Session D5 - New technologies for performance monitoring  

Water Sensitive Urban Design (WSUD) is central to integrated urban water management worldwide, underscoring the need for reliable monitoring. Being economical and flexible for operation and communication, low-cost sensor systems show great potential to mainstream digital water management. Yet, the implications of using no data, low-quality data, or high-quality hydrologic data for WSUD assessment and model development remain unclear. This study used a hydrologic model (PCSWMM) to evaluate a WSUD system in Singapore by comparing the uncalibrated model, model calibrated with low-cost sensor data, and model calibrated with traditional sensor data under present and future climate. The uncalibrated model generally underestimated water depth, runoff, and outflow volumes, while overestimating runoff and outflow reduction by about 2%–6%. These biases are not universal and depend on site conditions and model configuration. The low-cost sensor performed comparably to the traditional sensor in model calibration and hydrologic performance assessment. Overall, the study highlights the value of low-cost sensors in providing empirical evidence and improving hydrologic modelling for WSUD evaluation under present and future climate.

Urban stormwater is a major source of surface water contamination, carrying a complex mixture of organic micropollutants, including persistent, mobile and toxic (PMT) and very persistent, very mobile (vPvM) substances. Monitoring the dissolved fraction of these compounds is challenging, with sampling often being a major limiting factor. This study aimed to evaluate the applicability of passive samplers for selected hydrophilic organic contaminants in stormwater and to validate the approach under field conditions. Two passive sampling receiving phases were compared in laboratory flume experiments under controlled conditions, including variable contaminant concentrations and intermittent dry periods. Laboratory results indicate that passive samplers can provide reliable concentrations comparable to automated sampling, with only slight overestimation for some substances. A field validation study at a combined sewer overflow is currently ongoing, where passive and automated sampling will be compared under real stormwater conditions. Overall, the findings suggest that passive samplers offer a robust and practical approach for monitoring hydrophilic organic contaminants in urban stormwater.

Effective microbial monitoring in stormwater-constructed wetlands is essential for water quality and public health protection. This study evaluated passive electronegative membranes against automatic composite samplers for detecting CrAssphage at the inlet and outlet of the stormwater constructed wetland during rainfall events. Logistic regression showed no significant difference in detection frequency (p>0.05), but deployment duration strongly influenced passive sampler performance. Longer deployments (48-96 h) improved detection and provided consistent quantitative signals. Quantile regression revealed a significant linear uptake with an estimated accumulation rate of 7.44 mL d⁻¹ (95% CI). Back-calculated time-weighted average concentrations (TWAC) indicated that longer deployments over than 48 h yielded reliable estimates of water-column concentrations. This study demonstrates that passive samplers offer a practical, time-integrated approach for microbial monitoring in stormwater wetlands and that back-calculation of TWACs is feasible, although further validation is needed.