טכניון מכון טכנולוגי לישראל
הטכניון מכון טכנולוגי לישראל - בית הספר ללימודי מוסמכים  
M.Sc Thesis
M.Sc StudentSchur Ammon
SubjectA Model for Estimating the Crystal Size in a Single DTB
Crystallizer
DepartmentDepartment of Chemical Engineering
Supervisors Professor Emeritus Raphael Semiat
Dr. Grigori Zelmanov
Full Thesis textFull thesis text - English Version


Abstract

The goal of the present work was to develop a model for understanding and predicting the crystal size distribution in a draft tube baffle crystallizer. Two assumptions were involved in this attempt: The model should be based on fundamental understanding of the mechanism of crystallization; and the model should relate to the special design and flow characteristics of the system concerned.

The model was examined through two different systems on two different scales: A 22 liter pilot crystallizer and a 350 cubic meter industrial crystallizer used for the manufacture of Potash in the Dead Sea Works. Each of the crystallizers was operated at steady state and sampled. In addition, the controlling parameters of both systems were modified: The internal circulation was varied by changing the agitator speed, and the residence time was changed by keeping different levels of magma density inside the crystallizer.

Crystal growth is a combination of three simultaneously sub-processes: initial growth due to supersaturation as the driving force, agglomeration in consequence of turbulent collisions and destruction of agglomerates because of particle encounters with the impeller blade and other particles. An overall growth rate was formulated and integrated into the population balance equation.

Samples from both systems were examined by SEM analysis to find a similar, hierarchical structure that reinforces this model. In conclusion, it was found that the industrial crystallizer has a good correspondence with the presented model also when exposed to changes.  Increasing the magma density had stimulated the agglomeration process and enlarged the particles size while increasing the impeller speed didn’t have an impact in the checked range. Changes made in the laboratory crystallizer didn’t have the same results and have shown a lower level of matching to the model.

In general, two factors determine the particle size distribution in an industrial crystallizer and are critical for design. The first is the impact energy and is dependent on the flow characteristics inside the crystallizer.  The second is the residence time which is dependent on the operational conditions. Both parameters are expressed in the suggested model with a high level of success.