Real-world energy data of 200 feeders from low-voltage grids with metadata in Germany over two years

TL;DR

The FeederBW dataset addresses the lack of public, real-world LV-grid data by providing 200 low-voltage feeders in Baden-Württemberg, Germany, with two years of high-resolution measurements and rich metadata. The authors deliver three linked data records—feeder measurement data (minute-resolution electrical measurements), feeder metadata (installation and usage attributes), and weather data (hourly meteorology) with careful data linkage and quality controls. The dataset achieves high completeness (~99.36%) and documents limitations such as PEN-missing values, potential topology changes, and deviations from $S = \sqrt{P^2+Q^2}$ under multi-phase flows. It enables ML-based load forecasting, NILM, synthetic data generation, and weather–grid interaction analyses at a realistic scale, while adhering to FAIR principles and public access via Zenodo. The study highlights the increasing role of low-carbon technologies (PV, EVs, heat pumps, storage) in LV grids and demonstrates the value of including equipment-level metadata for interpretability.

Abstract

The last mile of the distribution grid is crucial for a successful energy transition, as more low-carbon technology like photovoltaic systems, heat pumps, and electric vehicle chargers connect to the low-voltage grid. Despite considerable challenges in operation and planning, researchers often lack access to suitable low-voltage grid data. To address this, we present the FeederBW dataset with data recorded by the German distribution system operator Netze BW. It offers real-world energy data from 200 low-voltage feeders over two years (2023-2025) with weather information and detailed metadata, including changes in low-carbon technology installations. The dataset includes feeder-specific details such as the number of housing units, installed power of low-carbon technology, and aggregated industrial energy data. Furthermore, high photovoltaic feed-in and one-minute temporal resolution makes the dataset unique. FeederBW supports various applications, including machine learning for load forecasting, conducting non-intrusive load monitoring, generating synthetic data, and analyzing the interplay between weather, feeder measurements, and metadata. The dataset reveals insightful patterns and clearly reflects the growing impact of low-carbon technology on low-voltage grids.

Figures (6)

• Figure 1: The data collection process of the feeder, meta and weather data. The collected data is combined in a data platform, the data is selected and again split into three data records. [icons in image: flaticon.com]
• Figure 2: Example for sensors and gateway used in the MV/LV substations to measure the LV feeders. (image: Netze BW GmbH)
• Figure 3: Selection criteria for the $200$LV feeders published in the FeederBW dataset. All the published feeders fulfill all conditions. Refer to \ref{['fig:data_process']} for the complete data process where the data selection is embedded. [icons in image: flaticon.com]
• Figure 4: Active power weekly quantile profiles from four LV feeders showing diverse characteristics. Constructed by shrinking the two-years measurements to one week and taking the quantiles for each minute of the week. The metadata of the LV feeders is in \ref{['tab:featured_feeders_infos']}.
• Figure 5: Active power plotted over two years for selected LV feeders. A seasonal effect in winter and summer is visible through higher loads and higher feed-in.