This article describes a systematic miniplant-based approach to rapid development of simulated moving bed (SMB) chromatography applications. The methodology involves analysis of single-column pulse tests to screen adsorbents and operating conditions and to determine initial values of profile advancement factors used to specify flow rates for an initial SMB miniplant experiment. A lumped-parameter linear driving force rate-based model is developed by fitting process data from a single miniplant run. The data are fit in a two-step procedure involving initial determination of effective adsorption isotherm constants as best-fit parameters with subsequent adjustment of calculated mass transfer coefficients to refine the data fit. The resulting simulation is used to guide further miniplant work and minimize experimental effort. The methodology is illustrated with miniplant data for a binary protein separation showing excellent agreement between model results and process data generated over a wide range of operating conditions.

Traditionally, olefin/paraffin separations are carried out using cryogenic distillation. The close relative volatility, subambient temperatures and high pressures required in cryogenic distillation make the cryogenic distillation an energy-intensive and capital-intensive separation scheme. Significant energy savings could potentially be realized if a nondistillation separation could prefractionate the olefins from the paraffins for more optimum placement in the downstream distillation unit or eventually completely replace the need for distillation. Reactive absorption with metal species carried in a solvent is one of the techniques that use a transition metal to selectively bind the olefin through formation of an olefin−metal π-complex. An experimental investigation into the selective absorption of ethylene over ethane using a novel copper metal solution containing aniline−CuCl− NMP was reported in the literature [Reine, Travis; Olefin/Paraffin Separation by Reactive Absorption. PhD Thesis, University of Texas at Austin, 2004]. The present work makes use of this work and models this reactive absorption−regeneration in AspenPlus for the selective absorption of ethylene over ethane using aniline−CuCl−NMP solution. The Aspen model predicted the experimental results for ethylene recovery and selectivity at varied liquid-to-gas ratios in reasonable agreement with the experimental results from the literature. The model can be used for a scale-up purpose and can also be extended for selective absorption of higher olefin/paraffin systems. It is also possible to use the simulation approach presented here to guide the research in this area, potentially reducing the energy consumption for the production of these key products.