I am going to try to give all the important info first. The jobsite is old factory building with a 480v ungrounded messenger supported overhead feeder assy consisting of 3 runs of 750 mcm copper travelling about 450 foot into a seperately tapped 200 amp 480v fused disconnect. the fuses are 200 amp 600v. Then, there is a 25' 100 amp tapped feeder going into a 100 amp rated fused 600v 3 phase disconnect. At this point, with no load, voltages are a-b 480, b-c 473, a-c 473. The 100 amp primary disconnect is fused at 90 amps and draws, with full load, 23.5 amps, 32.5 amps, and 29 amps 480v. The primary conductors then goes into a 45 kva 480/208y transformer with 1/0 secondary conductors to 200 amp 208 disconnect fused at 125 amp. The secondary volt readings(208 nominal) with no load are 208, 205, and 203. With 60 amp load the voltages decrease to 206, 203, and 199. My thinking was the loads are too much for the voltage drop but fixing the voltage drop by running new conductors back to the front panelboard service source was not economically feasible. My plan was to install a 30 kva delta delta mid point grounded 30 kva transformer to power all the 240v single phase loads off of to decrease the significance of the voltage drop for the 240v lights and a/c loads. This would then help the 45kva panel also due to significant decrease of loads.
The strange occurance happened at some point yesterday when all that had been completed is increasing the load side 45 kva transformer conductor to 1/0 copper and changing out the overheated initial 100 amp 240v fusible disconnect to a 200 amp 240v disconnect and increased the size fuses from 90 amp to 125 amp. Upon receiving a frantic phone call this morning stating that there was no power in his unit, both myself and my mentor went down to discover that phase 3 of the 480v primary disconnect had a blown fuse which was producing no power on leg 3. I examined the newly installed primary disconnect and saw no obvious signs of overheating or corrosion. Upon opening the 45 kva transformer, all connections were tight with no discoloration anywhere in the transformer terminal bar. There were no obvious signs of why the previously never blown primary 90 amp fuse had opened.
After 2 hours of continuous running, we could not replicate the failure and planned on continuing the initial plan of installing the new 30 kva delta delta transformer, Has anyone ever had a similar situation causing a primary overload device to fail with no obvious signs of any problems in the secondary side. In addition to turning off the secondary disconnect for testing, I removed all conductors from their respective terminals to verify the readings I was getting at the transformer termination lugs were the same that were at their respective terminals with no change.

OK, between all the words and details above, did you change BOTH the primary (480) fused disco, AND the secondary fused disco (208) ??

What was the physical condition of the 'blown' C-Phase 600 volt fuse? Was it a new fuse, or did you re-install the old fuses, IF you touched the primary fuse.

Did anyone possibly turn on the transformer and inrush blew the primary fuse?

BTW, does the 208 volt transformer have any taps to compensate for VD, which may need to be raised??

Yes to the question of changing out both disconnects to new ones. The existing 480v primary had a defective knife switch that meant the blades were only inserting half way in (lots of discoloration). new 600v 100amp disc9onnect with new 90 amp fuses took care of that problem. The secondary disconnect on the existing 45 kva wye transformer had formerly been a 100v 600v fused disconnect which kept blowing fuses as it was only fused at 90 amps 208v. I changed that existing secondary disconnect out to a new 200 amp 240v disconnect and replaced the secondary conductors (*which were previously too small 2 awg aluminum) with 1/0 copper and installed fuses for 150 smps. I am now changing out the rest of the feeder cable to 3/0 aluminum 5 wire to the existing 200amp 240vc panelboard.
in the whole history of this building, this 90 amp primary breaker had never triped and we could not replicate the problem. Everything today worked fine as well as yesterday after changing out the fuses (and extensive troubleshooting to be sure we did not overlook something).

lastly, it does have taps but no guide.I was looking at the delta transfromer which also has taps but since the input voltage is greater that 468 (which was the 5% MARK) (V a-b 480, b-c 473, a-c 473), I did not venture into changing the taps on the wye transformer, I assumed that tap 4 would be for 468 input voltage and currently the transfromer it is set up for 100%(tap 3). I did not know how adversely changing thios to tap 4 would be for the output voltages. I figured that all the 240v would run fine down to 208 which is why I went high leg delta.