What MCB Type and Rating do I need?

I am installing few Air-Conditioning (A/C) units in my home - cable length from A/C units to Consumer Unit (CU) range from 7m to 30m. The A/C units capacity range from 1.2KW to 1.9KW. I have used 2.5mm (singles: red, black, 1.5mm Earth) cables to run each A/C unit's 20A scoket direct, as separate circuits to the CU. My CU is Wylex - triple Phase Board - 12 ways (36 singles), I'm using single phase MCBs - Type C.

Questions:
1. Is it ok to use the 2.5mm cable to run the connections to the CU?

2. How do I protect the circuits - what rating of MCB should I use to protect each circuit? Note that the A/C units are: 1.2KW x 7 units, 1.2KW x 2 units, and 1.9KW x 5 units. 14 individual circuits.

3. Is Type - C the right MCB to use?

4. The CU has no RCD protection. I really want to have the house protected against electric shock. How can I do this? I am thinking of changing the 125A Incomer disconnector Switch (4Pole) to RCD Incomer. Is it possible to do this? I understand that the CU will trip off rather than the affected circuit only - switching off the entire power of the house. What is the alternative? If this is possible what rating of RCB do I need to use?

5. Because my CU board comes with Type C MCBs, its means that all the circuits may have to be protected using this. Is this ok? What rating should I use for the lights, Sockets, and as well as the A/C units as explained above?

I really will appreciate your advice and recommendations as always.

Thanks

Why did you use a reduced size earth and old colours? How big is the house
To be honest you'd be better employing a sparks to come and see what needs to be done, he can also inspect, test and certify the works.
One RCD covering the whole premises is a bad idea.

if you're running singles ( presumably in some conduit ) then the CPC has to be the same size as the line and neutral..

your best bet in these circumstances is to get a pro in to sort it.. he will have all the right test equipment to ensure that the circuits can support a type C breaker ( max Zs readings etc. ).
of you want rcd protection then RCBO's is the way to go in an install of tis size.. it's simply rediculous to allow some 36 circuits to go off because one circuit somewhere develops a fault..

Oriahig said:

I really will appreciate your advice and recommendations as always.

Click to expand...

My advice is that you get yourself a time machine so that you can return to a point before you started installing cables and MCBs and then spend some time learning about how to design circuits before you actually do it, so that you don't have to come here and ask how to choose cable and MCB sizes after you've already installed the circuits.....

The info here should be of use to you:

• //www.diynot.com/forums/viewtopic.php?p=76467#76467
• //www.diynot.com/forums/viewtopic.php?p=81696#81696
• http://www.tlc-direct.co.uk/Book/4.3.1.htm
• http://web.archive.org/web/20080213151445/http://www.kevinboone.com/cableselection_web.pdf
• http://www.tlc-direct.co.uk/Technical/Charts/VoltageDrop.html
  

ColJack said:

if you're running singles ( presumably in some conduit ) then the CPC has to be the same size as the line and neutral..

Click to expand...

Sorry to question this Col as I'd always use uniform conductor sizes when running singles but why does the CPC have to be the same size and the line and neutral conductors?
What about calculating the minimum size of a required CPC via the adiabatic equation (543.1.3)?

yes you can use the adiabatic, but for joe public on here they aren't going to know what that is or how to use it so it's easier to select form the table, and up to 16mm² it's same size CPC..

it's late, my memory sucks and it's been over a year since I've had to use it.... so where do we get the I²t from again???

Adam_151 makes a good post on that very subject here

yeah I read that last night and am still none the wiser..
it's not something I've ever really had to use.. I always go with the table and never rely in the SWA as the earth..

so the I is the current it takes to trip the device from the graphs, but what t do you use? 0.1s?

so for a 6A B type 60898 it's 30A for 0.1s?

so I²t is (30 x 30) x 0.1 = 90

and for a 32A B type 60898 it's 160A for 0.1s?
so I²t is (160 x 160) x 0.1 = 2560

Seriously, 14 AC units, all with their own outdoor condenser?! The outside of your house is going to look like a gym.

I appreciate you came here for electrical advice, and don't mean to preach at you, but you could have killed two birds with one stone by installing a number of multi-split units or a large VRF/heat recovery system with one outdoor unit. You then wouldn't need such a ridiculous number of circuits for the A/C.

ColJack said:

so the I is the current it takes to trip the device from the graphs, but what t do you use? 0.1s?

so for a 6A B type 60898 it's 30A for 0.1s?

so I²t is (30 x 30) x 0.1 = 90A?

and for a 32A B type 60898 it's 160A for 0.1s?
so I²t is (160 x 160) x 0.1 = 2560A

Click to expand...

I think you are meant to use actual fault level as opposed to the figures on the graphs, that is where the circuit breakers are a nuisance as a mechanical device takes time to open whereas a fuse tends to blow quicker as the fault level current increases. Once you know the fault level you can look at the graph (or manufacturers data) to get the value for time.
I believe the I²t basically gives you the value in joules let through, as power can be calculated by I²R * t and the fault is of negligible impedance just leaves you with I²t.

ColJack said:

yeah I read that last night and am still none the wiser..
it's not something I've ever really had to use.. I always go with the table and never rely in the SWA as the earth..

so the I is the current it takes to trip the device from the graphs, but what t do you use? 0.1s?

so for a 6A B type 60898 it's 30A for 0.1s?

so I²t is (30 x 30) x 0.1 = 90A?

and for a 32A B type 60898 it's 160A for 0.1s?
so I²t is (160 x 160) x 0.1 = 2560A

Click to expand...

Isn't 'I' not the highest current that can flow under fault conditions?
So, it'll be 230 / Ze (or Zdb) at the board (or PFC) where your fuse / mcb is connected.

t is the time taken to disconnect the fault (I).

Or refer to manufacturers data for I²t values when t is 0.1s or less otherwise you'll be prompted to use a CPC larger than what may actually be required.

You'll also need to verify that I²t is less than K2S2.

surely the I is not the PFC ( unless you are sizing the earthing conductor for the CU ).. the fuse / breaker operates in the required time to prevent the conductor temperature rising too much..

if you had a really good Ze at the origin of say 0.1 ohms, then the PFC would be 2300A right?
if you look that up on the charts for say a 32A BS88-2.2 fuse then you're into the 0.0xxxx seconds range or lower, which then starts to get iffy

there is no way of knowing the Ze until you visit the site so you cannot plan the circuits that way..

as long as the Zs are met for the breakers / fuses they will operate within the given times..

(sqrt ((320 x 320) x 0.1)) / 115 = 1mm² CPC ( rounded up ) for a 32A BS88-2.2 fuse? that can't be right can it?


or do we use the 0.4s values for sockets and the 5s values for fixed equipment?

that would be

(sqrt ((220 x 220) x 0.4 )) / 115 = 1.5mm² CPC ( rounded up ) for a 32A BS88-2.2 fuse ( for socket circuit )

(sqrt ((125 x 125 ) x 5 )) / 115 = 2.5mm² CPC ( rounded up ) for a 32A BS88-2.2 fuse ( for fixed equipment )