Search for a command to run...
• Mobility data from mobile phones can provide dynamic population estimates. • Seasonal and weekly population variations were easy to quantify • Mobility data enable dynamic load time series for model applications. • Person loads with lowered variance reduced the uncertainty in model-based design. • Nitrogen and phosphorous loads showed stronger correlation with population than BOD. Model-based design is an emerging tool for dealing with the uncertain dynamic loads entering wastewater treatment plants (WWTPs). But our understanding about the load-driving population-dynamics is limited. Therefore, we studied if mobility data (mobile telecommunications data) could be used to reduce uncertainties during design. Mobility data from Uppsala, Sweden between 2019–2022 clearly quantified population movement patterns that were useful for simulating load scenarios such as seasonal load-shifts, without data gaps from irregular influent sampling. Further, they showed fair correlations with the daily influent nitrogen load (R 2 = 0.49), which resulted in a more precise person load estimate than assuming a static population (23 % reduced variance). Unfortunately, BOD load variations showed little correlation with the population variations (R 2 = 0.21). Nevertheless, model-based reactor sizing based on mobility data successfully reduced the de-/nitrification volume safety factor with 5 %, which demonstrates their practical usefulness for WWTP design.