Evaluation of the Removal Efficiency of Perfluoroalkyl Substances in Drinking Water

Abstract: Per- and polyfluoroalkyl substances (PFASs) are chemicals that have been used for over 50 years. They are both hydrophobic and hydrophilic, which make them useful in a wide range of products, both in the domestic and industrial market. Recently, the global attention on PFASs has increased due to their possible harmful health effects on humans. Furthermore, PFASs have been detected in drinking water sources all over the world. Conventional treatment processes in drinking water treatment plants (DWTPs) are not able to remove PFASs. Therefore, more research is required to find efficient removal techniques for these compounds. The aim of this study was to investigate the removal efficiency of PFASs using two different adsorption techniques, anion exchange (AE) with the resin Purolite A-600, and granular activated carbon (GAC) of type Filtrasorb®400. The experiments were performed in laboratory batch-scale, at Swedish University of Agriculture (SLU), and column tests in pilot-scale, at Bäcklösa DWTP in Uppsala. The PFASs showed a high sorption potential to AE and GAC. However, the removal efficiency differed depending on the perfluorocarbon chain length, functional group, and concentration level. For the AE, in average 92 % of the PFASs were removed in the end of the batch experiments while the average removal efficiency in the column experiment was 86 %. In the batch experiments treated with GAC on average 55 % of the PFASs were removed in the end of the experiments while the column experiment had the average removal efficiency of 86 %. There was an increase in the removal efficiency with increasing perfluorocarbon chain length in the column experiments. However, in the batch experiments, the adsorption of PFASs decreased with an increasing chain length, except for the highest PFAS concentration level (5000 ng L-1) treated with AE and the lowest PFAS concentration level (200 ng L-1) treated with GAC. In the column experiments, the perfluoroalkane sulfonates (PFSAs) were slightly better removed than perfluoroalkyl carboxylates (PFCAs) with an average removal efficiency of 97 % for AE and 91 % for GAC compared to 67 % and 82 % for AE and GAC, respectively. In the batch experiments, there was no clear trend between the removal efficiency and functional group. Overall, the pilot-scale experiments removed the PFASs relatively well even after 42 days (on average, 86 % for both AE and GAC). The lowest removal capacity in the column experiments was seen for the shorter chained PFSAs (in average 46 % for ≤C6 PFCAs using AE and 75 % for ≤C7 PFCAs using GAC). More efficient treatment techniques are needed to minimise the PFAS concentrations in drinking water and the potential human.