Unfused spurs

[mikhailfaradayski]

would you care to expand on that?

Expand on what? - you quoted my entire message!

Kind Regards, John.

Just the bit immediately before my comment, the bit about you can envisage situations where your idea would be preferred over any other methods that are also available.

FGS, please don't expand on everything thats gone before, or it might end up equally as interesting as the double sockets epic

 

[EFLImpudence]

It does not meet the requirements for a 433.1.5 ring final.

Do you not make, or is there not, a distinction between unfused and fused (which 433.1.5 does not mention) spurs?

With radials.

Would you agree that this is illogical as a 4mm² radial may have a 2.5mm² unfused spur?

 

[JohnW2]

We seem to be repeatedly going over the same ground here (in relation to what I've already agreed is a pretty academic question).

Another suggestion was that a 1.5mm² cable from the ring to the FCU was a bodge. If it meets all requirements, how can it be a bodge?

It does not meet the requirements for a 433.1.5 ring final.

Indeed - but, as you recently agreed, that doesn't mean that it can't meet the requirements of 433.1.1 and be compliant with the remainder of BS7671.

433.1.5 refers to a RFC with or without 'unfused' spurs so I would suggest that if the spur were a fused one then the requirement for 2.5mm² cable need not apply to the spur.

On the supply side of the FCU it would, because the circuit has to be wired in a minimum of 2.5mm².

Only if one chooses to rely on the deemed-to-satisfy design described in 333.15, rather than design the circuit oneself.

Otherwise, when would 433.2.2 apply?

With radials.

I see nothing in 433.2.2 (or the corresponding 434.2.1 in respect to fault protection) that restricts it to any particular type of circuit. It is written in general terms and, as such, would appear to apply to any circuit - indeed, not even only final circuits.

I guess we need to remind ourselves again that 'the regulations' with which we wish to have compliance are represented by the entirity of BS7671, not just the examples and deemed-to-satisfy bits which it contains.

Kind Regards, John

 

[JohnW2]

Expand on what? - you quoted my entire message!

Just the bit immediately before my comment, the bit about you can envisage situations where your idea would be preferred over any other methods that are also available.

I didn't actually say 'preferred', merely that it could offer some, mainly practical, benefits in some situations. EFLImpudence has recently gone some way to answer this, when he wrote:

......John's original suggestion of a 4mm² spur on a 2.5mm² ring, I cannot see anything wrong with it as it saves the need for a FCU and it enables each socket to be used to ITS full rating rather than two or three (or however many) being limited to a total of 13A.
Also, if this spur should be, more or less, in a line from the ring. Rather than incorporating it into the ring, from one socket, with a cable joint and a second leg running back in the same channel as the first, a single run in 4mm² seems eminently preferable.

The design might be worth considering if, for example, one wanted to add a number of sockets to a circuit without restricting the total load on those sockets to 13A, particularly if those sockets were going to be in a circumscribed area somewhat separte from the remainder of the ring, and especially if they were sockets which one did not expect to be used all that much (but nevertheless wanted >13A max potential capacity). The alternatives would be to run a totally separate circuit (ring or radial) or to incorporate those new sockets into the existing ring. The former would often not be the preferred choice, whilst the latter could result in an appreciable increase in total length of the ring. There are obviously advantages in keeping the length of a ring as short as possible, and it would seem a particular 'pity' to appreciably extend its length for the sake of some sockets which were not often going to be used. Then, of course, the purely practical issues - e.g. less cables to run/terminate (ans somewhat less copper used!).

However, as both EFL Impedence and I have said, the main point being made is the general one that the examples and deemed-to-satisfy provisions in BS7671 are necessarily far from exhaustive, and that one would hope that those who design (and inspect) circuits would have the knowledge and ability to see beyond that and consider compliance with the totality of BS7671.
Kind Regards, John

 

[ban-all-sheds]

Do you not make, or is there not, a distinction between unfused and fused (which 433.1.5 does not mention) spurs?

No and no, when it comes to the cable on the supply side.

