Large number of 13A sockets in a steel building: Design considerations.

[Mark42]

This weekend I’m installing 20 2-gang 13A sockets around the inside walls of my new 20m x 8m steel building.

These are only for small car battery chargers, so total continuous load will be trivial. Plus one fused connection unit (directly in the ring) for the electric shutter at the far end of the building.

Cabling is surface 3-core 2.5mm SY girder-clipped to the horizontal steel members. It’s expensive, so cable cost is a limiting factor. I have one 100m reel only.

I have designed the install as below. It is not in accordance with ‘old-skool’ floor area guidance, but I believe it’s OK. Does anyone disagree: ie have I missed anything?

Basically it’s one huge ring around the walls of the building, with sockets equally spaced. Total ring cable length 60m. Considering volt drop at this length I will feed the ring from a 20A MCB. 32A would probably be OK, but I don’t need it, and 20A is a concession to ‘make up’ for exceeding the normal floor area suggestion.

Of course it could be split into two 20A radials down each side of the building, but I see no electrical reason for this.

I’ll lightly solder the stranded cables together in each box before connecting to the socket outlets, to reduce both ring resistance and the chances of the ring going open circuit for any reason in the future. Earthing will be excellent: in addition to the CPC, every metalclad box is bolted directly to the frame, boxes are of course strapped to the sockets internally, and the whole structure is bonded.

There is one small corner of the garage which will be used as a working area, with power tools etc. This has a separate 32A conventional ring, and a separate 16A radial to an industrial 16A socket for the arc welder. Both on a different phase to the ring above.

The main electric shutter is on a separate circuit – close to the CU so cable usage is insignificant. Should a battery charger fault pop the main RCD on the big ring circuit phase, it would not affect this shutter, so I could still get into the building without faffing around with manual overrides.

I’m a great believer in imaginative solutions, and in not always following guidance like an unthinking sheep. But I’m not so arrogant (yet!) not to ask for advice from those more experienced than I am. So, have I been daft and missed anything here? It’s my own place so I want to do it right, once, then not touch the thing for 20 years.

Cheers, Mark.

 

[malcolmsanford]

Personally Mark I would have installed a sub board for this as you never know what the future will hold. Then you would have no issues with disconnection times, length of runs etc etc.

It is a large ring final mate and your design though works by the sounds of things just seems to be a case of trying to do something a little off the wall, just to prove it can work, personally I like simple.

Not sure why your going to the trouble of soldering cables together, to be honest what advantage if any will you achieve, you may by doing this reduce the impedance slightly, but not sufficient to affect disconnection times on a 20amp MCB unless your thinking of a "c" type for the chargers, which you may have to consider with the start up on those babies. also if you need to fault find and split the ring there will be a few curses.

But with the RFC being RCD protected your ZS can be 1667, which negates the soldering really

Also by having such a large RFC on a single RCD your chargers are going to have Protective Conductor Currents and you may start getting leakage tripping on it, I'm assuming it's a 30mA additional protection one, so do you need the RCD protection?

Not sure about what you mean if you have a 20amp MCB for the RFC and then state the main RCD, is the shutter on a split board without RCD protection ? That sounds to me if you have an upfront RCD protecting the entire installation and if that is the case would not RCBOs be a better option, if you want RCD protection.

 

[micknew]

The best advice would be to get a qualified electrician in who would carry out the works and give you certification for insurance purposes.

 

[malcolmsanford]

The best advice would be to get a qualified electrician in who would carry out the works and give you certification for insurance purposes.

I suppose it would depend if this steel building is in the UK or Malaysia

 

[Mark42]

Personally Mark I would have installed a sub board for this as you never know what the future will hold. Then you would have no issues with disconnection times, length of runs etc etc.

Sorry, Malcolm I didn’t make it clear. Of course it’s on a sub-main: the building is 100m from the main house.

Supplied by a 4-core 10mm u/g SWA fed from a 32A 2-phase MCB in a main house board. Cores are L1, L2, N and PE, plus armour terminated with Piranah nuts (I don’t like banjos on plastic boxes) and bonded.

