Albedo Influence on the Performance of Solar Transposition Models
Abstract
Accurately estimating solar irradiance on tilted surfaces is essential for the performance assessment and optimization of solar energy systems. This work evaluates the accuracy and sensitivity to surface albedo assumptions of five widely used transposition models when estimating solar irradiance on tilted surfaces. Using high quality data from a site in Golden, Colorado, we assess model performance for the diffuse component of global tilted irradiance (GTI), defined as the sum of in-plane and reflected irradiance on the tilted surface by calculating the relative mean bias deviation (rMBD), the relative root mean square deviation (rRMSD), and the combined metric U 95. Then we evaluate sensitivity and robustness of the models for different surface albedo choices. Results indicate that the Hay-Davies and Reindl models exhibit greater robustness to variations in albedo, maintaining consistent accuracy, while the Perez and Temps-CoulsonKlucher models are more sensitive to inaccurate albedo assumptions. However, these two models are the most accurate when the true measured albedo is used. The isotropic assumption for the diffuse component (referred to as the isotropic model), while robust, systematically underestimates GTI and is therefore not recommended for precise aplications.
Keywords
tilted solar irradiance, global horizontal irradiance, surface albedo, transposition models, solar resource assessment