Would you agree that this is illogical as a 4mm² radial may have a 2.5mm² unfused spur?

Yes, but as observed, a 433.1.5 compliant ring final is an example of cookbook design.

 

[ban-all-sheds]

Indeed - but, as you recently agreed, that doesn't mean that it can't meet the requirements of 433.1.1 and be compliant with the remainder of BS7671.

Another suggestion was that a 1.5mm² cable from the ring to the FCU was a bodge. If it meets all requirements, how can it be a bodge?

I was pointing out that it doesn't meet all of the requirements.

I see nothing in 433.2.2 (or the corresponding 434.2.1 in respect to fault protection) that restricts it to any particular type of circuit. It is written in general terms and, as such, would appear to apply to any circuit - indeed, not even only final circuits.

433.1.5 refers to a RFC with or without 'unfused' spurs so I would suggest
that if the spur were a fused one then the requirement for 2.5mm² cable need not apply to the spur. Otherwise, when would 433.2.2 apply?

And a perfectly reasonable and consistent answer to what was, in a nutshell "If you can't use a downsized cable to supply a fused spur from a ring final when can you?" is "With radials".

 

[mikhailfaradayski]

Expand on what? - you quoted my entire message!

Just the bit immediately before my comment, the bit about you can envisage situations where your idea would be preferred over any other methods that are also available.

I didn't actually say 'preferred',

I never suggested you did. But I asked you when do you feel that your concept may be the preferred solution to a given problem

merely that it could offer some, mainly practical, benefits in some situations.

I'm not suggesting your idea is unacceptable, but i not convinced about any perceived practical benefits to either the installer or the user.

EFLImpudence has recently gone some way to answer this, when he wrote:

......John's original suggestion of a 4mm² spur on a 2.5mm² ring, I cannot see anything wrong with it as it saves the need for a FCU and it enables each socket to be used to ITS full rating rather than two or three (or however many) being limited to a total of 13A.
Also, if this spur should be, more or less, in a line from the ring. Rather than incorporating it into the ring, from one socket, with a cable joint and a second leg running back in the same channel as the first, a single run in 4mm² seems eminently preferable.

The design might be worth considering if, for example, one wanted to add a number of sockets to a circuit without restricting the total load on those sockets to 13A, particularly if those sockets were going to be in a circumscribed area somewhat sepseparateom the remainder of the ring, and especially if they were sockets which one did not expect to be used all that much (but nevertheless wanted >13A max potential capacity). The alternatives would be to run a totally separate circuit (ring or radial) or to incorporate those new sockets into the existing ring. The former would often not be the preferred choice,

Says who?

whilst the latter could result in an appreciable increase in total length of the ring. There are obviously advantages in keeping the length of a ring as short as possible, and it would seem a particular 'pity' to appreciably extend its length for the sake of some sockets which were not often going to be used. Then, of course, the purely practical issues - e.g. less cables to run/terminate (ans somewhat less copper used!)

Perhaps less copper, but try evaluating you idea against the 'cost of copper', not 'amount of copper' when you consider the the cost difference between 2.5 and 4mm. Also, a fused spur doesn't really cost very much, is it a big problem to just use one?

However, as both EFL Impedence and I have said, the main point being made is the general one that the examples and deemed-to-satisfy provisions in BS7671 are necessarily far from exhaustive

as well we know...

and that one would hope that those who design (and inspect) circuits would have the knowledge and ability to see beyond that and consider compliance with the totality of BS7671.
Kind Regards, John

Given your example above of supplying a number of sockets which may be remote to the rest of the RFC, avoidance of fusing down and constructed in such a way that it would increase the probability for the overload and disconnection of the OPD (albeit it in a safe manner). Don't you feel that although the circuit design may not fall short of safety levels called for in the BRB, some people may criticize your design choice when they consider 314.1?

 

[JohnW2]

Indeed - but, as you recently agreed, that doesn't mean that it can't meet the requirements of 433.1.1 and be compliant with the remainder of BS7671.

Another suggestion was that a 1.5mm² cable from the ring to the FCU was a bodge. If it meets all requirements, how can it be a bodge?

