Pedge wrote:


The 1.5mm cable can take upto 18amps and the 1mm 14amps.

Yes, that's the starting point.

Cable capacity is reduced by various factors, though. Since the first
consideration is 'will this cable get too hot carrying its normal load',
the first set of derating factors is to do with heat under normal load.
So, if a cable runs somewhere it's harder to get rid of the (natural,
inevitable, nothing wrong with it) warming, we allow less current
through it: examples of such spaces are running through thermal
insulation - the Regs have detailed tables and call out lots of
different ways a cable might run. Related to this is ambient temperature
(so derate the cable if it runs through an airing cupboard, or round the
back of a built-in oven, as it starts off warmer with no load), and
grouping with other cables (if it's close to other warm cables, they'll
warm each other up and can't be expected to lose as much heat as if they
were well apart.)

The second set of limitations isn't to do with the normal (design) load,
but rather performance of the cable under fault conditions. Here, you're
looking at what happens when there's a short-circuit across
live-to-earth and live-to-neutral. Although when thinking about the
'normal' case you treat cable conductors as having negligible, indeed
zero, resistance, when thinking about performance under these conditions
we examine the small but definitely non-zero cable resistances closely,
to work out (a) how much current will flow, (b) therefore how long it'll
take the fuse or MCB to react to that fault current, (c) how hot the
cable will get during that fault-clearing time. This is where not just
the thickness of the cable conductors but their total *length* comes
into play: the longer the run, the higher the resistance, the lower the
current flowing, BUT therefore the longer the fault takes to clear so
the longer the cable has to heat up. For these (rare, short-lived)
faults we allow the temperature to go quite a lot higher than for normal
loading (e.g. for PVC cables, the limit temp is 160 degrees C for the
fault condition, but 70 degrees for normal loading).

Finally, there's one more limit on cable length, which is working out
whether too much of the supply voltage will be dropped in the cable and
not make it to the load - the conventional limit is to lose no more than
4% of the supply voltage along the way.

All these factors need, in principle, to be taken into account each time
a 'final circuit' - a run from your consumer unit - is designed. This is
why it's often not possible to answer a simple-sounding query - 'will it
be OK to wire my lights/heater/shower/garage in 1/1.5/2.5/4/6/10
mmsq' - with a simple yes/no answer.

In practice, though, there are 'conventional circuits' where the
calculations are done in advance, and these are tabulated in the IEE
On-Site Guide, along with their maximum lengths. And in further
practice, lighting circuits are usually well over-specified - as you've
found yourself, the nominal capacity of 1 and 1.5 mmsq cable is up
around 14 and 18A respectively, given ungrouped, normal-ambient,
not-surrounded-by-insulation, yada yada yada. So in your case, it's
massively unlikely that 1mmsq would be at all wrong for your short
side-runs, even if the main part's done in 1.5mmsq; and the main part
being done in 1.5mmsq is itself unusual enough that it might be just
what whoever was wiring it had to hand, or becuase for a part of its run
it's surrounded by thermal insulation which won't be the case on the
bits you're adding in...

HTH - Stefek