New circuit design question

[David Prosser]

But if there is a fault then either the main fuse or the isolator fuse could blow..

If there was a fault it could blow the DNO fuse even if you put 80A or 63A fuses in the SFI.

 

[pc1966]

Now I'm confused.
So you are saying that a fused isolator is required, due to the distance from the meter to the first CU.
What's the point in putting in a 100A fuse. We have always recommended a fuse lower than the main fuse size?

You have every right to be confused, as the regulations on it are a bit convoluted!

Normally you must provide an OCPD when there is a change in the CCC of the cables (434.2) and you must provide a means of isolation (462.1) and it has to be appropriate for the type of customer/user (e.g. 537.2.8).

When you have tails or sub-main longer then 3m it is regarded as part of the installation so it has to be isolatable without pulling the DNO fuse, so you need a switch.

Now the DNO provides a fuse but it is for the protection of their network, not the customer's. However, if the DNO agrees that their fuse is acceptable protection you can omit the OCPD (433.3.1 section (iii) and 434.3 section (iv)) and get away with only a switch.

Generally getting the DNO to provide an agreement in writing as part of your system design is going to cost you time and effort and probably it is cheaper to simply ignore the omission rule and put in a switched-fuse instead of just a switch.

Now how big a fuse? Well you have a couple of factors to consider:

• The fuse must provide adequate protection of the sub-main under fault conditions, so (probably) 5s or less to meed ADS (assuming 411.3.2.3)
• The design musty have overload protection for the sub-main, probably by the incoming fuse, unless it is certain the nature of the loads (or totals of DB breaker currents, etc) cannot overload the cable so fault protection is sufficient (433).
• It is best if you achieve selectivity with the DNO fuse so after a major fault you just change the fuse under your control.

The first point has to be met, so if you have a high Ze and/or long sub-main so R1+R2 significant your choice is going to be either reducing the fuse size, or going to a delay RCD to earth fault protection.

The second point is usually met on long cables if voltage drop and fault-disconnection times are being met, but again is something that the designer has to have factored in.

The third point is not essential, as the DNO network is protected by their fuse, and really the design of the installation should not be such that is is likely to overload the supply. But if you want to achieve this by fuses you are looking to a 1.6:1 ratio or more (so 63A if DNO is 100A, etc) or going down the delay RCD route (as it is most likely going to clear before the DNO fuse goes expect for a high PFC, but of course only for faults to earth, not overload).

Here the OP has other issues, the tapping off of other supply kiosks allows for them to have fuses that are selective against the DNO/incoming fuse, so the total loss of supply would only be if the sub-main was damaged.

But the details of loads, etc, are not here and I am getting hungry so time to stop rambling!

 

[RyanAspire]

Thanks everyone for replies!

I don't expect the load at any kiosk to exceed 10A, except for the last kiosk which will be around 40A I would suspect, but again, there may be more in the future (there is a barn close to the final kiosk which will likely want lighting and sockets in eventually.

I have decided to go for an 100A switched fuse after the meter. 100A main isolator in kiosk 2 and 3. With the feed from kiosk 2 down to 3 via an 80A MCB. Then from kiosk 3 to the final kiosk via a 63A MCB. All lighting circuits on MCB and the rings on RCBOs. At the final kiosk will be a 30mA RCD which will feed a future ring and lighting,it will also serve a 32A MCB which will go onto a hookup point (which will have an RCD within too as regs state the static caravan must have its own RCD supply). My understanding from looking at the IET page someone kindly put up is from here I will run the SWA to the DB in the final kiosk as normal, from there I will only run L+N to the hookup point from where I can then connect an earth rod, with live, neutral and earth from here supplying the static caravan.

Does all this sound about right? Also with regards to Ze, DNO has kindly noted by their fuse that Ze is 0.14.

Again thanks everyone for the input. I don't think I was much different from people's responses but it's definitely given me some clarity on some things and help me make subtle adjustments, especially with regards to the static caravan having never supplied one. Also within BS7671 could it maybe be seen as a grey area? There is 708 (i think) which is regarding caravan parks and camping parks or 721 (again I think) which appears to relate to the actual internal installation of a caravan. In this instance, it isn't deemed as a caravan park as it is one residential static caravan on the property of its owner. Albeit I appreciate 708 is the closest regulation covering this application

 

[dodger421]

I’m willing to be proven wrong but I think you’ll struggle to find an MCB that will accommodate 70mm and fits in a standard CU or DB. Most MCB’s I’ve seen only accommodate up to 25mm.

