< Back to previous page

Project

Data-driven energy systems modeling for cost-efficient balancing

Our energy system is undergoing a major transition towards a more sustainable and less carbon-intensive system, characterized by an increased penetration of renewable energy sources, higher energy efficiency and reduced emissions of greenhouse gasses. In Belgium, the nuclear phase-out, which currently represents 40% of the installed capacity, fosters this integration of decentralized, intermittent generation, such as wind and solar power. In this context, ensuring a secure, reliable and cost-efficient management of our electricity network is significantly more complex than in the past. The success of new management strategies, hence, of the energy transition, strongly relies on the knowledge of the system state, which must be enabled through a better observability of the state of transmission grids and generation capacity. There is therefore an increased interest in collecting, maintaining, and sharing data of the electricity system, not only in terms of energy, but also for relevant weather and market data. Massive datasets are in that way appearing in a community not used to deal with such an amount of heterogeneous data with a high temporal and geographical resolution, leading to a need for new tools and expertise to fully leverage the underlying information. In this PhD, new data-driven approaches will therefore be developed in order to provide new methodologies for addressing the technical challenges arising in modern electricity systems and that require reliable predictive and/or descriptive models of influent variable quantities (e.g. electricity market prices, available cross-border capacities, renewable energy-based generation…) so that adequate policy, investment as well as operational decisions can be undertaken while ensuring a safe energy transition. In this context, the aim of this PhD is to improve the cost-efficiency and reliability of the electric power system by enhancing the reserve sizing (i.e., determining how much reserves are needed at each hour for the next day(s)), reserve procurement (i.e., ensuring the availability of the required capacity at the lowest cost) and real-time reserve activation (i.e., dispatching the procured reserves) procedures. To this end, data-driven probabilistic forecasts will be proposed of system imbalances, available regulation capacity (in Belgium and abroad) and import/export availability, which are to be embedded in state-of-the-art operational models of the electric power system and the electricity markets.

Date:11 May 2020 →  Today
Keywords:Data-driven reserve sizing, Power systems
Disciplines:Solar energy, Wind energy
Project type:PhD project