Swift hydraulic models for real-time control applications in sewer networks

Area of Science:

• Environmental Engineering
• Wastewater Management
• Computational Fluid Dynamics

Background:

• Real-time control (RTC) is crucial for safe and economical sewer system operation, addressing issues like sewer overflow and sulfide mitigation.
• Accurate prediction of sewage hydraulics is essential for model-predictive control in sewer networks, but traditional models are computationally intensive.
• Long transport times and dynamic flow rates in sewage systems necessitate efficient predictive models for effective RTC.

Purpose of the Study:

• To develop swift, data-driven hydraulic models for predicting sewage flow rates and cross-sectional flow areas in gravity sewers.
• To reduce the computational burden associated with traditional hydraulic models for sewer network analysis.
• To integrate these novel models into an RTC strategy for cost-effective sulfide control in a simulated sewer network.

Main Methods:

• Development of two data-driven hydraulic models tailored for different sewer feed types (rising main and gravity main).
• Validation of model predictions against solutions of the Saint-Venant Equations for flow rate and cross-sectional area.
• Integration of the developed swift hydraulic models into a real-time control strategy for NaOH dosing.

Main Results:

• The data-driven models accurately predict sewage flow rate and cross-sectional flow area.
• Computational time is reduced by up to four orders of magnitude compared to traditional methods.
• The integrated RTC strategy achieved cost-effective control of sulfide in a simulated sewer network.

Conclusions:

• Swift, data-driven hydraulic models offer a computationally efficient alternative to traditional models for sewer network analysis.
• These models are suitable for real-time control applications, such as sulfide mitigation, in sewer systems.
• The developed models have the potential for broader application in various sewer RTC scenarios.