I was pointing out that it doesn't meet all of the requirements.

All of the 'requirements' of what? Are you perhaps making the mistake of regarding 433.1.5 as representing a 'requirement'. The 'requirement' is to satisfy BS7671 and that could be with a design that does not correspond to any examples or deemed-to-satisfy provisions (like 433.1.5) which BS7671 contains.

I see nothing in 433.2.2 (or the corresponding 434.2.1 in respect to fault protection) that restricts it to any particular type of circuit. It is written in general terms and, as such, would appear to apply to any circuit - indeed, not even only final circuits.

433.1.5 refers to a RFC with or without 'unfused' spurs so I would suggest
that if the spur were a fused one then the requirement for 2.5mm² cable need not apply to the spur. Otherwise, when would 433.2.2 apply?

And a perfectly reasonable and consistent answer to what was, in a nutshell "If you can't use a downsized cable to supply a fused spur from a ring final when can you?" is "With radials".

No-one is disputing that 433.2.2 applies to radial circuits (amongst other things). However, the question asked when 433.2.2 applies, and you seem to have decide that it only applies in certain situations (like radial circuits), although I can see nothing in BS7671 which says anything like that.

Kind Regards, John

 

[JohnW2]

.... if, for example, one wanted to add a number of sockets to a circuit ...... The alternatives would be to run a totally separate circuit (ring or radial) or to incorporate those new sockets into the existing ring. The former would often not be the preferred choice,

Says who?

I would imagine that many people (particularly those paying for the work) would feel that creating a completely new circuit would not be the preferred way of adding a few sockets. Maybe I'm wrong.

Also, a fused spur doesn't really cost very much, is it a big problem to just use one?

The reason for not having a fused spur would be nothing to do with cost but, as I said, would be to avoid the total capacity of the new sockets being limited to 13A.

Given your example above of supplying a number of sockets which may be remote to the rest of the RFC, avoidance of fusing down and constructed in such a way that it would increase the probability for the overload and disconnection of the OPD (albeit it in a safe manner)....

Could you perhaps explain that bit?

Kind Regards, John.

 

[ajrobb]

Given that the original idea was to add a 4mm² unfused spur to an existing 32 A ring, the earlier comments about the extra current carrying capacity of 2.5mm² must be disregarded unless you have the design to show that the circuit was originally installed for more than 20 A. Grouping together the two 2.5mm² cables run to a socket under plaster is adequate and running them together in conduit is worse but not unheard of.

You might as well talk of running showers or cookers off two parallel 2.5mm² T+E cables.

Or if you want to show off, how about using a plumbing solution and connecting only the line conductor of the first leg and the neutral of the last leg? This should increase the current carrying capacity of 2.5mm² T+E ring to over 32 A (Ref C no grouping) as only one conductor in the first and last legs is generating heat and the circuit lengths for all sockets on the ring are the same. In the intermediate legs, the current is split between conductors so generates less heat...

 

[JohnW2]

Given that the original idea was to add a 4mm² unfused spur to an existing 32 A ring, the earlier comments about the extra current carrying capacity of 2.5mm² must be disregarded unless you have the design to show that the circuit was originally installed for more than 20 A.

My reasoning didn't rely on any of that 'extra current-carrying capacity' - and it was not me who introduced that. My reasoning was simple, based on just two things:

• 1...There is presumably nothing contentious about current-carrying capacity (or anything else) within the spur itself - i.e. I presume that no-one has any problems with a string of two or more sockets being fed radially by 4mm² cable.
  
  2...As far as currents in the ring are concerned, the arrangement I was suggesting would be similar or identical to having two or more 2.5mm² spurs, each feeding one socket, attached at the same point, or close points, on the ring - which is not 'forbidden'.

In passing, I would add that if your "unless...." condition (above) could not be satisfied, I suspect that BAS might well say that it was a non-compliant ring, on the basis that 433.1.5 'requires' the cable to have a current-carrying capacity no less than 20A.

