Short-term forecasts of wind and PV power levels are important parameters for ensuring optimum operation of combined power plants. As part of this WindFors project, the aim is to develop the necessary short-term forecasting (wind and PV) over a range of 0-60 minutes with a high degree of time resolution. This short-term forecasting data, in combination with long-term predictions of wind and PV grid feed data and of electricity load data based on weather forecasts, is then used as input parameters for the P2IONEER model developed by ZSW to optimise the design and operation of combined power plants. After development and trials, this new method is to be tested online over a period of several months at a test location with real wind turbines and PV systems; system operators and regional energy providers will also be involved here.
A SWE's long-range Lidar unit with a range of up to 10 km is to be used in trials for the wind power forecasting system. Forecasts of the wind power beyond the time frame of the Lidar measurements are provided using a machine learning (ML) system and then validated against other independent data. The Lidar unit will be mounted on the nacelle of a wind turbine during the online trials to provide measurement data for the online forecasting system.
A new cloud camera with a fish-eye lens is used for the irradiation and PV performance predictions. The data from the cloud camera is combined with quarter-hourly satellite data. Short-term forecasts of the solar irradiation and PV grid feed values will be generated using the ML process in conjunction with data from weather models. The cloud camera will be operated in the vicinity of a PV system for the online test. The forecasts (wind, PV) are used as input parameters to provide optimum control of distributed combined power plants that are interconnected using the electricity grid and gas network.
The project is funded by the Federal Ministry for Economic Affairs and Energy (0325740AB). It is realised by ZSW (coordination) and Stuttgart Wind Energy (SWE) at University of Stuttgart.