Modelling and optimization of energy consumption in the activated sludge biological aeration unit

Abstract

The biological aeration unit consumes the highest energy (67.3%) in wastewater treatment compared with physical (18.8%) and chemical (13.9%) treatment processes. The high energy consumption is caused by the supply of oxygen using air pumps/blowers and temperature that controls microorganisms' growth. The purpose of this study was to model and optimize energy consumption in the biological aeration unit. The multilayer perceptron (MLP) artificial neural network (ANN) algorithm was used to model energy consumption. The particle swarm optimization (PSO) algorithm was used to optimize the energy consumption model. Sensitivity analysis was performed to determine the percentage contribution of input variables towards energy consumption. The MLP ANN algorithm modelled energy consumption successfully and produced R², RMSE, and MSE of 0.89, 0.0265, and 0.00070, respectively, during the testing phase. The PSO algorithm optimized energy consumption successfully and produced a global solution of 0.993 kWh/m³. The percentage reduction between the lowest measured and optimized energy consumption was 38.4%. Aeration period (81%) and temperature (10.7%) contributed the highest towards energy consumption. In conclusion, the temperature played a significant role in energy consumption compared with the airflow rate (4.2%). When the temperature is conducive to allowing the growth of microorganisms, the removal of COD and ammonia will be rapid resulting in low energy consumption.