In recent years, simulated moving bed (SMB) technology has become increasingly attractive for complex separation tasks and emerged as the leading chromatographic separation technique in the fields of pharmaceuticals, fine chemicals, and biotechnology. Extensive usage of SMB technology has created a need for robust automatic control techniques to exploit its full economic potential. Automatic control of SMB units is a challenging task not only because of its nonsteady-state, nonlinear, mixed discrete, and continuous nature, but also because of long delays involved in exhibiting the effect of disturbances. This work proposes a new optimization-based control strategy, in which a linearized time-varying reduced-order model that accounts for the periodic nature of the SMB process is used for on-line optimization and control of SMBs applied to systems characterized by linear adsorption isotherms. Four internal flow rates, which can be adjusted via external flow rates, are used as the manipulated variables. On-line concentration measurements at the product outlets together with a periodic Kalman filter are used to reduce the effect of model errors. Optimal input adjustments that allow for the achievement of process specifications and optimal performance are calculated on the basis of the predicted evolution of the plant. The realization and implementation of the concept on a virtual eight-column SMB unit is described, and the capability of the controller to run the SMB plant at its economic optimum regardless of possible disturbances and model uncertainties is assessed.

A three-fraction simulated moving bed (3F-SMB) unit for separating feed mixtures into three fractions is presented. Criteria for the selection of the flow rate ratios in the four sections of the unit under linear chromatographic conditions are developed. These criteria translate into a region of complete three-fraction separation in the (m 2 , m 3 ) plane. The performance of the 3F-SMB unit is investigated through detailed simulations using as a model system the separation of the four nucleosides: 2′-deoxyadenosine, 2′-deoxythymidine, 2′-deoxyguanosine, and 2′-deoxycytidine. For nonlinear chromatographic conditions, simulations on the same model system allowed qualitative conclusions on the performance of the 3F-SMB to be drawn. Further, the possibility of using the 3F-SMB unit to implement a cleaning-in-place (CIP) step within a standard SMB is shown and validated through simulations, using as a model system a mixture of two nucleosides and one highly retained impurity.

No abstract

For over 150 years, slow sand filters have been an effective means of treating water for control of microbiological contaminants. Slow sand filters do not need constant operator attention, making them an appropriate technology for water systems that are small or that employ part-time operators. During the 1970s through the 1990s, research and field evaluations of slow sand filtration have demonstrated its efficacy for control of microbiological contaminants that were unknown in the 1800s. In addition, pretreatment processes such as roughing filters and pre-ozonation have been developed or adapted for use with slow sand filters, increasing the range of source waters that can be treated and the number of contaminants that can be removed in slow sand filters. Inclusion of a layer of granular-activated carbon in a slow sand filter bed has improved capability for control of synthetic organic chemicals. This paper reviews design concepts and process capabilities for slow sand filters and discusses recent innovations in slow sand filter design that now enable this technology to be applied more widely than would have been appropriate two decades ago. Key words: slow sand filter, design, operation and maintenance, microbiological contaminants, small systems, pretreatment.