Theoretical and experimental investigation of the effects of different cooling methods on the performance of photovoltaic panels

In this experimental study, the effectiveness of alternative cooling methods was investigated to reduce the power losses that photovoltaic panels are exposed to due to heating. In the study, primarily numerical analysis was carried out. As a result of the numerical analysis study, passive cooling methods to be tested in the experimental study were determined. Experimental work mainly includes cooling methods consisting of heat pipes (PV/T-IBAL) and finned heat sinks (PV/T-AK). These two systems were compared with conventional photovoltaic panels and analyzed theoretically and experimentally. The study showed that PV/T-IBAL and PV/T-AK panel temperatures were 8℃ and 3℃ lower than the reference panel temperature, respectively. In the study, the power generated in PV/T-IBAL and PV/T-AK panels occurred at 3.74% and 2.33% more than the power generated from the reference panel, respectively. In addition, the findings showed that although the average solar radiation decreased by 2.8% in a certain time period, the power increase in PV panels was approximately 8.5% as the average wind speed increased by 15.3%. As a consequence of the energy analysis made, the electric yield of the PV/T-IBAL panel was 12.3%, the electric yield of the PV/T-AK panel was 12% and the electric yield of the reference PV panel obtained 11.7%. In the experimental study, the thermal yield of the PV/T-IBAL panel was approximately 1% depending on the solar radiation and environmental status. Therefore, the sum yield of the PV/T-IBAL panel was 13%. As a result of the findings obtained in the experiment, an exergy analysis was performed. In this context, the exergy rate of the PV/T-IBAL panel using heat pipes was 13.7%, the exergy rate of the PV/T-AK panel using aluminum fins was 14.6% and the exergy rate of the reference panel was 14.9%. The findings obtained in the experimental studies showed that the PV/T-IBAL, PV/T-AK and reference panel surface temperatures were predicted correctly in the numerical analysis study with error rates of 4.5%, 9% and 14.5%, respectively.