CHESTNUT

[holmslaw]

..

 

[JohnW2]

(1)... it is only ok if "the load current in part of the circuit is unlikely to exceed for long periods the current carrying capacity of the cable. In other words you need absolute assurance that the spur cannot be overloaded, and without that assurance unfused spurs should not be installed.
(2)... Apart from that, it is irresponsible dangerous nonsense to say that cables <3metres do not need overload protection, without stating that they also need "to be installed in such a manner as to reduce the risk of fire and danger to persons." Which means they would never be acceptable in a normally occupied area or room.

Whilst, as you will be aware, I am to a large extent on your side in relation to this disagreement with BAS, I think you have overinterpreted the regs above. As per my highlighting, "unlikely to exceed for long periods" is not a requirement for 'absolute reassurance' that such cannot happen, and "reduce the risk" does not mean that it is never acceptable in a normally occupied room.

Another point not yet mentioned is that, if one regards the OSG as the IET's interpretation of BS7671, they seem to be potentially giving their blessing to unfused spurs which are >3m in length, hence not covered by the provisions for having downstream overcurrent protection we've been discussing. The OSG says that, 'as a rule of thumb' the length of unfused spurs should not exceed one eighth of the cable length from the spur to the furthest point in the ring - in some cases, that could exceed 3m.

Kind Regards, John.

 

[JohnW2]

I don`t think any of us would like to see a circuit with one twin socket in 2.5 T & E on a 32A MCB cos it don`t look right. Similarly non of us would like to see several radials, each with one twin socket connected to one fuseway even if the fuseway could comfortably accept that number of conductors , again it would not look right.

Agreed.

But both of those examples are no more or less onerous than the bog standard ring final that we do accept.

Again, agreed. However, it would be wrong to assume that the regs are logical or consistent. Hence, the fact that those things are, in engineering terms, no different (and no less safe) from what is allowed by the regs in the presence of a ring final circuit cannot necessarily be assumed to mean that they would still be compliant when the ring final itself is not present.

Indeed (per above discussion), if the cable to the socket is >3m in length, it's difficult to see how, in the absence of a ring final circuit, the regs can be interpreted as allowing a situation in which the OPD had an In greater than the Iz of the cable - regardless of what the cable is feeding. The fact that one could undoubtedly 'legitimise' it by adding a 'token' ring (are accessories required? could it be just a 2 inch ring of conductor within the CU?) is daft, but it wouldn't be the first time that such was true of regulations.

Kind Regards, John.

 

[ban-all-sheds]

When someone resorts to lying you know they've lost the argument,

holmslaw - when you stop behaving like a disgusting scumbag in making appalling and false allegations with no evidential basis whatsoever that I am deliberately writing things I know to be untrue in order to deceive people I might be inclined to deal with the other parts of your posts.

 

[ban-all-sheds]

Again, agreed. However, it would be wrong to assume that the regs are logical or consistent. Hence, the fact that those things are, in engineering terms, no different (and no less safe) from what is allowed by the regs in the presence of a ring final circuit cannot necessarily be assumed to mean that they would still be compliant when the ring final itself is not present.

The regulations do not differentiate between spurs from any point on a ring, spurs from the origin, spurs which become radials if the ring is removed or circuits which were always radials because there never was a ring.

In all cases the same considerations of downstream overload protection, upstream fault protection, length, risk of damage etc apply equally.

Indeed (per above discussion), if the cable to the socket is >3m in length, it's difficult to see how, in the absence of a ring final circuit, the regs can be interpreted as allowing a situation in which the OPD had an In greater than the Iz of the cable - regardless of what the cable is feeding. The fact that one could undoubtedly 'legitimise' it by adding a 'token' ring (are accessories required? could it be just a 2 inch ring of conductor within the CU?) is daft, but it wouldn't be the first time that such was true of regulations.

Would make no difference - 433.2/.3/434 apply to spurs as well as radials.

 

[holmslaw]

..

 

[JohnW2]

The regulations do not differentiate between spurs from any point on a ring, spurs from the origin, spurs which become radials if the ring is removed or circuits which were always radials because there never was a ring. In all cases the same considerations of downstream overload protection, upstream fault protection, length, risk of damage etc apply equally.

I agree that BS7671, per se, does not appear to make any such differentiation. However, as I recently wrote, if one assumes that the OSG represents the IET's interpretation of their own regulations (BS7671) (or, at least, is believed by them to be compliant with BS7671), appears to give its blessing to unfused spurs which could be >3m in length - which, as far as I can make out, would otherwise be non-compliant with BS7671 in relation to the issues you mention. What do you think?

