Removal of antibiotics in sand, GAC, GAC sandwich and anthracite/sand biofiltration systems

Abstract

Drinking water biofiltration offers the possibility of the removal of trace level micropollutants from source water. Sand, granular activated carbon (GAC), GAC sandwich (a layer of GAC loaded in the middle of sand bed), and anthracite-sand dual biofilters were set-up in duplicate at bench-scale to mimic the filtration process in real drinking water treatment works. During the 3-month system operation, removal of five antibiotics (amoxicillin, clarithromycin, oxytetracycline, sulfamethoxazole, and trimethoprim) and overall biofilter performance were evaluated. Natural surface water spiked with a mixture of the target antibiotics was used as feedwater to the biofilters. Results showed that the target antibiotics were substantially removed (>90%) by GAC-associated biofilters and partially removed (≤20%) by sand alone and anthracite-sand biofilters. In particular, the GAC sandwich biofilter exhibited superior performance compared to sand/anthracite biofilter, and the comparisons among all biofilters indicated that both adsorption and biodegradation contributed to the removal of the target antibiotics in the GAC-associated biofilters. Adsorption kinetics showed that sulfamethoxazole fitted with pseudo-first-order adsorption model, while trimethoprim, amoxicillin, oxytetracycline and clarithromycin fitted the pseudo-second-order model. All antibiotics fitted the Langmuir model according to the isotherm experiment. To date, this is the first study evaluating the removal of antibiotics by GAC sandwich biofilters. Overall, this research will provide useful information which can be used for optimising or updating existing biofiltration processes in industry to reduce antibiotic residues from source water.