Ground view factor computation model for bifacial photovoltaic collector field: uniform and non-uniform surfaces

The photovoltaic collectors in a field are subject to three types of solar radiation: direct, diffuse, and reflected irradiance. The reflected irradiance received by solar photovoltaic (PV) depends on the view factor from solar PV to the ground. This view factor component is dominant for bifacial PV due to additional reflected irradiance gain, which can be achieved from the module's rear side. This paper proposes and verifies a finite element method based view factor computation model, which can handle both uniform and non-uniform ground surfaces. The unique contribution of this work is that it introduces a geometric progression based finite element mesh generation process that forms the quasi-uniform grid. The generated grid values are fitted into the computation model to calculate the view factor from bifacial photovoltaics to the ground, known as the ground view factor (GVF). The proposed computation model can achieve an accuracy of 99%. To keep accuracy at this level, the smallest element size of the coarse mesh should be within 0.1%-0.4% of surface width or length. Moreover, the geometric progression ratio of the fine and coarse mesh should be in the range of 1.001-1.002 and 1.01-1.04, respectively. The model is analysed under six different PV field variables: multiple reflective ground surfaces, the height of PV, tilted ground surface, PV position in the ground, length and width of the ground, and PV string length. For the different string sizes considered here, the view factor model's computation time varies from 180 s to 257 min for the iteration size of 7.67 billion to 765 billion. The view factor computation model will contribute to analyse reflected irradiance at the rear side of bifacial PV, which is essential to predict the energy generation accurately. The proposed model is also beneficial for urban planning and addressing heat gain of the building-integrated PV system and energy usage. (C) 2021 The Authors. Published by Elsevier Ltd.