It’s a split-load 17[SUP]th[/SUP] Edn CU reconfigured as a split-phase DB. The single phase main switch removed, two 25A 2P RCDs installed, fed from the cable above via a separate 3P&N isolator acting as the local main switch. Non-standard I know, but it’s much cheaper than installing a 3P board for only a few ways. This is an unusual split-phase 240/480V ‘farm’ supply

Perhaps a photo is better, I’ll get one later.

It is a large ring final mate and your design though works by the sounds of things just seems to be a case of trying to do something a little off the wall, just to prove it can work, personally I like simple.

Yes, but I like saving expensive (and beautiful) SY too. I think putting in two rings is not more simple than my proposal, and just consumes wire and DB ways for no electrical reason. Unless I’m wrong, and that’s why I’m here asking questions.

Not sure why your going to the trouble of soldering cables together, to be honest what advantage if any will you achieve, you may by doing this reduce the impedance slightly, but not sufficient to affect disconnection times on a 20amp MCB unless your thinking of a "c" type for the chargers, which you may have to consider with the start up on those babies. also if you need to fault find and split the ring there will be a few curses.

OK I agree. I don’t like finely-stranded cable in socket plates that’s all. The terminal screws tend to displace the strands, the strands can annoyingly bunch up on insertion, especially when there’s 5 sq mm of them loosely twisted together, there’s always one bloody tiny live strand sticking out so you have to do it again, in poor light, and so on. I agree it’s overkill. As to faults? That’s what side cutters are for! But yes, point taken.

It will be a B-curve breaker: they are only very small ‘Accumate’ chargers for battery maintenance. I don’t know if they have excessive earth leakage from filter circuits or whatever. Time will tell, but yes, nuisance tripping would require attention, then I agree RCBOs on each circuit might be better.

The wiring may as well include the high-integrity earthing scheme as standard. Costs nothing.

But with the RFC being RCD protected your ZS can be 1667, which negates the soldering really

Yes I know, but why design down to worst-case when at no cost you can design up? My children will play with stuff in there, hoses, wet cars, no shoes on a concrete floor, wet clothes, who knows? If I don’t see Zs in the tens of ohms I’m not happy. Yes, overkill again!

Also by having such a large RFC on a single RCD your chargers are going to have Protective Conductor Currents and you may start getting leakage tripping on it, I'm assuming it's a 30mA additional protection one, so do you need the RCD protection?

Not sure about what you mean if you have a 20amp MCB for the RFC and then state the main RCD, is the shutter on a split board without RCD protection ? That sounds to me if you have an upfront RCD protecting the entire installation and if that is the case would not RCBOs be a better option, if you want RCD protection.

As above, you may be right.

I used two 2P RCDs, which I had in stock, as main switches on each phase. Therefore there are no unprotected ways. It looks fine to me so far IF there’s no nuisance tripping.

The 'important' shutter will be on the phase which will have no connected loads when the building is unattended, so nuisance tripping is unlikely (so long as my wiring's waterproof!)

I did use RCBOs in my main house for everything so will consider changing the 2P RCDs for 2P switches and then installing RCBOs where necessary.

Thanks for the advice. Hope this all makes sense. It's a bit hard in words!

The best advice would be to get a qualified electrician in who would carry out the works …

What, like you, you mean? OK, but tell me first that what I am proposing is electrically dangerous or in some way electrically incorrect. Then I’ll consider it.

 

[Engineer54]

So what sort of earthing arrangement are you on at this farm, TT system or PME??

 

[needasparks]

Did I read that correct, your having a spare un-used phase introduced (2 phase & N).

 

[Silly Sausage]

Instead of soldering use bootlace crimps.
If you're going for permanent connections (>20 years you say), use 2 in 1s....
Twin Cord End, 10.0mm - Red - Twin Cord Ends - Ferrules & Cord Ends - Crimping Tools | Crimp Terminals | Electrical Connectors

Leave plenty of slack cores in in case you ever need to cut them off.

