The aim of this research is to develop an accurate method for the analysis of signals composed of fluctuating harmonics. The results obtained of analysis are applied to the calibration of harmonic analysis instruments. A new method is presented suitable for the accurate analysis of smoothly fluctuating harmonic signals. The method is based on a model of signals with a known period, in which the harmonics are individually modulated by polynomial functions normalised over a sampled signal sequence time. Using this model, a decomposition method is developed such that the modulating polynomials can be recovered from a signal. The polynomial decomposition method leads to a piece-wise analysis of the waveform. Two methods based on least squares and splines respectively, are developed with the aim of giving continuity to the piece-wise analysis. Comparisons of the new method with the short time Fourier transform are given. Having defined a test signal and obtained and accurate analysis of it properties, it can be used to calibrate harmonic analysers. For a given applied signal, analysis with these devices can give rise to variation in results as a function of the phase between the signal and the STFT windows. This result distribution due to variable phase (RDVP) is discussed and examples are given for various signals. The RDVP complicates the calibration process due to the spread of results that occur when testing the device. A method is developed to find the RDVP for an applied signal that uses the polynomial decomposition method to find the modulation functions of each harmonic in the applied calibration signal. Having found the RDVP for an applied signal, it is necessary to fit the results of the analyser under test, to the distribution. The random nature of the phase makes the systematic comparison of the theoretical and measured distributions difficult to achieve. A novel method that uses multiple phase shifted modulated harmonics is presented. By comparing the results of the analyser under test to the distributions of each of the phase-shifted harmonics, a best-fit phase shift can be determined and the required calibration comparison made. Key words: time-frequency analysis, demodulation, harmonic analysis, fluctuating harmonics, waveform metrology, calibration of harmonic analysers.

The theme of the filter of compensating devices current harmonics analyzed at present on work is actual, because concrete requirements on quality of electric energy are established. One of such requirements – size of harmonics in a feeding electric network. In industrial electric networks to which converters are connected, high harmonics on size reach 20-25 %. Special devices for eliminating harmonics are used. Condensers which are used for indemnification of reactive electric energy, do not extinguish them, therefore are projected the filter – compensatory installations, using capacity of the battery of condensers as base of the power filter. Bases of designing of the filter – compensatory devices of the parameters of semi-conductor converters of the electric power intended for normalization with a power line are considered. The purpose work of the master – the analysis of a factory electric network on the maximum harmonics and to design the filter for eliminating harmonics. The electric network has been checked up by device PQ ANALYSER. Such harmonics 3, 5, 7, 9, 11, 13 are found. The greatest on a current is third harmonic. Condensers of reactive capacity connected, network does not act on a harmonic. Compensatory installation which is made of separate filters of clearing of harmonics is designed, will strongly reduce the specified harmonics in a factory network.

Within scope of this work, double-fourier analysis method of rapid calculation and detailed simulation method, which are used to investigate DC bus current harmonics in two level and three level three-phase inverters systematically, will be emphasized and two methods will be compared via applying different modulation techniques. In addition, DC bus currents will be investigated visually for various working conditions and modulation methods. After that, analysis methods will be applied and harmonic spectrums will be determined. After all, it will be showed that calculated harmonic spectrums could be treated as unified harmonics around certain frequencies and these unified harmonics could be reached easily via looking at predetermined table. Moreover, it will also be showed that unified harmonic values could be used to determine harmonic current components that are necessary for sizing DC bus capacitor and could be used in various inverter analysis.

