Pollution characteristics and spatial distributions of perfluoroalkyl substances (PFASs) in surface water samples were investigated in Taihu Lake in 2010 and 2019, respectively. A hybrid model based on the positive matrix factorization (PMF) and the health risk assessment model were employed for quantifying the contributions of sources to PFASs concentrations and the source risks. The method contained two stages:1 the sources of PFASs were apportioned using the PMF model, and 2 the contribution of health risks from each source was quantitively estimated. Three factors (source categories) were extracted using PMF, including:coating industry sources, textile and electroplating sources, and fluoride-processing industry sources. Their contributions to PFASs concentration were 29.59%, 25.68%, and 44.72% for 2010 and 67.69%, 10.26%, 22.05%, for 2019, respectively. The health risk of PFASs in the water assessed by the health risk assessment (HRA) model was 4.56E-07 for 2010 and 2.69E-07 for 2019, which was lower than 1E-06. The source contributions to health risks estimated by the PMF-HRA hybrid model were:64.86% (2010) and 92.48% (2019) for textile and electroplating sources, 31.30% (2010) and 5.04% (2019) for coating industry sources, and 3.84% (2010) and 2.48% (2019) for fluoride-processing industry sources. For the two years, the concentrations of PFOA and PFOS were reduced significantly, indicating the effective control of their emissions during the past ten years in Taihu Lake. However, it was also shown that the concentrations of PFBS and PFHxS were increased. These findings suggest that the above short chain-PFCs species should be the focus for further control and management, and their health risks should be studied in future research.
Keywords: PMF model; Taihu Lake; health risk; perfluoroalkyl substances(PFASs); source apportionment.