Research Windstore

The planned massive expansion of offshore wind energy poses growing challenges for the energy system. At times, large offshore wind farms generate enormous power over a relatively small area. This already leads to frequent congestion situations in the power grid when it cannot absorb the power. Due to the high volatility and output of offshore wind farms, forecast errors very often lead to large "schedule deviations" that must be compensated for by various players like grid operators and electricity traders at great expense and often short notice with expensive reserve power. Probabilistic forecasts in particular will play a central role in smoothing out these fluctuations and errors and for the planning and deployment of various flexibility options, as they take into account the uncertainties of weather events.

In this subproject, a novel concept for a forecast-based storage management system is developed and analyzed. This is intended to detect large forecast errors and large-scale power fluctuations at an early stage, utilize distributed battery storage and electrolyzers to compensate for them, and quantify the remaining uncertainty for the optimal planning of flexibilities. The goal is to make optimal use of offshore wind energy in the power system and to reduce system risks posed by large fluctuations. For the first time, artificial intelligence methods will be used to detect forecast uncertainties posing a risk from offshore wind energy early on for the optimal operation of battery storage and electrolyzers to improve grid operation. This will result in an optimized design of a forecast-based distributed storage concept and management for renewable energy, whose operating principle will be demonstrated in a field test.