mitchd wrote:

Why the earth rod for shed? would you like to quote a reg to justify it
please.
My preferred way would be 5 way consumer unit in garage rcd sockets
where nessary, no rcd protection to server (to prevent nusicence
tripping) agree with all mcb/cable sizes but would put small 2way
"garage consumer units" with rcd main switch in shed and summer house

Not an engineer etc...

I doubt there is a specific reg that says "you must do TT etc". It's going
to be a matter of making the design compliant with several relevant regs.

If you are going to export the house earth then it must be safe to do so.

See here first for some excellent background info:

http://www.iee.org/Publish/WireRegs/...s_answered.pdf

If you have a TN-S (elec co supply earth as a separate core back to
substation) then many would agree that exporting to outbuildings is OK
over short distances, for some definition of short that I'm not prepared to
define...

PME supplies are dodgey as it's not impossible for the earth/neutral to fall
off upstream (esp. in rural locations) then your neutral AND your local
earth is now being pulled up towards phase by attached loads (yours or
neighbours'). In the case of a metal greenhouse with supply and earth
bonding, your frame is now at 240V and the ground is still earthed - so you
get the idea. And you don't bank entirely on the main RCD to get you out of
trouble...

TT - you have your own electrode with a relatively high impedance. Phase to
earth fault in the house will float this up (and thus your greenhouse frame
in the previous example) to potentially dangerous voltages relative to the
real earth. Wet grass, bare feet leaning on the greenhouse you would not
need full mains voltages to potentially kill you.

--

If you have doubts as to exporting the supply earth, then the only sensible
option left is a local TT earth at each location (shed/greenhouse/garage
etc). This must be a proper TT installation, required RCD and MUST be
tested to prove that it has suffiently low impedance. Potential for messing
it up exists, so the testing is exceedingly important (Whifield's
exposition on the wiring regs does detail a method involving multimeters
and a car battery, but may as well hire a suitable tester).

With the greenhouse example, a phase-earth fault in the house has no effect
providing you isolate the two earthing systems, and a local fault is
hopefully going to achieve a lower differential voltage betwen the ground
and the frame (because the earth rod is in the vincinity). It's deemed
acceptable because there's not a lot else you could do to better it.

The whole thing is a bit of a minefield and it's far less easy to just give
a prescription answer, compared to many other aspects of domestic wiring.

All of the above is an attempt to offer general explanation to your query,
do not take it as a recipe to implement a specific solution!!!

Cheers

Tim