A scalable framework for high-resolution urban BIPV potential assessment: A case study in Tokyo

Estimating high-resolution solar irradiance in dense urban environments is computationally demanding because shadow casting and sky-view evaluation require massive queries, whose computational costs increase with urban complexity. This challenge limits the practical use of the building-integrated photovoltaics (BIPV) assessment for urban energy planning. This study presents a scalable, geometry-explicit framework that accelerates the two dominant bottlenecks in urban irradiance modeling: shadow calculation through sun-facing filtering and KD-tree candidate restriction, and sky-view-factor computation through vertical reuse combined with angular and facade-facing pruning. Using Chūō Ward, Tokyo, the framework is evaluated through a three-tier design (individual buildings, neighborhood blocks, and 1000 buildings), with additional validation performed using measured PV-output data from an independent university campus. Against Radiance benchmarks, the method preserves PV-accuracy, with 100% agreement for shadow casting decisions and low sky-view-factor error (a mean absolute error of 0.0072 and a symmetric mean absolute percentage error of 2.16%). It achieves a median 7.24-fold speed-up for sky view factor (up to 10.48-fold) and a mean 1.21-fold speed-up for shadow casting. Coupling hourly PV potential with electricity demand shows that annual BIPV can supply about 39.31%–47.67% of local demand across the study areas, while the surplus-hour ratio reveals stable midday surplus windows in some districts but stronger seasonal constraints in others. Energy-to-area ranking and coverage-dependent net present value screening further show that 15%–35% deployment remains economically favorable, whereas 50% coverage becomes unattractive. These results demonstrate that the proposed framework enables scalable, demand-aware, and deployment-oriented urban BIPV analysis for energy-transition planning.