Browsing by Author "Lewis, Carmen"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Modelling atmospheric transmittance for clear-sky spectral solar radiation in practical applications
    (Stellenbosch : Stellenbosch University, 2021-12) Lewis, Carmen; Strauss, Johann; Stellenbosch University. Faculty of Engineering. Dept. of Electrical and Electronic Engineering.
    ENGLISH ABSTRACT: Solar radiation technological advancements are contributing to increased solar power generation, energy access, and innovative use of solar power. However, developing countries still experience contention in adopting renewable energies as competitor energy generation technologies. A significant barrier for these countries is data-scarcity, which results in solar radiation research deficits. Developing countries, therefore, require innovative solar radiation characterisation, modelling and forecasting methodologies to circumvent this data-scarcity and still access the global technological advancements. Specific solar radiation methodologies often include several sub-models (or processes, which are formulated sequentially. Although the advantage of this approach is flexibility, the resulting model is influenced by the accuracy of sub-models. Considering that sky condition classification is an imperative component of clear-sky model conditioning, it is of interest whether the specific subjectivity and interdependence observed in literature can be removed from the classification process. As a result, a new solar resource classification methodology is formulated from fundamental frequency theory to overcome these interdependencies, while requiring few input parameters. A predominantly qualitative analysis of the classification results is pursued, due to the inherent subjectivity of classification and the validation method deficits in the literature. High-resolution sky images are used to demonstrate the character of each solar resource class, and the conservatism and flexibility of the classification methodology. In this dissertation, atmospheric transmittance methodologies for clear-sky spectral solar radiation modelling and assessment are further discussed. Technical applications of spectral models require a balance between rigour and accessibility - a balance not yet found in atmospheric transmittance literature. In general, transmittance methodologies formulate individual atmospheric extinction events to determine terrestrial radiation. These extinction events include Rayleigh and Mie scattering, and selective atmospheric gaseous absorption. Due to the varying influence (on solar radiation of the latter, the transmittance of these gaseous absorbers is inconsistently considered in literature, and often includes estimates and parametrisations. Transmittance methodologies generally also assume a single-layer atmospheric structure, which results in convoluted (atmospheric pressure and temperature adaptations. In this study, a minimal spectral clear-sky terrestrial model is developed for calculating the atmospheric transmittance of gaseous absorbers. The atmosphere is modelled according to a representative atmospheric layer, and the Bouguer-Beer-Lambert law is adopted to for- mulate gas-speci c transmittance functions. This approach can be used to calculate separate transmittances for speci c atmospheric layers, in simulation to more rigorous codes, while maintaining the simplicity expected for solar radiation applications. This approach also ex- cludes the use of uncommon ground-based data - accessible remote retrievals are adopted. However, this approach is exible and allows reference pro les and monthly averages of gaseous absorbers use. As a result, the proposed gaseous absorber transmittances based on a representative atmospheric layer, are well suited to data-scarce locations such as southern Africa.
  • Thumbnail Image
    Item
    A Spectral Guide: the analysis of solar irradiance and soiling through spectroradiometry
    (Stellenbosch : Stellenbosch University, 2017-12) Lewis, Carmen; Strauss, Johann M.; Stellenbosch University. Faculty of Engineering. Dept. of Electrical and Electronic Engineering.
    ENGLISH ABSTRACT: The influence of soiling on photovoltaic (PV) modules has been investigated in many research projects before, but this project extended the methods of soiling analyses on photovoltaic (PV) modules through the implementation of spectroradiometry. The objective with the implementation of a spectrometer was to remove the associative characteristics of PV modules within the measurements. Since spectroradiometric measurements are not often used for this application, several measuring and analysis methods are discussed in terms of PV applications. Previous research either failed to consider the module characteristics (such as temperature and manufacturing variances) or spectrometers have not been implemented in the measurement of irradiance and soiling under natural conditions. An automated measuring system was designed and built to ensure repeatable and accurate measurements of clear-sky irradiance and comparative measurements through glass samples. The combination of these two types of measurements allowed for the determination of relative transmission losses , as a result of the refraction and absorbance by dust particles and glass particles. Furthermore, lab measurements were acquired to indicate that the method used to determine relative transmission losses, was accurate. This project was used to generate a spectral measurement and analysis guide, specifically for the measurement of irradiance and determining the influence of soiling on the transmission of irradiance. Through relative transmission loss measurements, it was found that clean low-iron float glass (a good approximation for solar glass) induced a transmission loss of 8.6 - 9.2%. Furthermore, with automated measurements at a fixed angle, the transmission loss was calculated as 9:32% to 12:53% from clean to minor soiling solar glass samples. Also, an analysis on the shift in transmission within a measurement in the wavelength bands of UV, visible and NIR light was discussed - it was found that the soil and inherent glass particles supressed some photons within the UV and NIR ranges, causing a slight increase in content within the visible range. Several recommendations regarding the automated system design are supplied. A recommendation in regards to the measurement equipment was that a measuring range greater than 300 - 1000nm with good resolution is required for these measurements. This might mean the implementation of a second spectrometer specifically for measuring the NIR range. It was also found that measurements acquired in the direction of the sun location, rather than at a fixed measuring angle, increased the stability and accuracy of measurement, allowing for better approximations of transmission loss as a result of soiling.