الفهرس 
chapter 1Only 14 pages are availabe for public view
 |  | from | 126 |  |  |
| | | from | 126 | | |
Abstract
Egypt is planning to increase rapidly the share of solar energy in its electricity production and to have installed by year 2017 a capacity of 2.3 GW of solar powered plants. This goal can only be achieved by increasing the number of classical photovoltaic systems and by implementing new large scale concentrated solar power units. Contrary to PV technology, that exploits the global irradiance (GHI) reaching the Earth’s surface, concentrated solar power converters are sensitive only to its direct component received at normal incidence, the so-called direct normal irradiance DNI (Bn). Therefore, the evaluation of the potential of these systems production units requires an accurate assessment of this direct component. In this study, quality assured measurements of the global surface solar irradiance and its direct and diffuse components performed at three Egyptian sites (Cairo, Aswan, and Port Said) available at Egyptian meteorological authority (EMA) were used to test the ability of several methods allowing to estimate the hourly direct normal irradiance Bn. Firstly, in case lack some of surface solar irradiance measurements such as GHI and diffuse horizontal irradiance (DHI) used in estimating Bn, Radiative Transfer Model (RTM) (LibRadtran) has been used. In fact, this method needs aerosol optical depth (AOD) at 1000 nm derived from sunphotometer onsite measurements (β) and wavelength exponent (α) as an input by one hand and those(β, α) were retrieved from Copernicus Atmospheric Monitoring Service (CAMS) on the other hand in order to validate the latest for surface solar irradiance estimations. Statistical performance findings between two the inputs show that the difference between statistical indicators (rBias, rRMSD and R2) values are close, for example, it is less than 2% (for rBias) in case GHI estimation and increases to 6% in case of Bn estimation at Cairo and PSU. Secondly, if GHI and DHI measurements are available, indirect methods such as the basic method (GHI minus DHI) has been used to estimate Bn while if GHI is only available, three decomposition methods (Maxwell, Louche et al. and Lopez et al.) can be used. All methods used have been only tested at cloud-free conditions. The basic method proved as best among all selective methods, values of rBias, rRMSD and R2 are ranged between -4% to 4% and 5% to 6% and 0.95 to 0.97, respectively. However, three tested decomposition models are failed to reproduce the temporal variability of the measurements, this is due to large variability of the atmospheric content of aerosols. The present work proposes revised formulas from the decomposition models for Louche et al. (1991) and Lopez et al. (2000) that take into account into consideration AOD at 1000 nm derived from sunphotometer onsite measurements. It leads to a significant reduction to the error (Bias and RMSD) than the original models at the three target sites. However, because the AOD is rarely measured at the meteorological stations, we also quantify the performance of the revised models when the AOD is either derived from MODIS observations or obtained by CAMS products. Probably because of their finer temporal resolution that makes them more sensitive to reproduce the rapid variations of the AOD, the best results are obtained with the CAMS products. By using CAMS input for model 4, the bias decreased from – 57.7 to -13 W.m-2 and RMSD from 110 to 88 W.m-2 at PSU. Therefore, we recommend using a combination of the revised decomposition models and CAMS products to estimate the hourly Bn in areas such as Egypt where aerosols are ubiquitous. Note that the improved decomposition models are generally applicable in all-sky conditions, although their benefit has been demonstrated to be significant, and probably limited to, cloud-free conditions. For an example of application that requires an assesment of Bn, three quantities quantifying the turbidity: the broadband aerosol optical depth, δa; the Ångström turbidity coefficient, β; the Linke turbidity factor, TL are derived from the measured Bn by one hand and the estimated Bn on the other hand to be compared. They display similar in the seasonal patterns at the three sites, with an absolute turbidity maximum in spring, and a minimum in winter. These commonalities suggest that the main processes controlling the atmospheric turbidity are due to regional circulation.