Data driven modular methodology for assessing citywide scale rooftop solar photovoltaic potential

Abstract

The integration of solar photovoltaic (PV) systems into urban environments is essential to the energy transition. However, urban areas are complex environments for conducting effective deployment planning, as several constraints must be considered. Key considerations include selecting optimally irradiated locations, assessing the impact of building shadows, and addressing constraints on PV panel installation. In this framework, several solutions have been developed to estimate the photovoltaic potential of building rooftops. However, many approaches often fall short; some are too localised and lack scalability, while others are large-scale estimations based on numerous simplifications. This paper advances the field by introducing a more accurate method for estimating solar potential in urban areas through detailed rooftop identification. The proposed methodology is structured in four main blocks: rooftop identification, shading computation, solar panel placement, and yearly PV generation simulation. This modular approach enables independent, continuous upgrades for each stage. Moreover, new contributions to the state of the art are incorporated. For instance, the rooftop identification process is formulated as a clustering problem, and the silhouette score is proposed as an evaluation metric for comparing building rooftops using unsupervised identification algorithms, yielding easily understandable outcomes. Likewise, shading losses are calculated using a digital surface model generated from LiDAR data. Finally, photovoltaic energy generation is simulated using a standard open-source simulator. This tool has been validated with 54 sample constructions from Barcelona, Spain. The best identification algorithm yielded a median silhouette score of 0.720. Shading resulted in a 3.4% loss in yearly energy production.

This work has been partially supported by the Spanish Ministry of Science, Innovation, and Universities under the grants PID2023-147851OB-I00 (IndECom) and PID2023-152461OB-I00 (OMELET) and under the predoctoral grant FPU24/03792. It has also been supported by the Agència de Gestió d’Ajuts Universitaris i de Recerca (AGAUR) under grant ‘2023 INV-2 00044’.

7 - Energia Assequible i No Contaminant

11 - Ciutats i Comunitats Sostenibles

13 - Acció per al Clima

Postprint (published version)