Teoria do portfólio aplicada à previsão de recursos solar e eólico

Abstract

The generation of electricity from alternative energy sources have been increasingly used in the world. Solar and wind power generators are those which stand out in terms of growth. The implementation of wind and solar power plants is restricted as to its location, because they are sources that depend on specific climatic factors. Brazil, due to its vast territory and climate, is composed of regions that can alternative energy sources for electricity generation. The challenges for the integration of electricity production sources in large scale depends mainly on predictability, flexibility of other energy sources, due to its variability. From the perspective of costs, from March to October 2010, in Germany for example, due to solar and wind forecast errors, electricity production from these sources was 2 GWh lower than expected, resulting in an additional cost of 2 20 euros per MWh generated. This work aims to adapt the Portfolio Theory to improve the solar and wind predictability, finding a limit where the combination of energy sources will result in lower prediction errors. The theory is used in the financial sector so that investment risks are reduced by diversifying applications of investors. Developed by the american Harry Markowitz in 1952, the theory has the ultimate goal of methodize the definition of penetration percentage of each asset in order to improve financial returns. The Portfolio Theory adapted to improve the predictability of alternative energy resources is little used, so this thesis is a pioneer research in Brazil. By combining 70% of wind resource and 30% of solar resource, defined by the application of portfolio theory, resulted in a reduction of risk (standard deviation) of forecast errors. For solar resource, the risk was found to be 11.24 and wind resource 7.38. After the integration proposal, the risk obtained was 6.09. This study have use of solar and wind data measurements in Maracanaú - CE.