A smart grid cost/benefit analysis answers a series of economic questions that address the incremental benefits of each stage or decision point. Each stage of the economic analysis provides information about the incremental benefits of that stage with respect to the previous stage. With this approach stages that provide little or no economic benefits can be identified. In this study there are series of applications,-including quasi-steady state power flows over time-varying loads and costs of service, Monte Carlo simulations, reconfiguration for restoration, and coordinated control - that are used to evaluate the cost-benefits of a series of smart grid investments. In the electric power system planning process, engineers seek to identify the most cost-effective means of serving the load within reliability and power quality criteria. In order to accurately assess the cost of a given project, the feeder losses must be calculated. In the past, the feeder losses were estimated based upon the peak load and a calculated load factor for the year. The cost of these losses would then be calculated based upon an expected, fixed per-kWh generation cost. This dissertation presents a more accurate means of calculating the cost of losses, using hourly feeder load information and time-varying electric energy cost data. The work here attempts to quantify the improvement in high accuracy and presents an example where the economic evaluation of a planning project requires the more accurate loss calculation. Smart grid investments can also affect response to equipment failures where there are two types of responses to consider -blue-sky day and storm. Storm response and power restoration can be very expensive for electric utilities. The deployment of automated switches can benefit the utility by decreasing storm restoration hours. The automated switches also improve system reliably by decreasing customer interruption duration. In this dissertation a Monte Carlo simulation is used to mimic storm equipment failure events, followed by reconfiguration for restoration and power flow evaluations. The Monte Carlo simulation is driven by actual storm statistics taken from 89 different storms, where equipment failure rates are time varying. The customer outage status and durations are examined. Changes in reliability for the system with and without automated switching devices are investigated. Time varying coordinated control of Conservation Voltage Reduction (CVR) is implemented. The coordinated control runs in the control center and makes use of measurements from throughout the system to determine control settings that move the system toward optimum performance as the load varies. The coordinated control provides set points to local controllers. A major difference between the coordinated control and local control is the set points provided by the coordinated control are time varying. Reduction of energy and losses of coordinated control are compared with local control. Also eliminating low voltage problems with coordinated control are addressed. An overall economic study is implemented in the final stage of the work. A series of five evaluations of the economic benefits of smart grid automation investments are investigated. Here benefits that can be quantified in terms of dollar savings are considered here referred to as "hard dollar" benefits. Smart Grid investment evaluations to be considered include investments in improved efficiency, more cost effective use of existing system capacity with automated switches, and coordinated control of capacitor banks and voltage regulators. These Smart Grid evaluations are sequentially ordered, resulting in a series of incremental hard dollar benefits. Hard dollar benefits come from improved efficiency, delaying large capital equipment investments, shortened storm restoration times, and reduced customer energy use. The evaluation shows that when time varying loads are considered in the design, investments in automation can improve performance and significantly lower costs resulting in "hard dollar" savings. / Ph. D.

Este trabalho analisa a necessidade da aplicação dos compensadores estáticos de reativos em sistemas elétricos industriais e em sistemas de potência. Vários tipos de compensadores de reativos são apresentados e também é feita uma comparação entre eles. O trabalho está centralizado no estudo dos compensadores estáticos tiristorisados, principalmente o Reator Controlado a Tiristores (RCT) e o Capacitor Chaveado a Tiristores (CCT). São discutidas as influências de alguns parâmetros nas formas de onda do RCT. É feita a dedução das equações das componentes harmônicas de corrente e tensão do RCT. São também analisados os efeitos das harmônicas e os meios de atenuá-las, de modo a mantê-las dentro de níveis razoáveis. Modelos para a simulação de um RCT, para vários tipos de fenômenos em circuitos monofásicos e trifásicos, são apresentados. São Mostrados gráficos e tabelas de algumas simulações utilizando o programa CSMP (Continuous System Modeling Program) e um programa desenvolvido ciclo a ciclo, utilizando o método passo a passo. As simulações permitem avaliar com razoável precisão o comportamento do RCT associado a um capacitor fixo. Sugestões de estudos afins são apresentadas no final do trabalho. / This work analyses the need of using Var systems in industrial and in power electric systems. Various types of VAr systems are presented and also a comparison among them is done. The work is centralized in the study of thyristor static VAr systems. The thyristor controlled reactor (TCR) and the thyristor switched capacitor (TSC) are analysed in more details. Parametric influences in the wave forms of TCR are discussed. TCR voltage and current harmonic equations are deduced. Harmonic effects and also the methods of producing their attenuation are analysed. TCR simulation models for multiple types of phenomena in one phase and three phase circuits are presented. Some graphics and tables using the CSMP (Continuous System Modeling Program), and a cycle by cycle program that uses the step method are showed. The simulations can evaluate with a reasonable precision the behavior of a TCR with a fixed capacitor (FC). Suggestions of other studies are presented at the end of the work.

