Abstract not available

Reactive power is critical to reliable operation of the modern AC power system. There is a plethora of motor-loads, transformers, and power-electronic loads connected to the power grid, which consume reactive power for normal operation. Transmission lines also consume reactive power when they are loaded above their surge impedance loading (SIL). Reactive power can exact opportunity cost due to reduced capacity of the lines to carry real power, which in turn lowers revenue. Most transmission owners (TOs) levy large penalties against load serving entities (LSEs), industrial facilities, and other end-use customers, who consume more than their allotted amount, as measured by their power factor. These penalties are to incentivize their customers to meet their reactive power needs locally as well as to recuperate the TOs' financial losses. Harmonic pollution is another factor that prevents the optimal operation of the grid and the connected loads. Harmonics are attributable to proliferation of the diode-rectifier- or thyristor-rectifier-interfaced loads such as variable speed ac drives and power supplies in server farms, electric arc furnaces, and other non-linear loads, which are widely employed by the industrial sector. With wider adoption of harmonic-rich loads by the consumer sector as well, such as HDTVs and compact fluorescent lamps (CFLs), greater level of triplen harmonics associated with single-phase loads are also increasingly seen on the distribution grid. The increasing penetration of renewable resources and electrification of light-duty vehicles are expected to further aggravate the stresses and congestion on the utility grid. Reactive power compensation is necessary for supporting the AC grid and maintaining a healthy voltage stability margin. Compensation can also enhance the utilization of system capacity, lower system losses, provide fault ride-through, and enable a quick fault recovery. Existing VAR and harmonic compensation technologies are either too expensive or inadequate to meet the dynamic needs of the modern and the future power system. This dissertation presents a novel class of Dynamic VAR and Harmonic Compensators (DVHCs) for supplying or absorbing reactive power and providing harmonic filtering, where the compensation is in shunt with the line and the load. The underlying concept is based on augmenting a static or passive component like a capacitor or an inductor with a direct AC converter and imbuing the passive component with dynamic properties. The direct AC converter can be configured as a buck, a boost, or a buck-boost. A `fail-normal' switch is an integral part of the DVHCs that bypasses the converter when it fails, preserving the original functionality and the reliability of the passive component. The DVHCs are modular and scalable such that they can be employed in applications ranging from residential and industrial with voltages less than 480 V, to power distribution level with voltages as high as 35 kV. The Dynamic Inductor (D-IND) and the Dynamic Capacitor (D-CAP) are subclasses of the DVHCs. As the applications for supplying leading VARs are more prevalent, the primary focus of this work is on the buck, the boost, and the buck-boost configurations of the D-CAP. To understand the characteristics and operation of the DVHCs, this work has developed time-domain models for analyzing the transient and dynamic behavior; frequency-domain models for understanding the harmonic interactions and the steady-state relationships between switch duty and current harmonics; and small-signal models for studying the dynamics of the converter due to various perturbations. The small-signal models also enable extraction of transfer functions in designing controllers and assessing stability margins. Control architectures and techniques are presented for effectively controlling the D-CAP when commutating the semiconductor devices with both high and low switching frequencies. In modularly scaling the DVHCs to higher voltages, three medium-voltage topologies are discussed. They are based on series-connecting fractionally-rated devices, AC flying capacitors, and series cascading multiple two-level cells. These implementations allow direct connect to the medium-voltage grid, thereby obviating the use of transformers, and subsequently reducing the losses, cost, complexity, and footprint. A novel AC snubber concept is proposed to provide safe commutation of the AC switches, fault tolerance by managing the energy trapped in parasitics and filters, and to enable dynamic and static voltage sharing when integrated around the series-connected devices. Design equations for selecting and rating the devices and components in the buck, the boost, and the buck-boost configurations of the D-CAP are presented. Three sets of example designs, with one at low-voltage and two at medium-voltage, are discussed to demonstrate the typical size and ratings of the various components under realistic operating conditions. Measurements and the related discussions of a 40 kVA buck D-CAP prototype built to validate the effectiveness of the proposed concepts are presented.

