Modeling wind energy potential of Hatay region using probability distributions and artificial neural networks

Due to fluctuations in foreign exchange rates and global political-financial crisis, Turkey which provides demand of energy mostly by importation, has to diversify energy sources in order to carry on its economic development. According to framework of the 2023 vision, Turkey which keeps high the economic targets, has been taking supportive measures to evaluate the renewable energy resources for reducing energy import dependence. Especially, the use of wind energy and production are supported intensively as financially by laws and regulations. A precise determination of probability distribution for wind speed data is the most important step in statistically evaluating wind speed and wind energy potential of a region. In this study, wind characteristic and wind energy potential of the Hatay that is located in the Mediterranean coastal city of Turkey, are analyzed using the daily, monthly and annual average wind speed data recorded at meteorological satiation belong to Turkish State Meteorological Service measured at 10 m. Burr, Generalized. Gamma distributions which are previously untested in the area and Weibull distribution widely used in statistical analyses were applied to wind speed data sets for Antakya. Wind speed distribution curves, mean wind speed and the wind power density values of Antakya are obtained by using the Weibull, Burr (4P) and Gen. Gamma probability density functions. Kolmogorov-Smirnov (D), Chi-Square (χ²), the Coefficient of Determination and Root Mean Square Error Analysis are used to evaluate their goodness-of-fit of the distributions. As a result, it is found that the investigated location has a low wind power density. On the other hand, due to fluctuating weather conditions, estimating wind energy potential is still a significant problem. Artificial Neural Networks (ANNs) have been commonly used in Short-term and Just-In-Time modeling of wind power generation systems based on main weather parameters such as wind speed, temperature, humidity, etc. Two different data sets called hourly main weather data (MWD) and daily sub data (DSD) were used to estimate a wind turbine power generation in this study. MWD are based on historically observed wind speed, wind direction, air temperature and pressure parameters. Besides, DSD created with statistical terms of MWD consists of maximum, minimum, mean, standard deviation, skewness, and kurtosis values. Another purpose of this study in particular is to develop a multi-linear model representing the relationship among the DSD with the calculated minimum (Pmin) and maximum (Pmax) power generation values as well as the total power generation (Psum) produced in a day by a wind turbine based on the MWD. While simulation values of the turbine, Pmin, Pmax, and Psum were used as the separately dependent parameters, DSD was determined as independent parameters in estimation models. Stepwise regression was used to determine efficient independent parameters on dependent parameters and to removal the inefficient parameters in the exploratory phase of study. These efficient parameters and simulated power generation values were used for training and testing the developed ANN models. Accuracy test results show that interoperability framework models based on Stepwise regression and the neural network models are more accurate and reliable than a linear approach. In recent years, wind energy growing its use among resources of renewable energy, is encouraged by methods such as domestic production, purchase guarantee, tax exemptions in many countries. Reasons which do not require raw material, be a clean energy resource and reduce dependence on foreign, make it attractive to invest in wind energy. Wind energy investment feasibility study is a two-stage operation. An energy production estimation to define expected financial returns is the first stage. Costs of investment, operation and maintenance with unit energy cost are estimated in the second stage. As the last step of this study, an economical assessment was aimed using scenarios include variable investment credit interest rates, for a sample turbine at the beginning of the wind farm investment project. Moreover, according to variable interest rates that also affect the wind energy investment decision, unit energy generation costs were determined.