Managed aquifer recharge systems for drinking water reclamation are challenged by trace organic chemicals (TOrCs) since some of them are poorly retained. Although a lot of research has been done to investigate biological transformation of TOrCs in sand filter systems, there are still uncertainties to predict the removal. A laboratory column system with two different filter sands was set up to test TOrC transformation, the influence of low oxygen concentrations as well as the adaptation and influence of spiked TOrC influent concentrations. Bioactivity was quantified with the fluorescence tracer resazurin. In the experiment, a low elimination performance in the first column segment, defined as lag zone, was observed, implying incomplete adaptation or inhibiting co-factors. To assess these lag zones and to determine the dissipation time DT50 for 50% removal, a modified Gompertz model was applied. For acesulfame, formylaminoantipyrine, gabapentin, sulfamethoxazole, and valsartan acid DT50 of less than 10 h were observed, even when influent oxygen concentrations decreased to 0.5 mg/L. In general, TOrC transformations in technical sand with lower bioactivity and especially valsartan acid transformation responded very sensitive to low influent oxygen concentrations of 0.5 mg/L. However, in well adapted sand originating from soil aquifer treatment (SAT) with sufficient bioactivity, TOrC removal was hardly affected by such suboxic conditions. Furthermore, increasing the influent concentrations of TOrCs to 10 μg/L was found to promote adaptation especially for acesulfame and sulfamethoxazole. Benzotriazole, carbamazepine, diclofenac and venlafaxine were recalcitrant under the applied experimental conditions.
Read full abstract