Crystallization has been one of the important unit operations that combines both separation and purification step in a single operation. Many manufacturing industries rely on crystallization for the production of their materials. These industries include the pharmaceuticals, the agro-chemical, and general chemical manufacturers. The main objective of a crystallization operation is to produce crystals of uniform size with a narrow size distribution. This imparts aesthetic value for the product, which assumes significance when a particular product is marketed.

One such way of accomplishing the above mentioned objective has been by the use of Oslo crystallizers that operate on the fluidized-bed principle, in which a growing mass of crystals is suspended in an upward flow of the supersaturated solution through the crystallizer [1]. As the supersaturated solution flows upward, the liquor contacting the bed relieves its supersaturation on the growing crystals and subsequently, the supersaturation decreases along the upward direction. As a result, the crystals segregate in the bed with the small ones at the top and the large ones at the bottom. Consequently, the behaviour of a fluidised bed of crystals exhibits a crystal size distribution (CSD), which is a prime objective in the design and operation of crystallizers. The important feature of the fluidised bed crystallizer is the absence of impellor, as opposed to the mixed-suspension, mixed-product removal crystallization, which effectively reduces nucleation thereby producing large and uniform size crystals.

This paper is an attempt to model a fluidised-bed batch crystallizer. A rigorous model has been developed to describe the behaviour of crystal growth. It is assumed that liquid phase moves upward through the bed in plug flow and the solid phase in the fluidised bed is perfectly classified. The existing model is a modification of the Shiau et al. [2] model where the basis has been changed to solute free solution flow rate that remains constant throughout the operation. This modification effectively reduces one equation from the original model, thereby reducing the computation time.

The performance of the model has been analysed for potassium alum system, as it is known to exhibit size-independent crystal growth rate. The set of four differential equations thus obtained were solved using an inbuilt stiff equation solver ODE15S in MATLAB based on Gear's algorithm. The developed model is capable of predicting the variation of the bed voidage, solution concentration and crystal size along the bed height during the operation. Consequently, the resulting CSD has also been determined. The results were compared with Shiau et al. [2] model and were found to be in good agreement. The model presents significant information for crystallizer design and hence can be used for design purposes.

References

[1] Perry, R.H., Green, D.W., & Maloney, J.O. (1984). Perry's chemical engineers' handbook, (6th ed.). New York: McGraw-Hill.

[2] Shiau, L.D., Shueh, H.C., & Liu, Y.C. (1999). Modelling of a fluidized-bed crystallizer operated in batch mode. Chemical Engineering Science, 54, 865-871.