Design and analysis of different cooling effects on photovoltaic panels

Abstract

Solar radiation absorbed by the solar cells is not completely converted into electricity. The majority of the energy is utilised to increase cell temperature with the end result being reducing its electrical efficiency. The PV temperature can be lowered by a heat extraction method which can either be passive or active. Three different cooling methods, namely fins, air and water are investigated on a full size experimental setup. The first cooling setup consists of custom made aluminium fins installed at the back side of the PV panel parallel to the vertical position. Fins are designed to maximise the surface area in contact with the surrounding air which will act as the cooling medium. The second cooling set up is based on forced air flowing through the back side of the PV module. The thickness of PV module (40mm) is used as the available depth of the cooling duct. Therefore, the PV module is installed on a rectangular wooden insulator board painted black to increase its emissivity and which will in turn contribute to effective cooling. A 60W crossflow fan is installed at the bottom without any outlet diffusers at the top as it creates turbulent flow and enhances cooling. The third and final cooling setup was based on freefall water on the front of the module. A water tank was placed underneath the module as a reservoir and a 10W submersible centrifugal pump is used. A reference module is set up to compare the performance of each cooling method. The first experiment applies fixed cooling for six hours and positive gross energy increases are recorded for the three methods. However, when net energy is considered by decreasing the active cooling consumption, air results to a negative performance. It is noted that during low cell temperatures, active cooling methods were ineffective. Hence, an alternative setup is conducted with a thermostat control set to 40°C and the difference between the gross and net energy is decreased by half. In addition, it is noted that for all cooling methods there is a linear relationship of gross percentage increase with operating hours. Data simulation is conducted from an existing system following a T-test which confirmed its similarity between experiment periods. Hence, it is possible to determine the number of hours for which cooling will be required over a period of one year together with its respective gross percentage increase. Following the simulation of one complete year, it is noted that all methods result in positive income increase. Furthermore, the energy outputs and the estimated system costs per cooling method are introduced in an economic analysis, where payback time and ROI are calculated.