Increasing investor confidence could drive more investment, lower the cost of energy

Five wind turbines at varying heights with clouds in the background.

NREL researched more accurate ways to evaluate wind speed variability and average wind farm energy production to help reduce investment risk. Photo by Deb Lastowka, NREL 54460

Predicting the future is inherently challenging. The same applies to assessing wind resources, which creates challenges for wind power plant operators and owners. That’s because wind resources vary from year to year, and the variability of wind speed is a key part of the overall wind resource assessment process. This variability creates uncertainty in estimating wind plant energy production. Who can really say when—and how much—the wind will blow?

One of the outcomes of this uncertainty is what it does to wind investments. The uncertainty of wind plant energy production increases investment risks. Furthermore, evidence from wind plant owners and financiers suggests that the fleet of U.S. wind projects is underperforming compared to the preconstruction energy estimates that provide the basis of investment decisions. As a result, financiers are applying conservative adjustments or “haircuts” to financial risk models—sometimes without understanding the fundamental drivers of uncertainty—potentially leading to valuations with inaccurate risk probabilities. This raises the likelihood of increased costs associated with risks and represents a barrier to entry for investors without wind project experience.

Despite the significant effects of long-term variability, the wind energy industry lacks robust methods to quantify this uncertainty. To search for more accurate ways to evaluate wind speed variability and average wind farm energy production under different conditions, researchers at National Renewable Energy Laboratory (NREL) examined 27 different metrics across 607 wind plants in the United States. Their work, which is part of the WETO-funded Performance Risk, Uncertainty and Finance (PRUF) project, investigated the impact that different variability metrics have on the accuracy of risk estimation.

By applying several common approaches for calculating variability to the same 37-year monthly wind-speed and energy-production time series and then correlating the wind-speed variability estimates to the variabilities of actual wind farm energy production, researchers determined the accuracy of the variability calculations. Their findings included:

  • Of the 27 different metrics explored, robust coefficient of variation (RCoV) proved to be the most useful metric in quantifying long-term variability. RCoV is statistically robust, remains effective for practically all kinds of data sets, and leads to a more accurate depiction of wind-speed variabilities than other metrics.
  • Although widely used to calculate interannual variability, annual mean wind speeds mask signals from seasonal changes and diurnal cycles of wind, leading to unreliable representations of long-term wind-speed variability.
  • Wind resource assessments need to consider both the magnitude of wind speed and its variability, rather than focusing solely on wind speed.
  • Over the 37-year period, the central United States (the Plains and the Upper Midwest) recorded high wind speeds and low wind-speed variability, making the region ideal for wind energy development.

Researchers also found that estimates of energy-generation variability require about 10 years of monthly mean wind speed records to achieve a 90% statistical confidence.

Results of this research, which underscore the advantages as well as the importance of using a statistically robust and resistant method, were published in Wind Energy Science

Future NREL research under the PRUF project will assess variations using high-resolution wind speed and energy production data, quantify the wind resource relationship with long-term energy production, and explore the influence of climatic cycles on energy production. This research will improve the certainty of wind farm annual energy production estimates, provide a sound basis for industry decision-making, increase certainty for investors, and could bring new capital into the wind energy space.