Garage Board - achieving discrimination?

[CraigL]

Hi all,

New to the forum so thanks in advance for your help.

I've just fitted a garage board details below:
Main Consumer Unit is Dual RCD - whoever did the re-wire had already fitted 6mm t&e direct from cu into an external jb. I've connected the t&e to swa (house end) and at the garage end the swa is fitted direct to the garage board.

The garage board is a dual board with an rcd main switch. At present I have it connected with rcds on both sides, and it is working fine. Note the garage is only used for storage and rarely used, it's not being used for a workshop. Therefore imo there is no danger caused by having rcds at either end, merely an inconvenience due to any tripping- but I don't want to leave a job if there is a 'realistic' better/safer option.

My understanding of the best option for installing a sub-main in a garage is to have the garage board rcd protected and the feed from the cu from a non-rcd protected breaker - and therefore no inconveniece tripping. However the consumer unit is dual rcd, so how is possible to fit a non-rcd protected breaker?

It's been suggested that I could have changed one block to rcbos, except the garage board breaker - but then it's an additional expense. Alternatively is it possible or even allowed to remove one of the ways from the busbar, and therefore leaving a spareway not connected to any rcd and take a direct feed from the main switch to this breaker - but this would mean 3 lives leaving the main switch, which I don't think would fit? Or could I take a feed from the live side of one of the rcds. In doing this surely I would have to fit another neutral bar and therefore I would be changing the board from it's initial design, implications??

Thinking allowed as I'm typing, even if it is possible to achieve the above scenario, doesn't this mean that the t&e from the cu to the external jb is now non-rcd protected...

Thoughts??

Thanks

 

[Taylortwocities]

Michael, you have missed some essential points.

1. The new tails (heading towards the garage) MUST pass through a current limiting device. RCDs do not limit the current.
So a switch/fuse is needed at that point (I think I detailed this in this thread about 6 months ago. I CBA to read all the way back).
The fuse needs to be appropriate for the SWA.

2. A time delay RCD (or any RCD) is NOT necessary at the house end.

 

[Michaelwgroves]

Michael, you have missed some essential points.

1. The new tails (heading towards the garage) MUST pass through a current limiting device. RCDs do not limit the current.
So a switch/fuse is needed at that point (I think I detailed this in this thread about 6 months ago. I CBA to read all the way back).
The fuse needs to be appropriate for the SWA.

2. A time delay RCD (or any RCD) is NOT necessary at the house end.

Absolutely agree about switch fuse, schoolboy error.

I agree time delay "might" not be necessary, but it gives another level of protection I think is worth the extra small cost.

 

[Taylortwocities]

I think is worth the extra small cost

Time delay RCDs come in at £80+

Imo, bit more than a "small" cost! plus an enclosure.

 

[davesparks]

Dave, unsurprisingly you are correct. Had time to check my understanding this morning. Obvious when you take time to think about it.

So best solution would be pull tails from house CU into service block. Then new tails into house CU and new RCD enclosure with time delay RCD. Then SWA to garage CU with RCD.
Cheap simples and best soloution. Any fault in garage will only trip garage RCD, with SWA cable also protected by time delay RCD.

Why the time delay RCD?

 

[davesparks]

Re-reading the OP there is T&E for part of the run so it may need 30mA protection for that.

 

[davesparks]

I agree time delay "might" not be necessary, but it gives another level of protection I think is worth the extra small cost.

Level of protection against what? It does not provide shock protection and the fuse is more than adequate for fault protection.

 

[Midwest]

Level of protection against what? It does not provide shock protection and the fuse is more than adequate for fault protection.

Don't you start, had a similar long debate, a while ago.

 

[Michaelwgroves]

No, please do start as this thread has improved my knowledge which has got to be a good thing. And hopefully for others.

Agreed time delay RCD does not provide shock protection, but it does still provide earth leakage protection. So I'm still thinking it's a good thing ?

 

[davesparks]

No, please do start as this thread has improved my knowledge which has got to be a good thing. And hopefully for others.

Agreed time delay RCD does not provide shock protection, but it does still provide earth leakage protection. So I'm still thinking it's a good thing ?

Why does it need earth leakage protection?

 

[Michaelwgroves]

If someone working in the garden accidentally hits it, it will trip early. Rather than not at all if it was an Earth fault?

 

[davesparks]

If someone working in the garden accidentally hits it, it will trip early. Rather than not at all if it was an Earth fault?

That's why SWA is used, to ensure that a fault to earth occurs if the cable is damaged.
The fuse is providing fault protection.

 

[Michaelwgroves]

So that begs the question; what use does a time delayed RCD have?
We know it must be used for RCD's in series.

 

[davesparks]

So that begs the question; what use does a time delayed RCD have?
We know it must be used for RCD's in series.

Time delayed RCDs are used to achieve discrimination where the upstream device is providing fault protection because the required EFLI for the mcb or fuse cannot be met. For example TT systems where you can rarely achieve a low enough Ra to allow mcbs or fuses to operate in the required time.

 

[Ian1981]

As a s type max allowed tripping time is 500ms and 130ms min tripping time then I wouldn't particularly use it where you need a disconnect time of 0.2 seconds for final circuits upto 32 amps(depending on rcd tripping time when tested) 100ma standard device would be used for fault protection only.
Where a s type is useful is protecting tails entering a metal enclosure and distribution/circuits above 32amps which have generally a 1 second disconnect time and discrimination against downstream rcds.
For a TT system