(621a) Combined Materials Screening and Process Optimization for Large-Scale p-Xylene Separation Using Simulated Moving Beds (SMB)

Simulated Moving Bed (SMB) chromatography [1,2] has found significant applications in different areas including chemicals, petrochemicals, biotechnology, pharmaceuticals, and fine chemistry. The Parex process from UOP is based on SMB and is a leading technology for the recovery of p-xylene (1,4-dimethyl benzene) from a mixture of C₈ aromatic isomers. p-Xylene is used as an industrial solvent and an intermediate to produce polyethylene terephthalate (PET) for the production of films, fibers, resins and many valuable products. Because of its steady market, the demand for p-xylene increases 6-8% every year [3].

 

   The SMB process for p-xylene uses adsorptive separation and exploits the differences in affinity of an adsorbent material for p-xylene relative to the other C8 isomers. The use of SMB technology enables a simulated counter-current contact between the mobile fluid phase and the stationary solid phase while avoiding the difficulties in the movement of a solid phase. This makes SMB an inherently complex process. Unlike most adsorption processes it is a continuous process and operates in a cyclic fashion. An industrial SMB process utilizes as many as 24 beds which are divided into four different zones, namely feed, desorbent, extract, and raffinate zones. The flows to the zones are maintained using a rotary valve.

 

   Designing an SMB process with optimal bed length, diameter, and operating conditions, such as flows to each zone and switching times, is a challenging task. While significant studies [4-13] have been performed to optimize the operating conditions in order to either maximize product recovery, purity and throughput, or minimize the desorbent flow, the overall cost of p-xylene separation is usually not considered. Furthermore, most optimization studies for SMB-based p-xylene separation selects the adsorbent material a priori. While adsorption of xylenes on Y zeolites has been extensively studied in the literature, the potential of replacing Y zeolites with other adsorbents is often neglected or remains unexplored.

 

   In this work, we first present a novel grey-box constrained optimization approach for the optimization of a large scale and industrial SMB process. The process consists of 24 beds and 4 zones for the separation of p-xylene from a typical C₈ mixture containing 23.6% p-xylene, 49.7% m-xylene, 12.7% o-xylene, and 14% ethylbenzene. The objective is to minimize the total annualized cost (the combined investment, operating and material costs) of p-xylene separation, while achieving the specified product (p-xylene) purity and recovery of 99.5% and 97%, respectively. The constrained grey-box models are developed based on the calculations of purity, recovery, and separation cost from a discretized version of a detailed nonlinear algebraic and partial differential equation (NAPDE) model [10,13] that describes the overall SMB process. The fully-discretized NAPDE model is highly nonlinear and nonconvex, and it is often difficult to solve the model using commercial solvers. In this work, we optimize the SMB process using ARGONAUT (AlgoRithms for Global Optimization of coNstrAined grey-box compUTational problems) [14], which we have developed for solving grey-box and black-box constrained complex problems. While we develop efficient strategies for solving the discretized NAPDE model as an NLP problem using a local solver such as CONOPT, the constrained grey-box optimization approach shows the capability to find better solutions and outperforms CONOPT.

 

   We have also identified adsorbent materials for cost-effective xylene separation by combining in silico screening of zeolites and SMB process optimization. We not only select the most cost-effective materials, but we also attain the optimal process conditions while satisfying purity, recovery, and other process constraints. We previously demonstrated the applicability of combined in silico screening of materials and process optimization for post-combustion CO₂ capture [15–17] and natural gas purification [18] using zeolites. To the best of our knowledge, this work is the first to simultaneously identify cost-effective materials and optimized SMB process for p-xylene separation.

 

References

[1] Ruthven, D.M.; Ching, C.B. Counter-current and simulated counter-current adsorption separation processes. Chem. Eng. Sci. 1989, 44, 1011 - 1038.

[2] Gomes, P.S.; Minceva, M. Rodrigues, A.E. Simulated moving bed technology: old and new. Adsorption, 2006, 12, 375 - 392.

[3] Wantanachaisaeng, P.; O’Neil, K. Capturing opportunities for para-xylene production. UOP LLC., http://www.uop.com/?document=uop-capturing-opportunities-for-para-xylene....

[4] Mazzotti, M.; Storti, G.; Morbidelli, M. Optimal operation of simulated moving bed units for nonlinear chromatographic separations. J. Chromatogr. A, 1997, 769, 3 – 24.

[5] Kawajiri, Y.; Biegler, L.T. Optimization strategies for simulated moving bed and PowerFeed processes. AIChE J., 2006, 52, 1343 – 1350.

[6] Kawajiri, Y.; Biegler, L.T. A nonlinear programming superstructure for optimal dynamic operations of simulated moving bed processes. Ind. Eng. Chem. Res., 2006, 45(25), 8503 - 8513.

[7] Kawajiri, Y.; Biegler, L.T. Large scale optimization strategies for zone configuration of simulated moving beds. Compute. Chem. Eng., 2008, 32(1-2), 135 - 144.

[8] Bentley, J.; Sloan, C.; Kawajiri, Y. Simultaneous modeling and optimization of nonlinear simulated moving bed chromatography by the prediction-correction method. J. Chromatogr. A, 2013, 1280, 51 - 63.

[9] Bentley, J.; Kawajiri, Y. Prediction-correction method for optimization of simulated moving bed chromatography. AIChE J., 2013, 59(3), 736 - 746.

[10] Minceva, M.; Rodrigues, A.E. Modeling and simulation of a simulated moving bed for the separation of p-xylene. Ind. Eng. Chem. Res., 2002, 41, 3454 - 3461.

[11] Minceva, M.; Rodrigues, A.E. Two-level optimization of an existing SMB for p-xylene separation. Compute. Chem. Eng., 2005, 29(10), 2215 - 2228.

[12] Minceva, M.; Rodrigues, A.E. Understanding and revamping of industrial scale SMB units for p-xylene separation. AIChE J., 2007, 53, 138 - 149.

[13] Kurup, A.S.; Hidajat, K.; Ray, A.K. Optimal operation of an industrial-scale Parex process for the recovery of p-xylene from a mixture of C-8 aromatics. Ind. Eng. Chem. Res., 2005, 44(15), 5703 - 5714.

[14] Boukouvala, F.; Hasan, M.M.F.; Floudas, C.A.  ARGONAUT: AlgoRithms for Global Optimization of coNstrAined grey-box compUTational problems, to be presented in the AIChE Annual Meeting, 2014, November 16 – 21, Atlanta, GA.

[15] Hasan, M.M.F.; First, E.L.; Floudas, C.A. Cost-effective CO₂ capture based on in silicoscreening of zeolites and process optimization. Phys. Chem. Chem. Phys., 2013, 15(40), 17601 - 17618.

[16] Hasan, M.M.F.; Boukouvala, F.; First, E.L.; Floudas, C.A. Nationwide, regional and statewide CO₂capture, utilization and sequestration supply chain network optimization. Ind. Eng. Chem. Res., 2014, 53(18), 7489 - 7506.

[17] Hasan, M.M.F.; Baliban, R.C.; Elia, J.A.; Floudas, C.A. Modeling, simulation and optimization of post-combustion CO₂capture for variable feed concentration and flow rate. 2. Pressure swing adsorption and vacuum swing adsorption processes. Ind. Eng. Chem. Res., 2012, 51(48), 15665 - 15682.

[18] First, E.L.; Hasan, M.M.F.; Floudas, C.A. Discovery of novel zeolites for natural gas purification through combined material screening and process optimization. AIChE J., 2014, 60(5), 1767 - 1785.