Conference Agenda

Hydraulic models form the basis for far-reaching investment and operational decisions as well as emission- or immission-based considerations. Model calibration and validation are essential to ensure adequate model quality and to assess model uncertainties, particularly in large and complex networks. This study presents the methodology for the calibration of Munich's extensive hydrodynamic sewer network model, comprising 2,400 km of sewers and serving an area of 18,000 hectares. The calibration process is based on a comprehensive 18-month measurement campaign, which included flow, water level, and rainfall data collected from more than 500 temporary and permanent monitoring stations. The calibration process identified and corrected several model inaccuracies, including errors in flow paths, connection elevations, operational settings and assumptions on infiltration water discharge. The results demonstrate the critical role of measurement data in refining hydraulic models for large, highly meshed networks. In a future workflow, quality measurements will be included in the calibration process. The project impressively shows that (all) models are more or less wrong if not calibrated to measurement data. How wrong can only be answered within the framework of a model calibration by proofed measured data and thus whether the model is suitable for the respective application.

Urban drainage models like SWMM and CityDrain rely on high-resolution pollution data to simulate dynamic processes accurately. Optical techniques, such as spectrometer probes and multi-spectral cameras, offer real-time monitoring solutions but face accuracy challenges due to simplified light scattering models. In this study we improve our understanding of light transport processes in wastewater by addressing challenges associated with absorbance and reflectance spectroscopy. Specifically, we developed a 2D stochastic model, incorporating main light-water interactions: specular reflection, absorption, and scattering. The model is grounded in diffusion theory and accounts for plane symmetry, providing a robust framework for simulating optical measurements. Key optical parameters of wastewater, including the absorption coefficient, scattering coefficient, and anisotropy factor, were determined using a double integrating sphere. Measurements focused on the 400–900 nm range due to limitations at longer wavelengths. Initial results demonstrate that the model replicates the reflectance behaviour of wastewater, with promising implications for improving optical monitoring technologies. Current work focuses on validating the model through experimental measurement of light reflectance attenuation. This approach lays the foundation for the design of advanced optical sensors and real-time monitoring systems for wastewater treatment applications.

The need to monitor urban water quality is increasing due to the ecotoxicological and health risks urban contaminants pose to water resources. Moreover, the list of contaminants of potential concern is increasing, leading to monitoring challenges due to various urban water matrices and sites to be analysed by a range of analytical methods. However, without efforts toward standardizing datasets and metadata, the collected network of datasets will remain fragmented in space, time and context. We suggest a first draft for a standard for metadata for urban water quality for discussion and review among the urban drainage community to increase the potential for datasets reuse. This standard will also support the efforts of early career researchers, who are often the ones collecting data. The proposed data format will allow to reconnect datasets to generate new knowledge, favour reanalysis and bridge the gap between hydroinformatics and cheminformatics.

Urban runoff threatens freshwater quality making it important to understand pollution sources and pathways. Measuring every pollutant in real-time is unfeasible, hence the need for in-situ grab sampling. Integrating real-time sensor data, rainfall forecasts and runoff models can optimize automated grab sampling beyond the current capabilities of autosamplers. This study evaluates two sampling strategies to optimize sampling times: (1) rainfall-volume based strategy which triggers sampling by accumulated rain volume (mm) or intensity (mm/h), (2) rainfall-runoff based strategy that triggers sampling by peak flow trigger (m³/s) estimated using a hydraulic model. Both strategies account for “no-event days”, the number of days with no runoff. Events of interest, here defined as first flush events resulting from heavy rains following consecutive days without rain, are identified. Different combinations of trigger parameters were tested, the sampling times from each strategy are evaluated based on whether the events of interest have been sampled, and the number of samples generated in those events. We highlight the potential of optimized sampling strategies in improving water quality monitoring. The ability to remotely trigger auto samplings, enables tailored measurements based on the specific data requirements or research questions.

The COVID-19 pandemic has intensified chlorine-based disinfection, elevating risks from residual chlorine and disinfection by-products (DBPs) in sewers. Using a pilot-scale sewer system with MS2 bacteriophage (SARS-CoV-2 surrogate), we stimulated wastewater collection and transportation process, and compared process-based (reaction kinetics) and data-driven models (random forest, decision tree, deep learning, general additive model, stacked model) under static (collection) and dynamic (transport) scenarios. The experimental results showed that the changes in residual chlorine and DBPs concentration in the dynamic scenario were more complex than in the static scenario, and higher residual chlorine dose didn’t accelerate the inactivation of the virus. According to analyses, data-driven models showed superior accuracy for residual chlorine prediction (R² +0.03) but poorer robustness for DBPs (MAE +0.35 vs. process-based), while process-based models exhibited smaller RMSE increases (2.91 vs. 5.31) when predicting DBPs versus chlorine, reflecting their adaptability to complex chlorine-organic matter interactions driving DBP formation. Uncertainty analysis revealed data-driven models’ sensitivity to high initial residual chlorine and DBPs doses. As the global situation is still rapidly evolving with a more frequent outbreak of epidemic events, our study provides a tool for estimating hazardous substances production caused by sterilization behavior for pandemic prevention.

The COVID-19 pandemic has demonstrated the critical role of Wastewater-Based Epidemiology in monitoring infectious disease at population level. This work presents a comparative analysis of WBE data from Austria and the USA, focusing on temporal and spatial features in the data and highlight key aspects. Data dispersion and variability attributes, as well as general wave development are displayed and analyzed. Methodological differences between Austria and the USA are highlighted, concerning population size referencing with biomarkers and public health monitoring strategies. With clinical case and hospitalization reporting largely discontinued in early 2024, WBE has emerged as a critical tool for monitoring the epidemic's dynamics. Viral RNA concentrations in wastewater provide insights into temporal patterns and their relationship to hospitalization admissions. Cross-correlation analyses reveal that wastewater viral RNA peaks consistently precede hospitalization admissions by 2 to 12 days, highlighting the effectiveness of WBE in providing early signals. Data variability is shown to be significantly influenced by the size of the monitored wastewater treatment plant. Smaller plants exhibited higher variability in viral RNA concentrations as compared to larger plants. By stratifying the data into five groups based on plant size, a clear trend of decreasing variability with increasing plant size was observed.