Size estimation of wind/solar hybrid renewable energy systems without detailed wind and irradiation data: A feasibility study. (15th April 2021)

• Record Type: Journal Article
• Authors: Jamshidi, Siamak; Pourhossein, Kazem; Asadi, Meysam
• Is Part Of: Energy conversion and management. Volume 234(2021)
• Publication Date: 2021-04-15
• Subjects: Hybrid renewable energy system (HRES) -- Wind/photovoltaic/battery/diesel energy system -- Fast sizing -- Size estimation -- Meteorological parameters; Direct energy conversion -- Periodicals; Energy storage -- Periodicals; Energy transfer -- Periodicals; Énergie -- Conversion directe -- Périodiques; Direct energy conversion; Periodicals; 621.3105
• DOI: 10.1016/j.enconman.2021.113905
• Languages: English
• ISSNs: 0196-8904

Abstract

Highlights: Direct sizing of wind/solar energy systems via averaged meteorological parameters. No need for annual detailed wind and irradiation data for wind/solar system sizing. Implementation of size estimation function by polynomial regression and SVR. Abstract: Solar and wind energies are suitable alternatives to fossil-based electricity generation. Hybrid utilization of wind and solar generators are preferred to reduce the intermittency of output power. Hybrid renewable energy systems (HRES) have to be sized optimally in the design stage. In this regard, wind speed and solar irradiation data have to be measured and collected at short intervals and for at least one year at the site location, which means a delay in the design and construction of HRES. The purpose of this paper is to eliminate the need for detailed and long-term data in the HRES sizing process by replacing wind speed and solar irradiation data with averaged and usually-available values of meteorological parameters of the site, e.g., air temperature, elevation, relative humidity, roughness length, latitude, longitude, and precipitation. In this paper, the size of the wind/photovoltaic/battery/diesel HRES is estimated as a function of site parameters using polynomial regression and support vector regression models. 105 sites from Iran have been studied as training data to build these models and test them. Sensitivity analysis shows that estimation accuracy increases by increasing training data. Thus, it is expected that by using a sufficient number of training data with global distribution, the accuracy of the models approach the accuracy of current optimal sizing methods while the time-consuming wind speed and solar radiation data collection have been eliminated in the proposed method. For example, by increasing the training data from 40 to 100 sites, the mean absolute percentage error (MAPE) of estimating the HRES size reduced by 2.3 to 3.5 times.