Improved approach to separating sunlight types for renewable energy systems
Improved approach to separating sunlight types for renewable energy systems lead image
Not all sunlight is the same. When designing solar energy systems, researchers need to know what sunlight arrives at a location directly and what is reflected or scattered. While total irradiance is routinely estimated from remote-sensing imagery by satellites, researchers need to employ models to break the total into its direct and diffuse components.
“Resource assessment requires long-term, wide-area information on solar radiation, which can only be obtained from satellite-derived irradiance,” said Dazhi Yang, author of a paper on separation modeling. “However, most existing methods for deriving irradiance from satellite imagery only produce global horizontal irradiance, whereas resource assessment also requires diffuse and direct components of the irradiance.”
Over the years, hundreds of models have been developed to separate light fractions. An approach developed by Yang opens an unexplored way of constructing the separation modelling.
It builds on previous advances in separation modeling and uses the “temporal-resolution cascade.” This method deploys low-resolution, such as hourly or 15-minute, diffuse fraction estimates onto a one-minute model, where variabilities in diffuse fraction at different temporal resolutions are captured.
Yang tested the model with two years of data at seven different locations and found it could outperform or match all other existing models. Yang released his code and test data and would eventually like to embed the model in software packages and libraries that could be openly available.
“This paper does not just propose a single model but describes a new philosophy of separation modeling,” Yang said. “Since this philosophy is general, many of the previously published separation models can benefit from it.”
Source: “Temporal-resolution cascade model for separation of 1-min beam and diffuse irradiance,” by Dazhi Yang, Journal of Renewable and Sustainable Energy (2021). The article can be accessed at https://doi.org/10.1063/5.0067997 .
