chapter  14
Caustic-Side Solvent Extraction Chemical and Physical Properties: Equilibrium Modeling of Distribution Behavior
Pages 19

Introduction e solvent extraction process proposed and considered for cesium removal from the waste present at the Savannah River Site (SRS) is being investigated with respect to the behavior of system components under dierent conditions. A thorough understanding of the process is in part demonstrable by establishing a model that predicts the extraction of cesium based on the major components of the waste (or simulant). e ability to predict distribution behavior facilitates appropriate owsheet design to accommodate changing feed composition and temperature. It also provides greater condence in the robustness of the process overall. Finally, given the knowledge of the composition of any particular feed, a reliable model yields an immediate estimate of expected owsheet performance for comparison with process data. e scope of this modeling study was directed toward predicting the cesium distribution ratios obtained with ve dierent SRS simulants corresponding to ve real-waste tanks. Chemical analyses of the tanks provided the concentrations of sodium, potassium, cesium, nitrate, and free hydroxide. When preparing the simulants, the total concentration of cations could be as high as 5.6 M. e nitrate and hydroxide concentrations measured in the tanks could not balance the cation concentration. e quantity of anion still not accounted for by these analyses was lled either with chloride or with nitrite anions. Based on the total composition of the SRS waste, these four anions and three cations were determined to be the main components. e model will include species of these ions, and corresponding formation constants will be determined by the sequential modeling of simple systems containing rst one cation and one anion at the same time, then systematically increasing the number of components. A model representing the extraction of cesium from dierent media will then be established and cesium extraction behavior could be predicted by a simple input of the concentrations in the aqueous phase before extraction.