Experimental Validation of SMB Models

Chromatographic separations with product purities exceeding 99% and high separation costs require precise prediction of the optimal process design. This demands carefully validated models, especially in the separation sections of the SMB plant. Consequently, methods are needed to increase the number of measured data points. 

Importantly, sampling gives only a limited number of points in the axial profile. The complete curve cannot be reconstmcted from any temporal measurement, as the concentration fronts inside the columns change from the beginning to the end 

2 Isocratic SMB In the following some examples based on our own research are given for the validation of the SMB transport-dispersive model, using an online detection system in the recycle stream. All fiow sheet, column, and plant models were implemented in the gPROMS (PS Enterprise, UK) simulation tool and solved with OGEE methods (Section 6.4). 

In all cases good agreement between simulation and experiment was found. 

Isotherm data for sucrose in this concentration range were determined from pulse experiments and represented by a linear isotherm. Because of the low adsorp-tivity, this impurity is collected at the raffinate together with the glucose. 

---