Help please on a sub consumer unit fed from RCBO

[AJB Temple]

Long time lurker so this may be my first question. I am a qualified electrical engineer, but in aeronautics not domestic installations, and I may therefore ask silly questions. We have purchased a property that is unfinished, and where the electrician went bust before the installation was completed. 90% of the building is certified with BC, but there is a new kitchen in an attached outbuilding, that is substantially wired but not yet connected. My question relates to circuit protection.

There is a conventional Hager late 2018 twin RCD CU in the main house, along with a Henley block that splits to the meter tails to another Hager board containing a twin pole switch and 5 RCBOs. Each of these feeds a different out building via SWA in each of which there is a metal small CU with a handful of MCBs and usually a rotary isolator switch as well.

The new kitchen is fed from one of these Hager RCBOs via 16mm 3 core Doncaster SWA, which reaches a Hager CU in a cupboard in the new kitchen. It is presently not connected.

The new kitchen has a bunch of radial circuits. In essence some feed an Island with oven, hob etc - all radials. Others feed a steam oven, coffee machine, conventional oven, quooker tap with 7l cylinder, and a dozen sockets on a separate area in the same kitchen (basically wall side and island side).

Whoever specified the CU has bought and fitted a Hager which has two RCDs and numerous MCBs. My understanding, which may be wrong, is that if the CU is fitted with RCDs and fed by the RCBO, then nuisance tripping is inevitable? I think the developer may have made a wrong assumption or the ex electrician made a mistake.

I think the CU will need to have the RCDs removed? However, that would leave all of the appliances in the new kitchen prone to simultaneous failure in the event of a fault tripping the RCBO. Very inconvenient, especially as the user has to trek through five other rooms to get to the RCBO box near the meter.

I have yet to find a local electrician who is clued up (had a couple of false starts with people not capable of certifying). So, my questions:

1) If I wish to replace the dual RCD and MCB set up in the new kitchen with RCBOs (ie circuit dedicated), can the RCBO at the source box near the meter be retained, or would having an RCBO feeding a board containing RCBOs cause nuisance tripping? I am assuming it would.
2) Therefore, should the feed RCBO (ie the box near the meter) be replaced with a suitable switch in the CU box? There is only enough space for a single unit in the CU box at the meter end, if the RCBO is removed - ie there are no spare slots, although there are two RCBOs not presently in use so if necessary both could be replaced.

Many thanks. Adrian

(PS, I am looking for a good, friendly electrician near Paddock Wood / Tunbridge Wells in Kent, who can work with me on this, so if you know anyone, please PM me. It's a bit fiddly as I am finishing off the kitchen cabinets etc myself as time and limited budget permits.)

 

[loz2754]

The RCBO at the source protects the swa, and as swa doesn't need 30mA RCD protection even if buried in walls, then yes you can change the RCBO at the source for an MCB, as long as all the downstream circuits that need an RCD have an RCD.
If left as it is now, an earth fault would trip either or both of the RCD and the RCBO.

 

[AJB Temple]

Thanks. That was pretty much my thinking: swap out the RCBO for a MCB and then probably rather than use RCD and MCBs in the new kitchen board, use RCBOs for the critical circuits (refrigeration for example). Presumably it is OK to have RCBO's in the new CU as well as RCDs feeding whatever MCBs are needed?

As an aside, there is another sub CU in the attic of this room, also supplied with 16mm 3 core SWA, which is MCB populated only as just feeds a couple of sockets on a radial, and a couple of lighting circuits. Overkill, but potentially gives me another useful supply.

Earth system btw is TN-C-S. The electric car charge point has an earth rod - dedicated circuit.

 

[loz2754]

Presumably it is OK to have RCBO's in the new CU as well as RCDs feeding whatever MCBs are needed?

Yes, (if the consumer unit can be configured to allow this) but if cost is not an issue, most would now suggest all circuits be on their own RCBO.

 

[AJB Temple]

Thanks. I do agree, which is why I put in my original post that swapping out the MCB/RCD set up for RCBOs may be the way forward. Oddly though, both of the electricians I spoke to (neither of whom could certify once I enquired deeper) said it was pointless / very costly to install RCBOs. In practice I can buy trade and Hager RCBOs are under £20 each inc. and it is not as if I need loads of them.

 

[pc1966]

As above, if it is SWA from the main CU to the sub-CU then you don't need RCD protection at source as any damage to the SWA won't result in a shock since the armour will short out any nail, etc, that penetrates it. But you do need RCD protection somewhere on most circuits.

Changing the supply from RCBO to MCB would deal with the issue of residual currents/ weak faults to earth tripping out the whole sub-CU supply. However, selectivity between MCB/MCB or MCB/RCBO is quite poor, basically if a hard fault to earth (CPC) causes enough current to hit the instant magnetic trip point of the upstream MCB then it is likely to be committed to tripping before the down-stream one can clear the fault, leading to both tripping.

Most of the time that is not a big problem, but in some cases (e.g. workshop, maybe kitchen) a total loss of lighting on a fault is potentially serious. Options to work around that are either emergency lights (some baton style LED strip lights come with that built in) or supplying the sub-board from something that has a slight delay in going so the down-stream breaker can clear it.

