FiN: A Smart Grid and Power Line Communication Dataset
The increasing complexity of low-voltage networks poses a growing challenge for the reliable and fail-safe operation of electricity grids. The reasons for this include an increasingly decentralized energy generation (photovoltaic systems, wind power, etc.) and the emergence of new types of consumers (e-mobility, domestic electricity storage, etc.). At the same time, the low-voltage grid is largely unmonitored and local power failures are sometimes hard to detect. To overcome this, power line communication (PLC) has emerged as a potential solution for reliable monitoring of the low-voltage grid. In addition to establishing a communication infrastructure, PLC also offers the possibility of evaluating the cables themselves, as well as the connection quality between individual cable distributors based on their Signal-to-Noise Ratio (SNR). The roll-out of a large-scale PLC infrastructure therefore not only ensures communication, but also introduces a tool for monitoring the entire network. To evaluate the potential of this data, we installed 38 PLC modems in three different areas of a German city with a population of about 150,000 as part of the Fühler-im-Netz project. Over a period of 22 months, an SNR spectrum of each connection between adjacent PLC modems was generated every quarter of an hour. % and the voltage was measured every minute. The availability of this real-world PLC data opens up new possibilities to react to the increasingly complex challenges in future smart grids. This paper provides a detailed analysis of the data generation and describes how the data was collected during normal operation of the electricity grid. In addition, we present common anomalies, effects, and trends that could be observed in the PLC data at daily, weekly, or seasonal levels. Finally, we discuss potential use cases and the remote inspection of a cable section is highlighted as an example.