One possibility is it the motor shop was operating closer to 460 volts at the motor or possibly 460 unloaded, where at the site, he has 480 volts and 476 volts at a loaded motor... Maybe closer to 500 at the switchgear or transformer taps.
Redo the calculations just assuming for a moment for humor sake a 44 volt difference in input voltages, 460 vs 504. It seems to me that one would use plus percent taps on large motors with long runs of wire to them to compensate for voltage drop.
we know we have 475 at the motor, as measured
even if the shop is 500 or plus 5% that would only increase current 4-5%, we have a 30% increase
460 vs 504, 10% v increase, ~8-10% I increase, observed is 30%
it runs normal on lo, 80% x rated
it runs high on hi, 115%
on no load it runs 152 vs 195, shop vs field, but we do not know conditions: temp, spec, voltage (should only make a small difference), etc
if damaged it should impact all load conditions
do we evne know if the no load test was done in lo, hi, ???? at the shop
no one has offered a viable option for the 30%
same motor
same voltage ~
same load