In most domestic cases that is a switched-fuse instead of MCB. In commercial/industrial situation you also see the fancier MCCB that have a short-time trip delay before you hit the ultimate high current "instant" energy-limiting point, but they cost much more than a switched-fuse!

If you look in the Hager commercial catalogue downloadable from here:

Hager Brochures, Catalogues & Documentations | Hager UK

Discover all the product documentation you need in our brochures & catalogues! Download what you need in one click straight from Hager Official website!

hager.com

Then from around page 142 to 151 they give you selectivity tables for combinations of their own products (as well as generic BS88 fuses) which indicate the maximum prospective fault current where you will only have the downstream device trip as desired.

It might not be anything to worry about, but you should think about each area and if there is any specific risk or intolerable inconvenience that would merit any big changes.

TL;DR Probably change upstream RCBO to RCD, replace downstream with RCBO if possible, or add RCD(s) before the MCBs.

 

[AJB Temple]

That is extremely helpful and thought provoking. Thank you.

The SWA runs through protective trunking (drainage pipe as it happens) running underground and is fully protected where it emerges from the ground to run through a loft space, in a metal tray, to the new CU. I think risk to this cable is minimal.

All overhead (as opposed to under plinth) lighting to the new kitchen (and adjacent utility room) if from an entirely separate SWA protected by a RCBO and there are dedicated MCBs to the various lighting circuits. Switching is Quinetic for both interior and exterior lights.

The intention now (irrespective of whatever the original designer / developer thought) is for the kitchen to be split into two work zones. This would allow an RCD solution in the CU (dedicated to this room) but I am in fact inclined to minimise circuit failure inconvenience by having all the circuits on separate RCBOs. The utility room is really a glorified larder (no laundry) so contains mainly refrigeration and a freezer and dehydrator. This is better covered by an RCBO I feel.

The sockets (there are at a guess about 20 doubles) were actually designed as a single ring, but my intention is to split this lot into three radials: two work areas and a supply for the Quooker tap. The reason is the ring design is very convoluted (and already has installation continuity faults on it from the original electrician) and a failure on it will be a nuisance and troublesome for fault finding or later testing. I have test equipment (Megger MFT1711 etc) and will do all the 18th Edition / Part P certification tests myself before getting it professionally signed off. I will do a 100% test but I am very inexperienced as a domestic tester and it needs BC approval anyway.

I think the specification, especially cable sizes of both SWA and the various circuits, is more than sufficient. Over specced in fact.

 

[TJ Anderson]

Nothing wrong at all with a ring final circuit and if you have a MFT1711 available to use you can fully test it.

Good shout on going for RCBO for final circuits, better solution really and cheap enough these days.

 

[AJB Temple]

Thanks. I agree there is nothing wrong with a ring final in essence, but the design here is poor. The kitchen is divided into two work zones - an Island and a wall area. It is large. The Island has an induction hob, extraction, large Gaggenau rotisserie oven, Tepinyaki plate and pair of fridge drawers. Plus a number of sockets. The other area has two ovens, coffee machine, and water heater, plus numerous sockets. Ovens are all on radials. The design runs a ring around both areas, both above and below work surface and so if we get a fault anywhere on the ring, we lose all sockets and (I think) the hot water unit and the fridge drawers, plus all sockets in the rest of the room and the adjacent utility which contains the main fridge and freezer. It just seems an unnecessary risk to me to have so many eggs in one basket. From tracing the wire runs using continuity tests, I think it is easy enough to split the ring into two radials, and run a new radial into the utility.

It is all wired in 4mm T&E. I think 2.5mm would have been adequate for this ring so I don't know why they used 4mm except they clearly had lots of it as 3 new boxed Prysmian reels were left in a cupboard.

I can't think of any reason why I should not split the ring. Am I missing something here. I've traced the faults now, which was very silly poor terminations in fused supply boxes for the extractor units. They were very tight on space and put a junction box in a void beneath an oven (metal shielded for heat) and this was also badly terminated. All tests fine now for continuity and IR - but still assembled as a single ring currently.

Rather annoyingly, the SWE is a metre too short to reach the CU, without draping it in an awkward way across the loft, so that needs to be extended and clipped properly in my tray. I suspect this is not the electricians fault, as the kitchen layout actually installed, is different to the plan, and the CU was clearly intended to be much closer.

I sympathise with electricians who have to pick up another persons work, without having details of the original design.

 

[pc1966]

4mm certainly helps, but depending on the installation method (i.e. amount of thermal insulation around it) it is not fully 32A rated so you might still have to drop each radial to 25A and so a bit less diversity for loads on the various sockets.

However, I have to agree that it seems not a very good plan to run a single ring on all those sections!

Depending on the routes already in use and so on it might be worth completing it as two rings if a lot of insulation, but if "clipped direct" as not going through insulated regions then 32A radials might be best.

