Quantifying the accuracy of optical transmission loss techniques and identifying the best wavelengths for estimating soiling in a field study

Abstract

The optical transmittance loss (OTL) of glass coupons is frequently used to estimate the soiling losses of solar PVs. The soiling loss estimation methodologies (SLEMs) and selection range of wavelengths for OTL evaluation vary significantly, influencing the accuracy of soiling loss estimation. To our knowledge, no studies have quantified the accuracy of SLEMs (requiring hemispherical transmittances of glass coupons) with a standard soiling reference station in the field. Moreover, the data regarding the best single wavelengths for estimating the soiling loss with minimum error is significantly lacking in the literature. The present study focuses on quantifying the accuracy of SLEMs and single wavelengths by using multiple glass samples over a 150-day sampling period. The results showed that SLEMs that consider either the spectral irradiance of the solar spectrum or both the spectral irradiance of the solar spectrum and the spectral response of the polycrystalline PV material provided the best soiling estimates (no statistically significant difference between SLEMs: two-tailed p-values > 0.05). Moreover, the Ultraviolet (RMSE: 7.89 ± 6.39) and the Visible (RMSE: 1.49 ± 0.47) regions have low accuracy in estimating actual soiling losses. The results also show that the best single wavelengths for estimating the soiling loss are 760 to 850 nm (for polycrystalline), significantly different from the previous studies (Spain-600 nm and Golden Colorado-700 nm). Overall, the present study has implications for independent monochromatic optical soiling technologies. The use of multiple glass samples in the study provides reasonable statistical confidence in the correctness of the results.