I think there's a reg about stranded conductors, can someone point it out?

Have fun making off 40+ SY glands!

 

[lamb]

You may need to consider the flammable gases that can be produced when batteries are in charging mode. Depends on the battery type and how many.

 

[Engineer54]

Did I read that correct, your having a spare un-used phase introduced (2 phase & N).

No, .... he means that he has a 3 phase supply, but he's intentionally loaded up the other two phases so he can have his shutter door on on the remaining unloaded phase!! Not to worry he seems to know best!! lol!!!

 

[micknew]

Perfect example of a little knowledge being dangerous.

 

[Engineer54]

Perfect example of a little knowledge being dangerous.

Just look at the length of the sub-main from the house and then the 60m ring circuit!! Then he's talking about his kids playing with water hoses in there with bare feet!!! Very sensible with batteries being charged.

Still no mention about LABC or what type of earthing arrangement he has at this farm....

 

[Mark42]

Did I read that correct, your having a spare un-used phase introduced (2 phase & N).

No, .... he means that he has a 3 phase supply, but he's intentionally loaded up the other two phases so he can have his shutter door on on the remaining unloaded phase!! Not to worry he seems to know best!! lol!!!

LOL back, gentlemen. I’m not that bloody stupid. Maybe I was not clear. There are ONLY TWO phases on the site. It’s (antique) ‘spilt phase’ installed after the war to save copper. 240V x 2 at 180 degrees. Welcome to rural England, things move slowly here

The two phases will both be used and will have a reasonably balanced load when the building is IN USE. But when unoccupied (ie most of the time) only the maintenance chargers will be in operation, all on a single phase. Maybe 2A tops, intermittent, even the most absurd stickler can’t tell me to balance that across phases! (Can you?) Hence the ‘clean’ second phase, unlikely to trip out when nothing is in use.

Instead of soldering use bootlace crimps.

… I think there's a reg about stranded conductors, can someone point it out?
Have fun making off 40+ SY glands!

Brilliant idea: thanks! Pack of 6mm twin ferrules ordered already.

Is the stranded reg (which I also can’t find) about actually needing to use stranded where movement is expected?

Hmm… SY glands. I hesitate to admit it but I again use a non-standard method: unwind the braid, twist it, fit g/y sleeving, sleeve the two blacks with brown & blue, then add an inch or so of 12mm shrink-fit over the lot where the outer sheath ends. It looks neat, costs little, and can then be used with a standard cable gland. The sleeved braids can then be terminated in the metalclad box earthing lug. Flame away, gentlemen.

SY glands are a rip-off price, and I think only really indicated when EM protection is required.

You may need to consider the flammable gases that can be produced when batteries are in charging mode. Depends on the battery type and how many.

Good point, but it’s a big open, ventilated, area and only a few lead-acid car batteries. It’s only little maintenance chargers for rarely-used vehicles, to stop the batteries from self-discharging. No chance of hydrogen gassing off - I believe that only happens when really cooking batteries with a big charger, or with industrial batteries like in electric pallet trucks.

So what sort of earthing arrangement are you on at this farm, TT system or PME??

An excellent question, but one I hesitate to ask for advice about for fear of unleashing a torrent of conflicting opinions! (And urban myths.)

It’s TN-C-S and true PME. Two-wire HV supplying a pole trannie, split phase 480V from there to three houses in total. Overhead LV to the other two, with the poles spiked, and new u/g LV to my place with a spiked u/g joint in my yard. (It was a happy day when I could finally saw down those ugly poles in my garden.)

So what to do with a sub-main in a separate building 100m+ from the main house?

I assume standard practice would be to ‘run’ the PME on the SWA armour to protect the cable, terminate downstream in a plastic enclosure, sheath the nut, then install a TT system for the separate building. But I don’t understand why in my case.

