In this particular case, JP reported that the c.p.c. had been blown away by the fault - suggesting that the screw initially hit L + c.p.c. - so even the simplest of insulation test should have spotted that.

It could very easily have been different though - as you suggest I doubt an insulation test would spot a screw into L alone, and the outcome could have been identical.

Similarly the Danny Edwards case where L & c.p.c. had been reversed at the socket - had the socket been on a radial, the normal "R1+R2" test as frequently taught wouldn't have spotted the problem before energization.

The courts and public have this quaint idea that testing will find any fault - while the real world is often a little less perfect (just ask any software developer).

I think we might be able to do better though. In the case of screws through hidden cables, it seems to me that the 'double insulation' approach is flawed as it clearly isn't robust enough for service conditions (people will put nails & screws into walls where they want, regardless of any 'safe zones' we might dream up). So the obvious alternative is real ADS - i.e. put earthed metal between the live conductors and any potential victims. MICC, SWA or steel conduit - or my current favourite at the moment - BS 8436 cables. I see it as catching the problem as it happens, not hoping to catch it at some later date when some poor victim completes the circuit to earth and hoping the RCD has been exercised recently. At the very least BS 8436 cable makes the insulation test viable again.

- Andy.

Indeed it would.
I should have re-read JP's post.

If we know that test results are often made up and that even when they're done properly not all faults show up then I suppose what we install and how we install it has change to allow for that.

Another thought on this:
(I'm assuming the cable was flat twin & earth.)

In a situation like this wouldn't it have been better for both conductors to be blown away rather than the one being able to withstand the fault long enough to allow the other to be destroyed?

Would the outcome have been different if the cpc had the same csa as the live and was insulated in the same way?

All good points Mike,

What we've learned here is that inspection has fallen lower in the heirarchy than testing - there is a reason for the order of the words and I've always favoured inspection supported by testing at the correcty way to do things - rather than approach that makes inspection subordinate to testing.

We've also learned that proper installation methods and correct materials ie the selection and erection of systems is important.

As our IT chums like to say about "quality"

You can design it in
You can can install it in
But you can't test it in

So I would say that we need a focus on preventing the defects rather than quality testing in reality.

It's highly likely that full sized insulated CPC's would be an asset, however, in this case, the damage by the screw could easily remove "material" and result in unequal condictor sizes anyway with the same outcome.

As Andy mentioned above we need an earthed screen around live cores to ensure earth faults cause disconnection before live parts contact conductive unearthed components - we need more attention by installers to exactly what they do - I've seen enough "hanging gardens of Babylon" on new builds to say that the probability of this incident must be resonably high - I've also seen enough installations in metal frame systems to know that with just a little care and attention, coupled with appropriate programming of trades on site, can virtually eliminate the problem of screws through cables. Even a bit of loose plastic conduit can cure this issue of trapped cables.

Regards

OMS

-------------------------
Let the wind blow you, across a big floor.