Estimation of complete solar potential in urban areas based on three-dimensional building façade recognition

Abstract

Solar photovoltaic (PV) harvesting is a significant force leading to the rapid expansion of renewable energy. To facilitate the installation of PV modules at solar-abundant locations, an accurate estimation of solar PV spatial potential is indispensable. Solar energy could be reflected on high-albedo building surfaces inside the urban canyon. However, using constant albedos to represent the urban vertical surfaces or ignoring the indirect components in estimating received irradiation is the typical solution in current research, which leads to inaccuracies in final results. Using conventional ways to construct albedo datasets for different building surfaces is extremely labor-intensive. In this study, we address these challenges by proposing a novel framework that integrates facade material identification using street-view images. This framework incorporates the effects of multi-reflection, enabling both qualitative and quantitative analysis of the impact of facade albedo on solar energy distribution. To achieve this, we built a facade material dataset from street views and developed an segmentation model to effectively identify facade materials from street view images. Furthermore, this study provides the first accurate estimation of solar energy potential in complex metropolitan environments and elucidates how metropolitan environments with different albedo characteristics affect solar potential distribution. Due to the distinguishable features between materials in terms of the subtle texture and patterns rather than just their shapes and colors, identification requires more details from images, which makes a multi-scale inference structure a promising solution. Compared with existing methods combining scale features at the pixel level, we proposed a novel Multi-Scale Contextual Attention Network MSCA) using a Multi-Scale Object-Contextual Representation (OCR) block to exploit and combine contextual information from different scales in high dimensional layers. The experimental results show that the proposed model significantly outperforms the existing models, achieving a mean Intersection over Union (mIOU) of 70.23%. The results indicate that the MSCA can effectively obtain the materials information from street views and can be a reliable solution to providing urban albedo information for solar estimation. The segmentation results of the façade materials are further projected onto a 3D GIS model, which allows precise albedo values to be assigned to each urban surface. This enables the accurate simulation of solar potential, incorporating both direct and reflected solar radiation, as well as capturing the complex multi-reflection effects occurring in dense urban environments. By simulating how solar radiation interacts with various building surfaces, we provide a realistic estimation of solar potential distribution and comprehensively discuss the effects that the sophisticated albedo environment might bring to the solar potential. The experimental results show that the discrepancies in albedo significantly affect the overall solar potential by 8.0% to 9.1%. If multiple reflections among buildings are disregarded, the impact can reach 11.9% to 17.8%. The findings of this study offer valuable insights for urban planning by providing a scalable method for more precise solar potential assessment. The integration of real-world material data with 3D GIS enhances the decision-making process for optimizing photovoltaic (PV) deployment in urban areas, thereby contributing to more sustainable urban energy planning and efficient use of renewable resources.