Those will be harder to find and will probably have a higher price because they are not common. Besides a three phase unit with one core will likely cost less.
New 3∅, yes. But I'm finding good selection/pricing on 480->240 1∅, NOS, etc. We're willing to risk non shiny-new-in-box iron.
2 HP should only need around 3 kVA total. I would think 3 - 1 kVA units would work, but you seem to be against using delta configuration. If only using 2 transformers they will need sized larger - Not quite sure how much larger the 2 kVA units may work.
The AIM Table 4 tells me I need 4 KVA total; two @ 2KVA, or three @ 1.5 KVA. But I'm finding 2KVA units are out there; I'll get 3. I'd like to have tapped (not center-tapped) but we'll see.
If it were me I would have to make a decision between either single unit transformer or open delta made from two separate 1 phase units, OR pull pump and replace with correct voltage. All are going to be in about same price range.
My back of the envelope makes three single-phase 2KVA units to be far cheaper; the difference from two to three is not a killer.
I'm OK with corner-ground delta given your reassurances. As I said, my reluctance is just one of no experience with corner-grounded systems. First glance says the existing motor controller will work fine.
Here is a code question. The 240->480 transformers will be in a building ~100 ft away from the wellhead panel. Will we be required to have overcurrent protection on/at the xfmr secondaries, or only the 240V primaries?
I do like the derive 480 volts single phase and use a VFD. The VFD likely requires repair or replacement much more often than a transformer so that makes it a decision based on longer time period. Transformers could last well beyond 30 years if not undersized.
This to me looks less desirable. While it avoids the phase-imbalance issue, it may create RFI and will put unequal-over-a-cycle load on the house grid. Not sure how the solar grid tie inverters and other electronics will like that. [I can relate a major tale of woe re: that under another Subject...]
But by then they may be obsolete who knows? We might have something better for power transformation.
I keep wondering if/when we will see switcher supplies as pole-pigs & such. I recall reading that several percent of the total grid is wasted in distribution transformer losses. From what we've seen in laptop supplies/wall-warts, if you make enough switchers, the price falls dramatically. Maybe someday soon, the TCO lines will cross....
ps: found
this PGE note on power imbalance.