Wednesday Poster Session - Source control measures - Monitoring & modelling

This work presents the calibration and validation of a large-scale dripper-based rainfall simulator designed for experimental assessment of Sustainable Urban Drainage Systems (SUDS) by Canal Isabel II at Madrid. The infrastructure was developed to reproduce rainfall intensities between 5 and 70 mm/h with high stability and spatial consistency across a large test area (around 200 m2). Three flow-distribution meshes—differing in material and orifice size—were evaluated to determine the configuration providing the best hydraulic behavior. Calibration tests showed deviations below 3 mm/h for all intensities and Christiansen Uniformity Coefficients above 87%, confirming the robustness of the dripper-mesh system under controlled conditions. The best-performing mesh was subsequently installed for full-scale validation. Validation results demonstrated that the simulator reliably reproduces target intensities, with uniformities ranging from 75,6% to 83,2%, values consistent with expectations for large outdoor simulators where boundary and wind effects are more pronounced. Overall, the simulator provides sufficiently homogeneous and repeatable rainfall conditions for studies on runoff generation, pollutant mobilization and the hydraulic performance of SUDS under controlled and realistic scenarios. 

Standardization of particle size analysis (PSA) is crucial to ensure the proper blending of filter media used for stormwater treatment such as bioretention soil media (BSM). BSM typically contains of >80% sand and is amended with organic matter and fines (i.e., silt and clay) to support pollutant removal. However, there is currently no standardized PSA method to verify whether BSM meets design specifications before installation. This study compares three PSA methods—hydrometer, dry sieving with laser diffraction (DS + LD), and wet sieving with laser diffraction (WS + LD)—to evaluate their accuracy and repeatability for analyzing BSM. Twenty-seven BSM samples were collected from three bioretention cells in Ohio, USA. For the hydrometer method, substantial variability in results was observed depending on the BSM sample mass used for analysis. Triplicate BSM analyses by both DS + LD (83.9 ± 1.7% sand, 9.6 ± 3.0% silt, 6.5 ± 2.7% clay) and WS + LD (84.1 ± 1.7% sand, 10.6 ± 2.2% silt, 5.3 ± 1.7% clay) demonstrated high intra and inter-method agreement, especially for sand content. The hydrometer and WS + LD methods were the most consistent for measuring clay content. We recommend measuring the sand and fines fractions separately after appropriate dispersion when conducting PSA on sandy soils like BSM. Although PSA requires additional time and cost, ensuring the proper delivery of BSM which complies with design and regulatory specifications outweighs the costs of potential bioretention cell clogging and reconstruction. 

This study investigates how particle size distribution (PSD) varies across key stormwater treatment components using over 300 sediment and water samples collected from stormwater, baseflow, ponds, compact stormwater treatment systems (CSTS), bioretention forebays, and gully pots. Laser diffraction and sieving analyses show that stormwater and baseflow carry the finest particles, ponds retain a broad mix of fine–intermediate fractions, CSTS sedimentation units accumulate intermediate-sized sediments, and bioretention forebays and gully pots collect the coarsest material. PSD metrics (D10, D50, D90, Cu, Cc) clearly distinguish the retention behaviour of each treatment unit, demonstrating that system type is the primary driver of sediment characteristics. These findings highlight the limitations of relying solely on TSS measurements and underscore the value of routine PSD analysis for improving stormwater system design, performance evaluation, and maintenance planning. 

Urban watersheds, particularly in coastal areas, are increasingly vulnerable to intensified flooding driven by climate variability and the expansion of impervious surfaces. This study assesses the hydrological impacts of implementing linear parks as a nature-based approach to mitigating urban flooding in the Córrego Grande watershed, southern Brazil. A coupled 1D–2D hydrological–hydrodynamic model was developed in PCSWMM, calibrated and validated using observed events, and achieved satisfactory performance (R > 0.70, RMSE < 1.3, NSE > 0.3, ISE ≤ 6). Four scenarios were simulated, varying in the width of vegetated riparian corridors. The linear parks produced modest reductions in total runoff volume but substantially attenuated peak flows, with reductions of 19.7 to 22.7 per cent in the first peak and 11.4 to 13.8 per cent in the second. Importantly, the watershed became capable of responding to a 100-year rainfall event with hydrological behaviour comparable to that of a 50-year event in the pre-intervention condition, demonstrating a marked increase in resilience. Overall, the findings show that linear parks enhance hydrological performance and constitute a transferable nature-based strategy for sustainable urban water management and climate adaptation. 