O presente trabalho tem como objetivo comparar o desempenho de diferentes metaheurísticas bio-inspiradas aplicadas à resolução de problemas de otimização multiobjetivo do controle de tensão e reativos, ou controle Volt/VAr, em redes elétricas inteligentes. Entre os algoritmos implementados estão o algoritmo genético, o algoritmo memético, a otimização por colônia de formigas, a otimização por enxame de partículas e o strength pareto evolutionary algorithm (SPEA2). Aplicações dos algoritmos à resolução de problemas de otimização do controle Volt/VAr, em redes de distribuição de energia elétrica com dimensões reais, são utilizadas para comparação de seus indicadores de desempenho. A avaliação é feita tanto em relação à velocidade de busca quanto em relação à qualidade da solução encontrada. Os algoritmos mostraram resultados promissores para aplicação a redes de distribuição com dimensões reais, encontrando soluções de qualidade em tempos de busca aceitáveis. Parte deste desempenho se dá pelos métodos meta-heurísticos, parte por conta da modelagem adotada no processo de otimização. / The present work aims at comparing the performance of different bio-inspired metaheuristics applied to the Volt/VAr control multiobjective optimization problem in smart grids. Among the algorithms implemented are the genetic algorithm, the memetic algorithm, ant colony optimization, particle swarm optimization and the strength pareto evolutionary algorithm (SPEA2). Applications of the algorithms to solve Volt/VAr control optimization problems in distribution networks with real dimensions are used to compare their performance indicators. The evaluation is done both in relation to the search speed and in relation to the quality of the solution found. The promising results show that the algorithms are applicable to distribution networks with real dimensions, finding quality solutions in acceptable search times. This performance is obtained due to both the metaheuristic methods, and the modeling adopted in the optimization process.

Este trabalho analisa a necessidade da aplicação dos compensadores estáticos de reativos em sistemas elétricos industriais e em sistemas de potência. Vários tipos de compensadores de reativos são apresentados e também é feita uma comparação entre eles. O trabalho está centralizado no estudo dos compensadores estáticos tiristorisados, principalmente o Reator Controlado a Tiristores (RCT) e o Capacitor Chaveado a Tiristores (CCT). São discutidas as influências de alguns parâmetros nas formas de onda do RCT. É feita a dedução das equações das componentes harmônicas de corrente e tensão do RCT. São também analisados os efeitos das harmônicas e os meios de atenuá-las, de modo a mantê-las dentro de níveis razoáveis. Modelos para a simulação de um RCT, para vários tipos de fenômenos em circuitos monofásicos e trifásicos, são apresentados. São Mostrados gráficos e tabelas de algumas simulações utilizando o programa CSMP (Continuous System Modeling Program) e um programa desenvolvido ciclo a ciclo, utilizando o método passo a passo. As simulações permitem avaliar com razoável precisão o comportamento do RCT associado a um capacitor fixo. Sugestões de estudos afins são apresentadas no final do trabalho. / This work analyses the need of using Var systems in industrial and in power electric systems. Various types of VAr systems are presented and also a comparison among them is done. The work is centralized in the study of thyristor static VAr systems. The thyristor controlled reactor (TCR) and the thyristor switched capacitor (TSC) are analysed in more details. Parametric influences in the wave forms of TCR are discussed. TCR voltage and current harmonic equations are deduced. Harmonic effects and also the methods of producing their attenuation are analysed. TCR simulation models for multiple types of phenomena in one phase and three phase circuits are presented. Some graphics and tables using the CSMP (Continuous System Modeling Program), and a cycle by cycle program that uses the step method are showed. The simulations can evaluate with a reasonable precision the behavior of a TCR with a fixed capacitor (FC). Suggestions of other studies are presented at the end of the work.

