Synthetic Wind Generation Records for Australian Wind Farms
<p>In this study, we present methodology to create synthetic multi-year wind generation dataset at minute-scale granularity at the existing and future Australian wind farms. The purpose of the dataset&#160; is to assist studies of penetration of large scale and distributed renewable generation into the electricity systems and its impact on power system security in the National Energy Market (NEM).</p><p>Synthetic historical records are based on a spatial and temporal blend of reanalysis datasets with the minute-scale wind speeds observations at Bureau of Meteorology weather station network. Strengths and weaknesses of reanalysis data are illustrated and a correction methodology discussed. A method to introduce minute-scale and sub-hourly fluctuations absent in the reanalyses records is presented. Expected statistical properties of sub-hourly fluctuations in the wind generation records are derived from the characteristics of the background atmospheric state in the vicinity of the wind farms.</p><p>The accuracy of the dataset is validated in terms of&#160; power spectra and ramping frequencies in the simulated timeseries against existing minute-scale observations of wind generation at Australian Wind farms. The statistical properties of the observed and simulated timeseries match reasonably well, overall making the dataset suitable for the investigations of&#160; the implications of wind ramping on energy demand and generation at the existing and foreseeable infrastructure build in the NEM.</p>