With increasing focus on global clean water goals and stricter surface water quality legislation within the European Union, new strategies are needed to improve the performance of urban drainage systems. Wastewater treatment plants play a key role in nutrient removal for better surface water quality. However, the highly variable inflow from combined sewer systems--driven by both diurnal wastewater patterns and rainfall--can hinder optimal wastewater treatment plant operation, often resulting, amongst others, in elevated nitrogen and phosphorus levels in the wastewater treatment plant's effluent.

This study proposes a real-time control strategy that stabilises dry weather flow by actively using in-sewer storage, thereby minimising daily inflow fluctuations to the wastewater treatment plant. The method was applied to the Eindhoven wastewater treatment plant catchment, focusing on its two largest contributing catchments, Eindhoven and Riool Zuid. A simplified Storm Water Management Model model, based on an existing integral system model, was combined with a PySWMM-based control script and an empirical water quality model to estimate pollutant concentrations. Control decisions were based on current and forecasted rainfall, using both perfect (historic) and ensemble-based forecast data, determining four possible operational states: dry, transition, wet, and lingering wet weather.

This research demonstrates that dry weather flow control can improve daily wastewater treatment plant and surface water effluent peak concentrations without additional negative effects on the urban drainage system itself--for improvement in averages of effluent or surface water, additional treatment plant operational adjustments are required. The study also showcased the potential of real forecasts, with inherent uncertainty, in quantitative real-time control decision making. Lastly, the study underlines the potential of integrated, forecast-based control strategies to better align urban drainage and treatment plant operation, encouraging further development of a feedback-based optimisation framework.