I’m using 4-core 10mm: two phases, one neutral and a spare as a PE. Does this not count as a bonding wire? It’s (just) big enough (the DNO’s incomer is 35mm 100A x 2). So using this 10mm ‘bond’ should bring my steel building into the PME equipotential zone? Yes or no? Were I using ONLY the SWA armour, I agree it may not be enough since steel armour can get scuffed and rot over the years.

I plan to do some experiments: install a local earth spike as for a TT system, then measure the earth loop impedance at the sub-main with the PME only, the earth spike only, and with both in parallel. I suspect the PME, even down that long wire, will be better. What do others think?

My initial plan was to install two spikes in the wet earth on opposite corners of the building, bonded with 25mm to the structure, as if I was going for TT, but use the PME as well. Is this a naughty thing to do? Is there likely to be a PD between the DNO’s value for earth and my local true earth? And if so, would that cause a current to flow in the earth cables?

And the busted neutral saga? Would not my earth spikes take care of that? Or at least reduce any potential on the extraneous metalwork to a non-lethal level while waiting for a breaker to trip? But as I said my supply neutral is now underground, unlike my neighbours’ which looks like bits of bellwire on sticks, so I believe my local neutral is secure.

But it’s not my trade so I may have something completely wrong here!

 

[Engineer54]

Who built the steel framed building, you or was this building existing. I'm asking as i want to know if the concrete floor has any reinforcement ??

 

[Mark42]

Just look at the length of the sub-main from the house …

Yes, I know, it’s ridiculous. I really should move the house!

 

[Mark42]

Who built the steel framed building, you or was this building existing. I'm asking as i want to know if the concrete floor has any reinforcement ??

Good question: I built it.
No steel in the floor slab, no, so nothing in there which can be bonded.

The concrete is some new-fangled mix with fibre in it, which obviates the need for matting. At least that’s what the contractor said. It is guaranteed for 10-tonne vehicles …

 

[Engineer54]

Good question: I built it.
No steel in the floor slab, no, so nothing in there which can be bonded.

The concrete is some new-fangled mix with fibre in it, which obviates the need for matting. At least that’s what the contractor said. It is guaranteed for 10-tonne vehicles …

In that case you may have a problem using your PME earth in this building, especially if your kids are going to be playing with hoses bare footed. There could well be a potential difference between true earth and the PME earth in this instance. If you did have a grid in the foundation floor that could be bonded, then that would have been acceptable.

Looks like your going to have to go the TT route and your rods as you suggested, at opposite corners tied on to the structure and the structure to your MET. No need for 25mm cable 10mm will be fine!! Just make sure you go down a good depth with your rods, at least 2 rods deep at each corner, preferably 3 rods if your Ra is decreasing significantly the deeper you go with the second rod....

 

[malcolmsanford]

What I meant Mark was have a sub board in the building itself, in essence 3 boards, Main in the house, main in the out building and then your sub board say in the middle of the proposed ring, and then you could run 2 staggered rings one left one right, or then just 2 radials

 

[Mark42]

What I meant Mark was have a sub board in the building itself, in essence 3 boards, Main in the house, main in the out building and then your sub board say in the middle of the proposed ring, and then you could run 2 staggered rings one left one right, or then just 2 radials

Thanks, Malcolm, an interesting idea. But I don’t understand how that might be easier, or any more economic or better designed than running two radials or rings from the single sub-main in the building. Plus it would add complication.

Despite what others have said, 60m of 2.5 for a ring on a 20A breaker is absolutely not too long. Last night I looked up the specs for ‘Standard domestic circuits’ in the ECA 17th Ed guide. Max length of 2.5/1.5 T&E with a 32A breaker on TN-C-S and a 30mA RCD is 106m.

And I’m using SY, which has all the cores the same CSA so a 2.5mm CPC, and clipped direct to cold metal. Therefore 60m is clearly WAY within spec without even bothering to look up the cable specs and calculating volt drop.

The only ‘departure’ (if that’s what it is) is from the maximum floor area guidance. No one’s addressed that issue.

