Evaluating the benefit of probabilistic Lidar-based wind power forecasts in power systems management

Kurzfassung

The rapid expansion of renewable generation capacities increases uncertainty in power system operation due to their inherent intermittency and the challenges of accurate forecasting. While ensemble forecasts can quantify weather-related uncertainties, they are not yet fully integrated into power systems management. At the same time, minute-scale forecasts have enabled dispatch corrections with very short lead times. This work evaluates the benefit of applying probabilistic Lidar-based offshore wind power forecasts (Theuer et al., 2020) for shortest-term corrections in power systems and puts special emphasis on forecast situations which are expected to be of high relevance on the activation of short-term flexibilities in the power system. The probabilistic power forecast evaluation tool ProPower (Schyska, 2021 and Bents, 2023) has been developed to simulate a process where a day-ahead dispatch is cleared, followed by an intraday clearing for corrections, and finally a balancing stage to settle discrepancies between dispatch and observation. A key component of this tool is a stochastic clearing approach (Morales et al., 2014), in which dispatch is optimised to minimise both day-ahead and anticipated balancing costs. The latter are estimated from a set of scenarios equivalent to ensemble forecast members. Stochastic clearing has shown to outperform the deterministic clearing - resembling the status quo merit order system - by reducing the activation of short-term flexibility. ECMWF ensemble forecasts (Leutbecher and Palmer, 2007) are used for day-ahead and intraday market clearing in a German power system model. Additionally, Lidar wind power forecasts are used for corrections at a single offshore wind farm. The required balancing is then determined by the deviation between forecasted renewable feed-in and actual feed-in, derived from ERA5 reanalysis data or Lidar measurements. The study examines whether the use of probabilistic Lidar-based offshore wind power forecasts positively impacts key system indicators, such as energy not served, loss of load expectation, and wind power curtailment. The insights gained help to identify situations where such enhancements are particularly critical. This work is funded by the German Federal Ministry for Economic Affairs and Climate Action on the basis of a decision by the German Bundestag (Windramp II, ref. No. 03EE3101).