Posters

Abstract

The Weibull distribution has long been a fundamental tool for assessing wind resources. It describes the frequency distribution of wind speed and is defined by two parameters: scale (A) and shape (k). Typically, a Weibull fit is performed for various wind direction bins to create a Weibull wind climate, which is then used to calculate the Annual Energy Production (AEP). These methods are incorporated in industry-standard models like WAsP 12.9. In this study, we investigate the potential benefits of using the binned frequency distributions of wind speed directly, without performing a Weibull fit. We have modified the traditional generalization and downscaling methods to work with binned wind climates. This model is implemented in PyWAsP, an interface to the WAsP core routines. Instead of performing a Weibull fit and using the Weibull wind climate, we transform the wind climate assuming a Weibull distribution. This approach allows us to use the traditional geostrophic drag law and a newly implemented stability model, but with a binned wind climate as the output. This method better preserves the details in the distribution and allows for perfect self-predictions at a meteorological mast. We validated the PyWAsP model that provides binned wind climates at 151 masts. The results showed significant improvements in the prediction of wind speed. The improvements for predictions of power density were smaller, which is due to WAsP's design that better conserves the third moment of the distribution than the first. The most substantial relative improvements were observed when the vertical extrapolation distance was small. The methodologies we have introduced here can also be applied to time series, which is our next development goal. The future version of PyWAsP will generate microscale time series as output, using the observed time series at a measuring point as input.