RzB wrote:

OK - Should be OK now.... !!!

Yep. My comments on your calcs a

1. Current rating: You've used It = Ib/(Ca*Ci). Since overload
protection is required here this should be In/(Ca*Ci) - i.e. based on
the device rating, not the design current. Moreover in your case the Ca
factor (for loft temperature) and Ci factor don't apply simultaneously
(the cable thro' the wall won't be at 40 deg. and the cable in the loft
will be well-supported away from thermal insulation). With Ib = 41.3 A
and your intention of using the Wylex 108, we can select the 45 A BS
1361 fuse, hence It = 45 A / 0.81 = 56 A (10 mm^2 still OK). Or with a
60 A fuse we get It = 74 A (16 mm^2 required).

2. Shock protection: your calcs will be on the conservative side because
you've assumed 70 deg. conductor temperature (the 1.2 factor). In fact
the greatest part of the cable's length will be running well under its
rated current, so the loop resistance to be added to Ze is
correspondingly less.

3. CPC thermal compliance and disconnection time, with 6 mm^2 CPC and 50
A Type C MCB: fault current at 479 A[*] is less than 10*In, therefore 5
seconds disconnection time is not necessarily achieved. Design is
non-compliant, go no further (or retry with larger CPC). For the Type B
yes, it's OK to use 0.1 s (compliant), ideally use manufacturer's
published I^2*t data, which can be more favourable.
[*] The fault current will fall as the cable heats up, so you have to
err on the pessimistic side here when anything is marginal. Design
based on the use of measured values of Ze is discouraged, since this
value is outside your control. In principle the DNO could alter their
network, resulting in a higher value.

Incidentally, are there any metal service pipes or other
extraneous-conductive-parts entering the workshop? If so, you may need
local bonding to the earth bar in your dis-board - see Regulation
413-02-13 (a challenging read).

--
Andy