Development and application of a predictive model for advanced wastewater treatment by adsorption onto powdered activated carbon.

This work presents a mathematical method to describe adsorptive removal of organic micropollutants (OMPs) and dissolved organic carbon (DOC) from wastewater treatment plant effluent using powdered activated carbon (PAC). The developed model is based on the tracer model (TRM) as a modification of the ideal adsorbed solution theory (IAST) and uses the fictive component approach for organic matter fractionation. It enables the simulation of multisolute adsorption of OMPs considering competitive adsorption behavior of organic background compounds (OBC). Adsorption equilibrium data for DOC and seven different OMPs as well as kinetic data for DOC were derived from batch experiments performed with secondary clarifier effluent of two municipal wastewater treatment plants (WWTP 1 and WWTP 2). Two conventional PAC products were investigated as well as one biogenic PAC (BioPAC). Verification and validation of the fitting results based on operational data of WWTP 1 showed promising prediction of DOC and OMP removal efficiency. However, when applied to a static simulation of a full-scale PAC adsorption stage, the model overpredicts the removal efficiency of sulfamethoxazole and candesartan. For benzotriazole, carbamazepine or hydrochlorothiazide, predicted removal falls below operational removal. The model can be used to predict removals of good adsorbable OMPs but fails to accurately predict the removals of OMPs with variable or low PAC affinity. The model was further used for a dynamic simulation of DOC and diclofenac effluent concentrations of a full-scale PAC adsorption stage with varying operating conditions and influent concentrations. Results show that the hydraulic retention time (HRT) in the contact reactor is a decisive operational parameter for OMP removal efficiency besides the PAC dose.