Indeed (per above discussion), if the cable to the socket is >3m in length, it's difficult to see how, in the absence of a ring final circuit, the regs can be interpreted as allowing a situation in which the OPD had an In greater than the Iz of the cable - regardless of what the cable is feeding. The fact that one could undoubtedly 'legitimise' it by adding a 'token' ring (are accessories required? could it be just a 2 inch ring of conductor within the CU?) is daft, but it wouldn't be the first time that such was true of regulations.

Would make no difference - 433.2/.3/434 apply to spurs as well as radials.

See above. Could you please clarify - are you saying that you think an unfused (2.5mm²) spur from a 32A ring final is non-compliant if it is >3m in length? If so, how does the IETs statement in the OSG fit with that?

Kind Regards, John.

 

[ban-all-sheds]

However, as I recently wrote, if one assumes that the OSG represents the IET's interpretation of their own regulations (BS7671) (or, at least, is believed by them to be compliant with BS7671), appears to give its blessing to unfused spurs which could be >3m in length - which, as far as I can make out, would otherwise be non-compliant with BS7671 in relation to the issues you mention. What do you think?

I think I've not seen that in the OSG.

Where is it?

 

[JohnW2]

I think I've not seen that in the OSG. Where is it?

Second paragraph (sentence) of 7.2.2 on page 54 - of the red (original BS7671:2008) OSG - I haven't got the 'Amendment 1' one.

Kind Regards, John.

 

[ban-all-sheds]

Fair enough.

They missed out "or 3m, whichever is the smaller".

The OSG is not meant to reproduce the entire regulations, there is much which it omits - doesn't mean those things don't apply.

are you saying that you think an unfused (2.5mm²) spur from a 32A ring final is non-compliant if it is >3m in length?

Yes.

 

[JohnW2]

Fair enough. They missed out "or 3m, whichever is the smaller". The OSG is not meant to reproduce the entire regulations, there is much which it omits - doesn't mean those things don't apply.

Well, that's your interpretation. However, they may have missed out that qualifier deliberately, knowing (since they wrote it) that the intent of BS7671 was to allow unfused spurs >3m in length. As for whether or not that was the intent (we can only guess), don't forget that 433.1.103 of BS7671 does allow things in ring final circuits which would otherwise be non-compliant, and allows "ring final circuits, with or without unfused spurs", without any qualification about the length of those spurs (and this is 'the regs', not the OSG). I therefore think, but obviously do not know, that it is possible that the intent was to allow unfused spurs of any length. Even stranger, 433.1.103 does not explicitly restrict an unfused spur to supplying only one socket (Appendix 15 effectively does, but that's only 'Informative'), but one could argue that it is implicity caught by "...the load current in any part of the circuit is unlikley to exceed for long periods the [Iz] of the cable"

are you saying that you think an unfused (2.5mm²) spur from a 32A ring final is non-compliant if it is >3m in length?

Yes.

That would certainly be consistent with the remainder of the regs, but my point above about the 'intent' (actually word) of 433.1.103 again applies.

I may be missing something, but I'm also not sure that I understand the logic behind the 'rule of thumb' in the OSG. Why should the 'recommended' maximum length of an unfused spur increase if one is a long way from 'the furthest point on the ring' (the definition of which is not totally clear)?

Kind Regards, John.

 

[ban-all-sheds]

Well, that's your interpretation. However, they may have missed out that qualifier deliberately, knowing (since they wrote it) that the intent of BS7671 was to allow unfused spurs >3m in length.

Or they missed it out, along with countless other things, because they know that BS 7671 defines all of the relevant requirements properly.

Read the 2nd paragraph of the Preface and the 1st sentence of the Foreword.

As for whether or not that was the intent (we can only guess), don't forget that 433.1.103 of BS7671 does allow things in ring final circuits which would otherwise be non-compliant, and allows "ring final circuits, with or without unfused spurs", without any qualification about the length of those spurs (and this is 'the regs', not the OSG).

Do you think, when reading the regulations to determine the requirements for a particular installation component that it is a case of:

1) You find one regulation which defines every single requirement for that component

or

2) You need to look at several regulations to determine all of the requirements for that component

?

433.1.103 is but one part of 433. If you read the rest of it, what do you find about the requirement for the lengths of unfused spurs?

I therefore think, but obviously do not know, that it is possible that the intent was to allow unfused spurs of any length.

It's not. Read all of 433, and by reference within that 434.

