Inter-Storm Variation in Microplastic Occurrence and Composition in Urban Stormwater

Microplastics (MP) are a commonly reported pollutant in the freshwater, marine, and soil environment. Few studies to date have reported MP concentrations and polymer types observed in stormwater, particularly not for catchments with separate storm sewers.
Objectives
The objectives of this study were to determine the microplastic concentration, polymer fingerprints, and the inter-storm variation of MP in stormwater outfalls and in green infrastructure.
Status
To date, thirteen storm events from five sites have been sampled. All particles from 500-2000 μm from three storms (Bailey et al., 2020), all particles 250-200μm (Boni et al., in press) from nine storms, and 20-80 particles per sample from 63-250μm from eight storms have been analyzed (Parmar et al., in prep). Analysis will continue as will stormwater sampling this year.
Methodology
Composite stormwater samples (~5-10L) were collected for thirteen storm events from five urban areas across rainfall events of varying size. Sampling sites include urban runoff near a busy local highway (City N), bioretention basin next to a parking lot (BBR), stormwater outfalls near artificial turf fields (field P), and urban combined sewer runoff. Particles 63 to 2000 μm were collected on sieves, separated into three size classes, treated with a wet peroxide oxidation, density separated with NaCl, and buoyant particles (fragments, films, and spheres) were collected for analysis with attenuated total reflectance Fourier transform infrared spectroscopy (ATR-FTIR) or microscope FTIR for chemical composition. For the smallest size class, MP concentrations were estimated by multiplying the number of synthetic polymers observed by the total number of particles divided by the number of particles analyzed and normalizing to the volume of water sampled: MP composition was determined through spectral library searches and manual interpretation by a chemist. Statistical comparisons between sampling sites and dates were performed on microplastic concentration and polymer profile data.
Findings
#Microplastics were observed in all stormwater samples. Significant differences were observed in the total polymer concentrations between the sampling sites in the largest size range (250-2000μm), potentially due to differences in land use within the catchments sampled (Fig. 1). The smallest microplastics were observed at the highest concentrations based on our preliminary results.
Several different polymer types were observed, both manmade and natural in origin in all size ranges. For the 250-2000μm particles, manmade polymers observed included polyethylene (PE), polystyrene (PS), polypropylene, copolymer of ethylene-ethyl acrylate, polyethylene terephthalate, and acrylonitrile styrene-butadiene (Fig. 2). Of the 196 particles analyzed, 44 particles (22%) were identified as MP. The most prevalent polymer type in the 63-250 μm size range varied site-to-site and storm-to-storm for a given site. For example, PS was 55% of the particles analyzed for the first storm sampled at City N with no PE observed, whereas PE was prevalent in other samples. Preliminary results may indicate that a greater diversity of MP was observed in the smaller size particles analyzed here compared to the larger particles previously analyzed which for 5/9 samples had over 50% of the MP polymers composed of PS, PP, PE, and PETE (Boni et al., in press). Future work will include statistical comparison of the polymer fingerprints by site, size class, and storm via dissimilarity matrix-based techniques (e.g., nonmetric multidimension scaling, nMDS) to understand the relative importance of these factors.
Natural polymers observed included natural fibers (e.g., cotton), plant matter, and cellulose. It is not surprising that the oxidation step did not remove these natural polymers. We observed leaf litter even in water samples collected for another project that were treated with the cellulose digestion recommended in the current ASTM method.
A weak but significant negative correlation was observed between cumulative rainfall and MP concentrations in the largest size range, but no correlation was observed between antecedent dry days and MP concentrations. These analyses will be repeated including the smaller particle sizes when quantitative data is available.
Significance of the investigation or study
Overall, these results presented provide insight into the loading and character (size, morphology, polymer type) of buoyant MP particles in stormwater that may be useful in designing mitigation strategies.