3m max meter tails - DNO rule or BS7671?

[JohnW2]

That situation would be the same if the tails were not over three metres in length, in which case the DNO fuse is not adequate.

Exactly. Unless they were ridiculously long, their length would not have any appreciable effect on fault protection, so if a second fuse of the same rating of the DNO fuse would not provide adequate fault protection to tails >3m, then nor would the DNO fuse it usually give adequate fault protection for tails <3m.

Kind Regards, John

 

[JohnW2]

True! Although they don't say put the same rating, they actually say in that doc
""" ....... """ .... Which basically means you might not get 100a through one switch fuse if your electrician can't get the disconnection time low enough.

Whilst that obviously makes sense, I wonder if anyone ever thinks about that, or whether the DNOs have thought of the implications?

Whenever I've seen it done, an 80A (or sometimes 100A) is just put into to (single) 'switch fuse' (probably without any thought or calculation). ... and it would be very odd for the DNO to 'provide' an '80A supply' or '100A supply' if they knew that the Ze could be as high as 0.8Ω, such that there would be no way it could be 'safely' connected to the DB/CU through a single pair of conductors.

Kind Regards, John

 

[John D v2.0]

would be very odd for the DNO to 'provide' an '80A supply' or '100A supply' if they knew that the Ze could be as high as 0.8Ω

Maybe that's why the limit is 0.35 for new supplies (are all tn-c-s) and they will offer to convert you if they are doing other work.

Also, if the ze is 0.8 then that means you could indeed have a 100a supply without fault protection on the tails (hence the reg mentioned) therefore you could be compliant with bs7671 by keeping the length to 3m, making damage unlikely, and Henley blocking into two or more lower rated circuits to get your full 100a. For every bs7671 circuit you would check ads so there's no excuse for not doing it on a (sub) main.
Hopefully your external volt drop wouldn't be too high when you were taking the full 100a though....

 

[JohnW2]

Maybe that's why the limit is 0.35 for new supplies (are all tn-c-s) and they will offer to convert you if they are doing other work.

Maybe - but then, if the DNO's fuse provides adequate fault (and overload) protection for the tails, even if a 100A fuse, then that will often (usually?) remain the case for any likely length of tails, won't it? As I said before, the increase in loop impedance due to tails is unlikley to be 'appreciable' unless they are extremely long.

Also, if the ze is 0.8 then that means you could indeed have a 100a supply without fault protection on the tails (hence the reg mentioned) therefore you could be compliant with bs7671 by keeping the length to 3m, making damage unlikely, and Henley blocking into two or more lower rated circuits to get your full 100a.

I think there may be some confusion over these regulations. There are actually a pair of relevant ones, 433.2.2 (which has been cited and discussed) and 434.2.1, which are partially 'mirror images' of one another.

The reg which has been discussed (433.2.2) is about overload protection - which it says that (although still required) may be downstream of a reduction in CSA if EITHER there is adequate fault protection OR if the part downstream of a reduction in CSA is <3m and 'with damage unlikel'. I'm not sure that is really relevant to what we are discussing, since there is not usually going to be adequate overload protection of the tails at the DB/CU end of the tails.

Reg 434.2.1 is about fault protection - which it says that (although still required) and provided that 433.2.2 has not been invoked (i.e. overload protection provided downstream of a reduction in CSA) may be downstream of a reduction in CSA if the part downstream of a reduction in CSA is <3m AND 'with damage/fault unlikely'.

So, as I see it, neither of those regs allow for the omission of either adequate overload protection or adequate fault protection of the tails. they merely allow for one or other (but not both) of those protections to be downstream of a reduction in CSA.

For every bs7671 circuit you would check ads so there's no excuse for not doing it on a (sub) main.

Indeed, but with 16mm tails and a 0.8Ω Ze, that 'checking' is likely to show that if the (upstream) protection is from an 80A or 100A fuse (in cutout and/or switch-fuse) that the fault protection of the tails is inadequate. One could, as I think you are suggesting, split the distribution into two or more switch-fuses, each with a fuse ≤60A, but that would rather frustrate invocation of 'across installation diversity' - e.g. the tails supplying a shower circuit would probably have to be dedicated to that circuit.

