Pilot-Scale Validation and Predictive Modeling of Layered Adsorption Columns for Co-Removal of Cesium and Strontium

Cesium and strontium (Cs/Sr) are key radionuclides in radioactive wastewater. Dynamic fixed-bed adsorption is a practical approach for Cs/Sr removal, but engineering-scale design remains challenging because laboratory-derived breakthrough models have rarely been validated against independent pilot-scale data. Here, a mechanistic framework was developed and independently validated for a layered column containing silica-supported ammonium molybdophosphate (AMP/SiO2) and silica-supported potassium hexatitanate (K2Ti6O13/SiO2) as Cs- and Sr-selective beds. An axial-dispersion model with material-specific equilibrium relations and an effective liquid-concentration driving-force coefficient was calibrated solely against laboratory breakthrough curves. The resulting parameters were transferred without refitting to a pilot-scale column operated with a seawater-like feed, representing a 1500-fold increase in adsorbent mass. Laboratory calculations yielded R2 values of 0.997 for Cs and 0.998 for Sr. The pilot calculation reproduced the Sr curve with R2 = 0.974 and RMSE = 0.068; the predicted 5% breakthrough volume (V5) differed from experiment by 3.6%. Propagated V5 uncertainty limits were ≤9.3% for Cs and Sr. Cs remained below 5% breakthrough through 44.2 L. Sensitivity analysis identified equilibrium capacity and feed concentration as the principal controls on breakthrough bed volumes; geometry and transport affected bed utilization and front width. These results support model-based scale-up under the tested conditions.