In uk.d-i-y, Andy Wade wrote:

All quite correct, except for one thing no-one's mentioned yet: ambient
temperature correction factors (see Table 6A1 in your On-Site Guide). The
27A 'clipped direct' rating of 2.5mm^2 T&E only applies up to 30 degC, at 40
degrees the derating factor is 0.87, at 50 degrees 0.71. So in this
situation 2.5mm^2 cable is only suitable if the temperature won't exceed
about 40 degrees, and it's quite possible that that could be exceeded where
the cable approaches and enters the appliance.

So with 2.5mm^2 PVC cable Stefek's design is likely to be marginal or
non-compliant. With a 2.5mm^2 heat resistant flexible (3093Y etc.) you'd
probably be OK, provided you've got a bootlace ferrule crimper for making
the ends off in the wiring accessories.

Thanks for a calm exposition of just how and why the design is marginal.
"Marginal" is just right - i.e. (as I claimed) it's the sort of just-compliant
watch-the-pennies just-about-good-enough no-headroom apporach used in
practice on many mass-market builds; for one's own use at home 4mmsq would be
sounder, but the 2.5mmsq may be justifiable. Though I made passing mention of
the higher temperature of the oven/hob environment, I didn't make explicit the
necessary derating of the cable. Such higher ambient temp is particularly
likely to occur in the space behind the "built-in" oven, where there might
well be a metre or so of wiring slack of your (carefully, gently folded ;-)
final T&E. And I'd also briefly thought to mention the use of a heat-resistant
2.5mmsq flexible, and rejected it on the grounds of the poster being (a)
unlikely to have a bootlace (or ring) crimper to hand, to make the connection
to the terminations reliably, and (b) 2.5mmsq silicone-insulated or similar
flex being harder to source than T&E - the OP might've ended up with 1.5mmsq
"immersion heater" cable from some Saturday-morning trade-counter "sod the
weekend cowboys" type, which would be well under reasonable current-carrying
capacity...

NB (for the pedantic) your statements "nature of the load prevents
[overload]" and "diversity allows [...]" don't sit comfortably together in
the same paragraph without further explanation. Diversity is being applied,
quite validly, to the 'feeder' circuit from the MCB to the point where the
circuit splits after the cooker control switch. Since diversity has been
applied, overload, relative to the design current (Ib), _can_ occur -- but
that's OK since the 32A MCB will provide overload protection for the 6mm^2
cable. After the split point, if we drop down to 2.5mm^2, there's no
overload protection, only s/c fault protection provided by the MCB. Hence
no diversity can be applied after the split and the cables have to be rated
for the FLC of the appliances -- which of course is what we've done.

Again, thanks for the clarification. With the increased popularity of
split hob/cooker arrangements, will (or have) wiring-accessory mfrs
brought out a "dual-control" cooker control switch, with two 45A or
32-A-rated DP switches in the one (probably double-width) accessory?
Obviously you can mount two 45A control switches in a dual box, but
it's more of a pain wiring them up, with the heavier incomer needing to
go both in and out of the first one.

(Oh, and maybe the IEE will issue a guidance note on wiring split hob-n-ovens
from one existing 45A cooker feed, discussing the limits on going down to
smaller cable sizes for the post-control-switch portion, in as concise and
authoritative terms as Andy managed in the text quoted above? ;-)

Stefek