BR-PVGen - The Brazilian Photovoltaic Generation Dataset

BR-PVGen is an open-access dataset comprising synchronized, high-resolution operational and meteorological measurements from 51 distributed photovoltaic (PV) plants located across different climatic regions of Brazil. Data were collected via a cloud-based SCADA system between March 26, 2024 and June 9, 2025, and are standardized to a 15-minute temporal resolution. After preprocessing (temporal standardization, weighted moving average aggregation, and targeted interpolation of short gaps), the dataset contains: 14,400,480 valid inverter-level records 1,151,232 valid meteorological (solarimetric station) records The dataset is organized into three components: Metadata — plant-level technical specifications (installed capacity, bifaciality, panel efficiency, structure type, state); Inverter — active/reactive/DC power and internal temperature, at inverter level; Solar Station — on-site irradiance (POA, GHI, GRI), panel/ambient temperature, wind, precipitation, and station health data. All plant and inverter identifiers are anonymized, and precise coordinates are restricted to the state level to protect asset owners while preserving regional climatological resolution. BR-PVGen is intended to support research in photovoltaic generation forecasting, performance benchmarking and loss decomposition, fault diagnosis, power quality assessment, and microgrid control validation. Full methodological details, loss modeling equations, and benchmarking results are described in the associated data-descriptor article: De Jesus, C. D., Ferreira, R. N., & de Aguiar, E. P. "BR-PVGen - The Brazilian Photovoltaic Generation Dataset." IEEE Access (in press). Licensed under CC BY 4.0. Please cite both the dataset and the associated article when using this data (see README for full citation details).