A Multi-Parameter Algorithm for Wind Power Ramp Detection

The amount of wind power capacity being integrated to the grid is increasing as wind turbine technology improves and more importance is placed on mitigating climate change. However, the stochastic nature of wind power poses various technical and economic challenges to power system operators who must ensure that load and generation are instantaneously balanced. A phenomenon in wind power that is of primary concern to the wind power operators is wind power ramp events. It is vital to obtain a better understanding of these events to be able to manage it better. This paper proposes a new ramp detection model to improve upon the state-of-the-art ramp detection models. The aim of the algorithm is to segregate wind power signals into increasing and decreasing ramps to facilitate ramp detection as well as ensure that all possible ramps of varying duration are found. The ramp detection model is applied to measured wind power data of a utility size wind farm to evaluate its performance. Results indicate that the identification of wind power ramps with the segmentation method, correspond to visual ramp identification.