Anti-solvent crystallization of sucrose

Abstract

In 2011, the Wageningen University discovered and patented a new process for crystallization of sugar. In this process ethanol is added to the sugar solution and water is selectively removed by a zeolite. Royal Cosun renounced the idea of working with zeolites, but remained interested in the concept of anti-solvent crystallization. The study started out by analyzing the expected benefits of a process based on anti-solvent crystallization. The energy consuming process of water evaporation is no longer needed if only anti-solvent crystallization is applied. However, the concept introduces an energy consuming separation step to recover the anti-solvent. A literature study was conducted to determine the properties of the anti-solvent (ethanol) and the basic principles behind anti-solvent crystallization. Various models for the solubility, the viscosity and equilibrium data of sucrose mixtures were compared and validated. The model based on the modifed UNIQUAC method turned out to be in close agreement with a variety of solubility sources and validated, by the authors, for the expected ethanol concentrations and temperatures used in the process. Several process designs were considered for the new concept and a selection was made based on criteria provided by Royal Cosun. The selected process design consists of distillation columns to recover the ethanol, a crystallizer suitable for cooling and anti-solvent crystallization and recompression evaporators. Over the last years the Cosun Food Technology Center (CFTC) developed Mat-lab/Simulink based models of almost all the components of the sugar factory. These components can be linked together to represent the total sugar production plant. This model was also used for the study, however this meant that certain changes had to be made. Various new process blocks were developed that allow for anti-solvent crystallization. After several runs of the new model it turned out that only anti-solvent crystallization could not provide a high enough crystal yield. Therefor an additional cooling effect was considered. After optimizing the ratio cooling crystallization/anti-solvent addition it became clear that more cooling crystallization would result in lower energy consumptions. Finally two concepts were compared to the reference case, the traditional process. The results from the model showed that the new concept had a reduced energy consumption. However, the reduction is not enough to compensate the reduced electricity generation and the investment costs.