Sub main OCPD

[ferg]

I have seen rather frequently SWA sub mains taken from Henleys protected only by DNO fuse and a stand alone RCD.

I have seen it so often now that I am beginning to question my own knowledge.

My understanding is that the sub main should have it's own OCPD depending on the cable it is protecting.

Is there any possible justification or reason that would make this compliant?

The only possible justification I can think of is that if the Zs at the sub board is still low enough to take out the main fuse under short circuit fault conditions, the likely load on the sub board is lower than the CCC of the cable and that the RCD would operate if an earth fault occurred on the cable at a level that would not cause the main fuse to operate.

Have I missed something here or am I right in thinking this set up is non compliant.

 

[telectrix]

personally, i'd always fuse a sub main supply according to it's cable, and in any case ensure that the fuse i fitted was < the dno fuse, hopefully enough less to provide discrimination.

 

[davesparks]

There is no justification, it does not comply (unless the SWA is less than 3 metres long and being used as tails)
The suppliers do not permit the cutout fuse to be used as the OCPD for any circuit in the installation.

There's not much more to be said about it really.

 

[ferg]

There is no justification, it does not comply (unless the SWA is less than 3 metres long and being used as tails)
The suppliers do not permit the cutout fuse to be used as the OCPD for any circuit in the installation.

There's not much more to be said about it really.

I think you have highlighted an issue with my terminology there.
Technically speaking it's not a sub main just very long tails protected by a stand alone RCD.

 

[davesparks]

I think you have highlighted an issue with my terminology there.
Technically speaking it's not a sub main just very long tails protected by a stand alone RCD.

If its over 3m long it needs to be a distribution circuit (submain) and therefore needs appropriate OCPD/switchgear.

And unless it's a TT system or other special case an SWA protected by RCD would be causinge to question the design of the circuit.

 

[pc1966]

Even in a TT system there should be non-DNO OCPD suitable for the cable, etc.

I don't see anything wrong with using the armour to get the TT's earth rod connection from "outside" (i.e. at DNO cut-out location) into the CU location if it was a long and torturous route, but it would be a little bit unusual.
[automerge]1586185346[/automerge]
Also if the CU is a long way from the RCD it is not a very good design from the point of view of end user having to check two very different locations if, say, the lights go out to find what tripped.

As telectrix says, best to put a switch/fuse at the source and, if it is TT, then put the RCD in the CU as main switch so all likely trip resets are in the one location.

 

[davesparks]

Even in a TT system there should be non-DNO OCPD suitable for the cable, etc.

I don't see anything wrong with using the armour to get the TT's earth rod connection from "outside" (i.e. at DNO cut-out location) into the CU location if it was a long and torturous route, but it would be a little bit unusual.
[automerge]1586185346[/automerge]
Also if the CU is a long way from the RCD it is not a very good design from the point of view of end user having to check two very different locations if, say, the lights go out to find what tripped.

As telectrix says, best to put a switch/fuse at the source and, if it is TT, then put the RCD in the CU as main switch so all likely trip resets are in the one location.

If the Ra is low enough for the submain to not require RCD protection then that's OK, but in most cases the submain will need RCD protection in a TT system.

 

[pc1966]

If the Ra is low enough for the submain to not require RCD protection then that's OK, but in most cases the submain will need RCD protection in a TT system.

Yes, you are right!

Just after posting I thought about it and realised the RCD has to be at the origin of the system, and the SWA sub-main would be part of that system.
[automerge]1586186010[/automerge]
I guess the good design approach would be to make the origin RCD a 100mA or similar delay type and then to RCD/RCBO all circuits at the CU.

At least that way if there is some fault to earth that could trip an RCD you ought to pick it all up at the CU and hopefully never see the main RCD trip (short of a major fault on the SWA, etc).

 

[Andy78]

Are these new build houses by any chance ?

 

[ferg]

Are these new build houses by any chance ?

Not the ones I've come across.
It's predominantly outbuildings and additional DB's installed as an addition to an existing installation.

 

[radiohead]

Gotta say I cant recall coming across this ever but would agree that if I did I would see it as non compliant.
I suppose playing devils advocate if the distribution circuit went from the intake to a remote building as a 100a supply it could be argued that it would be fairly pointless having two 100a OCPD's in series about a foot from each other!

 

[barninks]

Could I just ask about potential issues with extending meter tails when moving a CU? Say if the cu was originally 3 metres away from the DNO fuse and has been moved to say 15 metres from the cut out fuse?

