Temporal extension of the wind speed in order to estimate the annual production of a wind turbine

The wind sector is facing the challenge of mini and micro generation, but the unpredictability of wind makes difficult to estimate the real potentiality of a possible wind site and consequently the correct sizing of the turbine. Usually, at least five years of on-site measurement of wind speed are considered adequate to permit a good estimation of a site potentiality, but many Authors suggest longer times. These waiting times are impractical for mini wind plants where small size investors are involved. For this reason, in the last years some techniques have been developed aimed at the extension of the data in order to give reliable estimation starting from short time measurements. In this paper, a MCP (Measure Correlate Predict) technique has been implemented able to integrate the available onsite measurements with those coming from one or more weather stations located near the site under analysis in order to forecast a typical annual speed trend. Then, this model has been integrated into a model based on Monte Carlo method able to evaluate the effect of wind variability. A probability distribution of the Annual Energy Production of a site can so be evaluated. An example of application of the method to different places is presented in the paper. Some considerations about the minimum measurement time to an acceptable estimation of annual production, and about the dependency of the reliability of the method on the site characteristics are reported, too.