UTILIZATION OF SUBSEASONAL WIND PREDICTIONS AT HUB HEIGHT IN SIDRAP I WIND FARM, SOUTH SULAWES

• Main Author: Feodora Santoso, Catherine
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/85288
• Institution: Institut Teknologi Bandung

Wind energy faces challenges that can affect its productivity, notably the intermittent nature of wind speeds. This issue was evident at the Sidrap I Wind Farm, where no electricity was produced in November 2019 due to decreased wind speeds. The variability of wind over time leads to uncertainty, making accurate weather predictions at various time scales essential for optimal wind turbine operation, including weekly (subseasonal) forecasts. However, weekly scale predictions have received less attention so far. Therefore, this study aims to utilize subseasonal wind forecast data and evaluate its performance using subseasonal to seasonal (S2S) forecasting methods. This research employs operational CFSv2 data from the National Centers for Environmental Prediction (NCEP) as the forecast model data and ERA5 reanalysis data as the observation data. Both data sets are used to estimate wind speeds at the hub height of the wind turbines at the Sidrap I Wind Farm using a logarithmic wind profile approach. Bias correction is applied to reduce the bias and RMSE between the model data and the observation data. The performance of the CFSv2 model data is then evaluated using the continuous ranked probability score (CRPS) and the Brier score. The results of this study indicate that the model performs quite well in predicting wind speeds at the hub height, as reflected by its CRPS and Brier scores, particularly in the third week of November. However, significant bias and error values remain even after bias correction, which need to be considered in future decision-making processes.