 

[pc1966]

If playing Devil's advocate I would say unless you have SWA rated for at least 100A then you could never be sure the DNO would not upgrade a cutout and fit 100A.

So I see the issue is more one of having control over the OCPD setup, and not depending on the DNO that (as already pointed out) is not something you should ever be using for your design.
[automerge]1586203893[/automerge]

Could I just ask about potential issues with extending meter tails when moving a CU? Say if the cu was originally 3 metres away from the DNO fuse and has been moved to say 15 metres from the cut out fuse?

The main issue, and the point of the wiring regs on this, is if you have long cables and it is not specifically designed to protect against damage then you don't have an OCPD under your control for those wires. Yes, you could argue that SWA is protected...

 

[Wilko]

Hi - my thought (to be corrected ) is the DNO cutout fuse is to protect the DNO cable. They have designed it and judged this to be the best place for it to be positioned, considering safety, functionality, cost and access etc.

The use of the DNO cutout fuse for protection of installation tails is mentioned in Reg 433.3.1 (iii) and if they have agreed then it may be compliant. Otherwise Reg 433.2.2 says 3m.

 

[Ian1981]

Hi - my thought (to be corrected ) is the DNO cutout fuse is to protect the DNO cable. They have designed it and judged this to be the best place for it to be positioned, considering safety, functionality, cost and access etc.

The use of the DNO cutout fuse for protection of installation tails is mentioned in Reg 433.3.1 (iii) and if they have agreed then it may be compliant. Otherwise Reg 433.2.2 says 3m.

I don’t believe that regulation 433.2.2 is applicable for the DNO fuse, more regulations 433.3.1 (ii) and 434.4 (iv)

 

[davesparks]

Gotta say I cant recall coming across this ever but would agree that if I did I would see it as non compliant.
I suppose playing devils advocate if the distribution circuit went from the intake to a remote building as a 100a supply it could be argued that it would be fairly pointless having two 100a OCPD's in series about a foot from each other!

Yes two 100A fuses in series is pointless, but it is the way it has to be done with the current regulations.
[automerge]1586246378[/automerge]

Could I just ask about potential issues with extending meter tails when moving a CU? Say if the cu was originally 3 metres away from the DNO fuse and has been moved to say 15 metres from the cut out fuse?

The tails should not be longer than 3 metres, so you install a distribution circuit to feed to CU.

 

[barninks]

The tails should not be longer than 3 metres, so you install a distribution circuit to feed to CU.

Yes fine, that is what I would normally do - but just trying to pick peoples brains on here and see what views there seem to be on the subject

 

[davesparks]

Yes fine, that is what I would normally do - but just trying to pick peoples brains on here and see what views there seem to be on the subject

What are you getting at here?
The rules are quite clear on the subject.

Are you trying to ask if it's acceptable to ignore the rules and just connect 15m long tails to the incoming supply?

 

[barninks]

What are you getting at here?
The rules are quite clear on the subject.

Are you trying to ask if it's acceptable to ignore the rules and just connect 15m long tails to the incoming supply?

Sorry, didn't meant to cause confusion - just was interested to see different peoples views on the protection offered by swa

 

[Andy78]

Sorry, didn't meant to cause confusion - just was interested to see different peoples views on the protection offered by swa

That's literally the confusing part.
When people ask "What does everyone usually do ?" or "What's everyone's different views on....." It makes it seem as if there aren't already regulations that cover this situation.

 

[barninks]

I am sorry

 

[davesparks]

Sorry, didn't meant to cause confusion - just was interested to see different peoples views on the protection offered by swa

SWA offers protection by connecting any metal object which penetrates it to earth (by the armour) before it makes contact with any live conductor, this causes the OCPD to operate.
It also offers protection by having a relatively tough outer jacket and a pretty string general construction thanks to the steel armour.
It also has protection against UV by having a black sheath

 

[radiohead]

Yes two 100A fuses in series is pointless, but it is the way it has to be done with the current regulations.

I realise that.

 

[Simon47]

As it happens, I have a submain in the house here - except it's T&E. From a brief chat I had with the vendor's ex's father, it sounds like it did have an MCB, but he changed it when he was told that the T&E needed RCD protection.
Thing is, it says 63A on the front, so it's OK isn't it ?
It'll get sorted at some point, but for the time being I know there's a 63A fuse in the cutout.