From Prediction to Sustainability: AI for Smart Energy Management in Wastewater Treatment Plants

Abstract Accurate energy prediction is vital in optimizing operations and self-consumption and ensuring sustainability goals in wastewater treatment plants (WWTPs) and environmental applications. Self- consumption refers to the proportion of energy produced locally that is used on-site to meet energy demands. To anticipate energy generation and consumption, this paper compares and evaluates the performance of Machine Learning (ML) techniques for energy self-consumption, including long- term memory (LSTM), support vector machines (SVM), recurring neural networks (RNN), gated recurrent unit (GRU), and XGBoost, to forecast energy generation (EG) and energy consumption (EC) in WWTPs. The performance of each model is evaluated using metrics such as mean absolute error (MAE), mean absolute percentage error (MAPE), and root mean square error (RMSE) using a solid dataset of daily operating records. The findings show that GRU achieves the highest performance with RMSE of 0.102, MAE of 0.085, and R² of 0.978, followed by LSTM, GRU, and RNN, showcasing reliable temporal prediction capabilities, as well as these models driving energy efficiency and reducing operational costs in WWTPs. This paper highlights actionable insight into adopting ML for sustainable energy management in WWTPs, transforming energy forecasting, improving energy self-consumption, and establishing the framework for increasing WWTP’s operational efficiency and environmental sustainability.