Unleashing wastewater heat Recovery's potential in smart building systems: Grey wolf-assisted optimization aided by artificial neural networks

This article presents an innovative and efficient way to address residential dwellings' substantial heating needs. The primary objective is to utilize the heat from wastewater to enhance energy efficiency through a control framework based on predetermined rules. This framework aims to increase the incoming air temperature at the air handling unit. A thorough evaluation is carried out to analyze all aspects of the proposed system compared to an identical system that does not incorporate the wastewater heat recovery process. The practicality of the concept is assessed for a residential building located in Beijing, China, employing the TRNSYS software. The most optimal operating condition is achieved via the grey wolf optimizer and TOPSIS decision-making approach equipped with the artificial neural network using MATLAB. Then, the proposed system's performance under optimal conditions is compared with similar works in the literature. According to the results, compared to the conventional system, a higher performance efficiency of 6 % and lower levelized cost of heating of 15.6 $/MWh is obtained by implementing the wastewater heat recovery process. The parametric study results also demonstrate a conflicting change in techno-economic and environmental indicators when altering the primary decision variables, highlighting the necessity for multi-criteria optimization. What stands out from the optimization outcomes is that the grey wolf method increases the efficiency and CO2 saving by around 5.2 and 531.2 kg/year while reducing the levelized cost of heating by about 13.6 $/MWh, respectively. The optimization results reveal that this condition is attained by raising the heat exchanger's effectiveness and the number of residences and decreasing the wastewater temperature. According to the scatter distribution of key parameters, the energy wheel effectiveness has low sensitivity, and the optimal points of tank volume are distributed within the range of 3 m3 and 4 m3. © 2023 Elsevier Ltd