Your description reveals a number of reasons for your process problems.  The most obvious reason for “product Inconsistency” is that the “mixer speed is not consistent.”  To obtain consistent results in any process, mixing or not, the conditions need to be repeatable and consistent.

Your definitions of “large scale 12.5 L” and “highly viscous (~400 cp)” are both at the extreme low end of most industrial mixing processes.  However, the conditions are important for laboratory studies.  Laboratory studies are difficult to use for mixing studies because the equipment is not necessarily geometrically consistent.  Unless the tests are carried out under closely controlled conditions results of similar tests may be different.  Effective scale-up, even from 500 ml to 15 L, would typically involve an increased rpm with geometrically similar equipment to give equivalent results.

To properly observe mixing results in a 500 ml beaker would require an accurate measurement of the stir bar dimensions, within 1 mm, and accurate, consistent, and measured control of the rotational speed, within 5 rpm.  The speed needs to be constant or changed on a programmed schedule.  The speed needs to be measured and recorded during the entire duration of the process.  These conditions are only sufficient to get consistent results.  To do scale-up, you need to vary the process conditions to decide how the mixer speed or stir bar length affect the process.  Ideally, you need to explore possible failures modes, such as too low of a stirrer speed, too rapid addition of powder, insufficient increased speed as viscosity increases, etc..  Understanding causes of failures may provide valuable information about what needs to be done on scale-up to 15 liters.

Moving from a stir bar in a beaker to a “reactor with paddle agitator (bottom) and turbine agitator (top)” will require an understanding of how mixing affects the process and what minimum conditions are necessary for success.  The scale-up problem involves two critical mixing factors:  mixer geometry and equivalent speed.  The mixer speed may need to be chosen to produce the same peripheral speed or the same power input or the same circulation pattern, or comparable blend time, or other factors influenced by the mixer.  Because of process unknowns, different effects need to be identified during testing.

The answers by this expert are based on the best available interpretation of the information provided. The consequences of the application of this information are the responsibility of the user. If clarification is needed, please submit a further question.