Just about everything about this mixer description is wrong.  The basic question is the input power requirement for this application.  The input power requirement for this anchor impeller operating at 50 rpm is about 445 kW (600 hp).  Even if the "fluid" is water based with a high heat capacity, the temperature rise due to power dissipation would be about 8C/minute.  At the lowest speed, 25 rpm, the power requirement would still be 110 kW (that is kilo watts, not watts).
Problems:  1) A viscosity of 700,000 cp represents a fluid that probably will not flow.  Nearly all fluids with such a high viscosity are non Newtonian, with some shear dependence and possibly a yield stress.  A single viscosity value does not adequately represent the fluid properties and mixing requirements.  Pouring this material from one container to another would take hours or even days, if it ever moved on its own.  2) A direct consequence of slow flow means that the rotational speed for this mixer is much higher than would be effective.  3) An anchor impeller is this application would be useless for mixing.  The anchor, if it could be turned, would only move material near the wall, leaving the center of the tank unmixed.  An anchor impeller provides no vertical fluid motion and will not mix the contents to any degree of uniformity in a reasonable period of time.  The only impeller type with a chance of working in this application is helical ribbon impeller or similar design that provides vertical movement of the material. 4) The anchor diameter is 99% of the tank diameter.  For such a tight fit, the inside diameter of the tank would have to be machined to a nearly perfect cylinder and the mixer would have to be mounted exactly on the center axis of the tank.  It would have to be centered without any angle from the vertical (less than a fraction of a degree). 5) The minimum shaft diameter for hydraulic loads is 5 cm.
This problem is so unreasonable as to offer no obvious path for practical improvement of evaluation.