Author(s): Fabiana Diniz Guimaraes; Tatiane Souza Rodrigues Pereira; Klebber Formiga
Linked Author(s):
Keywords: HEC-RAS; Flood mapping; LiDAR; Copernicus DEM; Digital elevation models; Urban flooding
Abstract: Urban wastewater pumping stations consume significant energy and are increasingly subject to complex, time-varying electricity prices. Many facilities still use ad hoc rules that neglect the tariff structure and the unpredictability of inflows, resulting in unnecessary expenses and the risk of overflow. This project creates and tests a multi-scale optimisation framework for level-based ON/OFF management in a large wastewater pumping station with dynamic tariffs. The study analyses 6 months of 5-minute SCADA data from the Águas Claras wastewater pumping station in Brasília, Brazil, whose seasonal fluctuation is caused by wet-weather input and infiltration. Reconstructing inlet flow via mass balancing utilising a two-stage wet-well design and reconciled SCADA data provides physically consistent forcing time series for simulation and optimisation. A multi-objective NSGA-II formulation finds simple hysteresis rules that minimise energy cost and overflow volume over four calibration horizons (daily average, daily maximum, weekly, and monthly). Validated non-overflow solutions undergo deterministic sensitivity and robustness study over six months, with ±20% perturbations of key parameters (ON/OFF levels, pumping rate, efficiency, tariff, and inflow). Optimised level rules lower energy expenditures by 7–13%, pump starts by 52–76%, and overflow without hardware changes. Monthly calibration horizons produce more solid rules, while daily optimisation around extreme inflows maximises savings but is more susceptible to inflow uncertainty. The SCADA-ready framework can be integrated into digitalised urban wastewater system PLC-based control schemes.
Year: 2026