Or if you want to show off, how about using a plumbing solution and connecting only the line conductor of the first leg and the neutral of the last leg? This should increase the current carrying capacity of 2.5mm² T+E ring to over 32 A (Ref C no grouping) as only one conductor in the first and last legs is generating heat and the circuit lengths for all sockets on the ring are the same. In the intermediate legs, the current is split between conductors so generates less heat...

What can I say? In an attempt to be serious, wouldn't the problem there be with those first and last legs, with all the current being carried by one pair of 2.5mm² conductors? The tables in Appendix 4 of the regs don't include figures for just one single-core conductor. However, 4D1A indicated very little difference between (Method C) 2 x 2.5mm² single cores (27A) and 3 or 4 x 2.5mm² single cores (25A), so I suspect that 1 x 2.5mm² would probably be appreciably less than 32A - does anyone have any appropriate tabulated data?

However, you have done a dangerous thing, particularly on a Friday night, by 'giving me ideas' If one slightly modified your 'proposal' by upgrading the cable for just those first and last legs to 4mm², one might actually have quite an attractive design - with, as you say, equal circuit lengths for all sockets and equal splitting of current between conductors within the ring!!

Kind Regards, John

 

[ajrobb]

Any increase in current carrying capacity using only one of the twin conductors would be a factor less than √2 (i.e. the same heat dissipation). If you use the reciprocal of the grouping factor for two cables of 0.8 you have a factor 1.25, so for 27 A original capacity for T+E, there might be at least 33.75 A but less than 38 A capacity using only a single conductor. If the cable is installed in thermal insulation, the additional capacity would approach √2 as the temperature drop within the cable becomes much less than the temperature drop in the thermal insulation. That gives a little but a minimum T+E current carrying capacity would be more than 22.6 A in order to allow 32 A in a single conductor.

The maximum cable length would be considerably smaller than for a normal ring as all the current has to flow around the whole ring. The current is more than a radial but the effective circuit length is half that of a radial circuit (consider a point load at the centre of the ring; the current flows up one leg and back down the other).

Any break in either conductor would kill the circuit (unlike rings and radials) and there are still two independent CPC routes back to the CU.

If you want to add additional safety, the disconnected ends of each conductor can be used to monitor voltage drop, so loose terminals can be detected before they burn. Alternatively, you can sleeve the unused conductor in the first and last leg and use both in order to increase the capacity of the cable to double that of nominal T+E rating. This will help where cables are grouped near the CU or have to pass through insulation (so your 4mm² suggestion is redundant.)

 

[JohnW2]

Any increase in current carrying capacity using only one of the twin conductors would be a factor less than √2 (i.e. the same heat dissipation). If you use the reciprocal of the grouping factor for two cables of 0.8 you have a factor 1.25, so for 27 A original capacity for T+E, there might be at least 33.75 A but less than 38 A capacity using only a single conductor. If the cable is installed in thermal insulation, the additional capacity would approach √2 as the temperature drop within the cable becomes much less than the temperature drop in the thermal insulation. That gives a little but a minimum T+E current carrying capacity would be more than 22.6 A in order to allow 32 A in a single conductor.

I can't disagree with the logic whereby you've arrived at those figures, but the tables in Appendix 4 of the regs don't seem to use the same logic! If you had considered using only two of four touching single 2.5mm² conductors (25A capacity for all four, per 4D1A), the √2 maximum increase would still apply. The grouping factor would be 0.65, but since the reciprocal of this (1.54) is greater than √2, you would probably say that the maximum possible increase in capacity would be limited to √2. This would presumably have led you to write (in analogy with above) "... so for 25A original capacity for 4 conductors, there might be as much as 35.35A capacity using only two of the conductors". However, as we know, 4D1A indicates only 27A for two conductors. I obvioulsy can't tell you how they arrived at that, but it's the reason I suspected that the figures they gave for one single 2.5mm² conductor (if they did) might well not be much more than their 27A figure for two of them.

The maximum cable length would be considerably smaller than for a normal ring as all the current has to flow around the whole ring.

