Uppskattning av vindklimat : – Implementering och utvärdering även metod för normalårskorrektion

Sammanfattning: The expected technical lifetime for a wind turbine is 20-25 years (Wizelius, 2007). In the process of planning a wind farm on a site an estimation of the average wind speed and the energy yield is required from the site. Due to large fluctuations in wind velocity from one year to another it is, from a climatologically point of view, not sufficient to measure the wind for a short period of time, e.g. one year. The year measured may have uncommonly high or uncommonly low winds and thereby generate an average not representative of the wind climate on the site. In the same time it is neither practical nor economically desirable to measure for a longer time period. Instead the measured data collected for a short period of time is scaled into a normal year before it is used to calculate the energy content. This normal year correction can be preformed using different methods. Principally, the methods relate the short time series, measured on the site, to one or more variables of a long time reference series. As a long time reference series the geostrophic wind on the site or a series of measured wind nearby can be used. If the correlation between the two series is sufficiently high a normal year correction may be done using the relation. Normal year correction has foremost been done using the relation between the velocities of the measured wind and a reference wind. The purpose in this study is to evaluate and implement a part of an algorithm for normal year correction considering additional variables besides wind velocity. The relationships studied are between measured wind speed and geostrophic wind speed, geostrophic wind direction and time of the year. For the purpose of evaluating the algorithm two wind data series from Näsudden, on the Swedish island of Gotland, for a period of 15 years has been used. Measured wind on 75 meters and geostrophic wind on 850hPa. Where the geostrophic wind has been used as a long time reference and the measured wind for one year at the time has been related to this reference. The relation has then been used together with the geostrophic wind data to create an estimation of the wind climate on Näsudden in three steps. Step one establishes the estimation due to the relation of geostrophic wind speed. Step two corrects the estimation due to the relation of geostrophic wind direction and step three corrects the estimation due to the relation of time of year. The conclusions from this study showed that with the method used for implementing the algorithm the normal year correction using only the relation between the velocities of the measured wind and the geostrophic wind gives the best estimation of the climatically mean wind speed. The standard deviation gives a 5 % risk for more than 0.436 m/s error for estimation of the mean wind on the site, which is comparable to earlier studies. The introduction of the wind direction dependence overestimates the mean wind on the site and amplifies the error. The introduction of the season dependence amplifies the error further and overestimates the mean wind additionally.