Comprehensive Methodology applied to solar radiation prediction for dual use for vertical agrivoltaics system
Résumé
Agrivoltaics aims to combine agricultural production with solar energy-based electricity, such approach offers a response to the water-food-energy nexus challenge which target to improve resilience both in terms of energy and food supply system. The emergence of this concept requires evaluating the synergies between agricultural and photovoltaic (PV) production. For Agrivoltaics, in terms of radiation, numerous coupling phenomena between the atmosphere, the ground and the PV surfaces are at play. This work focuses on the in-depth study and modeling of the physical interactions between vertically mounted bifacial PV modules and crops specifically in terms of shortwave solar radiation. The main objective is to propose a comprehensive solar radiation model which allows a better understanding of the respective contributions of downwelling and upwelling solar radiation for the front and rear faces of the modules and ground for dual use perspectives. The main challenge comes from the spatial variability of incident and reflected solar radiation. Indeed, it is necessary to consider the shading geometry including the PV structure, the hemispherical distribution of the downwelling radiance and the temporal profile of the albedo on several scales: from intra-daily to seasonal variation. In fact, the use of vertical modules makes the system more sensitive to reflected radiation, whether coming from the crops soil or from the PV modules. This will therefore allow to quantify efforts necessary to improve the precision on these components. To carry out this study, we propose to couple a systematic Cell to System type gain and loss methodology (adapted from the field of Cell to Module analysis) with a matrix-based Raytracing (RADIANCE) that enables the separation of Sky angular energy distribution from the geometry. We overcome one of the limits of this type of method which is that the properties of the surfaces constituting the geometry are fixed by carrying out a binning on the albedo then an interpolation to recover the temporal variations of such ground property. This produces an efficient transposition method compared to classic Raytracing. By doing so, it is possible to describe the solar radiation in the vertical plane of the PV modules according to the incident and reflected beam and diffuse components for each face whether on the scale of an instant at minute intervals or cumulatively at a monthly scale. This method is validated by in-situ pyranometer measurements at minute resolution which have been quality checked. A result that we can highlight is that for a constant albedo of 0.2, the share of the reflected component represents 14.8% of the total irradiance for the month of April 2023. An analysis of the effect of the temporal variability of the albedo compared to its average monthly value under clear sky conditions shows a deviation of 2.2% on the total irradiance.