An artificial neural network model for predicting volumetric mass transfer coefficient in the biological aeration unit

Abstract

The solubility of oxygen in a liquid is limited/restricted by the gas–liquid film that prevents gas from dissolving in wastewater. Oxygen in the biological aeration unit (BAU) is required by microorganisms to survive and eliminate organic and inorganic matter. This study developed a volumetric mass transfer coefficient (KLa) model using Artificial Neural Network (ANN) algorithm. The performance of the KLa model was evaluated using coefficient of determination (R2), mean squared error (MSE), and root mean squared error (RMSE). KLa model produced R2 (0.852), MSE (0.0006), and RMSE (0.0245) during the testing phase. Biomass concentration (22.29%), aeration period (20.55%), and temperature (19.63%) contributed the highest towards the KLa model. KLa model showed that the BAU should be operated at high temperatures (35°C), low biomass concentration (1.65 g/L), and low aeration period (1 h) instead of high airflow (30 L/min). Temperature should be included in the modelling of the BAU, to achieve optimum KLa.