Urban solar harvesting: The importance of diffuse shadows in complex environments

Abstract

Cities are arguably the epicenters of human and technological development, but as the world’s population increasingly concentrates in urban areas, meeting the intense energy demand sustainably and efficiently is becoming critical. Thanks to improving cost-effectiveness and versatility, solar harvesting technologies can be key in transforming cities into energy producers rather than just recipients. Every surface in the cities of the future may have the potential to harvest solar energy. But urban centers are complex environments with constrained areas, dynamic conditions, and many shadow-casting objects. To make the most of available resources, a decisive challenge is building modeling frameworks that are more geometrically flexible and balance realism with efficiency and practicality. In particular, some common simplifications like not modeling Diffuse shadows or not considering their anisotropic nature can be limiting in complex environments. Here, three shadow modeling approaches that consider all anisotropic components of solar radiation are developed and demonstrated in a variety of urban scenes. They are characterized and compared in terms of accuracy, precision, energy implications, and computational cost as a function of resolution and scene complexity. Not modeling Diffuse shadows leads to 3.9–30.9% RMSE in power estimates, and considering Diffuse shadows, but as isotropic-only, leads to 0.94–7.45% RMSE. All approaches converge within 5% deviations and the most accurate achieves <1% deviations. The rasterization-based approach is 7–30 times faster than the iterative analytical approach and 4–20 times faster than the ray-based approach. This work enables new technological capabilities for more diverse and widespread exploration of urban solar harvesting applications.