An adaptive real-time grey-box model for advanced control and operations in WRRFs

Area of Science:

• Environmental Engineering
• Process Control
• Wastewater Treatment

Background:

• Grey-box models are increasingly vital in wastewater treatment for combining mechanistic understanding with data-driven insights.
• There is a need for adaptive, intuitive grey-box models to optimize and control wastewater processes, particularly those with time-varying dynamics.

Purpose of the Study:

• To identify and utilize a grey-box model structure with an Extended Kalman Filter (EKF) for real-time estimation in a Modified Ludzack-Ettinger (MLE) process.
• To accurately estimate ammonia concentrations and track nitrification capacity within the MLE process.

Main Methods:

• Developed and implemented an Extended Kalman Filter (EKF) in Python, integrated with the SUMO (Dynamita™) process simulator.
• Employed a grey-box model structure designed for adaptive parameter estimation to capture time-varying system dynamics.

Main Results:

• The EKF accurately estimated ammonia concentrations across multiple tanks, even with limited input data.
• Real-time tracking of the system's nitrification capacity was achieved, offering valuable operational insights.

Conclusions:

• The developed EKF-based grey-box model provides an effective tool for real-time monitoring and operational guidance in wastewater treatment.
• This approach is crucial for advancing future control strategies, such as model predictive control, in wastewater management.