Generation of Series of High Frequency DNI Years Consistent with Annual and Monthly Long-term Averages using Measured DNI Data

Abstract

In recent years, many researchers are suggesting that statistical approaches are needed to achieve sound estimates of the economic feasibility of commercial Concentrating Solar Thermal Power (CSTP) plants. A key element of such statistical approaches is the capacity of generating large series of high-frequency years which are consistent with the estimated variability of the monthly and annual values of the relevant meteorological variables.This paper presents a method of using high-frequency solar Direct Normal Irradiance (DNI) measurements to generate high frequency DNI series consistent with the variability and distribution of natural series, and also matching monthly and annual long term averages. The method takes advantage of a novel technique for the nondimensionalization of measured high-frequency daily DNI curves. This nondimensionalization technique makes it possible to consistently use measured solar DNI data to generate new daily curves of high-frequency DNI data and the subsequent long series of high-frequency DNI years. The novel technique for the nondimensionalization of measured high-frequency daily DNI curves is based on the nondimensionalization of the temporal scale by dividing the elapsed Universal Time (UT) since sunrise by the total elapsed UT from sunrise to sunset and on the nondimensionalization of the solar DNI scale by dividing each actual solar DNI value by the corresponding DNI value of the clear-day solar DNI envelope curve.