Removal of poly- and perfluoroalkyl substances by North American water treatment practices, The

Abstract

Poly- and perfluoroalkyl substances (PFASs) are a group of environmentally persistent man-made chemicals that are being detected in water sources all over the world including surface waters, ground waters, tap waters, bottled water, and municipal wastewater influents and effluents. These chemicals are used as surfactants in a wide variety of commercial and industrial products, creating multiple pathways of exposure to human beings. Numerous studies, both epidemiological and laboratory-based animal studies, have been done to determine the toxicological effects of PFASs and have found correlations between these chemicals and an assortment of adverse effects. One of the most concerning pathways is exposure via contaminated tap waters resulting from the inability of conventional water and wastewater treatment systems to remove these chemicals. There were two main objectives for this project. The first was to measure the occurrence levels in raw water sources for 20 drinking water treatment utilities across the U.S., and evaluate the efficacy of various treatment processes in their removal of an extensive suite of PFASs including perfluorocarboxylic acids (PFCAs), perfluorosulfonic acids (PFSAs), and polyfluoroalkyl chemicals (i.e., PFCA and PFSA precursors). Detected concentrations of the stable end product perfluoroalkyl acids (PFCAs and PFSAs) for all samples collected were in the ng/L range for all utilities in this study with the exception of one, which had levels in the low micron g/L range. While the precursor chemicals FOSA, N-MeFOSAA, and N-EtFOSAA were detected in the low ng/L range in some surface waters and treated wastewater effluents, none of the precursor chemicals examined in this study were measured above reporting levels in ground water. More importantly, conventional water treatment techniques such as ferric or alum coagulation, granular/micro-/ultra- filtration, aeration, oxidation (i.e. permanganate), and disinfection (i.e., ozonation, chlorine dioxide, chlorination, and chloramination) were mostly ineffective in removing PFASs from drinking water. In many cases, the concentration of PFCAs and PFSAs were actually slightly higher following oxidative treatments, suggesting some potential formation of these chemicals from yet unidentified precursors. Advanced treatment technologies, such as anion exchange and granular activated carbon, demonstrated removal of PFASs under some operational conditions. In contrast, reverse osmosis consistently demonstrated significant removal of PFASs from contaminated raw water sources at full-scale drinking water treatment plants. The second objective was to evaluate two forms of advanced treatments at the bench scale including GAC and nanofiltration (NF). Virgin NF270 flat sheet membranes were tested at pressures ranging from 25 to 125 psi and using spiked deionized (DI) water and spiked artificial ground water (AGW). The effects of membrane fouling by humic acid in AGW was also tested under constant permeate flux conditions. The NF270 membranes, both virgin and fouled, demonstrated greater than 93% removal for all perfluoroalkyl acids under all conditions tested. GAC efficacy was tested using rapid small scale columns packed with Calgon Filtrasorb[RTM] 00 (F300) carbon and DI water with and without dissolved organic matter (DOM). DOM effects were also evaluated with F600 and Siemens AquaCarb[RTM]1240C with spiked and filtered natural river water. The F300 GAC had less than 20% breakthrough of all chemicals for the entirety of the spiked DI water experiment (125,000 bed volumes (BVs)). A dramatic effect was observed on the carbons when DOM was present, with greater than 20% breakthrough of all PFAAs by 10,000 BVs. PFASs are being detected in finished tap waters throughout the U.S., making it one pathway to human exposure because conventional water treatment practices are not effective in removing the chemicals. More advanced treatment techniques, such as AIX and GAC, have the ability to remove these chemicals with varying degrees of success depending on the life of the media and the specific chemicals. Membrane processes, such as NF and RO, have proven to be the most effective methods of treatment. Further research needs to be performed on these less employed techniques as well as the toxicological effects of these chemicals to determine if the cost of upgrading utilities is worth the risk of exposure.