Resolution of the cloud enhancement problem for one-minute diffuse radiation prediction
Files
Date
2018
Authors
Starke, A.R.
Lemos, L.F.L.
Boland, J.
Cardemil, J.M.
Colle, S.
Editors
Advisors
Journal Title
Journal ISSN
Volume Title
Type:
Journal article
Citation
Renewable Energy, 2018; 125:472-484
Statement of Responsibility
Conference Name
Abstract
For design and simulation of solar energy systems, quality information about all components of solar irradiance is crucial. In cases when only global irradiance measurements are available, separation models are a useful method to estimate DNI and diffuse irradiation. Several of such models have been developed since the 1960s, most of them aiming to deliver estimates in hourly resolution. For higher data resolution, such as in minute data, those models are not able to describe fast transient and cloud enhancement phenomena commonly observed in data with smaller time-steps. This paper proposes an adaptation of the BRL separation model, making it capable of delivering more precise irradiance estimates for higher resolution data. Two models result from this adaptation: one for Brazil and other for Australia. The proposed models yield a more precise DNI and diffuse fraction estimates to their respective countries, compared to other separation models commonly used in the technical literature. For example, using the recommended Combined Performance Index (CPI) as a single statistical indicator, the proposed model yields DNI estimates with CPI from 230 to 350% for Australia, and from 270 to 800% for Brazil, while the Engerer model, recently recommended as a ''quasi-universal" 1-min separation model, yields DNI estimates with CPI from 500 to 700% for Australia, and from 600 to 1800% for Brazil.
School/Discipline
Dissertation Note
Provenance
Description
Data source: Supplementary material, https://www.sciencedirect.com/science/article/pii/S0960148118302593#!
Access Status
Rights
Copyright 2018 Elsevier Ltd
Access Condition Notes: Postprint available after 1 April 2020