Predictive modeling of micropollutant removal in fixed-bed adsorbers

Fixed-bed adsorption is a promising technology for removing micropollutants from water, and its performance can be predicted using mechanistic models. Predictive modeling, however, is often hindered by the need for extensive parameter estimation. This study adopted a methodology that minimized experimental input and avoided calibration against breakthrough data. Three constant pattern models were evaluated: (1) an irreversible isotherm with film and intraparticle diffusion, (2) an irreversible isotherm with intraparticle diffusion only, and (3) a Langmuir isotherm with intraparticle diffusion only. Equilibrium parameters were obtained from batch adsorption experiments, and transport parameters were estimated from literature correlations. The models were tested against two published breakthrough datasets: ciprofloxacin on a polymeric resin and phenol on activated carbon. For the ciprofloxacin-resin system, only the model including both film and intraparticle diffusion resistances yielded quantitative agreement with the data. For the phenol-carbon system, none of the models reproduced the breakthrough curve, indicating underestimation of the intraparticle diffusion coefficient by the correlation. Agreement was recovered when a higher diffusion coefficient was assumed. These results demonstrate the limitations of current correlations and highlight the need for improved predictive tools for intraparticle diffusion in adsorption systems.