ViPES2X : Fully AI-driven Virtual Power Plant for Energy Storage and Power to X

As the complexity of the energy system grows, so does the need for dimensioning assets such as batte- ries, electrolysers, and other types of energy storage, as well as the need to operate them efficiently in the grid to keep CO2 emissions and energy costs as low as possible.
This project will develop a fully AI-driven Virtual Power Plant (VPP) using self-improving algorithms for operating energy storage and Power-to-X (P2X) systems in the most optimal way. The VPP will be able to forecast in real time the cost of operations (energy costs, losses, and degradation) and compare with income (energy import savings and revenue from balancing markets). The project will demonstrate a 15% cost reduction for hydrogen production by predicting system degradation, a 15-20% increase in battery life- time by predicting long term degradation, and a 45% reduction in the degradation of batteries by coaggregating and optimizing battery systems with slower reacting loads such as district heating and heat pumps. Project owner Hybrid Greentech will develop and demonstrate the VPP using cutting-edge research at DTU Energy, DTU Wind Energy and Energy Systems, and Imperial College London.

Green Hydrogen Systems and Lhyfe expects to implement the VPP in their electrolysers and Siemens Ga- mesa Renewable Energy to aggregate their battery systems with the VPP.