Assessment of Solar Radiation Data Used in Analyses of Solar Energy Systems

Abstract

Solar radiation data is essential to conducting performance analyses of solar energy systems. In the US, the National Solar Radiation data base has archived 30 years of hourly data for 239 US cities. Since the computational effort required to simulate systems for 30 years is excessive, it is more convenient to use typical meteorological year (TMY) data in performance analyses. TMY data provide hourly solar radiation and meteorological data representative of one 'typical' year for the same 239 locations in the database. Simulation studies have generally used these hourly values, although solar radiation can exhibit wide variations during an hour. Variations in solar radiation during an hour could result in inaccurate performance estimates for some types of solar systems, such as photovoltaic systems, that respond quickly and non-linearly to solar radiation. Short-term radiation data is not as readily available as hourly data. One year of 1- minute data for one US location and one year of 3-minute data for eight US locations were made available for this research. The impact of using short-term radiation data in performance analyses instead of hourly data and the accuracy of using statistically formulated TMY data instead of the actual long-term data was investigated. In addition, the behaviors of both data sets were studied in terms of diffuse fraction and frequency distributions. Correlations and distribution curves have been previously developed and comparisons of real data to these correlations were made to determine their accuracy. To quantify impacts, calculation of radiation on tilted surfaces using known models were made for both TMY data and short-term radiation data sets, and utilizability analyses were made to quantify the impact on performance analyses in a system-independent manner.