Tracing microplastics from raw water to drinking water treatment plants in Busan, South Korea.

The increasing amount of plastic waste has raised concerns about microplastics (MPs) in aquatic environments. MPs can be fragmented into nanoplastics that can pass through water treatment processes and into tap water; potentially threatening human health because of their high adsorption capacity for hazardous organic materials and their intrinsic toxicity. This case study investigates the identification, fate, and removal efficiency of MPs in Korean drinking water treatment plants. Two sites on the Nakdong River, two lake reservoirs (raw water sources), and four corresponding drinking water treatment plants were targeted to trace the amounts, types, and sizes of MPs throughout the treatment process. Monthly quantitative and qualitative analyses were conducted by chemical image mapping using micro-Fourier-transform infrared spectroscopy. MPs larger than 20 μm were detected, and their sizes and types were quantified using siMPle software. Overall, the number of MPs in the river sites (January to April and October to November) exceeded those in the reservoirs, but only slight differences in the number of MPs between rivers and lake reservoirs were detected from June to October. The annual average number of MPs in River A, B and Lack C and D was not distinctively different (2.65, 2.48, 2.46 and 1.87 particles/L, respectively). The majority of MPs found in raw waters were polyethylene (PE)/polypropylene (PP) (> 60%) and polyethylene terephthalate (PET)/poly(methyl methacrylate) (PMMA) (20%), in addition to polyamide (<10%) in the river and polystyrene (<10%) in the lake reservoirs. Approximately 70-80% of the MPs were removed by pre-ozonation/sedimentation; 81-88% of PE/PP was removed by this process. PET/PMMA was removed by filtration. Correlation of MPs with water quality parameters showed that the Mn concentration was moderately correlated with the MP abundance in rivers and lake reservoirs, excluding the lake with the lowest Mn concentration, while the total organic carbon was negatively correlated with the MP abundance in both rivers (A and B) and lake reservoir C.