Agreed.

The current is more than a radial but the effective circuit length is half that of a radial circuit (consider a point load at the centre of the ring; the current flows up one leg and back down the other).

The relative effective circuit lengths presumably depend upon exactly what arrangements of radial and ring one is comparing. I would have thought that, with some arrangements, they could be quite similar.

Any break in either conductor would kill the circuit (unlike rings and radials) and there are still two independent CPC routes back to the CU.

Indeed!

f you want to add additional safety, the disconnected ends of each conductor can be used to monitor voltage drop, so loose terminals can be detected before they burn.

Yep!

Alternatively, you can sleeve the unused conductor in the first and last leg and use both in order to increase the capacity of the cable to double that of nominal T+E rating. This will help where cables are grouped near the CU or have to pass through insulation (so your 4mm² suggestion is redundant.)

Yes, I thought of that soon after I posted my last message. As you say, one could make use of those unused 2.5mm² conductors, rather than use 4mm².

Seriously, though, although I presume this was all intended to be fairly tongue-in-cheek, it actually does sound as if it has a fair bit going for it, and the only major downside I've thought of so far is the restriction in maximum ring length which you have mentioned. Has such an arrangment ever been put forward (other than by you!) as a serious proposition?

Kind Regards,John.

 

[mikhailfaradayski]

.... if, for example, one wanted to add a number of sockets to a circuit ...... The alternatives would be to run a totally separate circuit (ring or radial) or to incorporate those new sockets into the existing ring. The former would often not be the preferred choice,

Says who?

I would imagine that many people (particularly those paying for the work) would feel that creating a completely new circuit would not be the preferred way of adding a few sockets. Maybe I'm wrong.

I can't be certain whether you are right or wrong here, as obviously, its impossible to canvass the opinion of 'everyone', but I'm not convinced that you are right.

Also, a fused spur doesn't really cost very much, is it a big problem to just use one?

The reason for not having a fused spur would be nothing to do with cost but, as I said, would be to avoid the total capacity of the new sockets being limited to 13A.

Sorry, its just that somewhere in this thread you said that you could see some practical benefits to your concept, I thought you may have been talking about costs as a benefit (this seems to be supported by your further comment about people paying the bill)

Given your example above of supplying a number of sockets which may be remote to the rest of the RFC, avoidance of fusing down and constructed in such a way that it would increase the probability for the overload and disconnection of the OPD (albeit it in a safe manner)....

Could you perhaps explain that bit?

sure, although it might be best if you told me what it is you want explaining further as it seemed quite straight forward to me.
But before we explore that, would you mind going back to my post to address the rest of it. Particularly the bit about 314.1

Kind Regards, John.

 

[ajrobb]

Seriously, though, although I presume this was all intended to be fairly tongue-in-cheek, it actually does sound as if it has a fair bit going for it, and the only major downside I've thought of so far is the restriction in maximum ring length which you have mentioned. Has such an arrangment ever been put forward (other than by you!) as a serious proposition?

It was originally a tongue-in-cheek suggestion, but it looks close to being a reasonable idea. I only came up with it last night in mid-post, so I cannot say it is thought through. Obviously with only a single active conductor, the heat generated will be localised, which can cause extra problems. However, we know that 2.5mm² T+E can carry 30 A in mid-air in both conductors, which is tantalisingly close to the required 32 A in a single conductor (dissipating 43% less heat overall). That suggests that the maximum insulation temperature (at the surface of the active conductor) might still be kept within bounds.

IF it works at all, it would also be better able to accept point loads like a 32 A unfused spur at any point rather than just in the middle for a standard ring.

I've only been considering PVC insulation so far. Thermosetting or mineral insulation would be less affected by concentrating resistive heat in a single conductor. It is clearly not a retro-fit to existing rings and could only be considered for new circuits.

Just considering a 5% voltage drop at 18 mV/A/m:

230V × 5% ÷ 32A ÷ 18mV/A/m = 20m (radial length)

Which equates to a maxim ring length of 40m, which is a lot less than 106m for a standard ring.