Simulation of a vertical bifacial PV system compared to measured values
Master thesis
![Thumbnail](/nmbu-xmlui/bitstream/handle/11250/3078962/no.nmbu%3awiseflow%3a6839571%3a54592232.pdf.jpg?sequence=5&isAllowed=y)
Permanent lenke
https://hdl.handle.net/11250/3078962Utgivelsesdato
2023Metadata
Vis full innførselSamlinger
- Master's theses (RealTek) [1826]
Sammendrag
Photovoltaic (PV) systems have gained great interest the recent years with an increasing rate of installed capacity. In addition to the monofacial solar panels, the interest in bifacial solar panels, with an expected market share of 30.0 % by 2030, increases. The bifacial solar panels utilize irradiance on both the front and rear sides of the panel to generate energy. Hence, they are able to generate more energy than monofacial solar panels, making them interesting due to the demand for energy-efficient solar panels.
Over Easy Solar AS offers vertically mounted bifacial solar panels with a height of approximately 22.0 cm. Compared to conventional solar panels, they do not need intrusive fasteners and heavy ballast. Thereby, the solar panels are suitable for many flat roofs and green roofs as light and water are provided to the plants underneath. At the Institute for Energy Technology (IFE) at Kjeller, a test site including solar panels from Over Easy Solar AS has been mounted. As the technology from Over Easy Solar AS was introduced in 2021, addressing the performance using software to simulate the irradiance and power output, with possible factors affecting the simulations has not been researched much.
In this thesis, the software bifacial_radiance was assessed by creating a 3D model and simulating the irradiance using backward ray tracing for the test site at IFE. Even though bifacial_radiance is a time-consuming software, simulations were performed for clear-sky days, overcast days, and days with varying weather, for all seasons in Norway. Overall, the results from bifacial_radiance show an overestimation of the irradiance compared to measured irradiance with a percentage difference of 5.06 % – 16.54 %, but the model gives a rough estimate of the expected irradiance. Still, there are some days where there is good compliance between simulated and measured irradiance, and other days with a large deviation, especially during the winter. There are many possible explanations for the deviation between simulated and measured irradiance such as albedo, shading, and snow. Differences between the 3D model and the test site, such as orientation, materials, surrounding objects, and positioning of objects are other factors that may affect the irradiance. This thesis thoroughly and systematically addresses factors that may affect the simulated and measured irradiance for a PV system.
The simulated power using pvlib shows an overestimation of the DC and AC power compared to the measured values. Possible explanations are system losses such as shading and cable losses. Snow covering the solar panels is also a significant factor that may affect the measured power and cause a larger deviation in the results. When simulating the DC energy for a longer period of time, the percentage difference between measured and simulated energy varies in the range of 28.29 % – 119.58 %. Hence, further work and implementation of the factors such as shading and snow should be done to develop a more accurate model for simulating power and energy.