Current imbalance in a motor could either result from source voltage imbalance, or more rarely from assymetries in the three phases of the motor.

If you have multiple motors failing and phase current imbalance is a factor, it would more likely result from source voltage problem. Imbalanced voltages give rise to imbalanced currents. NEMA gives a derating curve according to voltage imbalance which requires 1% derate for volt imbalance of 1%, 4% derate for volt imbalance of 2%, 9% derate for volt imbalance of 3%, 16% derate for volt imbalance of 4%, and 25% derate for volt imbalance of 5%.  I believe in this case volt imbalance is calculated as the maximum deviation from the average voltage.

The reason for this derating is increased heating which results from current imbalances.  This is due in part to localized heating (one phase carries more current than other phases) and also the negative sequence component of the unbalanced current creates reverse torque which must be overcome by increased positive sequence current.  If imbalance is truly the cause, then inspection of the motor might show localized overheating on the coils of one phase.  

Clamp-on current measurements would help to understand/quantify either possible imbalance or increased current due to high dissolved solids.

The wet environment you mention can be a direct threat to the motor insulation. Particularly bad if your space heaters aren't properly sized or aren't working properly.

Power system voltage spikes can also contribute to multiple motor failures.  There are certainly lots of possibilities.  Can you provide any more info?

From my experience, motors that drive agitators can be overloaded by a change in the material that is being agitated.  A clamp on meter can tell you if you are overloading the motors.  Of course proper overload protection should save you, but it sounds like you do not have the proper protection.

Wirenut

Source voltage imbalance: I agree but based on a few load readings this morning and the general history of operation of these motors, I think it maybe a combination of factors. Latest loads readings have the motors running 5-10A above the FLA rating.(IN SF) My approach is that the problem has to be global, since multiple motors are being affected

Harmonic distortion: This was also a concern but the concept doesn't really come into play unless solid state switching equipment is involved. Possible in this case since there exists a non-linear load.

Voltage spikes: Also possible. Need to determine power factor and analyze power quality

High resistance connections and heating: Also possible

My next course of action is to use a Motor Circuit Evaluator to see if I can eliminate some of these possibilities and narrow down the issue. Measuring the crest factor of the V & I signals should confirm the voltage spike issue. The same MCE can be used to measure extent of harmonic distortion, polarization index as well as dielectric absorbtion.

Of course inconsistencies with element and fuse sizes will be corrected and a subsequent load profile developed based on the readings. Maybe a pattern can be established which could lead to a solution

DATA: 35 TANKS-- 35 MOTORS (5 ROWS OF 7)
      40hp, 1460rpm, Mitsubishi 200L frame, N45 heater, SF-      1.15, AB size 3 starter
      Dedicated cable tray per row
      Problem on 3 out of the 5 rows
      measured supply voltage :  470-475V (480V system)
      

If you run a 40 hp motor @ 5-10 amps above FLA they will fail.  I think you need 50 HP motors.

Wirenut

Wirenut has a good point about the overload.  I'm guessing that about 7.5A above FLA corresponds to SF 115% loading.  Not a good place to operate continuously.  (and perhaps the fluid variability sometimes loads the motor even higher?).

It sounds like you've got a good plan to explore all of the possibilities (including multiple influences) with an open mind.

I would encourage you to add a post-mortem inspection of failed motors if possible to give clues as to the cause of failure.

OK. I'll post my findings when everything is complete. Thanks for the contribution.

Here is a handy link on unbalance & three phase motors.

http://www.franklinelect.com/Manual/AIM_29.htm