Modelling the biological treatment process aeration efficiency: application of the artificial neural network...

Area of Science:

• Environmental Engineering
• Wastewater Treatment Technologies
• Artificial Intelligence in Environmental Science

Background:

• Biological treatment processes (BTPs) are crucial for removing chemical oxygen demand (COD) and ammonia from wastewater.
• High energy consumption in BTPs is primarily due to inefficient oxygen transfer, stemming from low oxygen solubility and aeration efficiency (AE).

Purpose of the Study:

• To develop a predictive model for monitoring and improving aeration efficiency (AE) in biological treatment processes (BTPs).
• To identify key variables influencing AE for enhanced process control and energy reduction.

Main Methods:

• Utilized a multilayer perceptron artificial neural network (MLP ANN) algorithm to model AE.
• Evaluated model performance using R-squared (R²), mean square error (MSE), and root mean square error (RMSE).
• Conducted sensitivity analysis to determine the impact of various parameters on AE.

Main Results:

• The MLP ANN model demonstrated a high capacity for modeling AE, achieving R² = 0.939, MSE = 0.0025, and RMSE = 0.05 during testing.
• Sensitivity analysis revealed that temperature (34.6%), COD (21%), airflow rate (19.1%), and oxygen transfer rate/volumetric mass transfer coefficient (OTR/KLa) (15.7%) are the primary drivers of AE.
• Higher temperatures were found to increase AE due to reduced wastewater viscosity, facilitating better oxygen penetration.

Conclusions:

• The developed AE model effectively predicts BTP performance, offering a valuable tool for optimizing aeration processes.
• Implementing this model can lead to significant reductions in energy consumption within wastewater treatment facilities.
• Understanding the influence of key variables like temperature and COD is essential for efficient BTP operation.