Impacts des épisodes orageux importants: comment prendre en compte la variabilité temporelle des rejets non traités dans un système d'assainissement urbain?

[Departement_IRSTEA]Ecotechnologies [TR1_IRSTEA]INSPIREInternational audienceIn typical life cycle impact assessment (LCIA) studies of urban wastewater systems (UWS), average conditions are modelled but there are many annual flooding events with releases of untreated sewage. Such peak conditions are not considered and present a high temporal variability which is not currently accounted for. In addition, the aggregation of the loads from several storm events could bring an issue for the impact assessment on the aquatic categories of eutrophication and ecotoxicity. Hence we are investigating the contributions of these wet weather-induced discharges along with the inclusion of temporal variability in the life cycle inventory (LCI) for UWS. In the framework of the OPUR research programme (Observatory of Urban Pollutants) and in collaboration with the Paris public sanitation service (SIAAP), this work aimed at identifying and describing contributing flows from the UWS in the Paris area by a selection of routine wastewater parameters and priority pollutants. This collected data is organized according to archetypical weather days over a 24-hr span. Secondly, for each archetypical weather days and its associated flows to the receiving river water (Seine) the parameters of pollutant loads (statistical distribution of concentrations and volumes) are determined with statistical treatment of data. Then, the inventory flows (i.e. the potential loads from the UWS) can be used as inputs in a classical LCA to investigate the relative importance of episodic wet weather versus "continuous" dry weather loads coupled to further uncertainty analysis using a Monte Carlo method. Results analysis showed that a few severe events can be important contributors to the total annual pollutant load on some parameters (routine wastewater pollutants but also priority pollutants). The proposed method based on the definition and characterization of archetypical weather days has shown the appropriate level of temporal differentiation in the LCI to assess the impacts from unmanaged pollutant loads from UWS during intense storm events. With such significant contributions of pollutant loads at the LCIA level, further research is required to include temporally-differentiated emissions in the methodological framework of the aquatic categories of eutrophication and ecotoxicity, to better understand how the performance of an UWS system affects the receiving environment for given local weather conditions