Development of models for the estimation of global solar radiation over selected stations in India

Solar radiation data, a prerequisite for the designing and sizing of solar energy systems, is not available for many locations in India. However, the sunshine hour data is available for most sites from which an attempt can be made to compute solar radiation. Presently, new correlation models have been developed to estimate the monthly average global solar radiation on horizontal surfaces using the sunshine hours and compared with the measured data for 12 months (i.e. the whole year) incorporating the location parameters of a site. The models are used for computing values of six Indian stations (Palampur, Amritsar, Kodaikanal, Nandi Hills, New Delhi and Shillong) with different geographical locations. The estimated values are found to be in close agreement with their measured values. This estimation will lead to a better input for design and evaluation of solar energy systems performance. In this study, variation of the solar radiation on horizontal surface for clear-sky atmosphere was predicted and compared with measurements. The Hottel’s Clear-sky model was used in calculations of solar radiation on horizontal surfaces for clear-sky atmospheres. The usability of clear sky radiation and sunshine hour for predicting the global solar radiation was investigated. For this aim, a regression analysis was applied using S/So parameters. The results confirm that the model calculations are reasonably consistent with the measured data. In order to indicate the performance of a correlation equation, the mean bias error (MBS) and root-mean-square-error (RMSE) methods were used. Also, t-statistics is applied to determine which of the developed models give statistically significant results. © 2016, Springer International Publishing Switzerland.