SUB MAIN QUESTIONS. ADVICE NEEDED

[nicebutdim]

It's how I was instructed to do it by the French "Conusule" who are the regulating body in France, the incoming main is into a S type RCD at the meter this is EDF equipment you have no choice.

That would also be a TT supply?

 

[oscar21]

Mcbs are not likely to achieve selectivity. Using such devices for distribution is a poor design.

I had this argument with someone a while ago, he wanted to protect a sub main to a summer house with a 40A MCB in the consumer unit and I was saying that if the 6A lighting circuit has a problem it could well trip the 40A MCB before the 6A one, it just depends which is the quickest to go.

I was trying to explain that if you did a loop impedance test/IPF and it showed more than say 200A which it could well do, then if there was a dead short it could take out the 40A one instead, or maybe both but he was adamant thats what 6A MCBs are for and just couldn't understand it.

 

[pc1966]

I had this argument with someone a while ago, he wanted to protect a sub main to a summer house with a 40A MCB

Selectivity will be poor, but sometimes it comes down to how important it is. For an occasional summer house it is probably acceptable risk, but not for many circuits in a home where loss of stair lighting, etc, has other hazards.

Next time point them to the tables of selectivity as that will give figures for PFC/PSCC that are likely met, but in most cases you will find the end circuit exceed them so a real risk of supply loss.

That is why fuses or MCCB are the best choices, then the Ryefield or MCCB panel choice usually comes down to likelihood of a trip and need for non-skilled folks to reset.

 

[oscar21]

Are MCCB's less "sensitive" than an MCB then like fuses are, I'm no expert in commercial stuff.

 

[pc1966]

Are MCCB's less "sensitive" than an MCB then like fuses are, I'm no expert in commercial stuff.

Once you get pas the slow thermal trip side, many MCCB have a short-delay aspect (sometimes adjustable) of 20ms or so at modest overload currents, then once you hit really big faults in the kA range they go on an "instant" trip in the 10ms or less region to energy-limit the fault, just as MCB do.

It is this short-delay aspect that gives them decent selectivity with MCB that all go "instantly" as soon as you get past the thermal limit. That is one major advantage for applications like this where you are feeding MCB/RCBO boards.

The other aspect that help with MCCB is they often have adjustable settings, for the cheaper ones (and they are not cheap, this is a relative term) that is usually just the thermal trip point and you might get 80-100% adjustment on a given size, but that is enough so with a bit of choice you can design for any specific limit. For the fancy electronic ones you can dial in all sorts of trip points to get the sort of characteristics you need, and they can comes with RCCB aspects as well giving you a very fancy RCBO to practically any trip characteristic you want. At a price!

Fuses are old technology and cheap by comparison, but they naturally have a "delay" in how the wire heats up, and they offer exceptionally good energy-limiting. So if you really don't expect them to be blown/tripped, they are a very good alternative and will just sit there for decade after decade after decade, and still work if called upon.

 

[Rockingit]

Fuses are old technology and cheap by comparison, but they naturally have a "delay" in how the wire heats up, and they offer exceptionally good energy-limiting. So if you really don't expect them to be blown/tripped, they are a very good alternative and will just sit there for decade after decade after decade, and still work if called upon.

Hence my suggestion of a small Ryefield!

 

[pc1966]

Hence my suggestion of a small Ryefield!

If it is feeding sub-mains so 5s disconnect and all 63A or above it makes a lot of sense.

But I think there is at least on 3P load for a heat-pump, that would be better served by some 3P MCB or similar so on fault all phases go down. Still a DIN box and MCB from a set of fuses is still quite a viable option.

 

[Rockingit]

If it is feeding sub-mains so 5s disconnect and all 63A or above it makes a lot of sense.

But I think there is at least on 3P load for a heat-pump, that would be better served by some 3P MCB or similar so on fault all phases go down. Still a DIN box and MCB from a set of fuses is still quite a viable option.

Thing is, it's never just a heat pump, is it. It's some local sockets, a light or two...... so my choice would be to supply a small 4way 3ph board from said Ryefield. If you're building a plant room, build a plant room! Not some abomination of tails spaghetti.

 

[nicebutdim]

Not some abomination of tails spaghetti.

I think this part bothered me most about the proposed installation.

It's the sort of set up one might expect to be faced with in an old and regularly repurposed commercial property.

 

[Mark42]

Not sure if I’ve understood this properly, but it appears to have parallels with my own installation. I highly dislike the idea of feeding SP sub-mains from a TP board, unless for very small local loads - which are not likely to increase in the future - like, maybe, a garden shed. I would never even consider a single phase 80A sub main.
For anything else I run 4 (or 5) core SWA to additional TPN distribution boards. Then you have the ability to properly balance loads between the phases, and to run any future 3P loads.
And it looks a whole lot better.
I buy Hager 3P DBs when they come up on EBay, and populate them with Hager RCBOs from the same source. So it doesn’t even cost much.
OP: are these new 3-core runs long, buried, or otherwise non-negotiable?

Edit: Apologies -I just noticed this is a zombie thread. I came to it through a ‘recent posts’ link.