OK, what about this then: When my flat was being wired up the electrician ran earth wires to the enamel bath, the stainless steel kitchen sink, the hob, and to the central heating pipes. He also fitted special earthing bonds to a cluster of pipes in the cylinder cupboard, and ran another earth wire to the bonds. They have a tag attached: "Safety electrical earth - do not remove". He explained that this was to ensure that none of all this accessible metalwork could become 'live', and would always remain at earth potential. Thus a person (me) would never become an electical path: 'pipes-person-ground'.
As heating elements age, the insulating material which is packed around the wire inside begins to deteriorate, and becomes more and more electrically conductive when it is hot. This is what is happening when the cooker starts blowing the breaker at shorter and shorter intervals after you turn it on. The stray juice is shunted to earth via the earth wire and the system senses this and trips out. Without the earth wire it would shunt to earth through you.
With plumbing the risk is more from things like badly-routed cables chafing against pipes from thermal movement until the insulation is worn through, or from the immersion heater, which can either develop insulation problems the same way as cooker elements do, or in some cases can corrode through and expose the wire inside directly to the water.
But on our tramway, the rails (and thus the tram) are INSULATED from the earth. A dead short fault, say, in the innerds of the tram (that's ANY tram on the same section) would bring tram body, rails and 'other tram' up to overhead wire voltage (about 650 isn't it?). I shall 'Mind the Gap' with some care in future.
That would only happen if
all the returns were somehow broken somewhere (which would also cause the tram to come to a stand). I don't know if real systems do this but if I was designing a tram system I would include something to continuously monitor the integrity of the returns and cut off the supply in case of a fault for exactly this reason.
650V is nasty but not necessarily fatal, whereas 25kV is pretty much guaranteed to kill
It is said that third rail systems are in some ways the most dangerous. They make you grab on where a higher voltage would cause a more violent spasm and throw you off. Also DC is more apt to promote grabbing-on than AC.
Three rail power has operated under the floating rail system for many decades in the south of England and I'm not aware of any incidents involving touch potential.
I believe that the running rails can get up to 100V or so in worst-case conditions (maximum number of trains, all drawing power, with one of the feeds out, and at maximum distance from a feed). If a train is stopped, of course, it is not worst-case
I have recollections of cases on 25kV stock where crew were shocked to varying degrees whilst climbing aboard units that had been left with pan up on rusty rails in sidings. I don't know how commonplace such incidents are.
The inclusion of "rusty rails" makes me think that that is due to rust partially insulating the wheels plus rain and dirt creating leakage across the insulators on the roof, rather than stray ground currents.
'Course with 25kV AC we encounter another phenomenon: stray voltages cropping up on things due to reactive phenomena. Things To Do When Really Bored On Bescot Station: drag the back of your hand along the wire fence at the back of the platform and give yourself a tingle.
Whilst in Manchester recently (17th August) I had a look at the progress of the "Second City Crossing". Then, I made my way up Moseley Street where I saw that the now single tram track reverted to double by what I took to be a temporary set of points (they were on timbers on a ballast bed). One detail took my attention though, the overhead had been cut, and the loose tail looped down to rail level and there connected to one of the running rails. Safety, I assumed - all the residual (and new, of course) overhead down in St.Peters Sq. has to be dead whilst the ground works proceed. But what if there is a fault? A gap in the overhead bridged by a stray pantograph, say, or a metal pole - that sort of thing. Then, the 'insulated' track would become live with respect to the surrounding earth, would it not? Not being awkward - just something I saw, mused upon, still thinking . . .
That is why live sections have a short neutral section between them so the gap cannot be bridged by a pantograph. As well as the danger you describe that would also create untoward circumstances in normal running as the two sections would not necessarily be at the same voltage or phase and the result might be great big bangs and associated damage.