Evaluation of optimal cooling control for seeded batch crystallization inclusive dissolution with uncertainties
Crystallization process is generally controlled to be operated within metastable zone where the crystal particles will grow until the end of the operation. High supersaturation level resulted from the process operated too close with metastable limit at the beginning of cooling crystallization operat...
Saved in:
Main Author: | |
---|---|
Format: | Thesis |
Language: | English |
Published: |
2020
|
Subjects: | |
Online Access: | http://umpir.ump.edu.my/id/eprint/31466/1/Evaluation%20of%20optimal%20cooling%20control%20for%20seeded%20batch%20crystallization.pdf |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Summary: | Crystallization process is generally controlled to be operated within metastable zone where the crystal particles will grow until the end of the operation. High supersaturation level resulted from the process operated too close with metastable limit at the beginning of cooling crystallization operation causing high nucleation and crystal growth rate. As a consequence the target crystal product is able to achieve but there is unnecessary amount of fine crystals by the end of the operation. This excessive nucleation can be reduced by introducing dissolution step in the crystallization operation where the fine crystals generated by nucleation is dissolved back into solution which in return, reducing the amount of fine crystals in the overall CSD. Therefore the objective of this study is to develop the optimal cooling control inclusive dissolution phenomena for batch seeded crystallization using potassium nitrate crystallization as a case study. The mathematical model for potassium nitrate crystallization inclusive dissolution phenomena was developed and simulated in Matlab software. The open-loop simulation for nominal cooling of the process was validated against experimental data. Several other strategies pertaining to achieve desired CSD with minimum amount of fine crystals were deployed. The optimization algorithm was employed in order to determine the optimal set-point trajectory for closed-loop control. Based on this trajectory, optimal closed-loop control of cooling crystallization process using Proportional-Integral (PI) controller were able to maintain the crystallization operation at its set-point and successfully achieved the desired target of crystal product with minimum amount of fine crystals. The proposed optimal cooling crystallization process was able to improve mean characteristic length and amount of crystal fines by 67% and reduced by 71% respectively to compare with nominal cooling strategy. 4% larger crystal characteristic length and 25% lesser amount of crystal fines is successfully accomplished if compared with temperature swing strategy. Then the robustness evaluation of controller was tested through uncertainty analysis using Monte Carlo simulations in order to analyse the effect of input uncertainties towards the variation of the product distribution. Large variation on final CSD was observed. Sensitivity analysis was then deployed using Standardized Regression Coefficients (SRC) method in order to analyse the correlation of the input uncertainties with the model output CSD. Retuning of the controller was chosen as the action taken to improve the robustness. The final CSD show minimum variation after controller retuning step. Therefore robust optimal control for seeded batch cooling of potassium nitrate crystallization was successfully established. |
---|