Modelling of solar inputs and selection Of solar energy systems for different Geographical locations

Abstract

The variation of solar insolation at earth surface depends on solar position, geographical location, altitude, atmospheric scattering and attenuation. Solar insolation is obtained by summating instantaneous direct, diffuse and total solar irradiation components. Solar models need to satisfactorily predict Direct Normal Irradiation(DNI), Direct Horizontal rradiation(DHI) and Global Horizontal Irradiation(GHI) values for a given location. This model proposed in this thesis provides ways to capture the temporal and spatial solar resource variations for a geographical region. Methods to process the generated or available raster files and conversion of raster files into time series files were discussed and demonstrated. The data mining platforms for various data types are discussed. The instruments used in the cases of direct measurements were presented briefly. Algorithm to compute the solar position was implemented in FORTRAN and Python platforms. The measured National Institute for Wind Energy (NIWE) solar radiation dataset forms the core of this endeavor. The ways to access the downloaded raster files from Meteorological and Oceanographic Satellite Data Archival Centre (MOSDAC) was demonstrated. The Shuttle Radar Topographic Mission (SRTM) elevation dataset was used to analyze the terrain and altitude details. The solar radiation dataset from the National Solar Radiation Database (NSRDB) was accessed for training the forecasting models. The aerial image tiles were obtained from United States Geological Survey (USGS) earth explorer platform for area measurements and land use information. The methodology for extracting the Linke turbidity factor for Chennai from NIWE dataset was accomplished by FORTRAN programs newline