O presente trabalho tem como objetivo comparar o desempenho de diferentes metaheurísticas bio-inspiradas aplicadas à resolução de problemas de otimização multiobjetivo do controle de tensão e reativos, ou controle Volt/VAr, em redes elétricas inteligentes. Entre os algoritmos implementados estão o algoritmo genético, o algoritmo memético, a otimização por colônia de formigas, a otimização por enxame de partículas e o strength pareto evolutionary algorithm (SPEA2). Aplicações dos algoritmos à resolução de problemas de otimização do controle Volt/VAr, em redes de distribuição de energia elétrica com dimensões reais, são utilizadas para comparação de seus indicadores de desempenho. A avaliação é feita tanto em relação à velocidade de busca quanto em relação à qualidade da solução encontrada. Os algoritmos mostraram resultados promissores para aplicação a redes de distribuição com dimensões reais, encontrando soluções de qualidade em tempos de busca aceitáveis. Parte deste desempenho se dá pelos métodos meta-heurísticos, parte por conta da modelagem adotada no processo de otimização. / The present work aims at comparing the performance of different bio-inspired metaheuristics applied to the Volt/VAr control multiobjective optimization problem in smart grids. Among the algorithms implemented are the genetic algorithm, the memetic algorithm, ant colony optimization, particle swarm optimization and the strength pareto evolutionary algorithm (SPEA2). Applications of the algorithms to solve Volt/VAr control optimization problems in distribution networks with real dimensions are used to compare their performance indicators. The evaluation is done both in relation to the search speed and in relation to the quality of the solution found. The promising results show that the algorithms are applicable to distribution networks with real dimensions, finding quality solutions in acceptable search times. This performance is obtained due to both the metaheuristic methods, and the modeling adopted in the optimization process.

Power generation from solar energy has significantly increased, and the growth is projected to continue in the foreseeable future. The main challenge of dealing with solar energy is its intermittent nature. The received irradiation energy of the sun on the earth's surface can fluctuate in a matter of seconds and cause voltage issues to power systems. Considering the high growth rate of solar photovoltaic (PV) resources, it is essential to be prepared to encounter and manage their high penetration levels. Currently, simplified approaches are used to model the impacts of cloud shadows on power systems. Using outdated standards also limits the penetration levels more than required. Approximately 40% of the new PV installations are residential, or installed at a low voltage level. Currently, all components between utility distribution transformers and customers/loads are either ignored or modeled with oversimplification. Furthermore, large PV systems require a considerable amount of land. However, point sensor models are currently used to simulate those systems. With a point model, the irradiance values measured at a point sensor are used to represent the output of a large PV system. However, in reality, clouds cover photovoltaic resources gradually and if the solar arrays are widespread over a large geospatial area, it takes some time for clouds to pass over the solar arrays. Finally, before 2014, participation of small-scale renewable resources was not allowed in controlling voltage. However, they can contribute significantly in voltage regulation. The main objective of this dissertation is to address the abovementioned issues in order to increase the penetration levels as well as precisely identify and locate voltage problems. A time-series analysis approach is used in modeling cloud motion. Using the time-series approach, changes of the received irradiation energy of the sun due to cloud shadows are simulated realistically with a Cloud Motion Simulator. Moreover, the use of the time-series approach allows implementation of new grid codes and standards, which is not possible using the old step change methods of simulating cloud impacts. Furthermore, all electrical components between utility transformers and customers are modeled to eliminate the inaccuracy due to using oversimplified models. Distributed PV models are also developed and used to represent large photovoltaic systems. In addition, the effectiveness of more distributed voltage control schemes compared to the traditional voltage control configurations is investigated. Inverters connect renewable energy resources to the power grid and they may use different control strategies to control voltage. Different control strategies are also compared with the current practice to investigate voltage control performance under irradiation variations. This dissertation presents a comprehensive approach to study impacts of solar PV resources. Moreover, simulation results show that by using time-series analysis and new grid codes, as well as employing distributed PV models, penetration of solar PV resources can increase significantly with no unacceptable voltage effects. It is also demonstrated that detailed secondary models are required to accurately identify locations with voltage problems. / PHD

