A novel approach for power ramps classification in wind generation
Mohammed Saber Eltohamy, Mohammed Said Abdel Moteleb, Hossam E. A. Talaat, S. F. Mekhamer, Walid A. Omran
Abstract
Ramp events are characterized by large power changes in a short period and increase with increasing renewable generation. Even with hourly forecasts, their predictions are still unreliable. Thus, grid operators should classify these power ramps to understand their expected occurrence periods and range to balance them. Previous research was based on a binary classification of ramp events, which assumed that ramp events were similar to one another, which is not true. Some other studies used randomization and non-causative classification methods. Hence, a more accurate method is still needed. The paper presents two new methods for ramp event classification. The first method depends on the standard deviation score, and the second method assigns a score to each ramp, which depends on the maximum value of the historical power ramps that occurred within the studied time period. The new classification methods are applied to the output power of Belgium's aggregated wind farms from 2015 to 2019, and the relative frequency of each ramp category is determined. The results revealed that, even though the capacity of wind installations has doubled, ramping behaviour is nearly the same in all years.