Search

This study reports a new commercial simulated moving-bed (SMB) process for separating ~620 kt/year of p-xylene (PX) with 99.7% purity from the C8 aromatic mixture. An extended Langmuir isotherm was used to consider the binary interaction among six components. The binary interaction parameters were estimated from single and binary adsorptions experimentally measured on a faujasite X-type zeolite at 177 °C and 10 atm. Because the PX-SMB system included 24 bed lines connecting one rotary valve and 24 beds, three flushing sequences were necessary to prevent the deterioration of purity caused by the bed lines. An SMB dynamic model with port switching was created, and its solution for concentration profiles was compared to the plant data. The three flushing flow rates and the tertiary flushing port location were optimized by the parametric study. The PX production rate increased by 0.8% under the optimized condition, which corresponded to the growth of 5.1 kt/year of PX. [ABSTRACT FROM AUTHOR]

Using strong cation-exchange simulated-moving-bed (SMB) chromatography, a nitrogen−phosphate−potassium (NPK) fertilizer is produced in a cost-effective manner. The SMB process, operated in a nontraditional way, is divided into production and regeneration sections, for exclusion of undesirable ions, and is composed of six zones, including two wash-water zones. This paper addresses modeling, simulation, and optimization studies on this ion-exchange SMB process, based on experimental data obtained both from a pilot plant and an industrial plant. Model parameters that are initialized by empirical equations are identified, comparing the simulation results with the experimental data. Through sensitivity analysis of the model parameters, their effects on the process performance are examined. The simulation results show good agreement with in situ experimental data obtained in both the pilot- and industrial-scale plants. The objectives of this study are to optimize the SMB process in terms of (i) maximization of productivity in the production section and (ii) minimization of wash-water consumption, thereby resulting in (i) an increase in profit and (ii) a reduction in the overall operating cost in the downstream processing, respectively. The two objectives are sequentially treated within the framework of a multilevel optimization procedure (MLOP), which includes two pre-optimization levels, a productivity maximization level, and a desorbent (or wash-water) consumption minimization level. In this optimization study, it is demonstrated that wash-water consumption can be reduced by 5% at a 5% higher productivity. [ABSTRACT FROM AUTHOR]