Numerical Modeling and Prediction of the Significant Parameters for Wind Monitoring

Abstract

Romania has a high wind potential, representing more than 14,000 MW. After significant investments of over 5 billion euros made starting 2010, many wind farms were developed in regions with efficient potential, from the South-East part of the country. Nowadays, in January 2018, in Romania were registered 3025 MW produced by wind energy, representing around 30% of the total generated energy. To establish the efficient areas for future wind power plants a massive campaign of wind’s monitoring was developed, in the entire country. The paper presents a solution of the numerical modeling for the registered environmental data, significant atmospheric parameters. The complex realized database will allow future implementations of wind power plants. The data measured and stored refer at wind intensity and direction, pressure, temperature, humidity, solar radiation, and drew points, performed during four years with masts of height 70 m, situated at distance of 20 km each other. By numerical modeling is created a correlation and prediction of the measured data, plotted in correspondence to each elevation of the measuring stations. It was also analyzed the perturbations induced by the masts presence. Firstly, are mentioned some aspects referring to the masts installation, the solution adopted for a proper distribution through the analyzed area. The database elaboration was a challenge, due to the large amount of data recorded at intervals of 10 minutes (some parameters at 10 seconds) for a period of four years, for more than 12 parameters instantly. Besides these, there were stored and some other data referring at daily produced energy with some existent wind turbines. They will be considered as data input for future developments, with new generations of turbines, more efficient. It is created an original method to compact the database in order to use small amounts of computer memory. With the daily collected data was made and stored separately the average, maximum, and minimum wind velocity, for each day and month, from the measurements at time interval of 10 minutes. The relations between the values registered are within classification areas CL-4 and CL-8, allowing performing illustrations of over-prediction and under-prediction. The wind velocities under 4 m/s are stored in a separate folder because they are not useful in wind turbine functioning. These values are used only for estimation the future wind farms efficiency. The uncertainties are analyzed and are assessed the limits of errors, for the land classification CL-4. There are presented numerical results, some conclusions, and references.