A research group in India has experimentally compared the performance of PV and PVT systems under Real-World conditions and discovered that PVT installations offer significant stability benefits, in particular in hot climates.
The team put both systems on a roof in an unknown location. The setup included PV and PVT modules, Pyranometers, Thermometers and, in the PVT system, a cooling system with gravity. Researchers also used a machine learning method that is known as random forest to classify the operational efficiency of the systems.
“The novelty of the current work lies in the integrated approach, the combining of real-time field experiments, machine learning-based efficiency classification and detailed thermal electrical performance comparison between PV and PVT systems using perovskiet modules,” the researchers said. “In contrast to earlier studies that focused primarily on simulations or isolated lab experiments, this study was conducted under dynamic environmental conditions using calibrated sensors and an adaptive thermal collector.”
The PV system used a monocrystalline silicon module with a maximum power of 60 W and an efficiency of 15.2%. The module had 36 cells and measured 680 x 540 x 35 mm. The PVT system also used a 60 W monocrystalline silicon module. The thermal collector was a copper tube embedded in an aluminum absorb plate, with water as the work fluid. It achieved a peak thermal efficiency of 43.37%.
The analysis was performed for 10 days in July 2024. Data registered during cloudy or rainy periods were excluded due to inconsistent solar radiation. Average global solar strangement data came from the meteorological station of Natal-A304, which is operated by Inmet.
The results showed that the electrical power of the PV system varied from 35.85 W to 49.55 W, with efficiency from 6.62% to 8.43%. The PVT system produced between 36.21 W and 49.54 W electrical power, with an electrical efficiency of 7.25% to 8.67%. The maximum thermal efficiency was 43.37%and generated to 315.6 W thermal ability.
The random forest model achieved 97% prediction accuracy in classifying performance in low, moderate and highly efficient zones.
“This predictive modeling approach was experimentally verified with the help of data per hour, which enhances the applicability for real -time monitoring and control,” they said. “The study showed an increase of 291.6 % in the thermal energy output compared to electrical output in PVT systems. These results show the performance and stability benefits of PVT systems, especially in hot environments.”
The findings appear in “Improving solar energy efficiency through comparative analysis of photovoltaic and hybrid photovoltaic thermal systems“Which was recently published in Solar energy materials and solar cells. Researchers from the India’s Government College of Technology, Annapooranaa Engineering College, Hindusthan College of Engineering and Technology, Saveetha University and Chitkara University have contributed to the study.
