Correlative coherence modelling of South Australian wind farms

In this paper, a new method is presented for quantifying dependence in output among wind farms connected to an electricity grid. Overall stability of a grid is an important consideration. If output of wind farms is highly correlated, serious stress can be placed on the local infrastructure and as such it needs to be carefully managed. The correlative coherence analysis (CCA), first introduced by Getz [4] to capture variability of competing biological populations, is adapted to study the degree of movement in unison of wind farms. This method is applied to analyse stability of a system consisting of electricity output of five wind farms situated in South Australia. The results of this analysis are verified using principal component analysis. Distributional aspects of a correlative coherence measure on which CCA is based are discussed as well.

► A method is presented for assessing the overall correlation of groups of wind farms.
► Correlative coherence analysis (CCA) has been adapted from Getz (2003) for this purpose.
► The method has been cross validated by comparison with principal component analysis.
► CCA can be used to cluster groups of wind farms, thus identifying those that act together.
► This is an important tool for analysing the potential level of penetration of wind farm energy into the grid.