Wind power and photovoltaic prediction tools for balancing and grid operation

The integration of renewable energy sources (RES) into the electricity grid has become an important challenge for the utilization and control of electric power systems, because of the fluctuating and intermittent behaviour of mainly wind power and photovoltaic (PV) generation. By the beginning of 2009 almost 24,000 MW of wind power and about 6,000 MW of PV were installed in Germany. To integrate these large amounts of fluctuating RES into the grid accurate prediction systems are necessarily. Although the prediction is mainly used for balancing, other issues like the forecast of congestion, network losses, and the possible need for reserve power are growing more and more in the interest of the transmission system operators (TSO). Because the German TSOs are responsible for balancing renewable power production (as required by law) RWE Transportnetz Strom GmbH (RWE TSO) collaborates with research institutes in several projects to increase the accuracy of the prediction systems. Modern wind power and PV prediction systems commonly use numerical weather prediction model (NWP) as input. In Germany most of the balancing is done on the day ahead market. Therefore the forecasting horizon of most systems is between 24 and 48 hrs. As the intra day market has recently been developed further a requirement for even shorter forecast horizons (1 - 8 hrs) is receiving more interest. To this end we at RWE TSO in conjunction with several research institutes are developing several so called nowcasting systems. For nowcasting not only weather data of numerical weather models are used but also actual measurements of weather and/or power data are taken into account. One major step to enhance the forecast accuracy of the prediction is the combination of several weather models in order to exploit the fact that the models have strengths and weaknesses in different situations. On the basis of these single forecasts a combined power prediction is produced taking into account the characteristic- behaviour of the different forecasts and renewable resources. Though the main use is the balancing, other issues like the forecast of congestion, network losses, and possible need for reserve power are of growing interest. One example is the determination of the reserve power needed. High differences between the forecasts give a hint to an uncertain prediction. This information can be used to contract more or less reserve power on a day ahead basis and save costs compared to a fixed amount of reserve independent of the actual weather situation.