DNI and GHI Pyranometer readings under dewy and foggy conditions for CSP

Hello Paul

Our pyrannometer reads less than satellite derived solar data under dewy and foggy conditions, which is to be expected. But which data source: pyrannometer or satellite, should be used in SAM? In dewy conditions, the sun evaporates the dew early in the morning reducing the DNI and GHI readings. Could you let me know if SAM’s calculates allow for the effect of dew on the mirrors? In foggy conditions, the fog simple reduces the solar intensity. Could you let me know if SAM’s calculates allow for the effect of fog?

Regards

Paul

Hi Paul,

None of SAM's performance account for the effect of dew or other moisture evaporating from the solar collector. The solar radiation data in the weather file should account for the reduction in solar radiation due to fog.

As for which data source is best for use in SAM, the answer depends on the purpose of your analysis. For techno-economic analysis, I would recommend using a weather file that represents the solar resource over a period of many years because SAM's cash flow model uses hourly performance data for a single year to represent the system's performance over the project life, which is typically 25 or 30 years. If you are ignoring the financial model, and modeling the system's performance for a single year, then the pyranometer data may be more useful than satellite-derived data if the pyranometer is at the project site. If you have more than one source of data, you can run SAM using different weather files: For example you could run SAM using both satellite data and pyranometer data to gain insight into your system's design.

A useful resource for evaluating weather data from different sources is the solar resource data best practices manual ( PDF 7.3 MB ).

Best regards,
Paul.

Hi Paul,

The best approach would be to use site-adpated satellite-derived time series data. Site-adaptation procedure refers to the process by which systematic bias in satellite-derived data is removed. There are several ways in which this can be done. The basic requirement is to have multi-year time series data that covers the period for which ground measurements are available. An example of how this is done can be seen in the publicly available solar resource report for Upington solar park in South Africa (see pages 12-16): www.crses.sun.ac.za/files/research/publications/technical-reports/GeoModelSolar_SolarResRep_58-01-2011_Upington_rev2.pdf .

The result is a multi-year time series data - with lower bias, as well as better distribution of DNI values (important for accurate energy simulation of CSP power plants).

Regards,