I agree with waross, it's probably large current imbalances between the 3 legs.  Did you get the 210-220A off just one leg?  And then perhaps put the power monitor on a different leg?  There are power monitors that connect to all 3 and thus give "true power" accounting for the phase shift.  I can't remember the brand name. All three legs read between 210 and 220 with the clamp on.  


Just for clarification, I'll describe the set up.  We have a transfer pump connected to a tank that varies between 25'-4' depending on how long the pump is running.  When the tank is full, the pump draws about 220-225 amps, sometimes even higher.  When it is nearing 4', it draws about 200-210 amps.  

Matching up the pump curve with operation points, the power draw when the tank is full should be 150 hp, and 140 hp when the tank is almost empty. Converting with a power factor of .95 and motor efficiency of .9, the draws should be 157 Amps and 147 amps, respectively.

Pumping bypasses were checked and everything was closed.  If the draw was 220-225 amps, we would be pumping an extra 1000 GPM somewhere (operating point is around 2100 GPM) Vacuum pumps