MÉTODOS DE FLUXO DE POTÊNCIA PARA MICRORREDES CA E CC: CÁLCULO PROBABILÍSTICO DE ÍNDICES DE CONFIABILIDADE E DE DESEMPENHO

• Autor(a) principal: SILVA, Elson Natanael Moreira
• Data de Publicação: 2021
• Tipo de documento: Tese
• Idioma: por
• Título da fonte: Biblioteca Digital de Teses e Dissertações da UFMA
• Texto Completo: https://tedebc.ufma.br/jspui/handle/tede/4088
• topic: Microrredes CA;; Microrredes CC;; Ilhamento;; Fluxo de Potência;; Métodos Probabilísticos;; Confiabilidade; AC Microgrids;; DC Microgrids;; Islanding;; Power Flow;; Probabilistic Methods;; Reliability; Sistemas Elétricos de Potência

One of the main requisites to carry out several studies in microgrids operating in the islanded mode is to determine the microgrid state. The conventional power flow algorithms are not suitable for the microgrids operating in the islanded mode due to absence of a slack bus. Furthermore, the power electronics evolution has introduced significant changes in the distribution networks, such as: increasing of DC loads and DC renewable distributed generation. The connection of these DC technologies to distribution network adds new conversion stages that affect the efficiency of the power delivery process in islanded and connected microgrid operation modes. Recently, several researches has suggested that the operation of distribution networks in DC mode is more efficient due to: elimination of conversion stages, absence of frequency violations during the islanded operation of a DC microgrid and increase of the power transfer capacity for the loads. This thesis has as objective to introduce new methods of power flow for DC and AC microgrids. For AC microgrids, it is proposed a general structure for Power Flow with Multiple Slack Buses (FMS). From this structure, two new power flow methods were obtained: the MSF based on the Current Summation Method (MSF-CSM-AC) and the MSF based on the Gauss-Zbus Method (MSF-GZM-AC). The expansion of the MSF algorithms from AC microgrids to DC ones resulted in two new power flow algorithms: MSF-CSM-DC for radial DC microgrids and MSF-GZM-DC for meshed DC microgrids. Additionally, two linearized versions of the MSF for AC and DC microgrids were formulated. For AC microgrids, it is proposed a linearized MSF based on the nodal currents injections formulation. For DC microgrids, five versions of the linearized MSF are proposed. Three matrix versions are based on the following concepts: approximation of the nodal current injection formulation through Taylor series; (ii) approximation of the nodal powers injections formulation; (iii) approximation of the nodal currents injections formulation through linear regression. The other two linear versions, for DC microgrids, are based on the Admittance Summation Method (ASM). These two versions are originated from the representation of the loads DC and Distributed Generation (DG) DC by an equivalent linear model formed by a current source in parallel with a shunt conductance obtained via: (i) Taylor series or (ii) linear regression. Furthermore, two linear optimal power flow models are formulated for AC and DC microgrids aiming to determine reference values for the voltages and active/reactive powers of the droop control. The proposed power flow methods were applied in two probabilistic studies: the reliability indices calculation and the probabilistic power flow to estimate performance indices with uncertainties. The probabilistic power flow used the AC and DC linearized MSF to assess the impact of uncertainties associated with equipment failures and load forecasting errors on the following performance indices of the islanded microgrids AC e DC: losses, voltage profile and angular frequency. The MSF methods for pure and hybrid DC networks were used to estimate the impact of network constraints (voltage, frequency and loading) on the reliability indices of AC and AC-DC distribution networks. These constraints were modelled in the following power supply restoration strategies of distribution networks: microgrid formation and load transfers among feeders. The tests results demonstrated that the proposed power flow methods has good accuracy and low computational costs regarding to Newton-Raphson Method. Furthermore, the probabilistic power flow study showed that the uncertainties related to equipment failures and load forecasting errors has significant impact on the microgrid performance indices. Finally, the reliability indices evaluation reveals that the inclusion of network constraints in the islanding and load transfers affect significantly the reliability associated with the interruptions durations.