Bonding things together brings the potential down, especially to a level much lower than what contact across a person would lower the voltage. This will not make the stray currents go away, it just gives them a lower impedance path than through a person to flow through. Exactly what we do when bonding everything imaginable when constructing a swimming pool. We are not eliminating any stray currents we are just reducing points of potential within a specific area. Any EMF's associated with that current are still there, and possibly even increase to some extent since we made a lower impedance path which would allow more current to flow in that path.
Thanks for adding the detailed explanation for those who are following this thread.
My concern was exactly what you stated above, and to demonstrate the bold part, here is a numeric example:
Since this is in the close vicinity of a substation, it is possible that the earth voltage gradient at the soil surface might even exceed 120 volts. But let's pick 100 volts between points A and B somewhere under the house. Since we are talking for the moment about buried plumbing, I will assume that pipes at point A and point B each have a resistance to local earth of only 5 ohms. A perfect bonding wire between the two points would then be carrying a current of 10 amps.
If the bonding wire and its associated connections, terminations, etc. had instead a combined resistance of only .3 ohms, that would be enough to cause the measured voltage of 3 volts, and that 3 volts would be coming from a very low impedance source, not current limited in the way capacitive stray voltages would be.
The resistance of dry skin would be high enough that 3 volts would probably not be noticeable, let alone harmful. But with not just wet but saturated skin at both contact points, the current produced by a 3 volt potential with low source impedance could be quite harmful.
What this demonstrates is that when large stray voltages, specifically voltages caused by earth potential gradients, are involved, simple bonding will reduce the risk, but cannot eliminate it completely. Even an equipotential grid would have to be carefully constructed using large conductors to be effective.
A ground ring around the boundary of the property would not necessarily help if it was shallow enough that the sphere of influence of the ground ring was small compared to the width of the property on the line toward the substation.
My first impression when this story came up was "Obviously somebody screwed up on the wiring and bonding of the house, or there would not be a problem." More detailed consideration, like the above, has led me to a different conclusion.
If I had to try to remediate the problem and could not get POCO to fix their earth gradient problem, I would be inclined to make sure that any plumbing exposed inside the residence was isolated by a dielectric coupling or plastic pipe from all pipes in contact with the earth and then bonded to a single point. That way there would be no currents flowing in the bonding network (barring shorts in the house wiring) and near zero EMF could actually be established.
I am led to believe that the "unusual" raised tub or shower present when the woman moved in was in fact deliberately built that way to allow room for non-conductive drain pipe to be used between the tub and the buried piping. The plumber who put in the new shower followed conventional good practices in his work, blissfully unaware of the "underlying" problem.
