Continuous prediction technique for fast determination of cyclic steady state in simulated moving bed process

Abstract

In the simulation of cyclic processes, such as simulated moving bed (SMB), the system should be equilibrated to reach a cyclic steady state (CSS) before evaluating the process performance. However, the conventional method of successive substitution is quite time-consuming. In this work, a continuous predicting method (CPM) is developed for fast determination of CSS in SMB. In CPM, the direct prediction of state variable at CSS and solving model equation are conducted alternately until CSS is reached. In order to give a guideline for the selection of the acceleration factor, CPM is applied on SMB process for enantioseparation of 1,1'-bi-2-naphtol racemate and with the optimized acceleration factor, 59% of computation time saved compared with successive substitution. In addition, this method is further successfully used in a sugar separation process. Given its efficiency and simplicity, this method could provide a useful tool for SMB simulations.

Combination of space–time conservation element/solution element method and continuous prediction technique for accelerated simulation of simulated moving bed chromatography

“…The other type of technique aims to accelerate the convergence process, i.e. reduce the iteration numbers for reaching CSS [3,16,33,34]. Minceva et al [3] and Hidajat et al [33] used non-void initial conditions to reduce the iteration numbers, i.e.…”

“…It should be mentioned that the computational efficiency with CPM is closely related to the acceleration factor . Previously, we have optimized the  values for two SMB systems, which were used for the enantioseparation of racemic 1,1'-bi-2-naphtol and glucose-fructose separation M a n u s c r i p t [16], and the optimized  values are 0.5 and 0.55, respectively. The results suggest that the optimal  value generally relies on the specific features of a desired separation process.…”

“…Yao et al [34] employed "quasi-envelope" method to predict the CSS. Recently, we developed a continuous prediction method to accelerate the convergence process [16]. But due to the highly nonlinear feature of state variable propagating along columns, the improvement of this method is not so significant, such as the newly developed continuous prediction method (CPM) method which can only reduce iteration number by 50-60% [16].…”

“…In the previous work, we developed a CPM method for fast simulation of SMB process [16], in which a predicted state variable at CSS is constructed based on the last two iterations and used as the initial condition for the next iteration. The general function for the prediction is as follows:…”

“…In the design and optimization of SMB processes, the system performance is evaluated after reaching a cyclic steady-state (CSS) [16]. However, the time it takes to reach CSS could be very long.…”