This study examines the removal of microplastics and other anthropogenic particles (>10 μm) from surface water by a full-scale conventional drinking water treatment plant. The treatment process is composed of coagulation with aluminum hydroxide, flocculation, anthracite-sand filtration, and chlorination. Samples were also collected from pilot-scale biological filters consisting of anthracite-sand or granular activated carbon (GAC) media operated with or without pre-ozonation and at a range of different empty-bed contact times (EBCTs). Particles in 10 L water samples collected in duplicate using a fully enclosed sampling apparatus were separated using sieves with 500 μm, 300 μm, 125 μm, and 45 μm openings followed by filtration through 10 μm polycarbonate filters. Particles were counted using stereomicroscopy and characterized using μ-Raman spectroscopy. Full-scale conventional treatment removed 52 % of anthropogenic particles when comparing raw (42 ± 18 particles/L) and finished water (20 ± 8 particles/L). Coagulation, flocculation, and sedimentation accounted for the highest removal (70 %) of any individual unit process. Overall removal was reduced to 52 %, the difference being attributed to airborne particle deposition that occurred while water was detained in a clearwell (exposed to atmosphere via ventilation) that was used to achieve the required contact time for disinfection. The majority of the particles (>80 %) were identified as fibers 10-45 μm; microplastics were predominantly composed of polyester while the non-plastic anthropogenic particles were primarily cellulose. None of the pilot filter configurations examined resulted in significantly fewer microplastics when compared to full-scale conventional filtration. This study illustrates that the removal efficiency of conventional treatment may be limited when considering microfibers <45 μm in size.
Keywords: Anthropogenic particles; Biofiltration; Clearwell; Distribution system; Full-scale; Ozone.
Copyright © 2021 Elsevier Ltd. All rights reserved.