MAIN PUBLICATION :

• APPENDIX
• ABREVIATIONS
• GLOSSARY
• REFERENCES
• EVENTS
• Downloads

Variability versus Predictability of Wind Power Production

Accurate forecasts of the likely wind power output, in the time intervals relevant for generation and transmission capacity scheduling allow system operators and dispatch personnel to manage the variability of wind power in the system. Predictability is key to managing wind power’s variability and improved accuracy of wind power prediction has a beneficial effect on the amount of balancing reserves needed, so the accurate forecasting of wind power is important for its economic integration into the power system.

Today, wind energy forecasting uses sophisticated numerical weather forecast models, wind power plant generation models and statistical methods to predict generation at five-minute to one-hour intervals, over periods of up to 48 to 72 hours in advance and for seasonal and annual periods.

Forecasting wind power production differs from forecasting other generation forms or forecasting the load. Wind, being a natural phenomenon, is better suited to reliable statistical treatment and physical forecasting than conventional plants which are subject to physical faults.

Wind power prediction can be quite accurate for aggregated wind power, as the variations are levelled out; and the larger the area, the better the overall prediction. The extent to which prediction error decreases with the size of the region  considered is shown in Figure 2.6. It should be noted that the forecast accuracy is reduced for longer prediction periods.

Figure 2.6: Decrease of Forecast Error of Prediction for Aggregated Wind Power Production, due to Spatial Smoothing Effect

Source: Energy and Meteo Systems

The quality of the short-term forecast should be considered in relation to the gate closure times in the power market. Reducing the time needed between scheduling supply to the market and actual delivery (gate closure time) would allow shorter-term forecasts to be used, which could dramatically reduce unpredicted variability and lead to more efficient system operation without compromising system security. Changing from day-ahead to intra-day commitments has a dramatic impact on accuracy and the cost of balancing the system. It is important to understand that for system operation, it is not just wind forecasting accuracy that is relevant for balancing the system, but also the sum of all demand and supply forecast errors relevant for system operation.