Non-linear multivariate analysis of the global solar radiation received across five cities in South Africa.

Abstract

South Africa is considered one of the most developed countries in Africa, however with more than 80% of its electricity being generated from coal, this country is considered one of the highest contributors to greenhouse gas emission throughout the continent. The impacts of this fossil fuel dependency are prominent in the environmental degradation experienced - climate change conditions as well as the current state of emergency faced by the national power utility, ESKOM. While provisions such as load shedding are being made to avoid the country from facing blackout, the consequences of these resolves significantly influence the economy of the country. Although the cost of applicable renewable energy technologies has decreased considerably over the past few years, South Africa continues to lag in the adoption of renewable energy systems in a global comparison. Most applications of potential solar renewable energy systems are currently in the investigation stages, leaving this readily accessible resource's capacity idle. This makes research in solar renewable energy highly significant with regards to progressing the country's uptake of green energy technologies. Our study proposes linear and non-linear analysis of multivariate models for the estimation of global solar radiation (GSR) received across five major cities in South Africa. The significance of this study is to allow for effective GSR estimation in the application of solar technologies, while increasing implementation of these alternatives. Measured quantities such as sunshine duration and solar radiation for certain regions are limited due to the expensive equipment required and maintenance thereof. Local meteorological sources are unable to provide historic data which is complete, as these quantities are scarcely quantified. The dependency of GSR on meteorological variables such as air temperature, relative humidity and relative sunshine duration was evaluated for the period January 2007 - June 2018 to realize estimation models for each of the study sites. The Hargreaves-Samani and Angstrom-Prescott empirical models served as the foundation for our single variable analysis of GSR reliance on each meteorological parameter and their relative variations. Our results have indicated that our proposed multivariate, non-linear equations perform better than the empirical models as well as single variable, linear regression equations. Our suggested models are site-specific and demonstrate a strong correlation to historic GSR values with low, acceptable error indicators. Further to this, we have recognized that second and third-order relationships between H/Ho and multiple meteorological variables provide a more accurate description of GSR for most of the cities under study. This analysis could potentially contribute significantly to the investigation of solar radiation alternatives and photovoltaic (PV) technologies in South Africa. We believe that integration of estimation models within the design and installation stages of PV technologies will be largely beneficial in ensuring their optimum intake. The models discussed in this study verify the reliability and accuracy of GSR estimation through readily accessible meteorological factors in a cost-effective manner.