The study evaluated the capability of the Storm Water Management Model (SWMM) to simulate, on an event-by-event basis, the behavior of the storage layer of a bioretention cell, using parameters obtained from the characterization of the materials composing its filling. Based on the analysis of six rainfall events, the initial simulations with standard parameters inadequately represented the drainage dynamics of the structure. From this, it was hypothesized that preferential flow paths were not being captured by the model’s standard configuration and that the discharge representation might also have been impaired by the absence of evapotranspiration data. To overcome this representational limitation, a fictitious drain was introduced to represent these processes contributing to the structure’s emptying, which significantly improved model accuracy, increasing the mean Nash–Sutcliffe Efficiency (NSE) from 0.32 to 0.90 during calibration and from 0.32 to 0.76 during validation. 

Biofiltration treatment systems facilitate interactions between plants, soil media, and microbes to treat nitrogen-rich waters like stormwater and greywater. Novel nitrogen removal processes such as anammox-assisted Simultaneous Nitrification, Anammox, and Denitrification (SNAD) can be effective in directly converting ammonia to nitrogen, but these have not been adequately researched within the biofilters. This research aims to study the role of biofilter media amendments that would influence the treatment of high strength real combined sewer overflows (CSOs). For this study, conventional sand biofilters are amended with zeolite, scoria, engineered media, biochar and coconut coir and a long term (40 week) evaluation under varying operational conditions was conducted. It is found that bioretention systems with media exhibiting higher specific surface area, porosity, and cation exchange capacity had enhanced nitrogen removal. Media properties such as high porous surface area and ammonia buffering capacity further influenced the presence of nitrogen-transforming microbial communities. The study will be significant in providing engineered solutions and suggesting design modifications to treat uncontrolled CSOs. 

Urban stormwater runoff from roadways contains contaminants in both dissolved and particulate forms, necessitating effective depollution strategies that account for the behavior of both phases. Many pollutants possess significant mass fractions bound to particulates, exceeding 50% and, in some cases, nearing complete particle association, underscoring the necessity of accurately modeling particulate transport and retention. Particle size is critical to filtration performance, as larger particles are typically removed near the surface of biofilters. In contrast, smaller particles may penetrate deeper into the media or bypass filtration altogether. The objective of this study is to enhance an existing model of micropollutant treatment in stormwater biofilters (MPiRe) by incorporating a physical filtration process for suspended solids. Currently, MPiRe consists of a hydrological module and a reactive pollutant transport module but addresses only dissolved-phase processes. The proposed filtration module simulates depth-dependent particle removal, enabling the model to capture particulate-bound pollutant removal mechanisms that are currently missing from the modeling framework. 

This study presents the first-year monitoring results of a Sustainable Urban Drainage Systems (SUDS) pilot facility implemented in northern Madrid to inform stormwater reuse strategies within the forthcoming Madrid Nuevo Norte urban development. A permeable pavement area and a bioretention zone were monitored alongside a conventional gully used as control. Twelve rainfall events were analyzed through water-quality monitoring, continuous hydraulic measurements, and soil characterization. Results show that both SUDS significantly attenuate peak flows and enhance stormwater quality through filtration, sorption, and biogeochemical interactions. Most parameters fall within reference values drawn from relevant Spanish regulatory directives which, although not legally binding for this type of system, provide an informative benchmark for assessing reuse suitability. This suggests that, with appropriate post-treatment, SUDS-treated runoff could constitute a non-potable water resource viable for reuse. The findings offer evidence-based guidance for integrating decentralized stormwater capture and treatment into new urban districts, supporting Madrid’s transition toward a circular and resource-oriented approach to sustainable rainfall management. 

Regular performance assessment of blue–green infrastructure (BGI) is critical for sustainable urban stormwater management, yet it remains difficult due to the complexity of urban settings, variability across BGI systems, and limited financial and staffing resources. Traditional approaches for estimating underperformance rely on physics-based hydrological models, which either oversimplify real-world conditions or require extensive data and computational effort. This study introduces an integrated, data-driven framework for evaluating BGI performance risk by linking two predictive components: (1) machine-learning models that forecast BGI inspection scores, and (2) models that estimate ponding duration following storm events. Using a comprehensive dataset from the City of Philadelphia, we develop ML models that predict inspection scores based on historical inspections and local environmental and sociodemographic variables. In parallel, we leverage a unique set of observed hydrologic data from multiple BGI sites—combined with weather inputs—to predict ponding time as a direct indicator of system performance. By coupling these two models, we generate event-specific forecasts of underperformance risk for BGI across the city. The ability to predict ponding duration using readily available data represents an important advancement toward proactive, scalable, and sustainable management of urban BGI systems.