Protection of TT submains

[JohnW2]

When I achieved a situation in which all final circuits in my house’s (TT) installation were explicitly protected by RCDs or RCBOs, my initial thought was that the up-front Type S RCDs had become essentially redundant. Neverthelss, I retained them, not the least because they served as useful isolators (and appeared to do no harm, other than the very small risk that they would take out a whole phase if there was a N-E fault on one of the couple of final circuits protected by SP RCBOs).

However, on reflection, I think that those up-front RCDs probably are still required, to protect the (several) sub-mains. 411.3.2.4 requires that a distribution circuit in a TT installation has an L-E fault disconnection time not exceeding 1 second. Certainly if I had anything approaching a ‘pure TT’ installation, there’s obviously no way that such a disconnection time (indeed, any disconnection at all!) could be achieved by the 60A/80A fuse protecting the sub-main – so, it seems to me that I still do need those RCDs. Do people agree?

In practice, ‘incidental paths to earth’ in my TT installation are currently such that EFLI is very low, probably low enough for the sub-main fuses to provide the <1 disconnection time. However, it doesn’t feel right that I should rely on that, since a man from the water supplier with a bit of plastic pipe could change the situation dramatically ‘overnight’. Again, do people agree?

Kind Regards, John

 

[ricicle]

Are the sub-mains buried anywhere less than 50mm deep (assuming t+e) ?

 

[JohnW2]

Are the sub-mains buried anywhere less than 50mm deep (assuming t+e) ?

They are singles, not T&E - but, no, they are not buried anywhere - essentially all 'clipped direct' (largely under floorboards). However, although being buried <50mm would itself create a need for RCD (or mechanical) protection, 411.3.2.4 would appear to always apply, regardless of cable installation method - or am I missing something?

Kind Regards, John

 

[ericmark]

On a building site to protect SWA cables with 1A, 300ma, or 100ma delay type RCD may make a lot of sense. Also varying delay times would seem to make sense.

However when the guy put a nail in the wall to hang his coat on it took the lot out so the idea of discrimination does not work.

With a gradual increase in leakage yes does work but not the nail in the wall.

What may be a problem is SWA glands where using the SWA as earth. You really want a metal enclosure to gland to and to have that metal enclosure you need the RCD protection before the enclosure.

 

[ericmark]

Singles in what? I would assume conduit and again to earth conduit need the RCD before the metal conduit.

 

[JohnW2]

Singles in what? I would assume conduit and again to earth conduit need the RCD before the metal conduit.

6181Y singles in nothing. No conduit.

Kind Regards, John

 

[ban-all-sheds]

6181Y singles in nothing. No conduit.

 

[JohnW2]

6181Y singles in nothing. No conduit.

Even though I inherited (rather than installed) it, what's the problem? It's no different from T+E (and, as I've said, is protected by an RCD as well as a fuse) - would you be spanking me for T+E, as well?

Kind Regards, John

 

[mfarrow]

They're singles in sheath, that's the protection.

 

[JohnW2]

They're singles in sheath, that's the protection.

Exactly - that's what 6181Y is! I really don't understand BAS's problem - does anyone share his unhappiness?

Kind Regards, John

 

[Iggifer]

maybe he's thinking you meant 6491X?

 

[JohnW2]

maybe he's thinking you meant 6491X?

That would be the most obvious explanation, but that would be a very unusual error for someone as precise as BAS to make!

Kind Regards,John

 

[Iggifer]

*fingers crossed* hahahaha

 

[JohnW2]

*fingers crossed* hahahaha

Quite so! He appears to be slightly conspicuous by his absence at the moment!

Anyway, interesting though this discussion about the physical nature of my submains may be, does anyone have any thoughts/views about what I wrote/asked in my OP?

Kind Regards, John

 

[EFLImpudence]

If the only way you can satisfy the disconnection times is with an RCD then one will have to be fitted for each distribution circuit.