As for the run to the caravan point, yes you can run a 2-core SWA but you still need to earth the armour at the supply end but isolating it from the enclosure at the hookup.

 

[RyanAspire]

I’m willing to be proven wrong but I think you’ll struggle to find an MCB that will accommodate 70mm and fits in a standard CU or DB. Most MCB’s I’ve seen only accommodate up to 25mm.

As for the run to the caravan point, yes you can run a 2-core SWA but you still need to earth the armour at the supply end but isolating it from the enclosure at the hookup.

I don't believe you will be proven wrong, I've looked and looked and looked! The only two options I see I have is either terminal blocks in the DB (70mm in 25mm out) or the preferred option, reducing pins.

 

[westward10]

Are these single phase boards, 70mm? good luck.

 

[Spoon]

If there was a fault it could blow the DNO fuse even if you put 80A or 63A fuses in the SFI.

So because there may be a very slim change of the main fuse blowing with a 63A fuse in the SFI, then you choose to do no selectivity whatsoever... Interesting.... Not something I would do.

 

[RyanAspire]

80A switched fuse is what I meant too

 

[RyanAspire]

Are these single phase boards, 70mm? good luck.

They are, i have looked to buying ones on the larger side to help with the cable size. Is there an easier/better alternative?

 

[pc1966]

They are, i have looked to buying ones on the larger side to help with the cable size. Is there an easier/better alternative?

How about using a branch joint to bring off some 25mm SWA that connected to your kiosk:
https://www.cef.co.uk/catalogue/pro...-resin-joint-kit-for-cables-up-to-70mm-4-core

Wrangling 70mm won't be fun and few terminal blocks, etc, are going to take that large a cable in double. Also that kit means no joints along the length to slightly less chance of an overheating terminal, etc.

 

[westward10]

It is a mission to get a 16mm into a single phase boards and a 70mm will be mission impossible. Just the gland for a three core would be 40mm.

 

[pc1966]

I don't expect the load at any kiosk to exceed 10A, except for the last kiosk which will be around 40A I would suspect, but again, there may be more in the future (there is a barn close to the final kiosk which will likely want lighting and sockets in eventually.

I have decided to go for an 100A switched fuse after the meter. 100A main isolator in kiosk 2 and 3. With the feed from kiosk 2 down to 3 via an 80A MCB. Then from kiosk 3 to the final kiosk via a 63A MCB. All lighting circuits on MCB and the rings on RCBOs. At the final kiosk will be a 30mA RCD which will feed a future ring and lighting,it will also

If you expect 10A max or so then why a 80A MCB???

Also you won't get very good selectivity with MCB/fuse, and none guaranteed if 100A fuse and 80A MCB!

Though to be fair you probably won't have a PFC/PSSC high enough by the end to cause problems (e.g. Hagar say 63A B MCB of the NBN commercial B-curve is selective to 4.3kA with a 100A BS88 fuse).

Really you would be much better putting in 63A switch-fuses for each kiosk and then your choice of 32A/6A RCBOs, etc. Then you get selectivity with your incoming & the DNO fuses, can safely run 16mm or 10mm from the output of the switched fuse to the kiosk, and the final circuits are selective to around 1.1kA for faults on the ring not taking out the 63A fuses (again, based on Hager guide and that by the far end I doubt you will be seeing much more than 1kA or so PFC).

 

[pc1966]

How about using a branch joint to bring off some 25mm SWA that connected to your kiosk:
https://www.cef.co.uk/catalogue/pro...-resin-joint-kit-for-cables-up-to-70mm-4-core

I think they so a MS70 that might be better here for taking off a thinner cable, but not obvious on the usual suspect's web sites:
https://www.birkett-electric.com/v3/multi-service-cable-joints/

 

[David Prosser]

So because there may be a very slim change of the main fuse blowing with a 63A fuse in the SFI, then you choose to do no selectivity whatsoever... Interesting.... Not something I would do.

A fault that is going to blow the 80A fuses is also likely to blow the 100A fuses you get no realistic discrimination . There's no point in limiting yourself further down stream by starting with 80A fuses.