Avaliação do ponto de conexão de geração intermitente através de Aritmética Affine e solução da curva de carga

Abstract

The insertion of distributed generation into worldwide power systems has been increasing at a rapid pace in the last few years. Large investments in clean and renewable sources, especially wind and solar, are being made by countries to minimize the environmental impacts caused by fossil fuel power generation. Connecting a distributed generation using intermittent sources can bring some benefits for the operation of the distribution system, such as improved voltage levels, increased reliability and reduction of electrical losses. However, from the point of view of the transmission system, the connection of this type of generation in some areas of the system can end up having a negative impact on its operation, causing voltage variations which can affect network equipments and increase the operation complexity. In addition, impacts on the load frequency control and voltage control among other aspects must be evaluated. Thus, this type of connection must be studied with greater level of depth. In this context, the systemic impacts of the connection of a solar plant to electrical power systems, evaluating the levels of electrical losses and variations in voltage levels are analysed. Two methodologies for solving the power flow in the presence of uncertainties are evaluated. The uncertainties considered are the generation and system load data. The first methodology is based on sequential Newton power flow solutions along the load curve, in which load and generation curves obtained from the IEEE24 Bus Test System and from the UFJF Solar Laboratory, respectively. The impacts on the electrical losses are analyzed, as well as the variation on bus voltage levels in the bars for various generation conditions. The second methodology is based on the use of affine arithmetic to determine the impacts of these uncertainties. The results of losses obtained will be treated considering the average point of the intervals for comparisons with other methodologies. The results obtained from the two methods are compared with those obtained by the Monte Carlo Simulations, in order to validate them.