Generally you want to avoid joints in cables if at all possible. However, if the SWA is too difficult/expensive to replace then you could join it using a metal adaptable box and glands, along with suitable joints for the conductors (proper crimp and heatshrink would be my choice) or you can get the professional resin filled SWA joint kits for underground use that are pretty tough and reliable:

S16 16-25mm Straight Resin Joint Kit for Cables up to 25mm 2 Core

Proven underground cable jointing system

www.cef.co.uk

You would need to check the data sheet to get the right one for the size of SWA in use as well as properly following the instructions. Given that installing the sub-CU is certifiable work it is probably best to speak to the electrician doing it about the best options to have the cable installed without strain and the CU in a usable location.

 

[Rockingit]

There is a slightly unorthodox heath robinson method that will cost you nothing..... use a ramp test to see which RCD trips first - if it's the downstream ones then whilst not strictly speaking ideal at least your concerns of nuisance tripping are eliminated!

 

[DPG]

There is a slightly unorthodox heath robinson method that will cost you nothing..... use a ramp test to see which RCD trips first - if it's the downstream ones then whilst not strictly speaking ideal at least your concerns of nuisance tripping are eliminated!

But on a N-E fult greater than 30mA they will still both probably trip.

 

[Rockingit]

But on a N-E fult greater than 30mA they will still both probably trip.

Depends on which has the quickest response time (NB - not that I'm really suggesting this as a serious course of action!)

 

[AJB Temple]

Thanks both. I am stuck with the SWA because it goes through and underground duct and is clipped inside trunking on two exterior walls as well. It would be a real pain to replace it. The joint will be in a fully dry attic space, so it does not really need a resin filled joint. My plan was to fit the joint inside an inspection box, with glanded in and out. The box will be mounted safely to a board just above the metal tray I already put in.

The wiring in the kitchen does not run through insulation anywhere. Some of it runs through 40mm circular conduit buried in the walls - this supplies the wall sockets. But even this is super protected as it has a stone plinth in front of it and you would need to be a real idiot to damage the cables. The majority of it is clipped under the units. This is fully accessible for inspection as the units are 30cm off the floor with access below. It is all additionally protected with plastic trunking.

I like the ramp test idea. I hadn't thought of that. However, the drawback is I would beed to do the install first and then test it. If I do have a problem then I would need either to deal with the RCBO at the meter end, or do something different at the new kitchen end. I am pretty much decided on doing all circuits at the kitchen end with RCBOs anyway, so I might as well put in a suitable MCB at the meter end.

The only think holding me back is a) working out where it would be smart to install surge protection (needs to be Hager as everything is Hager at the moment) and b) whether I ought to install (or have someone install) combined RCBO / AFDD devices. Surge protection is cheap enough, so I suppose I may as well stick one on the board.

I understand broadly how AFDD circuits work and have watched the JW and David Savery vids on this (and the AFDD controversies) , but I have zero domestic knowledge. It is not clear to me whether the 2nd Amendment to 18th edition which goes live on 28th March this year will mandate AFDDs, the local electrician I spoke to has zero knowledge about them, local wholesalers do not stock them as far as I can see, and the cost is eye watering! My understanding is mandate will apply for designs after 28/3/2023. If I do need to specify AFDDs for future proofing, then I shall want to minimise the number of circuits as AFDDs are costly. In fact, that cost aspect might justify changing the board totally to Wylex or whatever. This installation will be finished and commissioned within 2 weeks so will pre-date Amendment 2, but if it is a good idea then I will bite the bullet. No cables are in any way vulnerable at all in this installation though.

 

[GDF144]

Jumping on this thread rather than starting another. Ive been asked to change a landlord's board, that supply's a block of flats. The supply to each flat is in twin and earth. The flats consumer units have rcd protection so if I make the circuits on the landlord board rcbo they're as likely to trip if there's a problem in a flat. Other than rewire the supply to each flat in swa (which they won't go for). What are my options?

 

[DPG]

Jumping on this thread rather than starting another. Ive been asked to change a landlord's board, that supply's a block of flats. The supply to each flat is in twin and earth. The flats consumer units have rcd protection so if I make the circuits on the landlord board rcbo they're as likely to trip if there's a problem in a flat. Other than rewire the supply to each flat in swa (which they won't go for). What are my options?

Don't think there is one. Unless the T&E is surface or conduit all the way then it will need 30mA RCD protection

 

[pc1966]

Other than rewire the supply to each flat in swa (which they won't go for). What are my options?

The critical issue here is if there is any sub-main cables buried at less than 50mm from the surface. If you can establish there is not that risk, then supplies from fuses (like a Ryefield board) or MCCB/MCB become a reasonable solution.

Otherwise the only compliant solution is 30mA RCD (or RCBO) protection and that is likely to have nuisance trip issues due to the total leakage of the flats, as well as very poor selectivity.

 

[pc1966]

TL;DR Can someone accidentally nail* a cable?

[*] using nails no longer then 50mm

 

[GDF144]

Will have to look and see if I can trace where the cables run. Can't see them going for the work if requires running in new feeds so will be left on mcb's anyway

 

[mainline]

Pme ?

Will have to look and see if I can trace where the cables run. Can't see them going for the work if requires running in new feeds so will be left on mcb's anyway

Pme earthing ?