I think I should reply to my admirers.

1) Go out and read the 16th edition regulations IN DETAIL before committing
finger to keyboard. Together with the background clauses in the IEE
discussions.

2) Just to give you a clue, 2.5sqmm cable in free air WITH CONSTANT LOADING
can be adequately protected by a 30A fuse in a radial circuit.

3) The IEE regulations are a code of practice and not a Bible. The IEE
specifically states that the code is a guide.

4) Changing wire gauge is not bad practice. Sorry, but that is the real
world.

5) This application is an intermittent loading. Thus cable ratings change.

6) Both of the admirers I noticed had failed to look at the application in
sufficient detail to spot the really weak points in the installation.

a) If solid cable is used to connect to a vibrating pump( shower pumps
normally are on anti-vibration mountings), it shears! The pump should be
connected with stranded wire.
b) The fusing problem is not in the CU but locally. The pump should be
protected with a local 5A fuse.

7) If you ask what would I have done:

a2) Left the fuse @30A
b2) Split the 6mm feed in the loft into a 13A socket( for local convenience)
and a fused 13A spur using the local 2.5mm2 cable, both mounted in a surface
box.
c2) Terminate the 2.5mm2 adjacent to the pump in a 13A socket.
d2) Fit a 13A plug flexible lead to the pump fused @5A.

The inherent problem in the setup was the ability of the source to over
supply energy to the pump under failure conditions. A 15/16A CU fuse helps
in this respect but is not ideal, hence the preference for a local fuse @ 5A

Hope this explains the reasoning behind the statements.

Regards
Capitol