Road Map for Creation of Advanced Meteorological Data Sets for CSP Performance Simulations

Lourdes Ramirez (1) , Kristian Pagh Nielsena (2) , Franck Vignola (3) , Manuel Blanco (4) , Philippe Blanc (5) , Richard Meyer (6) , Stefan Wilbert (7)

1 CIEMAT - Centro de Investigaciones Energéticas, Medioambientales y Tecnológicas
2 DMI - Danish Meteorological Institute
3 University of Oregon
4 CyI - Cyprus Institute
5 O.I.E. - Centre Observation, Impacts, Énergie
6 Suntrace GmbH
7 DLR - German Aerospace Center

The best way to determine the production from a solar facility is to take the 10 – 20 – 30 years of data that are available and run a simulation with that data. If time of day or capacity matter, those parameters can be factored in at this time. Now the variation in generation over the time span under consideration can be seen and it is possible to select the typical year’s production and get a feeling for the variation from year to year and month to month. If available, use of one year of high quality measured data can be used as input and a more detailed analysis of the plant generation can be performed. Adjustments can be made if the more detailed analysis shows that it is necessary. Trying to select one year that is typical before running the data might be fine for an initial survey, but a universal typical year dataset that is applicable for all types of generation probably is not be the best solu-tion because different facilities can react differently to the same input irradiance. Storage and the ability to adapt to rapidly changing irradiance are likely factors that can influence the selection of the most typical solar year. The annual DNI is likely to vary less than 10% from year to year, so the characteristics of the incident solar radiation and how the plant handles variations in the resource are likely to play as important a role in the annual production of the solar facility as the annual DNI. In addition, I have never been a fan of breaking the production down into monthly blocks and selecting the most typical monthly production. Weather comes in waves and patterns and cutting and splicing break up these patterns. This is how some TMY files can create annual irradiance estimates that are above or below any experienced at the site. A moving average approach can yield a full range of generation estimates although the beginning and ending year are underrepresented.

Mots clés: Performance Simulations; Meteorological Data

Domaines: Energétique; Météorologie

Lourdes Ramirez, Kristian Pagh Nielsena, Franck Vignola, Manuel Blanco, Philippe Blanc, et al.. Road Map for Creation of Advanced Meteorological Data Sets for CSP Performance Simulations. [Research Report] SolarPACES Report, IEA SolarPACES. 2017, 15 p. ⟨hal-01561666⟩