Effect of dead volumes on the performance of an industrial‐scale simulated moving‐bed Parex unit for p‐xylene purification

Abstract

The cyclic steady state (CSS) of the industrial-scale, seven-zone, simulated moving-bed (SMB) unit for p-xylene (p-x) purification (Parex unit) with three types of dead volumes-bed lines, push-around and pump-around circulation lines, and bed heads-is analyzed. In particular, the effects of the size and level of hydrodynamic dispersion of each dead volume on process performance and on its CSS are studied in detail. The circulation lines change the CSS behavior from t à -periodic to Nt à -periodic, where t à is the switching interval and N512 is the number of columns in each adsorbent chamber. A high level of axial dispersion in the bed lines, characterized by P eclet numbers smaller than 100, affects the p-x purity. Moreover, the bed lines lower the average p-x concentration in the extract, which reduces the p-x recovery. If the small time lags introduced by the circulation lines are neglected, it is possible to develop a detailed process model that considers the operation of the Parex unit over a single switching interval as opposed to a full cycle, and whose CSS solution can be efficiently computed using a full-discretization approach. Finally, it is shown that the volume of the bed heads influences significantly the performance of the Parex unit, and that its impact on the location of the operating point with respect to the boundaries of the separation region can be approximately taken into account using the standard true moving-bed-SMB equivalence rules if they are corrected for the presence of extra interparticle fluid.

Design and optimization of JO‐IEX process for highly efficient quaternary separation of 5′‐ribonucleotides

“…1,2 Incipient SMB processes included Molex ® process for the separation of n-alkanes from branched and cyclic alkyne mixtures, Parex ® process for the separation of p-xylene from C8 isomer mixtures, and Sarex ® process for the separation of fructose from glucose in the sugar industry, and so on. [3][4][5][6][7] Subsequently, the application of SMB expanded to the fields of medicine, life sciences, and so on. [8][9][10] The separation of chiral drugs and proteins based on SMB is the most striking application.…”

“…Simulated moving bed (SMB) was developed originally by Universal Oil Products Company to separate some specific products in the petrochemical industry 1,2 . Incipient SMB processes included Molex® process for the separation of n ‐alkanes from branched and cyclic alkyne mixtures, Parex® process for the separation of p ‐xylene from C8 isomer mixtures, and Sarex® process for the separation of fructose from glucose in the sugar industry, and so on 3–7 . Subsequently, the application of SMB expanded to the fields of medicine, life sciences, and so on 8–10 .…”

Modelling of a Simulated Moving Bed in case of non-ideal hydrodynamics

“…Since its conception, numerous major improvements have been made in the field of the SMB modelling, regarding the mass transfer modelling (Mendes et al, 1996), CSS prediction (Pais et al, 1997a,b), SMB optimal design (Storti et al, 1995 andMazzotti et al, 1997a,b) and column dead volume modelling (Migliorini et al, 1999, Minceva and Rodrigues, 2003and Silva et al, 2016). Yet, plug flow models (axially dispersion or not) have been generally used to model the flow adsorption columns and their suitability has not being questioned.…”