I think the 50v or less is for normal dry conditions, special locations need a bit more protection.
It's obviously essentially arbitrary, but one has to pick some number.
Even with wet skin etc., <50V will probably avoid death in most people, but not all. In fact, no figure would be guaranteed to avoid death in
all people - there are plenty of people out there who (often without knowing it) have hearts that are so electrically unstable that there are constantly at risk of spontaneously developing lethal rhythm disturbances i.e. even with no electric shock at all) - so, in them, a shock resulting in less than 1mA though their body, for a very short time, could sometimes be the 'final straw' that resulted in their death.
However, if we required touch voltages a lot less than 50V, it could result in practical problems. Even with SB, the touch voltages are determined by the length (and CSA) of the SB conductors, so we could end up with prohibition of installing things with exposed-c-ps "further than X cm from a pipe to which had to be bonded" (and/or a requirement to use massive bonding conductors)
All in all, I suspect that 50 V is not an 'unreasonable' figure to go with - but the precise number chosen is obviously essentially arbitrary.
Has anyone ever been touching a 50v bog standard phone line whilst ringing signal is active? It does bite even in normal dry conditions!
Indeed so
However, although, as above, it theoretically could kill, I wonder if that has ever happened (in the last century or whatever)?
.... I became a fan of supp bonding and remained so even when RCDs were mandated on most circuits. Even the speed of working RCDs could still leave one uncomfortable. Then factor in the possible failure rate of RCDs ....
See what I've just written to secure. I agree that if one is not happy to trust an RCD to always clear a fault fairly quickly, then local SB is the only way of preventing high touch voltages.
(not just complete failure to actually disconnect at all but the failure to do so within the required time.
In terms of 'real' L-E faults in a circuit (rather than just an L-E path trough a victim), an RCD does not really have to perform 'on spec' to achieve what one wants (clearance of the fault before anyone has a chance to get a shock). Even if it took 10 or more times the In to make it triop, or even if it took 10 or more times longer to trip, it would still serve that purpose, and only 'complete failure to trip' would really be an issue.
As I see it, the only situation in which 'out-of-spec' could be a problem is if the only residual current was that flowing through a victim - and, as you know, I do wonder how often RCDs actually do trip as a result of that.
Putting aside the age old game of "catch this capacitor" I have only had a mains shock once at about 12 years old and a BT line zap once in adult life.
Both incidents taught me to avoid such stings.
Ignoring 'tingles', it's also literally many decades since I last had a significant shock of any sort, but my experience was a bit worse than yours during my (probably foolish- or, at least, 'less cautious') youth ! Back then I had several mains voltage shocks and at least some that would today be classed as "HV" ones (>1,500 V DC) - but, like you I soon learned the desirability of behaving in a manner which avoided such events!