Towards a model for road runoff infiltration management

Water resource preservation is a critical concern, and road runoff, rainwater leaching from roads, presents a significant challenge due to its micropollutant load. Road runoff carries various pollutants, including fuel, oils, brake fluids, tire particles, de-icing agents, heavy metals, polycyclic aromatic hydrocarbons (PAHs), salts, and microplastics. Phytosanitary products used in road management also contribute to groundwater contamination. While collection and storage methods exist (underground infiltration, porous pavements, rain gardens, bioswales, wetlands), infiltration is a promising approach for groundwater recharge and flow reduction. However, concerns exist regarding soil and groundwater contamination, necessitating risk-benefit evaluations. Infiltration ponds play a filtration and adsorption role, but knowledge on how soil physico-chemical properties affect micropollutant trapping remains limited, with research primarily focusing on specific pollutant families. This study aimed to measure micropollutant abundance in a road runoff treatment facility, monitor their distribution in sedimentation and infiltration ponds, and correlate micropollutant abundance with soil physico-chemical properties to develop a model for effective road runoff management.

Previous studies have highlighted the presence of various pollutants in road runoff, including PAHs, metals, and chlorides. However, research using broader pollutant lists or non-targeted approaches has revealed the presence of unexpected substances like drugs and pesticides. Existing literature lacks comprehensive understanding of how soil properties influence micropollutant trapping during ground infiltration. While some studies have shown promising results for reducing environmental impact and increasing aquifer recharge, these often focus on limited pollutant shortlists. Therefore, a more holistic approach considering the diverse range of micropollutants and the complex interplay with soil properties is necessary.

The study utilized a road runoff treatment facility consisting of a sedimentation pond and an infiltration pond. Water, sediment, and soil samples were collected from the facility and analyzed using mass spectrometry coupled with gas chromatography and liquid chromatography to identify and quantify 2406 micropollutants. Soil physico-chemical properties (density, texture, pH, granulometry, water content, organic matter, porosity) were also measured at various depths in the infiltration pond. Chemical enrichment analysis, using the ChemRICH tool, was employed to cluster micropollutants based on structural similarity. Spearman's correlation analysis was used to investigate the relationships between micropollutant abundance and soil properties. Statistical analyses (Wilcoxon's rank-sum test) were performed in Metaboscape 4.0 (Bruker) using peak areas as the reference unit. Metal and chloride concentrations were analyzed using standard methods by EUROFINS. Detailed extraction methods for micropollutants from water and soil are described in the paper, along with specifics of LC-HRMS and GC-MS/MS analysis and data analysis procedures.

The sedimentation pond effectively trapped a significant number (610) and abundance (79%, reaching 91-98% for PAHs) of micropollutants, primarily in the sediment. The infiltration pond's surface layer (0-10 cm) retained a substantial number and abundance (86%) of micropollutants, significantly reducing their presence in deeper layers. Chemical enrichment analysis revealed six consistently impacted clusters regardless of the pairwise comparison: aniline compounds, organothiophosphorus compounds, benzene derivatives, pyrethrins, benzopyrenes, and fluorenes. Correlation analysis showed consistent positive or negative trends across all micropollutant families, irrespective of hydrophobicity. Specifically, higher sand proportions with large particle sizes (>100 µm) were positively correlated with micropollutant abundance, while silt and small particles (<20 µm) showed negative correlations. Organic matter content and soil density were also positively correlated, while water content and pH were negatively correlated. Root systems in the top 10 cm of the infiltration pond contributed significantly (10.9-35.5% of soil dry weight) to the overall soil composition.

The results demonstrate the crucial roles of both the sedimentation and infiltration ponds in preventing micropollutant spread. The sedimentation pond acts as a highly effective primary barrier, trapping a large fraction of pollutants, thereby preventing their transfer to the infiltration pond and ultimately the environment. The surface layer of the infiltration pond further demonstrates significant retention capabilities, limiting the infiltration of micropollutants to deeper soil layers. The observed correlations between micropollutant abundance and soil properties highlight the importance of soil composition in pollutant trapping. The strong positive correlation between sand content, particularly larger particles, and micropollutant retention suggests that increasing the sand proportion in infiltration ponds could significantly enhance their effectiveness. This is potentially attributable to both adsorption and biofilm formation. The role of soil biodiversity and plant root systems should also be further investigated as they contribute to pollutant trapping and biodegradation, potentially increasing the overall effectiveness of the system.

This study provides a robust model for predicting micropollutant abundance in road runoff infiltration ponds based on key soil physico-chemical properties. Increasing the proportion of sand with large particles (>100 µm) is identified as a simple and cost-effective strategy to improve micropollutant retention. Future research should focus on further investigating the role of soil biodiversity and plant root systems, and optimizing infiltration pond design based on this model to improve efficiency and sustainability.

The study was conducted at a single site, limiting the generalizability of the findings to other locations with potentially different soil types and traffic conditions. The analysis focused on a specific set of micropollutants, and other emerging contaminants may exhibit different behavior. The relatively small sample size for the correlation analysis may influence the strength of some correlations. A more comprehensive investigation into the interactions between soil organisms and micropollutants is also recommended.