Abstract
Wastewater treatment plants (WWTPs) are essential for purifying wastewater to acceptable standards to benefit public health and the environment. Various treatment processes are used to purify wastewater, such as physical, biological, and chemical. Copious quantities of energy are consumed during wastewater treatment, presenting a challenge. Most of the energy is consumed in the biological aeration unit (67.3%) compared to physical (18.8%) and chemical (13.9%) processes. The driver of energy consumption in the biological aeration unit (BAU) is temperature and airflow supply. The BAU is responsible for the removal of chemical oxygen demand (COD) and ammonia in wastewater, and this is achieved using microorganisms. Microorganisms require a conducive environment for survival and oxygen for respiration; therefore, temperature and airflow supply are vital. Temperature controls the environment, allowing the growth of microorganisms while enhancing the metabolic rate in the BAU. Low temperatures (18 °C to 22 °C) impede the metabolic rate of microorganisms; as a result, an extended aeration period is required, causing high energy consumption. Airflow supplies the oxygen required by microorganisms for respiration in the BAU. High airflow rates produce sufficient dissolve oxygen (DO) required by microorganisms; however, high airflow rates consume high energy owing to using large air pumps. This is a global concern that should be resolved. The study aimed to optimise the energy consumption in WWTP, focusing on the BAU. This study was conducted in three parts. First, laboratory experiments were conducted. The focus was on experimental design and setup, wastewater collection, disposal, and analysis of the influence of temperature and the airflow rate in the BAU, energy consumption, and aeration kinetics. Laboratory experiments ensured data to model the energy consumption in the BAU. Second, the modelling of energy consumption in the BAU was conducted using a multi-layer perceptron (MLP) and artificial neural network (ANN) algorithm. Modelling was essential because it developed the optimization model that was used to optimise energy consumption in the BAU. Third, optimisation focused on applying the particle swarm optimisation (PSO) algorithm. Optimisation was essential to solve the global concern in this study. The study results indicated that COD removal (157 mg/L) consumed 2.019 kWh/m³ energy at a temperature of 35 °C and airflow rate of 10 L/min. Ammonia removal (14.5 mg/L) consumed 1.611 kWh/m³ energy at a temperature of 32.5 °C and an airflow rate of 5 L/min. High COD and
vii
ammonia removal in the BAU were achieved at low airflow rates of 10 L/min and 5 L/min and high temperatures of 35 °C and 32.5 C, respectively. The P-value of temperature (0.033) was less than the significance level of 0.05 compared to the P-value of airflow rate (0.525), which was greater. This implies that temperature had a significant impact on energy consumption compared to the airflow rate. The MLP ANN algorithm performed well in modelling energy consumption. The mean squared error (MSE) during the training, validation, and testing phases were 0.00040857, 0.00026596, and 0.00070446, respectively. The root mean squared error (RMSE) during the training, validation, and testing phases were 0.2, 0.016, and 0.0265, respectively. The R² during the training, validation, and testing phases were 0.9338, 0.958, and 0.891, respectively. The ANN energy consumption model predicted the observed energy consumption data accurately, confirmed by the analysis of variance (ANOVA) test. During the training, validation, and testing phases, the P-values were 0.366759, 0.8662, and 0.5427, respectively, which were greater than the significance level of 0.05. This implies there was no statistical difference between the observed and predicted data. The PSO algorithm optimised energy consumption to find the global optimum solution subject to constraints. PSO algorithm produced a global optimum solution of 0.993 kWh/m³. The optimum measured energy consumption during COD and ammonia removal was 1.611 kWh/m³. The percentage reduction between the optimum measured and the optimum optimized energy consumption was 38.4%. The reduction in energy consumption was analysed using the ANOVA test. The P-value of 0.0017 was less than the significance level of 0.05; therefore, there was a significant difference between the optimised and observed energy consumption data. The BAU should be operated at a temperature between 32.5 °C and 35 °C and an airflow rate of 5 L/min to maintain low energy consumption. Future research should be conducted on how the temperature in the biological aeration unit can be increased to a range of 30 °C, with minimal energy being inputted by external devices, such as heat pumps. Although heat pumps consume less energy, an improved way to heat wastewater in the biological aeration unit will be beneficial.