A Database of Direct Normal Irradiation for Predicting Long-Term Performance of Concentrating PV Systems in Europe and Africa

Purpose of work: We present a database of direct normal irradiation (DNI) giving long-term monthly averages of DNI with a high spatial resolution over the area of Central and Western Europe and Northern Africa. The availability of this database makes it possible to perform estimates of the energy production of sun-tracking concentrating photovoltaic systems. Approach: The method for calculating DNI is based on the use of data from geostationary meteorological satellites. In the present case, the satellites used are the Meteosat series of satellites, situated at 0° longitude. From 1998 to 2005, the data are from the Meteosat First Generation satellites (MFG), while the Meteosat Second Generation (MSG) are used from 06/2006 to 05/2010. The calculation method for the MFG images is a modified version of the Heliosat method[1], while the approach for the more advanced MSG is based on radiative transfer calculations as described in Mueller et al. [2]. From the global horizontal irradiance the direct and diffuse components are calculated using a variant of the Skartveit-Olseth model[3]. Scientific innovation and relevance: DNI has been calculated from satellite data by other groups (see for instance XX). However, in nearly all cases access to the data has been on a commercial basis. There is a need for high-resolution DNI data that are freely available for policymakers and interested individuals as well as for pre-feasibility studies of concentrating PV power plants. Free DNI data in high quality also supports the development of small and medium size solar energy business and is therefore important for climate protection and economic growth. Finally, the quality of the CM-SAF data is very traceable due to reviewed validation reports available at the CM-SAF web page and a data access free of charge via a web user interface, enabling a broad user feedback on data quality and needed improvements. Results: The calculation of the solar irradiance has been performed for the period 1998-2005 for MFG and June 2006 to May 2010 for MSG. From the hourly global horizontal irradiance the direct and diffuse irradiance are calculated. The resulting maps are projected onto a regular latitude-longitude grid with a resolution of 1.5 arc-minutes (MSG) or 1.8 arc-minutes (MFG). Validation using ground-based measurements of DNI from 10 locations show that the standard deviation of the annual bias in the estimates is 4.3% for the MSG data and 10.2% for the MFG data. Considering only regions with high annual DNI (>1800kWh/m2) , the four ground stations show STD of 5.5% for MSG and 5.2% for MFG. Conclusions: The new dataset for DNI has a spatial resolution that is sufficient for pre-feasibility studies for CPV power plants. The validation shows that the older data set (MFG) has a rather large uncertainty in the annual average DNI, which however is somewhat diminished if we only consider areas where CPV is likely to be deployed. The dataset presented here will be made available as long-term monthly averages in the PVGIS web application (http://re.jrc.ec.europa.eu/pvgis/). [1] Cano D., Monget J.M., Albuisson M., Guillard H., Regas N., Wald L., "A method for the determination of the global solar radiation from meteorological satellite data", Solar Energy, 37(1): 31-39 (1986) [2] Mueller R., Matsoukas C., Gratzki A., Hollmann R., Behr H., "The CM-SAF operational scheme for the satellite based retrieval of solar surface irradiance - a LUT based eigenvector hybrid approach". Remote Sens. Environ., 113(5): 1012-1024 (2009) [3] Skartveit A. and Olseth, J.A., "An hourly diffuse fraction model with correction ofr variability and surface albedo", Solar Energy, 63(3): 173-183 (1998)JRC.F.7-Renewable Energ