Martin Angove wrote:
Given the opportunity (and let's ignore backwards compatibility, at
least to begin with), how would you redesign our domestic final circuits
to be completely idiot proof?

I don't think that's possible, or desirable. You might better ask how we
can redesign motor cars to be idiot proof, as they cause far far more
deaths and serious accidents than electrical wiring (or even appliances)

* a ring circuit with a break in one of the live conductors will
overload one leg as it is fused at 30/32A and the cable is only good
for 20A (27A under ideal circumstances)

A current-balance MCB that trips the ring if the imbalance between the
two legs exceeds a certain proportion of the total is probably feasible,
and could probably be combined into a single-circuit RCBO.

Ring overloading and high point load really only occurs in kitchens,
domestically. Treating all hob/ovens, washing machines, tumble driers
and dishwashers as fixed heating appliances and mandating points for
these to be provided with each point rated at 10A or actual load if
greater (in the same way we rate lighting points at 100W) would in
almost all cases provide at least 4 sockets (they could be on their own
32A ring, or radial, or multiple 15A radials) in the kitchen, and take
these appliances off the general purpose ring. The gp ring should be
able to cope with other loads, deep fryer, combi microwave, as these are
fairly intermittent anyway.

All those appliances are plumbed in or otherwise non-portable anyway, so
there is no disadvantage to restricting their use, unlike a
toastie-maker or microwave whcih might be used in a teenager's room.
Such provision is also more easily enforced because these appliances are
usually *fitted* by tradesmen and in kitchens comes under (or can be
brought under) Part P for the fixed wiring.

* unfused appliance plugs (less problem with tampering) would mandate
radials with a very low MCB rating and hence a large number of
radials; one per room or even more.
* unfused appliance plugs will *always* take out the circuit MCB on a
fault, and hence tracing the appliance which caused the fault can be
tricky since there are likely to be several candidates.

Yesm that is a step back to the 'good old days'

... Sockets with built-in MCBs, or RCBOs?

Is an alternative, but the difficulty with removing the plug fuse is
backwards compatbility, which cannot be ignored. We can't have new
unfused appliances being plugged into old high-current rings, and any
use of adapters or requiring householders to change plugs is going to
send the death rate shooting up.

Talking of RCDs, what about changing the way we use these things?

Like, using them correctly instead of whole-house?

Owain