Current practice in U.S. underground coal mine high-voltage distribution systems is to attempt to limit ground fault current to 25 Amperes and de-energize the circuit at 10 Amperes. However, the significant amount of system capacitance due to the use of shielded cables can cause ground fault current to be two or three times the intended ground fault limit. Consequently, this practice can cause several issues such as ground fault currents significantly exceeding the neutral grounding resistor current limit, loss of relay selectivity in the distribution system, and transient overvoltages in certain ground fault situations. These issues are solved to some extent by using a resonance grounded system, currently used in some other countries. However, a shortcoming of traditional resonance grounded systems is the inability to deal with the harmonic components existing in ground fault current. With the increasing use of nonlinear sources such as variable frequency drives, the proportion of harmonic components in ground fault current can be significant. Consequently, although the fundamental component can be almost fully compensated in a traditional resonance grounded system, the harmonic components can still be large enough to maintain arcing and cause personal injury and equipment damage. In this dissertation, a novel method is developed to perform real-time prediction of the harmonics in ground fault currents. Methods for neutralizing the ground fault current harmonics and identifying ground fault location are also developed. Results indicate that the combination of traditional high-resistance grounding and active current injection to neutralize harmonics in the ground fault has the potential to significantly reduce the total ground fault current and reduce arc and flash hazards during ground faults in high voltage distribution systems.

Submitted by Haroudo Xavier Filho (haroudo.xavierfo@ufpe.br) on 2016-03-01T13:09:14Z No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) Silva Filho, Luis Rodrigues da.pdf: 5874183 bytes, checksum: 60b4246d84e077fb2131d50faca19bd3 (MD5) / Made available in DSpace on 2016-03-01T13:09:14Z (GMT). No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) Silva Filho, Luis Rodrigues da.pdf: 5874183 bytes, checksum: 60b4246d84e077fb2131d50faca19bd3 (MD5) Previous issue date: 2015-01-30 / CAPES / Atualmente, um grande número de computadores e outras cargas elétricas sensíveis ligadas à rede elétrica são diretamente afetados por problemas de qualidade de energia. Uma das questões mais importantes está relacionada com as correntes harmônicas geradas pelo aumento do número de cargas não-lineares conectadas ao sistema elétrico, tais como retificadores a diodo e tiristores. Como consequência, estes harmônicos podem causar distorções na tensão, perdas adicionais no sistema de energia e mau funcionamento de equipamentos eletrônicos sensíveis. A literatura apresenta diferentes topologias de filtros híbridos de potência visando a compensação harmônica de corrente. Este trabalho apresenta o estado da arte de algumas topologias, realizando uma análise comparativa entre elas. Em comum, todas as topologias possuem um reduzido número de chaves. Adicionalmente, é apresentada uma nova topologia chamada six-switch two-leg hybrid power filter, que além de apresentar um número reduzido de chaves quando comparada a outras topologias com dupla compensação, é capaz de fornecer compensação total, mesmo para cargas com um alto conteúdo harmônico. Os resultados experimentais são apresentados para um protótipo, a fim de demonstrar que o desempenho de compensação harmônica da proposta cumpre a norma IEEE 519. / Nowadays, a large number of computers and other sensitive electrical loads connected to the power grid are directly affected by power quality problems. One of the most important issues is related to current harmonics generated by the increasing number of nonlinear loads connected to the power grid, such as diode and thyristor front-end rectifiers. As consequence, these harmonics can cause voltage distortions, additional losses in the power system and malfunction of sensitive electronic equipment. The literature presents several hybrid power filters topologies, aiming current harmonic compensation. This work presents some state of - art topologies with reduced number of switches and performs a comparative analysis among them. Additionally, a six-switch two-leg hybrid power filter topology, that presents a reduced number of switches when compared with dual topologies, is proposed. The solution is capable of providing fully compensation even for loads with high harmonic content. Experimental results, obtaneid by a prototype, are presented in order to demonstrate that the harmonic compensation performance meets the IEEE 519 standard.