However, my personal installation illustrates a couple of other issues.

Firstly, as has been somewhat discussed, I don't think it can be said that there is any 'reduction in CSA' at the origin of my tails. The supply enters my house (and remains the same for tens of metres (overhead) upstream of my installation) in ~16mm² and all my tails are at least 16mm². That being the case, it appears that neither of the above regulations apply.

Secondly, mine is a TT installation. Only two of my many CUs are within 3m of the meter, most of the distribution circuits being pretty long, and, as things have been done, are protected at their origin by 60A or 80A switch fuses. However, although all final circuits are RCD/RCBO protected, I have retained the 'up front' TD RCDs in order to provide adequate fault protection for the distribution circuits. That being the case, I think I could theoretically invoke 433.2.2 and omit overload protection (i.e. NOT have switch-fuses) because, despite their length, the tails enjoy adequate fault protection (from TD RCDs).

Kind Regards, John

 

[EFLImpudence]

The reduction in c.s.a. is a red herring.

If their were a reduction then a switch-fuse would be fitted as a matter of course to give overload protection if required - i.e. 16mm² tails with 100A DNO fuse.

The discussion is about the demand for a switch-fuse by the DNOs solely because of the length of tails.

 

[JohnW2]

The reduction in c.s.a. is a red herring.

I agree, but people have introduced that by talking about 433.2.2 - which does NOT apply if there is no reduction in CSA. Furthermore, it does not apply unless one is trying to make use of downstream protection - which, again, is not the case in relation to what we're discussing.

The discussion is about the demand for a switch-fuse by the DNOs solely because of the length of tails.

Indeed - and, as we've both said, it makes little/no sense. If the DNO fuse does not give adequate protection to tails just over 3m in length then (except in very marginal cases) nor would it give adequate protection to tails <3m in length.

As has been said/implied, in the case of a 100A (and probably also 80A) service fuse, if the Ze may be as high as 0.8Ω, then the requirements for ADS could never be satisfied, no matter what the length or CSA of the tails - so all one can really hope to achieve with a switch-fuse is to provide overload protection to the tails - and even that it probably really only relevant when the DNO fuse is 100A. However, again, that would be regardless of the length of the tails.

Kind Regards, John

 

[EFLImpudence]

Agreed, but 0.8Ω is just a bad case of 'what if'.

If it is a more usual 0.2Ω then all is well.

 

[JohnW2]

Agreed, but 0.8Ω is just a bad case of 'what if'. If it is a more usual 0.2Ω then all is well.

Also Agreed.

However, if they don't feel able to guarantee less than 0.8Ω, then they have essentially trapped themselves into having to accept/admit that their service fuse might not give adequate fault protection to the tails - but that remains the case regardless of the length of the tails, so cannot explain a '3 metre rule' (or, indeed, any 'X metre rule').

What about what I said on the basis of my installation? Do you think that (for a TT and/or TN installations), regs-wise one could avoid the need for any additional protection (overload or fault) for tails (of any length), by providing fault protection by means of an RCD (possibly TD, like mine) at the origin of the tails?

Kind Regards, John

 

[SUNRAY]

Hi all, I was aware DNOs usually have a 2 or 3m rule on tail length, which I assumed for their own reasons (simplicity/flexibility and consistency I guessed). However I was having a read of UKPNs doc and they pin the blame on BS7671 for that rule. That was unexpected.



From:
http://library.ukpowernetworks.co.u...Customer+Supplies+up+to+100A+Single+Phase.pdf

The part of the reg they cite apparently applies only to circuits without adequate *fault* protection.

To determine the adequacy of fault protection purely for BS7671 you would just need the total Zs, the adiabatic parameters for the cables involved and the fuse characteristics?

