New 4mm T&E feed to existing remote 2.5mm ring. OK or not?

[Mark42]

I'm building a small rental annexe at my own house, which will be supplied by a new local sub-main, with local metering. This includes one existing bedroom which will become part of the annexe.
This room has an existing ring final in 2.5mm, all in good condition. Rather than run two new 2.5s, a long way, in parallel to the new DB, may I run a single 4mm, then connect the existing 2.5mm 'tails' (now pulled back from the main house DB) to this locally? One less joint, and one less awkward run of cable. Protection would be a 25A RCBO.
I can think of no reason why this is not OK electrically, but I've never seen it done. It feels a bit 'bodgy' but so is running two great lengths of 2.5 in parallel for no particular reason.
Have I missed anything? Are there any regs about the CCA change within a circuit which might apply?

 

[GBDamo]

mmm, lollipops, I like lollipops.

 

[ipf]

Seems to me you should be getting someone experienced in to sort this. Messing about with existing 3 phase, especially involving an old farm supply, sounds pretty dodgy. Local authority will need informing too, presumably.

 

[Charlie_]

“fed by 5-core 16mm NYY. It's being done properly”
A contradiction in terms

 

[Pete999]

“fed by 5-core 16mm NYY. It's being done properly”
A contradiction in terms

Agree for fun

 

[Mark42]

“fed by 5-core 16mm NYY. It's being done properly”
A contradiction in terms

Why?

 

[Charlie_]

I wouldn’t use nyy for that purpose.
I would use swa and do it properly.
What is the installation method?

 

[Mark42]

I wouldn’t use nyy for that purpose.
I would use swa and do it properly.
What is the installation method?

Hi Charlie. Rather than my go through all the arguments again , please see my other thread:
NYY for a sub-main
I have yet to hear a convincing reason why one should always follow 'convention' and use SWA, especially in internal-only and loft space domestic installations. And especially not to follow the other 'convention' of bloody horrible T&E for sub-mains, which I think is amateurish. Even if it is compliant (sometimes).

 

[Mark42]

Seems to me you should be getting someone experienced in to sort this. Messing about with existing 3 phase, especially involving an old farm supply, sounds pretty dodgy. Local authority will need informing too, presumably.

Oh dear, here we go again. But yes, you're probably right. I have been 'messing about' with this for only 50 years, so need a rest and perhaps it is time I gave way to someone with more experience. Who do you know? Preferably someone who's still alive.

 

[Charlie_]

Maybe this will help

 

[Mark42]

Maybe this will help ...

Yes, I know, but if NYY is so clearly non-compliant (and in this case only according to the opinion of one bloke who wrote the paper), then why is it so widely sold?

And can you say you have never used NYY (or, previously, 'Hi-tuf'), or SY anywhere, ever, in an installation, and that you would fail on inspection, any installation which contains it? If not then no offence please, but your argument rather fails by counter-example.

I prefer to make technical judgments based on science and direct observation rather than on specious arguments based often on commercial protectionism between suppliers rather than on technical merit.

 

[Mark42]

OK, back to the original question.

The reason I enquired about the 'lollipop' circuit is that although it's obviously OK electrically (and I didn't want to suggest that outright so as not to be accused of technical immodesty) I had, and still have, a sneaking suspicion that there's a reg somewhere about changing conductor size mid-circuit, which may inadvertently prohibit the practice. I've forgotten a lot of things, and hoped someone here might refresh the memory

Alternatively, I'll give up, remove one leg of the RFC from the 'final' socket, join the other leg to more 2.5mm, protect it at 16A and make it a radial.

I've never liked rings anyway.

That solution is clearly 100% compliant, and easy to understand by those who come after me, but maybe the potentially non-compliant alternative would have been a better-engineered way of doing the job.

 

[ipf]

Oh dear, here we go again. But yes, you're probably right. I have been 'messing about' with this for only 50 years, so need a rest and perhaps it is time I gave way to someone with more experience. Who do you know? Preferably someone who's still alive.

Well, you know it all, anyway. Why bother, your minds made up. Right or wrong, if it works, you'll be fine, I'm sure.
Still messing, eh? 50 years for a piddling little job says it all.

 

[Charlie_]

It’s not just one bloke.
Its basec and many others.
Those cables have their uses but not for what you are proposing.
Yes I do fail installations based on their non compliant use.
You mentioned yy and I offered you an opinion.

Based on your technical expertise I’m surprised you are stumped by such a very simple rfc problem.

 

[Mark42]

Well, you know it all, anyway. Why bother, your minds made up. Right or wrong, if it works, you'll be fine, I'm sure.
Still messing, eh? 50 years for a piddling little job says it all.

Ha Ha! You completely misjudge me. I'm a scientist by nature and training: it's always a pleasure to be proved wrong and hence to learn something new. That's how science works.

But rather than receive an ad hominem attack from behind anonymity, I'd prefer to hear why, technically, what I propose is 'wrong'. Explain that, clearly, and I'd be the first to thank you for teaching me something new.

Go ahead please.

Cheers, Mark

 

[Mark42]

... Those cables have their uses but not for what you are proposing.
Yes I do fail installations based on their non compliant use.
You mentioned yy and I offered you an opinion....

OK, fair enough, and thank you for the opinion.

Here's the spec of what I ordered:
NYY spec
Are you sure you think it's non-compliant?

It's interesting to get opinions on this forum, all from experienced and qualified electricians, but those opinions are often in direct opposition. All cannot be right.