Simulated moving bed (SMB) chromatography represents one of well-established separation technologies and its operation for a given separation typically undergoes startup, production and shutdown stages. Usually only the desired cyclic steady state (CSS) is employed for production purposes and transient behavior is neglected. However, improving transient performance is of significant advantage for applications of different scales. In particular, for small-scale separation campaigns where very often the same SMB unit is used repeatedly to process small batches of different mixtures and transient operating time remains comparable to production period, developing efficient startup and shutdown strategies is of great significance. However, most of literature deals with the design and optimization of cyclic steady state. Furthermore, CSS conditions are often used as transient operating conditions. Although easy, this operating regime can not guarantee satisfactory transient performance. Only little attention was given to consider transient behavior [1,2]. Systematic studies of optimal startup and shutdown operation have not been reported so far in the literature.
Recently, we have first proposed a multistage startup policy for the SMB process , which allows to dynamically adjust the operating conditions during the startup period. The optimal startup operation is then formulated as a dynamic optimization problem for which a specially tailored decomposition algorithm is developed. In this work, we will present such an operating concept in detail and generalize it to the shutdown process. Moreover, we will also apply it to a newly suggested SMB modification which integrates fractionation and feedback (FF-SMB, ). Several case studies are used to demonstrate the feasibility of the solution algorithm and to evaluate the potential of the new operating regime. The quantitative results reveal that with the optimal transient operation, the process can achieve a significant reduction in both transition time and desorbent consumption. The transient performance in terms of product concentration and purity is also improved considerably.
 B.G. Lim, C.B. Ching, Sep. Technol. 6 (1996) 29
 Y. Xie, S.Y. Mun, N.H.L. Wang, Ind. Eng. Chem. Res. 42 (2003) 1414
 S. Li, Y. Kawayiri, J. Raisch, A. Seidel-Morgenstern, 9th International Symposium on Dynamics and Control of Process Systems, Belgium, 2010
 L.C. Ke▀ler, A. Seidel-Morgenstern, J.Chromatogr.A 1207 (2008) 55