Concept of a probabilistic digital twin for the monitoring and control of low-voltage grids

Abstract:
The transition to a climate-neutral energy system has prompted the European Union to promote the expansion of renewable energy sources. This change has had a profound impact on low-voltage (LV) grids, in particular an increase in distributed generation, driven mainly by photovoltaic systems. In addition, the emergence of high-demand technologies, such as electric vehicles and heat pumps, has placed further stress on LV grids. To mitigate potential grid congestion, flexibility options such as energy management systems (EMS) for controllable consumption devices are being investigated. However, the implementation of appropriate control measures requires reliable and accurate grid monitoring, which is currently hampered by the limited metering infrastructure in LV grids. This study presents a probabilistic digital twin (PDT) for LV grid monitoring, designed to integrate grid, meteorological and socio-economic data to address and compensate for uncertainties arising with this lack of measurement data. The probabilistic framework enables the assessment of the grid state in scenarios where real-time feedback from the grid is not available. The PDT calculates the probabilities and likelihoods of specific grid events, eliminating the need for the distribution system operator to collect additional data. By using the grid state assessment provided by the PDT, an EMS can implement preventive grid controls, in opposition to the currently applied reactive controls, to manage load flows and mitigate potential congestion. The effectiveness of the PDT concept is evaluated through a case study of a German low-voltage grid.