An analytical approach for estimating the global horizontal from the global tilted irradiance. (August 2019)

• Record Type: Journal Article
• Authors: Halilovic, Smajil; Bright, Jamie M.; Herzberg, Wiebke; Killinger, Sven
• Is Part Of: Solar energy. Volume 188(2019)
• Page Start: 1042
• Page End: 1053
• Publication Date: 2019-08
• Subjects: Solar irradiance models -- Analytical inversion -- Decomposition in plane of array -- PV as solar sensor; Solar energy -- Periodicals; Solar engines -- Periodicals; 621.47
• DOI: 10.1016/j.solener.2019.06.027
• Languages: English
• ISSNs: 0038-092X

Abstract

Highlights: A new analytical approach to compute GHI from GTI is developed. A new decomposition model in POA and an inverse transposition model are introduced. The methodology was applied to 12 sensors and a one year of minutely data. The new analytical approach performs better than an iterative procedure. Abstract: Knowledge of horizontal solar irradiance is crucial for the nowcasting and forecasting of generated photovoltaic (PV) power. High quality irradiance measurement devices, however, are typically not collocated with PV systems. The lack of measurements can be compensated by numerical weather models or satellite-derived products, but they provide only limited temporal and spatial resolutions. Another possibility is to directly use PV systems as irradiance sensors, since the measured PV power is a good indicator of incoming solar irradiance. The challenging part in this procedure is the computation of the global horizontal irradiance G h from the global tilted irradiance G c, because a combination of decomposition and transposition models is not analytically invertible. Hence, the majority of existing solutions to this problem are numerical procedures. In this paper, an analytical approach to compute G h from G c is presented. The comparison of the proposed approach with one of the existing iterative (numerical) approaches shows promising results. When applied to 1-min data at four different locations, the new approach outperforms the iterative procedure by up to 9% in terms of the relative root mean square error (rRMSE) for east/west module's orientations, and performs slightly better with the south orientation. Moreover, the new approach provides results in less than 1 s, whereas the iterative procedure requires more than 20 min for a one year of data. An open source R-script of the new approach is also publicly available and provided as supplementary material.