| No, I don't share your motivation and may we may share a lack of a solution. | | I believe that the problem in the dairy industry was the so called stray | currents. Considering that such currents, of a higher magnitude, can be | produced from the LV system, additional, smaller currents from the MV system | may not be the problem. I once read a book which dealt with this and this | contained a detailed technical analysis which brought the concept of ground | currents from MV systems into question. A greater problem was poorly | maintained LV wiring. However technical opinions and legal opinions are two | different things.
Given a constant resistance, MV is going to push current harder than LV. Certainly in many cases it can be an LV issue. When it is, it is easier to deal with and get fixed. But when it is an MV issue, then it is a lot harder to get fixed. Perhaps more initial incidents might be from an LV based cause. But those caused by MV persist longer because the utilities tend to not want to take the steps to fix it.
| If the MV system is grounded properly, there will be unbalance ground | currents-true. A separately grounded LV system, including one using your | isolation scheme will have much higher ground currents and, since the earth | is a common conductor, there will still be coupling between MV and LV ground | systems. Rather than isolate or try to do so, it might be better to spend | money on ground rods- even a ground grid and tie everything to this grid.
This depends on where the return point on the LV system is. If the system is derived at a pole transformer, which will be grounded, then there will be two return paths, the neutral of the service drop and the ground itself by means of neutral current flowing over the neutral to ground bond, to the entrance based grounding electrodes, and through ground to the pole grounding electrode, and up the pole grounding wire to the transformer.
The problem is the system has multiple points of grounding the neutral. The NEC doesn't allow this and for good reason. The NESC forces it on the service drop.
The fix is to derive an entirely new system with a transformer where the primary is connected to the two line wires at 240 volts for single phase or delta for at 208 or 480 volts for three phase. The neutral on this system will have exactly one bonding point (unlike the service drop system that has at least two). As long as there are no faults in that system, then any neutral currents can only flow back to the source by means of the neutral conductor. Even if the service drop neutral/ground wire is connected to the secondary bonding point (e.g. all grounds/neutrals are connected together), it introduces no additional paths on the secondary system because the secondary neutral has exactly one point for it to reach ground (even though after that point is then can branch out to get to ground by the electrodes at the building and through the service drop neutral to the pole grounding electrode).
However, this doesn't prevent any MV current from coming down the neutral and flowing through the building grounding electrodes, unless the primary is fully isolated (not even ground connected) from the secondary.
| Note that in HV substations a grid is used in order to make the ground as | nea an equipotential as possible and the fence is either outside the grid | far enough to have a low step potential in case of a fault or inside by a | distance such that anyone or anything outside and touching the fence is well | within the grid. This is a bit extreme but the same idea of an equipotential | grid to which all neutrals, enclosures, etc are connected may be a better | alternative for locations considered critical. This will mean that the | ground under a cow is at the same potential as the milking machine hardware | attached to the udder. Your isolation scheme doesn't do this.
I don't know if they do this, or not, but the milking machine could be made to work where there is absolutely no metal contact to milk where the milk flow is continuous back to the udder. This isn't hard to do, so I want to believe they have done this.
Still, this doesn't address the fact that these problems exist outside of the scope of milking activity. Apparently some step potential exists near any point of grounding electrode that has a metallic path back to an MV circuit. Apparently livestock can sense it, at least when standing or walking bare-hoof on at least wet ground. MV neutral currents flow out each grounding electrode and radiate outward from there.
If I were a dairy farmer, I would at least isolate the milking operation, which is where the livestock would be near grounding electrodes and when their sensitivity would have a significant impact. I don't know offhand what kind of power they need. But a plastic natural gas pipe, or a propane tank, connected to a generator, might do the job. The area might still need a ground ring to divert any exterior ground currents passing by. Or maybe I could "go green" and power everything from solar power sources. Or maybe from indigenous methane :-)
Adding a transformer at the end of a service drop I do think will improve things (because of the singular bonding). It's not all I want, but it is more than most have.