Simulation And Optimization Of Simulated Moving Bed Process For P-xylene Separation

C8 aromatic isomers(consisting of p-xylene,as well as o-xylene,m-xylene,ethyl-benzene)are produced from catalytic reforming,gasloline pyrolysis,or toluene disproportionation.p-xylene(PX),as the most important isomer of mixer-xylenes,is often used for terephthalic acid(TPA)and dimethylterephthalate(DMT)synthesis.The PX made TPA and DMT could be used as raw materials for polyethylene terephthalate(PET),which are widely used in polyesters fibers and plastics production and promoting of chemical fiber industry.Mixed xylenes separation is the core unit of an aromatics complex.SMB has already been proved to be a promising option in this field because of its low solvent consumption,small apparatus size,low investment cost and high yield.SMB process often includes a system of multi-variable operations and complex dynamics nonlinear model.The simulation and optimization of SMB process has been unprecedented significant and meaningful for both SMB design and industrial production.To achieve such goal above,the following work has been organized as follows:Based on PX SMB process mechanism model,two mathematical models has been proposed to simplifying the four-zone system: the true moving bed(TMB)approach and the simulated moving bed(SMB)approach.The effect of operational variables and adsorbent’s deactivation on the process performance was studied though simulation and results showed that purity and recovery of PX in products are sensitive to switching time and circle flowrate changes while flux ration of zone II has a conspicuous impact of the trade-off between purity and recovery for final products.The decline in adsorbent capacity also has a large impress on PX recovery but little on purity.The industrial-scale,seven zone,simulated moving bed unit for p-xylene purification with three type of dead volumes-bed lines,circulation lines and bed heads is analyzed.The results show that the raw material tapped in bedline cause a singnificant decrease in PX purity.The seven zone operation including line and secondary flushes would enhances purity of PX while maintain the same recovery.The flow rate of line flush(QLF),secondary flush-inlet(QSFi)vary within a certain range for parametric study,where QLF,QSFi set for being able to flush 110%,140% of the BL volume,the purity of PX reaches 99.9% with 0.14% improvement of recovery,while desorbent consumption increase by 0.77%,compared with the industrial operation conditions.The drian-out flow of TF was enlarged to enhance PX recovery,where removed the raffinate that is a mixture of MX,OX,EB and less PX tapped inside in bed line.The TF flow rate set to flush 90% of the BL volume,the productivity of the eight-zone operation increase by 0.85kg/m3﹒hr-1 over that of the seven zone,which corresponds to about 2682 tons of PX/year in this industrial SMB unit.On the basic of the sub-models and sensitivity analysis,reduced sequential quadratic programming(r-SQP)was employed to optimize the four-zone and seven-zone SMB process for PX purification.The optimization goal is to maximize the productivity of PX and minimize the desorbent consumption under the condition of meeting the required of product purity and recovery.To maximize the PX productivity of four zone SMB process with fixed feed flow rate,decision variables are optimized by r-SQP method within given constraints.Results showed that under optimal conditions,the PX productivity increase by 2.43%,while the desorbent consumption decreased 2.87%.The same optimization strategy was used to optimize seven zone industrial-scale SMB processes.A 2.43% of increase in PX productivity and a 4.85% of decrease in desorbent consumption were reached under optimal conditions,compared with initial conditions.All results showed that this method is suitable for the optimization of simulated moving bed process.