Otimização de redes elétricas de distribuição: planeamento e operação
Ficheiros
Data
2015-10-22
Título da revista
ISSN da revista
Título do Volume
Editora
Resumo
Nas últimas décadas o setor da energia elétrica tem assistido a profundas alterações quer na sua vertente técnica quer na vertente económica. Hoje, no setor elétrico, assiste-se à
crescente influência económica na sua gestão relegando para um plano secundário as questões
técnicas.
O crescente aumento da carga, o aproveitamento da produção distribuída gerida por
entidades privadas, a evolução tecnológica ao serviço dos equipamentos das redes, os meios
sofisticados de controlo, a liberalização do setor elétrico, a atualização da regulamentação que
conduz a requisitos cada vez mais exigentes, as tendências futuras tais como a evolução dos
veículos elétricos e as smart grids fazem com que a gestão da rede elétrica, cada vez mais
complexa, seja um desafio constante por parte de todos os seus intervenientes.
Neste contexto, o planeamento e a exploração das redes elétricas, em especial a rede de
distribuição de MT, apresentam-se como fulcrais de forma a garantir a satisfação dos requisitos
técnicos impostos pelo natural funcionamento do sistema e pela regulamentação mas, também,
na obtenção do retorno financeiro dos investimentos efetuados em torno da rede elétrica.
Dada a complexidade do problema e atendendo à necessidade de tratar informação
imprecisa, vaga e em certos casos ambígua, de satisfazer diversos objetivos que, em muitas
situações, encontram-se em conflito entre si, é necessário recorrer a avançadas técnicas de
otimização que utilizam complexos algoritmos matemáticos auxiliados por potentes meios
informáticos (hardware e software) com vista à obtenção da melhor solução para o problema.
Este problema do planeamento e da operação das redes de distribuição em MT, dada a
sua relevância para a sociedade, tem sido do interesse das operadoras e de diversos
investigadores desde meados do século XX. Em numerosas referências de revistas técnicocientíficas
são sugeridas metodologias, mais ou menos aproximadas à realidade, e abordagens
algorítmicas com vista à melhoria das soluções do problema. Nesta tese, o problema é formulado com as não linearidades inerentes ao tratamento
matemático e é resolvido por recurso a técnicas determinísticas. É apresentada a formulação
matemática do problema do planeamento e do problema da reconfiguração bem como todo o
conjunto de restrições técnicas a que a rede deve obedecer. Dado que o tipo de problema de
decisão tratado pode ser relacionado hierarquicamente também pode ser decomposto em subproblemas.
A técnica de decomposição utilizada foi a Decomposição de Benders. Para o problema
do planeamento consideraram-se três objetivos: minimização do custo das perdas, minimização
do custo da energia não fornecida e minimização do custo de investimento em baterias de
condensadores; no caso do problema da operação, e de forma a obter-se a reconfiguração ótima
da rede de distribuição, na sequência de uma avaria num dos seus componentes, considerou-se
a minimização das perdas e da energia não fornecida.
Os resultados apresentados nesta tese refletem a aplicação da metodologia definida para
o planeamento e para a reconfiguração em duas redes: uma rede de teste 70 barramentos e uma rede real de 201 barramentos. Face aos resultados, comprovou-se a adequação, robustez e versatilidade da metodologia que poderá ser um auxílio muito importante para a tomada de decisão quer ao nível do planeamento quer ao nível da operação de redes de distribuição de energia elétrica em MT.