Este trabalho apresenta o projeto e a implementação prática de um método completo para ser utilizado na conexão de UPSs monofásicos em paralelo. Existe um algoritmo que executa o droop de fase/frequência, e um novo método que trabalha com e sem comunicação de dados entre os inversores. Quando a comunicação está ativa, um novo algoritmo é utilizado, inserindo um resistência virtual variável, junto com o compartilhamento de potência ativa, obtendo um baixo valor de THD (Total Harmonic Distortion) na tensão de saída e bom compartilhamento de potência. Quando a comunicação de dados não funciona, uma resistência virtual constante é inserida, aumentando a THD de saída com carga não-linear, mas ainda proporcionando um bom compartilhamento de potência ativa. A vantagem é poder obter um bom desempenho quando a comunicação de dados está operando, mas, no caso de sua falha, o sistema ainda funciona, proporcionando maior confiabilidade. A implementação possui um algoritmo de emulação de carga eletrônica, com o propósito de executar testes de produção, baseado no mesmo algoritmo de paralelismo, apenas mudando algumas variáveis. / This work presents the design and experimental implementation of a complete paralleling method to be used for parallel single-phase UPSs connection. There is a algorithm that performs a phase/frequency droop, and a new method to work with or without data communication among the inverters. When communication is working, a new algorithm is used, inserting a variable virtual resistance in the output, along with active power sharing, obtaining a low output voltage THD (Total Harmonic Distortion) value and good power sharing. Without communication a constant virtual resistance is inserted, increasing the output THD with non-linear load, but still allowing a good active power sharing. The advantage is to obtain a good performance operation with communication, but, in case of communication failure, the system still works providing more reliability. The implementation have a electronic load emulation algorithm, with purpose to execute factory tests, based in the same parallelism algorithm, just changing some variables.

Este trabalho apresenta o projeto e a implementação prática de um método completo para ser utilizado na conexão de UPSs monofásicos em paralelo. Existe um algoritmo que executa o droop de fase/frequência, e um novo método que trabalha com e sem comunicação de dados entre os inversores. Quando a comunicação está ativa, um novo algoritmo é utilizado, inserindo um resistência virtual variável, junto com o compartilhamento de potência ativa, obtendo um baixo valor de THD (Total Harmonic Distortion) na tensão de saída e bom compartilhamento de potência. Quando a comunicação de dados não funciona, uma resistência virtual constante é inserida, aumentando a THD de saída com carga não-linear, mas ainda proporcionando um bom compartilhamento de potência ativa. A vantagem é poder obter um bom desempenho quando a comunicação de dados está operando, mas, no caso de sua falha, o sistema ainda funciona, proporcionando maior confiabilidade. A implementação possui um algoritmo de emulação de carga eletrônica, com o propósito de executar testes de produção, baseado no mesmo algoritmo de paralelismo, apenas mudando algumas variáveis. / This work presents the design and experimental implementation of a complete paralleling method to be used for parallel single-phase UPSs connection. There is a algorithm that performs a phase/frequency droop, and a new method to work with or without data communication among the inverters. When communication is working, a new algorithm is used, inserting a variable virtual resistance in the output, along with active power sharing, obtaining a low output voltage THD (Total Harmonic Distortion) value and good power sharing. Without communication a constant virtual resistance is inserted, increasing the output THD with non-linear load, but still allowing a good active power sharing. The advantage is to obtain a good performance operation with communication, but, in case of communication failure, the system still works providing more reliability. The implementation have a electronic load emulation algorithm, with purpose to execute factory tests, based in the same parallelism algorithm, just changing some variables.

