Researchers at Dublin Institute of Technology (DIT) have developed a new software tool that estimates more accurately the power generation of wind energy technologies based on measurements of the wind velocity and associated variables using new statistical modelling methods. Accurate analysis of the wind speed is critical in determining the ideal location for constructing a wind or wave farm, monitoring the power output of the farm and providing estimates of future power quality. Current analytical methods of wind speed data are based on the use of a normally distributed model for the velocity gradient or wind force. However, it is well known that normal or ‘Gaussian’ distributed models are inaccurate. Thus, any power quality estimate and/or prediction based on this model is prone to inaccuracy. This has been shown to be the case in many instances relating to the location and construction of wind farms when the output power has been significantly less than predicted. DIT’s approach is based on Levy distributions to model the statistical characteristics of the wind. Research shows that the average power output is inversely related to the Levy index computed from the wind velocity. Coupled with advanced statistical modelling methods, this fundamental result can be used to accurately monitor and predict the ‘quality of power’ generated by wind farms.

Project Details


Dr. Derek Kearney


Dr. Brian Kearney


February 2015 


Niall McCoy
Thomas Wolmington



Technical Information

The sustainable power output over time of any wind turbine is determined by many design and environmental factors, but time dependent variations in the wind speed are critically important and are also the most difficult to monitor and predict. This novel approach for evaluating wind velocity data uses Levy-type distribution models for the statistical characteristics of the wind velocity. These models facilitate the derivation of a relationship between the power output from a wind turbine and the Levy index. The average power output as a function of time is (inversely) related to the Levy index for the wind velocity. This approach is the basis for a toolbox of statistical analysis software that brings a number of advantages based on a fundamental shift in the way that time dependent environmental variables such as the wind velocity are analysed. These advantages include consistent performance indicators for the ‘quality of power’ produced by wind farms and improved power output forecasting allowing optimum choice of wind and wave farm scale and location.