Samenvatting

The generation of electrical power from renewable sources such as wind and solar power is on the rise. However, the nature of this power generation is characterized by a relatively high level of unpredictability, whereas the demand for electrical power continues to grow steadily. These factors – among others – lead to increasing volatility in the energy markets, which in turn increases the reliance on (the flexibility of) the power grid. Given that supply and demand of electricity must always be balanced, this results in significant price fluctuations on the day ahead (EPEX) and imbalance energy markets. In the Wattflex project, we investigated how the control / management of batteries can be used to reduce the imbalance on the electricity grid, while ensuring a valid business case for the battery owner / operator. We optimized for two scenarios: home batteries, typically combined with solar panels, and large scale batteries (1MW+) that are placed near solar parks. By exploring different control mechanisms (algorithmic, linear programming and reinforcement learning), we get a better understanding of where and how AI can be applied in the domain, and what the limitations are.