Evaluation of direct solar irradiance from 18 broadband parametric models: Case of Algeria

In this work, eighteen clear sky broadband models for estimating the direct irradiance at normal incidence (DNI) have been investigated. The used solar data has been recorded from two sites of different weather conditions; Bouzaréah site (Algiers), located on Mediterranean coast and Ghardaïa site, about 600 km, southern of Algiers in the desert region. Three input parameters were measured by two weather devices, namely; hygrometer and manometer systems for calculating the temperature, relative humidity and local pressure. In addition, two Pyrheliometer devices were fixed on the sun-tracker system for measuring the DNI. Moreover, three statistical indicators were used, in the goal to make comparison between the estimated and measured DNI results. Accordingly, the relative bias (rb), the relative root mean square error (rRMSE), the coefficient of determination (r2), and the relative error (re) have been calculated. Results obtained from both sites, show that the rb and rRMSE of MAC, Yang, Dogniaux, and ESRA models are typically smaller than 2%. in other hand, a further consideration has been taken to select the best model by evaluating the relative error between the measured and estimated DNI versus the solar elevation angle. The results averred that the ESRA, Dogniaux, MAC and Yang models (in decreasing of performance) performed better than the remaining models, with the rRMSE of 0.443, 1.066, 1.237 and 1.283 for Algiers, and 0.390, 1.100, 1.223 and 1.443 for Ghardaïa. This fact is important to take into account, in the case where measured data are not available, for the selection of the most suitable locations for the installation of the future concentrating solar power plants in Algeria.