Input characterization of perfluoroalkyl substances in wastewater treatment plants: source discrimination by exploratory data analysis

Abstract

This paper presents a methodology based on multivariate data analysis for identifying input sources of perfluoroalkyl substances (PFASs) detected in 37 wastewater treatment plants (WWTPs) across more than 40 cities in the state of Minnesota (USA). Exploratory analysis of data points has been carried out by unsupervised pattern recognition (cluster analysis), correlation analysis, ANOVA and per capita discharges in an attempt to discriminate sources of PFASs in WWTPs. Robust cluster solutions grouped the database according to the different PFAS profiles in WWTP influent. Significantly elevated levels of perfluorohexanoic acid (PFHxA), perfluorooctanoic acid (PFOA) and/or perfluorooctane sulfonate (PFOS) in influent have been found in 18 out of 37 WWTPs (49%). A substantial increase in the concentrations of PFHxA and/or PFOA from influent to effluent was observed in 59% of the WWTPs surveyed, suggestive of high concentration inputs of precursors. The fate of one precursor (8:2 fluorotelomer alcohol) in WWTP was modeled based on fugacity analysis to understand the increasing effluent concentration. Furthermore, population-related emissions cannot wholly explain the occurrence and levels of PFASs in WWTPs. Unusually high influent levels of PFASs were observed in WWTPs located in specific industrial areas or where known contamination had taken place. Despite the restriction on the production/use of PFOA and PFOS, this paper demonstrates that wastewater from industrial activities is still a principal determinant of PFAS pollution in urban watersheds.