Atualmente, as concessionárias de energia enfrentam vários desafios na tentativa de fornecer energia com qualidade e confiabilidade, por exemplo, redes de distribuição com pouco monitoramento e controle, envelhecimento dos ativos, atender a crescente demanda dos consumidores e satisfazer padrões de qualidade, entre outros. No Brasil, para atingir os índices de qualidade no fornecimento de energia estabelecidos pela Agência Nacional de Energia Elétrica (ANEEL), tem-se como alternativa aumentar a inteligência da rede com sistemas de tecnologia de informação (TI), além de utilizar Dispositivos Eletrônicos Inteligentes (IEDs), medidores inteligentes e sensores. Adicionalmente é necessário o emprego de uma infraestrutura de comunicação que forneça uma visibilidade abrangente das condições de carga e do estado do sistema de distribuição, permitindo fazer uso de técnicas de otimização e tomada de decisão para aumento da inteligência. O controle da tensão e o gerenciamento da potência reativa (Volt/VAr) são práticas de automação que podem ser empregadas nos sistemas de distribuição. Estas práticas ajudam a aumentar o benefício, a confiabilidade e a eficiência no fornecimento da qualidade do produto, minimizando os custos de operação e manutenção nas concessionárias. Na presente dissertação foi desenvolvida uma metodologia de controle Volt/VAr descentralizado (subestação), por meio da coordenação e controle dos transformadores com mudança de tap sob carga - LTC, dos bancos de capacitores localizados na subestação e ao longo dos alimentadores, bem como dos reguladores de tensão instalados ao longo dos alimentadores. Destaca-se a utilização da técnica de Estimação de Estados, a qual possibilita o uso da metodologia de controle Volt/VAr em tempo real. O algoritmo de otimização desenvolvido neste trabalho foi implementado através de simulações numa rede teste e numa rede real que permitem mostrar o potencial da formulação matemática desenvolvida. Os resultados da metodologia proposta comprovam a melhoria do perfil de tensão dos alimentadores da rede testada ao longo do dia, assim como uma diminuição das perdas técnicas, as quais chegaram a diminuir 7% em energia. Estes resultados foram alcançados sem ultrapassar a quantidade de operações definidas por dia nos diferentes dispositivos de controle. / Currently, power utilities face challenges in order to provide energy considering minimum requirements of power quality and reliability due to the absence of control and monitoring infrastructure, aging assets, growing demand, amongst others. In Brazil, the Agência Nacional de Energia Elétrica (ANEEL) establishes specific power quality and reliability indices. Power utilities should attend the limits for those indices by improving intelligence of network using IT systems, installing Intelligent Electric Devices (IEDs), smart meters and sensors, amongst others. Additionally, communication infrastructure is needed to provide the monitoring of load conditions and distribution system state. Optimization and decision-making tools are important for developing smart grids. Voltage control and reactive power management (Volt/VAr) are automation techniques that can be used in distribution systems. These practices lead to benefits for the utilities such as improvement of reliability and efficiency of the system, minimization of operation and maintenance costs, amongst others. In the present dissertation a methodology for decentralized control (Volt/VAr) was developed, through the coordination and control of load tap changers (LTCs) in power transformers and capacitors banks located at the substation, as well as capacitor banks and voltage regulators scattered throughout the feeders. State Estimation was utilized, which allows the application of the Volt/VAr Control methodology in real time environment. The optimization algorithm developed in this work was implemented through simulations in a test network and in a real network which validated the proposed methodology showing the potentiality of the mathematical model. The results of the proposed method showed voltage profile improvement over the day for the network tested as well as a 7% decrease in power energy losses. These results were achieved without exceeding the number of operations per day defined for all control devices.

