two ring mains joined

[JohnW2]

If there is a break in the second ring, the cable could be overloaded.

Hi there; I trust all is well with you.

Why is what you say peculiar to a 'second ring'? If there were a break in the one-and-only ring of a standard rung final, the cable could also become overloaded. In fact, as I've said, the main effect of adding any cross-connections to a ring is to reduce the risk of any part of the cable becoming overloaded [whether in normal use, or in the presence of a break anywhere in the ring(s)].

Kind Regards, John

 

[JohnW2]

I don't know any 'BS7671-trained cable installer' who would install this circuit. It must have been done by a 'real' electrician.

That's surely the point that was being made - that only someone who actually 'understood' could possibly contemplate installing such a circuit or know how to test it - since it's not in the "cable installer's cookbook"?

Kind Regards, John

 

[PBC_1966]

On a related issue, I've heard some who go by the "cable installer's cookbook" comment that they would code 1 a 2.5 sq. mm cable feeding a single 13A socket protected by a 30/32A fuse/MCB. Their view is it's not a standard BS7671 circuit arrangement and is dangerous because of the protective device rating exceeding the cable rating. Yet they're quite happy to run a 2.5 sq. mm spur from a 30/32A ring circuit, sometimes even directly from the fuseway, because "BS7671 says you can."

That, to me, demonstrates a complete lack of logical thinking at the basic electrical level.

 

[ericmark]

There is no reason that a normally protected cable of 2.5mm² of less than 3 meters length should not feed a 13A socket. 434.2.1 does state 3 meters, but many seem to think this does not apply to a spur from a ring final. I agree it should apply to a ring final or a spur direct from a consumer unit in the same way.

 

[EFLImpudence]

434.2.1 does state 3 metres but not in respect of the subject being discussed, i.e. a 2.5mm² spur with 32A OPD, therefore it (the 3m limit) is not relevant.

How many more times?

 

[JohnW2]

On a related issue, I've heard some who go by the "cable installer's cookbook" comment that they would code 1 a 2.5 sq. mm cable feeding a single 13A socket protected by a 30/32A fuse/MCB. Their view is it's not a standard BS7671 circuit arrangement and is dangerous because of the protective device rating exceeding the cable rating. Yet they're quite happy to run a 2.5 sq. mm spur from a 30/32A ring circuit, sometimes even directly from the fuseway, because "BS7671 says you can." ... That, to me, demonstrates a complete lack of logical thinking at the basic electrical level.

... and to me.

Kind Regards, John

 

[PBC_1966]

What does 434.2.1 actually say in the current edition?

 

[EFLImpudence]

It relates to the position of devices for fault current protection.

 

[scousespark]

If a break occurs in a 'standard' ring final, it can be detected at the CU. If the circuit has an interconnection, or added ring a break not show up in an e2e test. This is why interconnections are not accepted in bs7671.

 

[JohnW2]

If a break occurs in a 'standard' ring final, it can be detected at the CU. If the circuit has an interconnection, or added ring a break not show up in an e2e test.

As I always say, the one 'downside' of cross-connections in a ring is that it requires a bit more time/effort, and proper knowledge/understanding, for testing and fault-finding. AFAICS, in almost every other respect, one or more cross-connections make a ring 'better'/'safer' - in terms of current sharing between conductors, EFLI, VD and 'redundancy' (particularly CPC redundancy).

With or without cross-connections, a break in a ring will be undetected by users, and the possible consequences of such a break beyond a cross-connection will never be 'more than', and will nearly always be 'less than' (since they involve less sockets) a break in the same circuit arranged as a single ring - again, AFAICS, the only 'issue' relates to testing/fault-finding.

This is why interconnections are not accepted in bs7671.

Can you help me understand what regulation in BS7671 indicates that such interconnections are 'not accepted' (i.e. non-compliant)?

Having said all that, given that there are undoubtedly some (perhaps many) 'electricians' out there who do not have the knowledge or understanding that would enable them to adequately test or fault-find on such a circuit, it probably is reasonable and desirable that they are not installed, even if they would be, in most cases, 'better'/'safer'.

Kind Regards, John

 

[ban-all-sheds]

I don't know any 'BS7671-trained cable installer' who would install this circuit. It must have been done by a 'real' electrician.

That's surely the point that was being made - that only someone who actually 'understood' could possibly contemplate installing such a circuit or know how to test it - since it's not in the "cable installer's cookbook"?

Why are you ignoring the real possibility that it was installed by someone who didn't have a clue, and didn't do any testing?

 

[JohnW2]

Why are you ignoring the real possibility that it was installed by someone who didn't have a clue, and didn't do any testing?

I'm not ignoring that possibility - it is so obvious that I didn't think it needed saying.

I would never install such a circuit, even in my own home (even if I do believe that it is generally 'safer' than a standard ring final). However, I am questioning the view that many/most appear to have that it is necessarily non-compliant [with what reg(s)??], and therefore would need to be 'coded' on an EICR.

Kind Regards, John

 

[ban-all-sheds]

I would say that what makes it non-compliant is the lack of an exception which allows it. There is one for rings - there are none for any other topologies.

 

[JohnW2]

I would say that what makes it non-compliant is the lack of an exception which allows it. There is one for rings - there are none for any other topologies.

...but it is a ring, which is specifically allowed - and there's nothing in the regulation which allows it which says that there cannot be any cross-connections.

That seems reasonable to me since, as I keep saying, in relation to the reason why that 'dispensation' is required (because one is using cable with a CCC less than the In of the OPD), adding cross-connections can but improve the situation (making it less likely that any of the cable will be overloaded for significant periods of time). I therefore see no reason why they would have felt the need to explicitly exclude cross-connections (which they haven't).

Kind Regards, John

 

[ban-all-sheds]

but it is a ring, which is specifically allowed

Topology 101

This is a ring:

This is not:

- and there's nothing in the regulation which allows it which says that there cannot be any cross-connections.

It's the other way round.

With the cross connections it is not a ring, and therefore it is not exempted by 433.1.204 from the requirements of 433.1.1.

So yes, you may have that layout if you like, but you'd need to drop the MCB to 20A.

That seems reasonable to me since, as I keep saying, in relation to the reason why that 'dispensation' is required (because one is using cable with a CCC less than the In of the OPD), adding cross-connections can but improve the situation (making it less likely that any of the cable will be overloaded for significant periods of time).

Be that as it may...

I therefore see no reason why they would have felt the need to explicitly exclude cross-connections (which they haven't).

They have, by virtue of explicitly requiring that the circuit be a ring.