Orientador: Ernesto Ruppert Filho / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Eletrica e de Computação / Made available in DSpace on 2018-08-04T17:39:56Z (GMT). No. of bitstreams: 1 Villalva_MarceloGradella_M.pdf: 5074196 bytes, checksum: 39f4e09d9f1f624c909a5ee7f2d3fa4d (MD5) Previous issue date: 2005 / Resumo: o tema da qualidade de energia elétrica adquiriu grande importância em anos recentes. O número crescente de cargas baseadas em conversores eletrônicos criou a preocupação com a distorção harmônica em sistemas de energia elétrica. Qualidade de energia é um assunto importante tanto para as companhias distribuidoras e geradoras de energia elétrica como para os consumidores. A poluição harmônica pode causar sérios problemas às companhias e aos consumidores. Alguns dos possíveis problemas são a instabilidade na operação do sistema, distorções de tensão, perdas de energia, interferência eletromagnética e ressonâncias perigosas. A utilização de equipamentos para eliminação de harmônicos tornou-se necessária e são versas as soluções possíveis. Tradicionalmente se utilizam Itros passivos do tipo série ou paralelo. Esses Itros resentam como desvantagem suas grandes dimensões, seu custo elevado, elevadas perdas de energia, pouca e cácia, dependendência da interação com parâmetros do sistema e risco de introdução de novas freqüências de ressonância. O desenvolvimento recente da eletrônica de potência e do processamento digital de sinais tornou possível a utilização de Itros ativos de potência, que são condicionadores eletrônicos de energia. Filtros ativos de potência podem ser utilizados para realizar a Itragem de harmônicos em sistemas elétricos, sem os problemas apresentados pelos dispositivos passivos. Filtros ativos série são utilizados para eliminar distorções de tensão e Itros ativos paralelos são usados para eliminar correntes harmônicas originadas por carga não lineares. Esses últimos podem ainda ser usados para corrigir o fator de potência e para equilibrar as correntes de linha em sistemas desbalanceados com quatro os. Este trabalho estuda a aplicação de Itros ativos paralelos a sistemas trifásicos com quatro os. São estudados teorias de potências, métodos de compensação de harmônicos e sistemas de controle. A principal contribuição do trabalho está no emprego de um método seletivo de compensação de correntes baseado em redes neurais adaptativas, além do estudo do emprego de redes neurais no controle de correntes. Esses e outros assuntos foram experimentalmente comprovados com um protótipo de Itro ativo construí do em laboratório / Abstract: Electricity power quality has gained increased importance in the past few years. The growing number of power electronicsbased loads has created a preoccupation about harrnonic distortion in electric systems. Electricity power quality is an important issue both for energy utilities and consumers. Harrnonic poIlution may cause severe problems to electricity utilities and consumer networks such as system instability, voltage distortions, power losses, electromagnetic interference and harrnful resonances. The use ofharrnonic ltering equipment has become necessary and many solutions have been studied. Traditional passive devices such as tuned shunt lters and series reactors present disadvantages such as bulky sizes, high costs, increased losses, smaIl effectiveness, dependence on the parameters of the electric system and yet more risk of resonance with other elements of the network. Recent deve10pments in the major of power electronics and digital processing have made possible the pplication of electronic power conditioners such as series and shunt active lters. Active power lters may be used to overcome the drawbacks of passive devices. Series active power lters are used to reduce voltage distortions of electric systems. Shunt active power lters are used to mitigate harrnonic currents originated by nonlinear loads. The latter may be also used for increasing power factor and for balancing line currents in four-wire systems with unbalanced loads. This work is concerned with the application of shunt active power lters to three-phase four-wire electric systems. Power theories, compensation methods and control systems are discussed. The application of a se1ective compensation method based on adaptive neural networks and the employment of a neural network in the current controlIer are the main contributions of this work. These and other subjects were experimentalIy tested with a prototype of an active. power lter built in laboratory / Mestrado / Energia Eletrica / Mestre em Engenharia Elétrica

Recent decades have witnessed a significant increment in power contribution from wind generators. This increment in penetration requires power engineers to tackle multi-fold challenges concerning operational and stability aspects. There exists a significant attention among the researchers in analyzing the impact of wind generation on various system aspects. This thesis focuses on steady state voltage stability aspects with penetration of Variable speed wind generators. Traditionally, ancillary services are supplied by large conventional generators. However, with the huge penetration of wind generators as a result of the growing interest in satisfying energy requirements, and considering the benefits that they can bring along to the electrical system and to the environment, it appears reasonable to assume that ancillary services could also be provided by wind generators in an economical and efficient way. Certain types of wind generators can support reactive power for the Power Market. Fixed and Semi-Variable speed wind turbine generators were predominantly employed during the early installations of wind generators. These units require reactive power support from the grid and are usually equipped with capacitor banks to provide the necessary reactive power. Further, STATic synchronous COMpensator (STATCOM) and Static Var Compensator (SVC) with various configurations were proposed to enhance the system operations under normal and disturbed conditions. On the other hand, Variable speed wind turbine generators provide flexibility in control and hence are becoming increasingly popular. Popular among this class of wind turbine generating units are Doubly Fed Induction Generator(DFIG) and induction/synchronous Generator with Front End Converter(GFEC). Contrary to Fixed and Semi-Variable speed wind machines, Variable speed wind turbine generators are capable of providing reactive power to the grid. The converter and control schemes associated with these machines permits controlling the active and reactive power output to desired level. It is possible to control the reactive power output of these machines independently of the active power. Researchers in the past have investigated the impact of reactive power output of variable speed wind machines on system stability. In literature, approaches are proposed to utilize the flexibility in reactive power generation of DFIG to reduce system losses, improve reliability in static and dynamic system operation. Approaches in literature investigate the impact on voltage stability of system by considering the flexibility of reactive power output of wind machine in isolation. However, significant improvement in static voltage stability, voltage profile, system power losses etc. can be attained if the Reactive Power output of Variable Speed wind machines can be properly coordinated with other reactive power controllers. The prime objective of the thesis is to propose algorithms to coordinate the reactive power output of Variable Speed Wind Generators with other reactive power controllers for Enhancement in voltage stability margin, system losses and voltage profiles.