In message , at 21:34:48 on Tue, 21 Sep
2010, Roy Badami remarked:
Also the Plaistow signalling irregularity, which was quickly found to
have been caused by 'a defective component'.

ITYM 'an incorrect component'

Thanks - sorry about that - probably thinking of the previous
discussions as I wrote it...

In many ways an incorrect component is rather less disturbing than a
defective component, as the latter would indicate a wrong side failure.

I'm not sure why an incorrect component (which would seem to have
performed, or rather failed to perform) in the same way as a defective
one) is any better.

Obviously incorrect installation or maintenance is still a concern,
but at least we now have a better idea as to the underlying cause.

I'd say that fitting an incorrect component is much worse, because the
incorrect component had to be procured and fitted, and pass
commissioning checks and subsequent maintenance checks. Some of these
could be 'excused' if the documentation clearly specified the incorrect
component, but that in itself would be a worry.

Also, in that context, the comment about checking that similar
components are not present at other sites makes more sense.

Checking other sites for an incorrect component is easier than checking
for [potentially] defective correct components, I agree.
--
Roland Perry