Capacidade de hospedagem de geração fotovoltaica em redes de distribuição: modelagem e estudo de caso

Abstract

Photovoltaic systems are increasingly integrated with the low voltage (LV) and medium voltage (MV) distribution network. This situation has already started to produce significant challenges for the distribution system operators due to the energy quality impacts associated with their integration with the network, such as increases in voltage levels, harmonics, intermittence, bidirectional power flow and frequency variations . In addition, the absence of centralized control raises other issues, such as the need to modify protections, network security issues, generation/- consumption imbalance, among others. In this context, Hosting Capacity (HC) studies focused on distribution systems are increasingly relevant. HC is an indicator of the maximum amount of distributed generation (DG) that can be hosted at a given point in the power grid without violating predefined operating limits. The main challenges related to studies of HC in distribution systems are related to the specific characteristics of these systems, such as their size, absence of accurate data and their high X/R values. Aiming to contribute with studies related to this area, this work presents a methodology for obtaining HC in distribution systems using the open source software OpenDSS. At first, the study will be carried out for a test distribution system provided by the IEEE, the IEEE European Low Voltage Test Feeder (STDE-IEEE). After a HC study at STDE-IEEE, HC studies will be prepared in the distribution system of the Federal University of Lavras (UFLA). For this, the campus electrical network will be modeled, by surveying the parameters of the feeder and the main characteristics related to transformers, cables and loads, creating a model of this system for power flow studies. The data obtained will be applied to possible reinforcements in the UFLA electrical network, as well as in the contribution to energy management and allocation of new distributed generation systems. For STDE-IEEE the results show that the insertion of twice the amount of generation required for each node/load, the HC was achieved and a specific section of the circuit was responsible for most of the limitations. The modeling and study of each scenario was an important factor for understanding and analyzing the power flow in all scenarios and studies of HC in the UFLA electrical network. Voltage was a limiting factor in all scenarios that reached HC, even in scenarios in which there was low DG insertion. The results showed high voltage variations even at low DG powers. For the UFLA system, the results show that there was no significant variation in voltage levels and the HC was reached by the maximum current supported by the electrical conductors.