Even stranger, 433.1.103 does not explicitly restrict an unfused spur to supplying only one socket (Appendix 15 effectively does, but that's only 'Informative'), but one could argue that it is implicity caught by "...the load current in any part of the circuit is unlikley to exceed for long periods the [Iz] of the cable"

No, it's explicitly "caught" by other parts of 433.

That would certainly be consistent with the remainder of the regs, but my point above about the 'intent' (actually word) of 433.1.103 again applies.

Well - I've already made my point which is that you can't take one regulation in isolation and expect it to be a self-contained definition of every aspect of compliance with BS 7671.

I may be missing something, but I'm also not sure that I understand the logic behind the 'rule of thumb' in the OSG. Why should the 'recommended' maximum length of an unfused spur increase if one is a long way from 'the furthest point on the ring' (the definition of which is not totally clear)?

I have no idea either.

 

[JohnW2]

Or they missed it out, along with countless other things, because they know that BS 7671 defines all of the relevant requirements properly. ... or 2) You need to look at several regulations to determine all of the requirements for that component .... 433.1.103 is but one part of 433. If you read the rest of it, what do you find about the requirement for the lengths of unfused spurs?

Well, provided that the MCB and/or RCD (for TT) in/around the CU provides adequate fault protection (434.2.2) for the spur cable (which I imagine it virtually always will), I would imagine that one (particularly you) could argue that 433.3.1(ii) allows one to have no overload protection for a spur cable (of any length), per se, since you would presumably argue (as you have done) that the fact that there is a '20A maximum' socket on the end of the spur means that "it is not likely to carry overload current".

If one invoked that reg to argue that no overload protection of the spur cable is required, 433.1.1 (which is what would impose the 3m length limit) clearly would become irrelevant and inapplicable.

Even if you're now nervous about invoking your '20A maximum load' view of a double socket in the manner described above, the above argument could surely be used to conclude that a spur feeding a single socket would not usually be subject to a 3m length limit?

Kind Regards, John

 

[ban-all-sheds]

Well, provided that the MCB and/or RCD (for TT) in/around the CU provides adequate fault protection (434.2.2) for the spur cable (which I imagine it virtually always will), I would imagine that one (particularly you) could argue that 433.3.1(ii) allows one to have no overload protection for a spur cable (of any length), per se, since you would presumably argue (as you have done) that the fact that there is a '20A maximum' socket on the end of the spur means that "it is not likely to carry overload current".

I'm beginning to despair.

The fact that you have a socket at then end means that you have BS 1362 fuses providing the protection against overload as referred to in 433.2.2 (as was, if Amd 1 has changed the number). You don't have to argue that it is not likely to carry overload current - you have a device protecting the cable against overload installed along the run of it.

If one invoked that reg to argue that no overload protection of the spur cable is required, 433.1.1 (which is what would impose the 3m length limit) clearly would become irrelevant and inapplicable.

But I'm not, and never suggested I was.

Even if you're now nervous about invoking your '20A maximum load' view of a double socket in the manner described above, the above argument could surely be used to conclude that a spur feeding a single socket would not usually be subject to a 3m length limit?

Not only am I still not nervous about that, in the context of proper standards-compliant design, I would never need to argue that because I would never argue for a spur without overload protection.


We should get a point of accuracy resolved - yes, in theory 434.2 can be satisfied without invoking the 3m limit if you want to do the adiabatic calculation in 434.5.2, and if you have 6kA devices or manufacturers data for 10kA ones (k²S² for a 2.5mm T/E when considering a L-N fault is 82,656 which is less than the requirement for 10kA devices in BS EN 60898).

For L-N faults k²S² drops to 29,756 - too low even for a 6kA device, so you're stuffed if there's no RCD (and possibly if there is - I don't know what the let-through is for those).

 

[JohnW2]

The fact that you have a socket at then end means that you have BS 1362 fuses providing the protection against overload as referred to in 433.2.2 (as was, if Amd 1 has changed the number). You don't have to argue that it is not likely to carry overload current - you have a device protecting the cable against overload installed along the run of it.

Indeed, you didn't need to argue that with the approach you were suggesting. However, my point is that one could use the presence of the BS1362 fuse(s) (certainly for a single socket, possibly/probably for a double one) to indicate that that the spur cable was "not likely to carry overload current". If one did that, then, provided fault protection was OK, there would (per regs) be no need for any overload protection, hence 433.2.2 (number hasn't changed), with its 3m limit, would become irrelevant.

As for the fault protection, I need to think about what you've said but, on the face of it, I can't see why achieving an adequate degree of fault protection should be any more difficult in this situation than it is for any other circuit (e.g. a 20A socket radial) using 2.5mm² cable.

Kind Regards, John.