Wind Plant Performance with Age in the United States

Accurate knowledge of how the performance of wind farms evolves over time is needed in order to assess the financial viability of wind projects and understand their lifetime energy output. Though the change in performance with age has been analyzed in several European countries, this research is the first to quantify it for the U.S. wind fleet. This study analyzed the performance of 917 wind projects across the United States to determine how the output of wind plants changes over time. To accurately assess long-term performance, recorded generation for each wind plant from the Energy Information Administration was weather corrected with a time series of modeled generation. Modeled wind speeds from publicly available reanalysis datasets were used to model wind speed at turbine hub heights, which were combined with detailed plant metadata and power curves to estimate generation. Several reanalysis products were tested and ERA5, which provides hourly data at approximately 30km resolution, was selected as the most useful model for this context. Extensive metadata for each plant was also used to investigate the influence of a number of plant characteristics and meteorological variables on performance.

Trends in plant performance were analyzed by individual regression on a plant-by-plant basis and a fixed effects regression over the entire fleet. We found that performance declined at an average rate across the full fleet of roughly 0.4% per year, so that after 15 years of operation, the average plant was producing about 94% of what it would have produced early in its life. However, we also found evidence that newer plants (i.e. plants less than 10 years old) exhibited little to no degradation in performance, in contrast to older plants which did show significant degradation in performance over the first 10 years of operation. Unlike the findings of previous studies on European wind fleets, this degradation does not occur smoothly over the lifetime of a plant. Fixed effects regression showed that a significant drop in performance occurs after 10 years of operation. This drop in performance occurs at the same time that plants age out of eligibility to claim the Renewable Energy Production Tax Credit. This step change in performance seen after year 10, and the lack of performance decline seen in newer plants, highlights that plant operators have flexibility in their choice of options to mitigate performance decline and that strategies to reduce performance decline have become more effective with newer plants.