Assessment of Wind Atlas Analysis and Application Program and computational fluid dynamics estimates for power production on a Jeju Island wind farm

Abstract

The accuracy of power production predictions made using a linear model, Wind Atlas Analysis and Application Program, and the computational fluid dynamics model, Meteodyn WT, was assessed for a Seongsan wind farm on Jeju Island, South Korea. The actual data on power production from 10 Vestas V80 2 MW wind turbines were collected through a supervisory control and data acquisition system. Wind data were collected over a period of 1 year from a 70-m-height met mast which was about 1.6 km away from the wind farm. The two applications were used for estimating the annual energy production for both an entire wind farm and an individual wind turbine and accomplished this by assessing five different wind farm sizes ranging from 5 km × 5 km to 20 km × 20 km. The results showed that Meteodyn WT software performed better than the Wind Atlas Analysis and Application Program in predicting power production for all the five domain sizes studied. For the entire wind farm, the relative error in annual energy production prediction was from 2.5% to 3.5% when running Meteodyn WT, while ranging from 3.8% to 11.5% when running WindPRO. While a little overestimation in the capacity factor occurred using the WindPRO application, a little underestimation was found using Meteodyn WT.