ANNOUNCEMENTS
Wind is a Renewable source of energy that possess enormous amount of potential. Harnessing power from wind, along with other sources of renewable energy can replace conventional or traditional sources of energy consumption, such as fossil fuels to a far extent.
The potential of wind at a site varies from place to place, and accordingly sites may be called as windy sites. However, its intermittency or variable nature poses a challenge while predicting its power that could be generated from wind turbines. The variations in wind could be diurnal, hourly, seasonal, monthly, etc. depending on climate and other factors influencing it.
In compliance with the regulations laid down by the government, wind power ahead of the generation from wind farms has to be forecasted and scheduled into the grid. This is done in order to match the demand for power from Distribution utilities with that of supply from RE generators so that grid could be strengthened and more resilient towards its operation while creating balance between demand and supply.
An attempt has been made through this study, to predict wind speed at sites suited for wind power generation. The sites of wind power generation researched in this report are arid and coastal. The variables that can account to variations in wind speed have been studied at three of the sites.
The Significance and contribution of variables in predicting wind speed are determined and through forecast error metrics, the extent of error reduction in NWP output was examined. An analysis for wind behavior in different seasons throughout the year has been through this study. Statistical downscaling approach has been used to understand the error in wind speed. Different wind seasons show varying error differences and thus regression model has been tried to develop for predicting final wind speed and accuracy of those models have been compared through metrics.
Keywords: NWP, wind speed, variations, error metrics, regression
