You don't need a floating neutral on the high side when the secondary is a wye voltage. Essentially in a 120/208 3 phase bank each can is hooked up individually just like a single phase can, the 3 phase voltage is taken from leads out of 2 cans . Of course on a single phase can the primary neutral is grounded. The fella that made the drawing must have been confused by the cans having 2 insulated primary bushings. Now if you are in an area that used to have delta primary, then maybe that is an old spec that has never been updated. There is no advantage to floating the primary neutrals on a 120/208 bank , in fact the specs your using cost more to build due to unnecessary equipment.
The only time I have ever built a 3 phase bank with a floating neutral was by the rule of thumb, "wye primary/delta secondary, float the neutral". The way it was explained to me,( I haven't researched the theory) is the floating neutral helps balance the primary and could ,in theory be at primary voltage. That is why we always insulated the floating neutral and had to use cans with 2 primary bushings.
Wye primary has a grounded primary side connection
If you see "floating bus" that is a Wye Delta bank
The "ferro switch" is nothing but a ground switch....you could do the same with a jumper wire.
Ferro Resonance is a situation were an inductive circuit offers capacitance and inductance that cancel each other...result--- the transformer windings (in series) don' offer enough resistance to current and that allows the voltage to rise until the windings are all energized. When you close in the first switch the other two transformers are feed through the series connections to the first transformer... close number 2 and it multiplies higher and we see the "show"
Two ways to manage ferro resonance.... limit the capacitance or increase the impedance
Plain talk... add some load on the secondary's or drain the capacitance with the "ferro switch".... oh... third choice--- feed the bank from some gang switch that closes all 3 phases at the same time.
Only explanation I can come up with is some hotshot young engineer. You know the kind, that can't even figure out semi-strain guying. Reminds me of a trouble call I was on when I was 4th year. Customer was renovating his bathroom and had installed GFI plug that would not stay on. After amp checking his service, we got permission to isolate it , then I found 5 amps on the service neutral! Happened to be 3 or 4 spans from a 44kv substation dip. Well I concluded that there had to have been a leaking 44kv underground cable. The utilities engineer came up with some bafflegab explanation nobody could understand. About a year later one of those same station cables blew up.
Question to this specific situation... is the bank at the end of a lightly loaded line section?
Some lightly loaded phase sections will produce ferro resonant symptoms and cause mystery fuse opening and other weird stuff.
On a WYE DELTA... most standards require that floating H2 buss to be "ungrounded". If the feed falls down or one TX opens up some how... the 3 pots will operated as an open delta.... until it fails from over load.
If a feeding phase fails.... any single phase TX will operate properly between the break and the bank.... If the Primary is not bonded on that closed delta bank is bonded to ground... that wire is hot at full phase voltage and none of those single phase customers will call in .... if it's floating- that voltage will be about one half and those singe phase pots should produce enough trouble to get a trouble man out ... if the 3 phase customers did not call already.
Open Delta (2pot) bank feed with Y get primary grounded on purpose, if you can find a set of diagrams for the banks...and think this through it makes sense.