand.

Perhaps start off considering what bits of BS 7671 effectively limit the length of a circuit. Things like Zs, voltage drop (and occasionally L-N loop impedance) come to mind. There's no single place in the regs that define these requirement - its a combination of lots of separate requirements.


Then consider the effect on circuit length of adding a spur to various points on a ring - both a ring that's already at its maximum circuit length and a ring that's somewhat shorter.


Remember that this is all ultimately driven by the laws of physics - rather than what would be convenient to an electrician.


   -  Andy.

I must thank you all once again. All of your helpful ideas have nudged my thinking into a different direction. I am beginning to hope that I may now understand what that sentence in the On-Site Guide may mean. 


To begin again, ‘As a rule of thumb for rings, unfused spur lengths should not exceed 1/8 the cable length from the spur to the furthest point on the ring.’


My nascent understanding of electricity and electrical wiring includes the belief that a ring circuit operates somewhat differently to a radial circuit. In a domestic situation, an installed circuit is supplied with a nominal 230V of electricity which can be drawn off at socket-outlets and suchlike to provide power for electrical equipment. Would it be true to say that for a radial circuit one equation (more or less) can be used to determine the limits to how that circuit may be used? Presumably, that same equation could not be used alone to determine the limits to a ring circuit because it is supplied with power at both ends rather than just one. Hence, the IET have designed Table 7.1(i) to save electrical installers the time and effort of performing these various calculations for both radial circuits and ring final circuits. Please be patient, I am not pretending to know but to get the gist.


Looking at Table 7.1(i), if I install a ring final circuit with a 32A RCBO and 2.5/1.5 mm2 cable installed correctly (according to an allowed method), I could run the circuit to a maximum length of 106m. The beginning and end of the circuit will be at the consumer unit or distribution board. Applying the rule of thumb given in 7.2.2, I can install an unfused spur at the consumer unit, not longer than 1/8 the cable length from the spur to the furthest point of the ring, which just happens to be the cable length of the ring i.e. 106m. Therefore, the maximum length of any spur at the consumer unit will be 13.25m (106/8). Should I wish to install a spur half way around the ring circuit, at the far end of the dwelling say, the maximum length of the spur would be half i.e. 6.62m (53/8). However, if I were to install a spur a quarter of the way around the ring, the furthest point of the ring from the spur is three quarters of the full length i.e. 79.5m (106 x 0.75). Therefore, the maximum length of a spur at that point would be 9.94m (79.5/8).


Say, my ring final circuit is 80m long. The maximum length of a spur at the consumer unit is 10m; half way around the circuit it is at its shortest, 5m; a quarter of the way around the circuit it is 7.5m. All of these are useful lengths for installing unfused spurs in a practical situation. The difficulty being that unless you have installed the whole circuit yourself or have the electrical diagrams of the installation at hand, you would be left guessing the locations of the spurs. But that would be missing the point, because the idea of the rule of thumb is to provide you with a guesstimate. What 7.2.2 is saying to me is that, if I wanted to put a spur longer than 5m half way around my 80m ring circuit, I should consider extending the ring further (maximum 106m), rather than attempting to do any calculations myself.


By Jove, I think I’ve cracked it! Unless you know differently, of course.


Thank you, thank you, thank you!

Would it be true to say that for a radial circuit one equation (more or less) can be used to determine the limits to how that circuit may be used?

Sort of. There'll be a number of different requirements, each of which can usually be reduced to a simple equation. Each calculation will have a slightly different basis - loop impedances will just depend on the cable size & length, so relatively straight-forward, but voltage drop also depends on the load current - which means making some different assumptions especially for circuits like socket circuits that have many different points where loads may or may not be connected. You get quite different results if you assume that the loads are concentrated at the far point, or spread evenly, or spread unevenly in some way. If calculating in full manually you'd often have to do several calculations and then pick the most limiting (shortest) result. The OSG used to note against some circuits what the limiting factor was - e.g. zs for earth fault loop impedance, sc for L-N loop impedance or ad for adiabatic (usually due to the smaller c.p.c. size), otherwise it was voltage drop.


In many cases you can treat as ring as being similar to a radial with half the cable of twice the c.s.a.. - which would suggest to me that a spur from the MCB could be up to a quarter of the overall length of the ring for the same result (I'm still puzzling where 1/8th has come from, unless someone thought that a 2x safety margin would be a good idea)

 

Applying the rule of thumb given in 7.2.2, I can install an unfused spur at the consumer unit, not longer than 1/8 the cable length from the spur to the furthest point of the ring, which just happens to be the cable length of the ring i.e. 106m.

I would have said that the furthest point of the ring was half way around (after that you start getting closer to the start again) - so even with the spur taken from the MCB you've only 1/8th of 53m rather than 1/8th of 106m to play with.


Don't discount the fact that these things are written by humans, and to err is human... Even IET texts aren't necessarily word perfect.


    - Andy.