Remember that the main supply is only 25A per phase (no more is needed, so a 10mm incomer could be specified). So multiple 20A radials or even 32A rings on the same phase are pointless, since by definition they could never be fully loaded anyway. I forgot to mention that one reason I was protecting the ring at 20A was to provide discrimination (probably) between that and the 25A main switch.

One can think of it as two 30m radials (regs allow 40m of 2.5/1.5 T&E), both terminated in the same 20A breaker, but with the far ends of the radials then linked (so it then becomes electrically a ring circuit), which I considered good design since if then offers huge redundancy in the cable spec. What is wrong with that??

Enjoy KSA – I was in Dhahran in 1991, and subsequently, and hated the place (As the hookers say, just close your eyes and think of the money!)

Perfect example of a little knowledge being dangerous.

Perfect example of not reading someone’s posts properly then jumping to conclusions.

Enjoy the exercise!

 

[Mark42]

In that case you may have a problem using your PME earth in this building, especially if your kids are going to be playing with hoses bare footed. There could well be a potential difference between true earth and the PME earth in this instance. If you did have a grid in the foundation floor that could be bonded, then that would have been acceptable.

The ‘children with bare feet and hoses’ was my over-the-top example of why it is always better to design for worst-case scenarios, that’s all. It’s unlikely to happen in real life! I mean, how could I allow my little children to spray water all over an electrical installation? It’s completely unthinkable.

They might make it rusty!

Looks like you’re going to have to go the TT route and your rods as you suggested, at opposite corners tied on to the structure and the structure to your MET. No need for 25mm cable 10mm will be fine!! Just make sure you go down a good depth with your rods, at least 2 rods deep at each corner, preferably 3 rods if your Ra is decreasing significantly the deeper you go with the second rod....

OK, and thanks for the advice, and sorry for the delay, I had to sleep on this one.

I still don’t understand why it’s not OK (electrically) to use the ‘exported’ PME, in parallel with a TT system. I’ll come clean: I’ve never really understood this. But I’ve never met an electrician on any site who could explain it to me either, without going round in circles, or using the ‘that’s how it’s done’ response and getting annoyed. (Which means of course he had no clue either.)

In accordance with modern practice, the floor slab is underlain by a 1000-gauge PVC sheet on sand blinding, so is probably electrically isolated from the earth. Rag bolts go from the concrete to the metal stanchions, but these do not pierce the DPC. But then of course the stanchions will be bonded to earth via the new TT system.

In the case of a lost Neutral on the DNO’s side of the N-E link, presumably the problem is that load current will attempt to return to the star point via the earth infrastructure
which includes all the structural metalwork, appliance casings and so on, which someone might be touching, so the current path would pass through the victim and into the general mass of earth. Or is the point of breaking the PME to prevent this happening in the remote building? (Since breaking the PME is really the same as breaking the N-E link at the service head, through which the fault current would need to flow.) BUT if there are additional low-impedance earth paths present on the consumer’s side – ie my stakes, what is the problem? I just can’t see it. Perhaps I’m being thick.
Plus anyone touching exposed metalwork would divert some of the current and trip RCDs.

 

[Mark42]

Job done.

Pic of the shed shows what I was on about – I wanted to run an RFC all around the perimeter.

But OK, I’ve listened, and changed my mind. I installed it as two separate radials, on two 20A breakers, rather than as a single ring on one 20A.

I believe it's electrically not as sound as my ring proposal (the CSA is lower), but I agree more ‘standard’. An example of where strict compliance with regulations is actually worse electrical design than ‘breaking’ the rules.

Also a pic of using SY for the sockets but without the correct SY glands, designed, I believe, primarily for where EM screening is specified.

Unravel the screen, twist, sleeve g/y, then shrink-fit the lot. Outer sheath ends in a normal cable gland. Is this OK? BTW, the holes in the back of the metal boxes will be sealed to keep insects out. Not fitted the accessories yet, and though I’d better consult this wise oracle before proceeding …

Thanks, Mark.

 

[mogga]

Im assuming this may be for FLT batteries iv seen them done with busbar a few times dependant on the load

Only concern I may have would be the "wosk area" and any welding ect in the presence of battery fumes ect

 

[Silly Sausage]

Massively envious of your 'shed'!

