Wouldn't the required bonding in the NEC prevent this type of situation? If the neutral was bad at the transformer, then the electrician would be able to isolated this I would think.
You think GFCI's are going to be required everywhere now just like AFCI's ?
I doubt that either a GFCi or an AFCI would make any difference at all in this case. Proper bonding certainly might have prevented it. But only if the drain pipes were bonded as well as the plumbing.
And if the grounds were made to separate ground electrodes without proper bonding conductors between the ground rods, it might not have helped at all.
The whole allegation of "stray voltage" is a catch all for anything that has not been investigated well enough to determine the underlying cause.
Some possibilities in no particular order, and taking into account that the proximity of the substation might have something to do with it.
1. The shower piping was not grounded and had a voltage induced in it either capacitively or magnetically.
2. There was an AC wiring or appliance problem in the house which resulted in leakage to unbounded water pipes.
3. The shower pipe was grounded but the drain pipe was not, with either possibility 1 or 2 only the drain pipe.
4. Enormous ground currents related to the substation caused a voltage gradient near the surface which caused two grounds electrodes to be at different potentials.
5. A faulty service neutral bond could have caused the service neutral to deviate so far from ground that the local ground electrode bond could not pull it all the way to earth ground. This would leave a potential difference between metal which is grounded by earth contact but no bonded and metal which is bonded to the elevated ground electrode potential.
Any of these should be detectable, quantifiable and repairable.
I am sure other members will come up with other possible explanations, as well as opinions on the competence of the workers who tried to fix this.