In last decades the electricity sector has witnessed profound changes both in its technical and economic aspects. Nowadays, we watch the growing of economic power influence in management of the electricity industry. The load increasing, the use of distributed generation, managed by private entities, the technological evolution of network equipment and the sophisticated means of control, the liberalization of the electricity sector, the updating laws, the future trends such as the development of electric vehicles and smart grids make the management of the power grid a hard and complex work for all stakeholders. In this context, the planning and operation of electrical networks, in particular the distribution network, are presented as the key to ensure the satisfaction of the technical constraints imposed by the operating and regulation of the system. Also, the planning and operation help in getting the return of financial investments made in the network. Given the complexity of the problem which in many cases have to deal with fuzzy information, vague and in some cases ambiguous and also to satisfy several objective, in many cases, are in conflict with each other, it is necessary to use advanced optimization techniques. Mathematical complex algorithms helped by powerful computer tools (hardware and software) are used to solve that optimization problems in order to obtain the best solution. The distribution networks planning and operation problem given its importance to the society, has been of the interest of operators and researchers since the mid-twentieth century. Numerous references in technical and scientific journals propose methodologies close to the reality and several algorithmic approaches in order to improve the solutions of the problem. In this thesis, the problem is formulated with the nonlinearities inherent to the mathematical treatment and is solved by using deterministic techniques. A mathematical formulation for the planning problem and for the reconfiguration problem, as well as the full set of technical constraints of the network, are presented. Once the type of considered decision problem may be related hierarchically, so it can also be decomposed into sub-problems, the decomposition technique used was the Benders Decomposition. In the planning problem were considered three objectives: minimize the system´s power losses cost; minimize the capacitor placement and size cost, and minimize the energy not supplied cost. In the reconfiguration problem the objectives considered were the minimization of losses cost and minimization of energy not supplied cost in presence of a fault. The results presented in this thesis reflect the application of the proposed methodology for the planning and reconfiguration of two networks: a test network with 70 bus and a real network with 201 bus. The results, proved the adequacy, robustness and versatility of the proposed methodology which can be an important support tool for the decision-making, both in terms of planning and operation of the power distribution networks.
In last decades the electricity sector has witnessed profound changes both in its technical and economic aspects. Nowadays, we watch the growing of economic power influence in management of the electricity industry. The load increasing, the use of distributed generation, managed by private entities, the technological evolution of network equipment and the sophisticated means of control, the liberalization of the electricity sector, the updating laws, the future trends such as the development of electric vehicles and smart grids make the management of the power grid a hard and complex work for all stakeholders. In this context, the planning and operation of electrical networks, in particular the distribution network, are presented as the key to ensure the satisfaction of the technical constraints imposed by the operating and regulation of the system. Also, the planning and operation help in getting the return of financial investments made in the network. Given the complexity of the problem which in many cases have to deal with fuzzy information, vague and in some cases ambiguous and also to satisfy several objective, in many cases, are in conflict with each other, it is necessary to use advanced optimization techniques. Mathematical complex algorithms helped by powerful computer tools (hardware and software) are used to solve that optimization problems in order to obtain the best solution. The distribution networks planning and operation problem given its importance to the society, has been of the interest of operators and researchers since the mid-twentieth century. Numerous references in technical and scientific journals propose methodologies close to the reality and several algorithmic approaches in order to improve the solutions of the problem. In this thesis, the problem is formulated with the nonlinearities inherent to the mathematical treatment and is solved by using deterministic techniques. A mathematical formulation for the planning problem and for the reconfiguration problem, as well as the full set of technical constraints of the network, are presented. Once the type of considered decision problem may be related hierarchically, so it can also be decomposed into sub-problems, the decomposition technique used was the Benders Decomposition. In the planning problem were considered three objectives: minimize the system´s power losses cost; minimize the capacitor placement and size cost, and minimize the energy not supplied cost. In the reconfiguration problem the objectives considered were the minimization of losses cost and minimization of energy not supplied cost in presence of a fault. The results presented in this thesis reflect the application of the proposed methodology for the planning and reconfiguration of two networks: a test network with 70 bus and a real network with 201 bus. The results, proved the adequacy, robustness and versatility of the proposed methodology which can be an important support tool for the decision-making, both in terms of planning and operation of the power distribution networks.
Descrição
Tese de Doutoramento em Engenharia Eletrotécnica e de Computadores
Palavras-chave
Rede elétrica , Otimização , Redes de distribuição , Modelo fuzzy , Simulação Monte Carlo