 

[Mark42]

Im assuming this may be for FLT batteries iv seen them done with busbar a few times dependant on the load

Only concern I may have would be the "work area" and any welding ect in the presence of battery fumes ect

I was asked this before.

Naa, they are only for ordinary batteries: cars, lawnmower, motorbike, children’s quads, that sort of thing. Nothing big. And only little ‘Accumate’ maintenance chargers to stop the batteries self-discharging when left for a long time.

No hydrogen gassing off to speak of.

Massively envious of your 'shed'!

Thanks Archy. I like my shed too. The only way I can afford admittedly ridiculous projects like this is by doing everything myself. Which is one reason I’m on here asking questions …

Oh, and the other thing is to buy nearly everything on EBay, and to throw away your Tv.

 

[Silly Sausage]

........... is by doing everything myself. Which is one reason I’m on here asking questions …

Absolutely, go for it! (as long as you comply and get by the Notification Police (sarcasm can be taken as read there))

Oh, and the other thing is to buy nearly everything on EBay, and to throw away your Tv.

Ebay is my friend too. Need to get unhooked from trawling Electricians Forum next!

ghjdjuyhdtyjgutdrse6vbaew

 

[UKMeterman]

Hi,
Don't lightly solder where screw terminals are concerned, the screw pressure over time will cause the solder to cold flow. Have a look at farnell part number 1121755 for the correct item to use.

 

[Mark42]

Hi,
Don't lightly solder where screw terminals are concerned, the screw pressure over time will cause the solder to cold flow. Have a look at farnell part number 1121755 for the correct item to use.

Thanks - Very good advice about the cold flow.

I already ordered some 6mm (yellow) twin-entry ferrules, which two x 2.5mm SY cores fit nicely, but once crimped they won't go into the terminals!

The Farnell site is not clear on the spec of 1121755. Is that TWO x 2.5mm cables or 2.5mm TOTAL CSA?

I don't want to remove strands if it can at all be avoided.

 

[Silly Sausage]

Don't cut any strands!
Try the next size down.

 

[malcolmsanford]

Most likely me Mark, but I see a load of yellow and green cables there but not a brown in sight. I see where you have heat shrank the braid ok, I see a couple of Blues and then 4 what look like CPC coloured cables ..........

 

[johnboy6083]

malc, there is 2x brown, but it looks like a black in the flash. 2x neutrals, 2 x CPC and 2x braiding

 

[malcolmsanford]

Lol I still can't see them, but take your word for it mate

 

[UKMeterman]

It is designed for 2*2.5mm to sit side by side.

 

[Mark42]

Thanks, John! Yes, Malcolm it is you: I reckon you've had too much desert sun Or you need a new monitor ...

It is pretty poor sleeving, from TLC. All a bit stuck-together and floppy, I don't like it much. I prefer to shrink-fit cable ID markers, but it's a waste of money for all that lot, which no one will ever see again.

 

[Mark42]

It is designed for 2*2.5mm to sit side by side.

OK. Ordered on-line just now!

Malcolm, joking apart, this is a massively interesting psychological effect, especially relevant when quality-assuring one's own work in many trades. Can you really not see those two brown cables? They are quite clear and absolutely in your face!

I can't remember what the effect is called, but I know psychologists have studied it. There is a book by Prof Richard Wiseman called ‘Paranormality: Why we see what isn't there’ which touches on it. Something about the 'gorilla in the room'.

Subjects are shown a video of people arguing in the street or something, but no one, when debriefed afterwards, remembers seeing the guy in a gorilla suit walking by. But he's there and absolutely clear. I reviewed the book on another forum where I annoy people:

On all things wondrous strange: ghosts, mediums, and rubber hands - Mark Ribbands - RichardDawkins.net

It’s all very fascinating.

 

[malcolmsanford]

To me they look like 2 conductors covered in black heat shrink/sleeving/neoprene, they don't look brown at all but there again I have old eyes