Advancing Wastewater Treatment Efficiency: Empowering Data-Driven Modeling with Domain Knowledge Integration

In Canada, ensuring clean and safe water while promoting environmental sustainability is a top priority. While wastewater treatment plants (WWTPs) has been playing a crucial role in public health and environmental protection, but their operational efficiency still needs to be improved. To this end, this research project focuses on harnessing the power of data-driven modeling (DDM) and the wisdom of domain experts to enhance WWTP operations. Current WWTPs often operate with excess caution, consuming substantial energy, space and resources. We aim to change that by leveraging data from various sources, including SCADA systems, water quality measurements, and regulatory data, to predict and optimize WWTP performance. Specifically, we’re exploring the integration of domain knowledge with DDM to make predictions more accurate and models more interpretable.
Our research, supported by comprehensive datasets spanning four years, aims to predict critical parameters, such as total phosphorus concentrations, essential for meeting water quality regulations. By embracing DDM and domain knowledge, we aim to make WWTPs more efficient, eco-friendly, and adaptable. This research aligns with federal sustainability strategies and supports the Ontario Clean Water Agency in achieving optimal performance while addressing environmental concerns.

Faculty Supervisor:

Usman Khan;Satinder Brar;Stephanie Gora

Student:

Partner:

Ontario Clean Water Agency

Discipline:

Engineering

Sector:

Construction and infrastructure; Utilities

University:

York University

Program: