On-site Guide 17th Edition

not got mine yet...

Nor me, but Amazon sayeth:

- Order Placed: 3 May 2008

- Dispatch estimate: 5 Sep 2008

- Delivery estimate: 11 Sep 2008

- Arrived this morning.

Max. circuit length for a 32 A ring protected by a Type B MCB is now 106 metres.

Where Part M applies (re. height of sockets and switches) the consumer unit should be readily accessible.

Fig. 7.3 indicates what could be considered the preferred core colours for 2-way switching: brown for the permanently live line strapper, black for common-common, grey for the switched line strapper.

What was it?

So, a watering down :( From 84m to 106 is hell of a percentage hype.

I suppose nobody plugs lights into a ring circuit so the extra 2% drop doesn't matter. And since it doesn't matter why the difference in the first place? I would love the committee to answer that one.

Cynic :-) A little more voltage drop won't matter too much for mood lighting. In any case most domestic rings are nowhere near fully loaded for the great majority of the time.

Surely it's obvious that lighting - tungsten lighting at least - is considerably more voltage sensitive than most other loads.

Slightly ironic they think of this now, given the govt sponsored push to try and get people using other forms of lighting ;-)

Yes, but perhaps we should have gone back to the old 2.5% voltage drop limit anyway in the interest of energy efficiency. Except that it might use more copper. What's the embedded energy in a 100 m reel of 2.5 T&E, I wonder...

I seem to think there is a problem in the regs, as if you work out using 2.5mm for a ring main, its just not quite big enough for the MCB protecton, so you will have to use 4mm. Its only very marginal.

The only trouble is i cannot remember where, but I remember on the regs course it being highlighted.

The only times I can think of that being a problem is either with old cables that have 1.0mm^2 CPCs on circuits with rewireable fuses, where spurs may be insufficiently protected, and perhaps on current installs where the supply impedance is *very* low and a fault close to the CU would result in a prospective short circuit current that could exceed the breaking capacity of a typical MCB.

I wonder if the poster was thinking about this...

Jim A

Well, yes, that's old hat now. Nothing's changed in the 17th ed. apart from the RCD requirements. So far as cable current rating is concerned there are two principal requirements to meet:

(i) the as-installed current rating must be at least 20 A everywhere in the circuit, with all relevant derating factors taken into account, and

(ii) under the reasonably foreseeable load conditions the _actual_ as-installed rating at any point should not be exceeded for long periods.

The second of these only comes into play when you have a cluster of heavier current loads (i.e. a kitchen, usually) at one end of the ring.

It's worth noting that some of the tabulated ratings for 2.5 T&E in Table 4D5 (Table 6F in the new OSG) are below 20 A. The two particular cases are:

- ref. method 101 - cable above and in contact with a plasterboard ceiling with >100 mm thermal insulation above (rating 17 A);

- ref. method 103 - cable in a stud wall with cable not touching the inner wall surface (rating 13.5 A).

In these cases 4 mm^2 cable would have to be used unless a different cable installation method could be chosen. 30/32 A rings with 4 mm^2 cable (1.5 mm^2 CPC) are now included in table 7.1 in the new OSG.

Shame that most modern houses would be like that. So now we have been building houses for 30 years where rings are dangerous according to the 17th edition?

What does it say about dropping it behind the plaster board when its stuck to brick work using dabs?

BTW I suspect that the rating assumes the stud wall is full of insulation and that the cable is therefore wrapped in it, not all stud walls have insulation in them but someone could put foam in them for some obscure reason.

There's nothing new in the 17th ed. there, except for the rating being explicitly tabulated. 13.5 A is exactly half of the clipped direct rating and a 50% derating has long applied for cables completely surrounded in thermal insulation. Doubtless there are quite a few non-compliant installations but I'd hope that the great majority have the cabling clipped to the studwork (rating 21 A) or run in oval conduit in contact with the studwork or the wall (rating 20 A).

The closest installation method I can see for that is no. 40 (assuming it's not insulated plasterboard) - multicore cable in a building void. Ref. method B applies, giving a rating of 23 A according to Table 4D2A.

Some contain a form of fibreglass quilt to reduce sound transmission.

If RCBOs are fiited in a CU on all circuits, is there really any real need for an earth wire, apart from backup in case the RCBO fails?

Of course, if there were no earthing, then a live to normally earthed metal casing fault in, say, a toaster would not trip the breaker until you touched it and provided a path to earth!

Yes.