From an entirely practical point of view - do you have any control of what will be plugged in to the sockets and how many start/stops might happen? 

I have at least one 110V 16A extension lead that is probably 50m long. 1.5mm three core yellow flex.

All to often I have had a clever wally connect his large chop saw to a 110V lighting circuit (Often 110V work stand lights will come with two 16A 110V sockets on them) with the intension of doing hundreds of chop saw start/stops per day, and then is surprised when it trips the lighting circuit or melts the lighting cables if the lights are plugged into larger overcurrent protection. 

All too often I have seen melted 110V cables because of 110V machinery start/stopping or a 110V  floor sander. People will take off a 32Amp for a 16Amp to make it work. I do understand that the thinking that a 32Amp should protect the 4mm circuit, but not the extension lead or the actual plug socket its self. (fire is the risk I'm most concerned about here)

 if you cannot work out which formula applies and try and memorize them but just get it all wrong, its a recipe for disaster, so lets  chuck all that and  go back to first principles.

so in big loose handfuls, 1m length of 1mm2 is 16 milliohms cold, and more like 18 when hot.

but you have 8 such 1mm strands, in parallel, (2 paths of 4mm) going 50m there and then 50m back.So that is ~ 2 to 2.2 milliohms per m, and 100m there and back call it 0.2 to 0.22 ohms, for the round trip  at 16A that is 3,2V to 4V and at 32A that is 6,4V to 8V.

So, is the max load at the far point 16A or 32A ? that depends how the sockets are laid out.

Now OK. what is the regulation of the transformer ? don't know but could be up to  5% drop at full load, Then there is the drop in the primary circuit, where of course the voltage is more or less doubled, but the current is halved and you have not mentioned that at all.

you could be approaching or even exceeding 10%. and then you are getting to a place where a C type breaker will not prompt trip.

Mike.

I don't understand that bit about voltage drop. Ok, on one leg of the ring which is 50m there is 8.8V drop but this is a ring arrangement so this will be compensate with additional 50m 4mm2 cable to the furthest socket so we should ended up with about half of this amount so 4.4V sounds better now. Current me if I am wrong.

Sorry, I did a bit of simplifications (or sleight of hand) there - which I would probably done better to explain. I presumed there was 50m of cable to the furthest point (so ring had 100m of cable overall) and presumed the ring was balanced and the total load could be up to 32A, with the possibility that all the load could be close to the furthest point. In those circumstances both legs would be carrying 16A each over the 50m - hence the calculation above. You could equally use the /4 formula, but then you'd have to plug in the whole 32A and 100m rather than 50 - so you'd end up with the same figure.

   - Andy.

This is the biggest concern I have because the customer is not sharing any data regarding transformer neither protection fuse. It's only guess from previous project made by my colleagues. I am not ever sure if they have RCBO/RCD fitted or just MCB. Basically, I was told to provide 1x 16A socket in each module. There are 8 modules so 8 sockets are required. They are need to be done in ring arrangement and they finish the installation once the unit on site. I am a little bit lost now. Should I increase the cable size from 4mm2 to 6mm2 to decrease VD or split this CCT to 2 smaller ones? I am also waiting for the customer to sent me more details about what is provided on their side so this may clear the situation out a little bit.

well, as I kind of said above,  then you are not designing it, they are, and they are adopting all attendant responsibility, and that needs to be made clear to them. You can only design with full data. ;-)

Do you have a web link to the sockets with build in fuses/breakers that are to be used ? I'd be interested.

Mike.

I was told to provide 1x 16A socket in each module.

Where did the idea of a ring come from?  Normally they're only used for 13A sockets (because the plugs are fused) or much much larger thing in the distribution world (true ring mains) otherwise radials are much more usual.

+1 for needing proper requirements if you're going to design it (you'll need Zs for starters even to decide if an MCB will be suitable.  I suppose you could try a "robust" design that could cope with just about any input and load (rather like some people have to do for mobile units) - e.g. a Class-II DB with MCB+RCD for each outgoing circuit arranged as say 8 radials in 10mm² (to get v.d. below 6.875mVA/m for 16A over 50m and stay below 5% of 110V) - but that's likely not to be economic and still potentially be vulnerable to some edge case issues.

   - Andy.

According to all the information and things need to be taken into the consideration in terms of this 110V ring circuit I am not sure what to do now.

The customer in now willing to share any more info to me apart from the fact that the transformer is 5kVA and protection device they use for this circuit is 32A MCB type C

I am not going to argue with them as end of the day they are going to do the final connection to the LVAC unit so their expectation from my side is to fit 8 of 16A sockets into each room. I am going to use 6mm2 cable instead of 4mm2 to decrease VD on this distance. Do you also thing it would be a good idea to use 16A or 20A RCD sockets to protect every individual socket and have some discrimination as well?

Do you also thing it would be a good idea to use 16A or 20A RCD sockets to protect every individual socket and have some discrimination as well?

In general, if the sockets are fed from a 32A circuit, I'd certainly have 16A overcurrent protection at each outlet, so you're probably looking at sockets with a modular DIN rail enclosure built-in - in which case adding an RCD in there too (either as a separate device or an RCBO) might well make sense. That said, the shock risks from a 55-0-55V system are much reduced anyway, so there's much less of a need for 30mA RCDs than for 230V sockets.

But you probably first have to decide how the main circuit provides ADS (if you're going this to BS 7671 so need at maximum 5s disconnection time). Presuming it's a typical 110V CTE RLV system - i.e. 55V to Earth, 32A C type has to have Zs below 0.16Ω (BS 7671 table 41.6) - if you have a 6mm² copper c.p.c. as well as line conductor you're talking about a resistance of around 7.3mΩ per metre - so 100m worth divided by 4 gives 182.5mΩ or just over 0.18Ω for R1+R2 if you like - so you're over the limit even before you start thinking about adding on the transformer impedance and the L-N impedance of the 230V supply circuit (the equivalent of Ze or Zdb). If you used T&E with a reduced c.p.c. it would be even worse of course. So you might have to think about an RCD at the transformer end ... which then begs the question of do you need some at the sockets as well? Of course the supply end one needn't be 30mA instant type - it could be say a 300mA time delayed (S-type) so that 30mA units at the sockets could discriminate with it - but then your costs are increasing and there's a trade-off between capital costs service convenience to be considered. Or a radial per socket with 16A overcurrent protection (and perhaps an RCD) at the supply end might start looking attractive.

If you're going to do the work and they aren't supplying a detailed electrical design, then I'm afraid you're the designer and you'll have to do the maths. The OSG and many of the usual rules of thumb aren't much use for 110V circuits as they're all based on 230V and often the results are very very different.

   - Andy.

"transformer is 5kVA " what is the duty rating?

Is it five minutes on and fifteen minutes off or continuous?

Could it be Intermittent Rating: 5.0 Kva and Continuous Rating: 2.8 Kva?

well then they need to spec the cable as well, if they are holding the design data.

by the way a "5KVA transformer" may mean one where the output has dropped by a few % at 5KVA not one rated for 24hour load at that level.
This example is 25% duty cycle, i.e. 5 minutes on / 15 minutes off.. and is more than 10% voltage drop at that load point.

The way transformers are specified application note

Mike.