Session B4 - At source SCMs: spatial prioritisation

Master plan for stormwater management for the 35 municipalities in the metropolitan area, based on a qualitative and quantitative assessment of the system. This is based on an analysis of land use in urban areas and hydraulic modelling of the networks, which subsequently enabled the development and testing of development proposals to solve overflow problems and mitigate the risks of pollution of the receiving environment. The specificity of this master plan lies in the analysis of pollution flows into the receiving environment and the impact of two types of management strategies: upstream solutions (integrated water management / nature-based solutions) versus transfer and storage solutions (civil engineering). A tool has been created to assess the pollution flows generated by urban stormwater runoff and discharged into the natural environment, considering climate change and different urban configurations. In a context of strong urban dynamics, these quantified elements on hydraulic and pollution aspects enable the water utility to make informed decisions on its overall stormwater management strategy.

Combined Sewer Overflows (CSOs) deteriorate waterbodies. Furthermore, due to extreme events related to climate change, water discharges will even increase. The implementation of stormwater disconnection solutions helps to mitigate discharges, but their effectiveness depends on where they are conducted. This article aims to study the relevance of Machine Learning approaches to identify which factors related to the urban sub-catchments operating mode explain CSOs. One unsupervised learning method has been evaluated on precipitations to distinguish them according to their properties. One supervised learning method – decision tree – has then been applied on a general database created using the hydrological modelling of the Écully catchment. The first results, provided with an accuracy larger than 80%, show the role of precipitations data. They also indicate the first sectors to disconnect and the hydrological parameters influencing CSOs, such as Fimp or the lag time K. These processes need to be applied on more diversified data so that a tool for disconnecting stormwater using Machine Learning methods can be further developed.

Guidance on how to address hydromodification impacts on urban streams is only available in a limited number of jurisdictions. In the case of the Nose Creek watershed in Calgary, Alberta, Canada, the combination of its continental, semi-arid climate and tight glacial tills means that it has an evapotranspirative hydrologic regime. A relationship between the widening of the creek and the increase of the average annual runoff volumes was reflected in staged runoff volume targets in the 2008 Water Management Plan. However, the recognition that the resulting discharges exceeded the morphologically most sensitive flow window led to the adoption of flow duration curves with the premise that increases outside of the most sensitive flow window might be considered. Using PCSWMM, matching the target flow duration curve is accomplished using a water balance approach. First, the overall volume under the curves is “matched” through the implementation of landscape-based LID practices. Next, the shape of the resulting discharge curve is finetuned using a series of flow control provisions within the outlet control structure of the end-of-pipe wet ponds. The use of a dual-layer approach offered by the Bioretention LID Control within PCSWMM allows for the representation of Calgary’s evapotranspirative regime. While the adoption of this approach is a step forward in representing the urban water balance and thus reducing impacts on hydromodification considerably more effort is needed as the LID Controls in SWMM are not optimal for representing the urban hydrology of the Canadian prairies.

The Water Framework Directive (2000/60/EC) promotes an integrated approach to water management, treating water as a shared resource and defining both ecological and physicochemical quality targets. In this context, Sustainable Drainage Systems (SuDS) offer practical solutions to limit urban stormwater runoff, improve water quality and reduce urban flood risk. This study focuses on Sesto Ulteriano, a 290-ha industrial area in San Giuliano Milanese (Metropolitan City of Milan), where the drainage system, connected to artificial channels, is undersized and subject to frequent CSOs. Two hydrologic-hydraulic models were developed in SWMM5: one representing current conditions of the drainage network and one including SuDS interventions distributed across 24 ha. The models were run using a full year of rainfall recorded at a nearby gauge, with outputs evaluated at event scale. For each precipitation event, total overflow volume, maximum peak discharge and the number of CSO activations were extracted and compared between the two scenarios to evaluate SuDS performance. SuDS improved system behaviour, reducing overflow volumes and peaks by 31% and 42%, respectively, and lowering CSO activations by 13%.