So it stands to reason that they are saying their fuse doesn't provide adequate fault protection for any length of tails. They are also saying it's not the "overload protection device" for the tails for BS 7671 purposes either.

In that case what is it for? Unspecified DNO protection purposes? Or is it just there to allow meter operators to safely isolate the supply by removing it?

Thoughts?

A new block of 6 flats 2016 with multi-way head, meters and isolators (Deffo not fused) in a cupboard and 25mm² singles (not DI) in trunking to each flat.
Found recently when we altered the communal area electrics which had been supplied from one of the flats CU.
In steel all the way.

 

[EFLImpudence]

However, if they don't feel able to guarantee less than 0.8Ω, then they have essentially trapped themselves into having to accept/admit that their service fuse might not give adequate fault protection to the tails - but that remains the case regardless of the length of the tails, so cannot explain a '3 metre rule' (or, indeed, any 'X metre rule').

True.

What about what I said on the basis of my installation? Do you think that (for a TT and/or TN installations), regs-wise one could avoid the need for any additional protection (overload or fault) for tails (of any length), by providing fault protection by means of an RCD (possibly TD, like mine) at the origin of the tails?

Well, it depends what you want. That wouldn't be omitting fault protection, would it?
Overload may be covered by the CU - possibly not in a large installation.

RCD won't protect against short-circuit (is the neutral loop alright) and domestic tails don't usually have the liklihood of an earth fault - although with metal CUs I suppose they do now.
Switch-fuse and RCD? or lower rated DNO fuse.

Lots of ifs and buts.


What is TD?

 

[JohnW2]

Well, it depends what you want. That wouldn't be omitting fault protection, would it?

Well, as you go on to say, it would be omitting protection against L-N faults. ... and that is the flaw in the suggestion I was making!

... and domestic tails don't usually have the liklihood of an earth fault - although with metal CUs I suppose they do now.

Yep - something else for which we have to thank the LFB !

RCD won't protect against short-circuit (is the neutral loop alright) ... Switch-fuse and RCD? or lower rated DNO fuse.

The L-N loop impedance is low enough for the DNO fuse to provide adequate protection against (L-N) faults (and adequate overload protection for the tails) - so all my switch fuses (same rating fuse as DNO's) are presumably actually redundant.

What is TD?

'Time-delayed', a.k.a. Type S.

Kind Regards, John

 

[EFLImpudence]

The L-N loop impedance is low enough for the DNO fuse to provide adequate protection against (L-N) faults (and adequate overload protection for the tails)

Yes, I suppose I still had in mind the example of 0.8Ω Ze.

so all my switch fuses (same rating fuse as DNO's) are presumably actually redundant.

I suppose so.

'Time-delayed', a.k.a. Type S.

D'oh.

 

[JohnW2]

I suppose so.

And that takes us back to the start of this discussion. Even though, in a situation like mine, the DNO fuse gives adequate (L-N fault and overload) protection to my 'tails' (distribution circuits), I have little doubt that they would nevertheless try to insist on my installing switch-fuses (which could have the same rating fuse as theirs), if I didn't have them, because of the (considerable) length of some of those distribution circuits - even though that does not really make any sense (even though I've done it!).

Kind Regards, John

 

[EFLImpudence]

After reading what was originally stated as be the reason (apart from the ludicrous 'or isolator') I think it highly probable that a misreading lead to the rule in the first place and no one can question the instructions of the DNOs.

 

[JohnW2]

After reading what was originally stated as be the reason (apart from the ludicrous 'or isolator') I think it highly probable that a misreading lead to the rule in the first place and no one can question the instructions of the DNOs.

Yes, that's quite possible.

What seems particularly odd to me is that I didn't think DNOs had any particular interest in anything downstream of their cutout or, at least, downstream of the meter - at least, unless it was something that could adversely affect their network. As far as I can see, their only conceivable interest would be in the health of their fuse, but no additional fuse likely to be installed (for tails >3m) would be likely to discriminate against theirs, anyway (particularly if, as would probably often be the case, it was of the same rating as their fuse)!

Kind Regards, John