Atualmente, as concessionárias de energia enfrentam vários desafios na tentativa de fornecer energia com qualidade e confiabilidade, por exemplo, redes de distribuição com pouco monitoramento e controle, envelhecimento dos ativos, atender a crescente demanda dos consumidores e satisfazer padrões de qualidade, entre outros. No Brasil, para atingir os índices de qualidade no fornecimento de energia estabelecidos pela Agência Nacional de Energia Elétrica (ANEEL), tem-se como alternativa aumentar a inteligência da rede com sistemas de tecnologia de informação (TI), além de utilizar Dispositivos Eletrônicos Inteligentes (IEDs), medidores inteligentes e sensores. Adicionalmente é necessário o emprego de uma infraestrutura de comunicação que forneça uma visibilidade abrangente das condições de carga e do estado do sistema de distribuição, permitindo fazer uso de técnicas de otimização e tomada de decisão para aumento da inteligência. O controle da tensão e o gerenciamento da potência reativa (Volt/VAr) são práticas de automação que podem ser empregadas nos sistemas de distribuição. Estas práticas ajudam a aumentar o benefício, a confiabilidade e a eficiência no fornecimento da qualidade do produto, minimizando os custos de operação e manutenção nas concessionárias. Na presente dissertação foi desenvolvida uma metodologia de controle Volt/VAr descentralizado (subestação), por meio da coordenação e controle dos transformadores com mudança de tap sob carga - LTC, dos bancos de capacitores localizados na subestação e ao longo dos alimentadores, bem como dos reguladores de tensão instalados ao longo dos alimentadores. Destaca-se a utilização da técnica de Estimação de Estados, a qual possibilita o uso da metodologia de controle Volt/VAr em tempo real. O algoritmo de otimização desenvolvido neste trabalho foi implementado através de simulações numa rede teste e numa rede real que permitem mostrar o potencial da formulação matemática desenvolvida. Os resultados da metodologia proposta comprovam a melhoria do perfil de tensão dos alimentadores da rede testada ao longo do dia, assim como uma diminuição das perdas técnicas, as quais chegaram a diminuir 7% em energia. Estes resultados foram alcançados sem ultrapassar a quantidade de operações definidas por dia nos diferentes dispositivos de controle. / Currently, power utilities face challenges in order to provide energy considering minimum requirements of power quality and reliability due to the absence of control and monitoring infrastructure, aging assets, growing demand, amongst others. In Brazil, the Agência Nacional de Energia Elétrica (ANEEL) establishes specific power quality and reliability indices. Power utilities should attend the limits for those indices by improving intelligence of network using IT systems, installing Intelligent Electric Devices (IEDs), smart meters and sensors, amongst others. Additionally, communication infrastructure is needed to provide the monitoring of load conditions and distribution system state. Optimization and decision-making tools are important for developing smart grids. Voltage control and reactive power management (Volt/VAr) are automation techniques that can be used in distribution systems. These practices lead to benefits for the utilities such as improvement of reliability and efficiency of the system, minimization of operation and maintenance costs, amongst others. In the present dissertation a methodology for decentralized control (Volt/VAr) was developed, through the coordination and control of load tap changers (LTCs) in power transformers and capacitors banks located at the substation, as well as capacitor banks and voltage regulators scattered throughout the feeders. State Estimation was utilized, which allows the application of the Volt/VAr Control methodology in real time environment. The optimization algorithm developed in this work was implemented through simulations in a test network and in a real network which validated the proposed methodology showing the potentiality of the mathematical model. The results of the proposed method showed voltage profile improvement over the day for the network tested as well as a 7% decrease in power energy losses. These results were achieved without exceeding the number of operations per day defined for all control devices.