Energy saving and performance assessment of a split-type air conditioner integrated by an evaporative cooling unit with nanofluids

Temperatures inside inhabited buildings rise at an increasing rate during the summer, which leads to the urgent need for air conditioners. At the same time, their performance decreases during the peak summer period, which leads to a loss of thermal comfort, and energy consumption increases accordingly. Therefore, it is necessary to inspect ways to enhance the thermal performance of air conditioners considering high external temperatures to restore air conditioning performance and minimize energy consumption. This research was directed to experimentally evaluate a split air conditioning device for cooling mode with an evaporative cooling unit (ECU) employing single and hybrid nanomaterials. Four types of evaporative cooling pads were examined: aspen wood wool, honeycomb paper, coconut fiber, and plastic sponge. An ultrasonic atomizer was utilized as a mist maker to energize the air-cooling process. Four different diameters of spray nozzles were tested for the evaporative cooling unit: 0.2, 0.3, 0.4, and 0.5 mm. Two types of TiO2 and MgO nanomaterials were used by weight ratios of 0.25, 0.5, 0.75, and 1%, in addition to the hybridization between them, TiO2/MgO, in equal mixing ratios of 50/50%. The crucial results reflected that the wood wool cooling pad provides a higher performance index and EER than other types. Higher COP is obtained with a hybrid TiO2+MgO/H2O by approximately 15.8%. The overall energy savings rise when considering MgO+TiO2/H2O instead of MgO/H2O and TiO2/H2O. The ECU combined with the mini split A/C achieves the highest COP compared to the traditional A/C device by 35.9% and recovers its initial cost within 4.7 years with the hybrid nanomaterials. Using the MgO+TiO2/H2O mixture, the net life-cycle cost saving is US$ 430.16. Multiple nonlinear regression methods were used to propose a mathematical model for COP. The Adjusted R2 is about 0.964287, and the root mean square error value is about 0.1717628, indicating the prediction model's accuracy and reliability.