Sorry for the late reply to your transformer Q.
First off, the load per line [X1, X2, or X3] to common [X0] will have a maximum of 2 KVA on this 6 KVA Wye system. Right now, it looks like your less than 1 KVA with the 7 amp load on X1.
Got a few questions to throw your way,
- Is this 7 amp load Linear? - Did you check with a multi Hz / true RMS ammeter??.
- Was this transformer's connections altered in any way??
- When you mention the "TEE" connected transformer, is this a reference to an Open Delta TEE, or Scott connection??
- Is there any excessive current flowing through the grounding electrode conductor from the X0 termination?? - excess of 0.25 amps?
Circulating Harmonic currents might be causing an excessive heat in either the Secondary winding, or the Primary winding[s] - possibly both, depending on the source which is reflecting the Harmonic noise.
Other than this, incorrect polarities would be a cause for excessive heat. Another would be poor ventilation, or just a bad winding / insulation failure [short between winding segments, or to the core].
As you know, transformers will normally be hot - even at idle [due simply to losses consuming true power (wattage) and result in heat output]. When they are abnormally hot [maybe followed by letting some smoke out
..], then there is definitely a problem!
There should not be a real problem with this system, as it's a 4 wire Wye secondary. It should handle upto 2 KVA L-N without too much of a power quality issue, since each secondary is equal to a simple 2 wire circuit.
The primary coils will react to the change when a secondary coil, or coils, "drain" current - which will allow the primary to draw in more KVA to match the requirements on the secondary. The problem on a dual voltage system - such as this 208Y/120 4 wire Wye is the primary side will have protection set for "X" KVA @ "X" voltage. If using one coil [lower of the two voltages], it could easilly be loaded up higher than it's real maximum rating, but never trip the primary OCPD. Having fuses on the secondary as you described is good.
Any particular reason why the load wasn't distributed, instead of all on one line?
P.S. Thank you for the suggestions regarding schematics terms!!!