Setting intelligent city tiling strategies for urban shading simulations

Assessing accurately the solar potential of all building surfaces in cities, including shading and multiple reflections between buildings, is essential for urban energy modelling. However, since the number of surface interactions and radiation exchanges increase exponentially with the scale of the district, innovative computational strategies are needed, some of which will be introduced in the present work. They should hold the best compromise between result accuracy and computational efficiency, i.e. computational time and memory requirements. In this study, different approaches that may be used for the computation of urban solar irradiance in large areas are presented. Two concrete urban case studies of different densities have been used to compare and evaluate three different methods: the Perez Sky model, the Simplified Radiosity Algorithm and a new scene tiling method implemented in our urban simulation platform SimStadt, used for feasible estimations on a large scale. To quantify the influence of shading, the new concept of Urban Shading Ratio has been introduced and used for this evaluation process. In high density urban areas, this index may reach 60% for facades and 25% for roofs. Tiles of 500 m width and 200 m overlap are a minimum requirement in this case to compute solar irradiance with an acceptable accuracy. In medium density areas, tiles of 300 m width and 100 m overlap meet perfectly the accuracy requirements. In addition, the solar potential for various solar energy thresholds as well as the monthly variation of the Urban Shading Ratio have been quantified for both case studies, distinguishing between roofs and facades of different orientations