The proliferation of Electric Vehicles (EVs) poses new challenges for the Distribution System Operator (DSO). For example, the rectifiers that are used for charging EV batteries could significantly influence Power Quality (PQ), in terms of harmonic distortion. The emissions from individual EV, are well regulated by current harmonic emission standards. But what the aggregation of multiple EV loads will look like is still uncertain and the research being made in this area is still in its early stage.  The DSO responsibilities include ensuring grid code compliance confirmed by PQ metering.  In general, 10 minute RMS values are sufficient. However, the large scale integration of non-linear loads, like EVs, could lead to new dynamic phenomena, possibly lost in the process of time aggregation. In this thesis PQ and, in more detail, the concept of harmonics, and how this is related to EVs, is presented. A current-harmonic load model using power exponential functions and built from actual measurement data during the individual charging of four different fully electric vehicles was constructed. The model was based on individual emitted current harmonics as a function of state of charge (SOC), and was used to deterministically simulate the simultaneous charging of six vehicles fed from the same bus. The aggregation of current harmonics up to the 11th was simulated while randomizing battery SOC, the start of charging, and the kind of vehicle. Additionally, an investigation of the impact on aggregation in time was conducted. The analysis clearly shows the importance of phase angle information, its correlation to SOC, and how the aggregation of EV loads is influenced by these factors. The analysis also shows that 10 minute RMS aggregation could lead to significant deviations from the “actual” (200ms) data.  This indicates that 10 minute value monitoring could lead to information losses.

An electrical distribution network supplies power to many households at a time. Each house has various types of electrical appliances which consume different levels of power depending on its function and the load it supplies. An ideal case would be when no appliances cause power disturbances, however, in reality, many electrical appliances (e.g. refrigerators, computers, heat pumps, etc.), due to their inherent characteristics. The distribution network is now fed back with non-sinusoidal power, called harmonics, causing inefficiencies and potential failure in other appliances connected to the same network. Harmonics are the additional frequency components to the fundamental sinusoidal voltage or current (e.g. 50Hz fundamental frequency in Europe). To avoid the malfunction of the network, each electrical appliance in any household has to comply with the standard IEC 61000-3-2 (current rating $<$16A) defined by the International Electrotechnical Commission (IEC). According to these standards, current harmonics have been observed up to 40th harmonics. If the fundamental frequency is 50 Hz, then harmonic currents up to 2000 Hz have to be considered. Usually, the harmonics are quantified by the measurement when the appliance has been constructed at which point changes are costly and time-consuming.  This thesis investigates the possibility to model these appliances at an early stage of development enabling cost-effective and quick improvements. The model is achieved by reducing a heat pump into a set of chokes, an EMI filter, a rectifier, and a load represented by a variable current source to simulate the compressor which draws the majority of power the system uses. The state-of-the-art, referred to as white box modeling, uses schematics to model these components while this thesis presents a methodology to model without schematics, referred to as black box modeling.  For the black box model, impedance of the components is measured and equivalent circuits are modeled accordingly in Simulink to simulate the harmonics. The validity of a black box simulation model from the equivalent circuit extracted by the impedance measurements is investigated in this thesis. The black box modeling is compared with the white box modeling and measurements for the purpose of verification.   The result is good but more work is needed. The black box model is verified using two different systems which are similarly accurate. Simulating harmonics at a higher power level yields better results. The black box model is more accurate than the white box model, mainly due to the inclusion of parasitics in the EMI filter. Possible additional harmonics generation in unmodeled components is investigated and found to be insignificant.

Harmonic propagation studies of public distribution networks require accurate models of aggregate residential customers (groups of customers) that simulate the harmonic emission of the multitude of household appliances in the network. Most of the present models were developed with the component-based approach, where models of individual household appliances are combined to build the model of multiple customers. This approach requires high amount of input data, like models of individual household appliances and detail information of customer behavior and device composition, which is usually not easy to acquire. However, with the increasing number of PQ-analyzers in the networks, the measurement-based approach is now more and more considered for the modeling of aggregate customers. The measurement-based approach uses measurements of the network in combination with top-down methodologies to obtain models of the aggregate customers. Compared to the component-based approach it has several advantages, like inherent consideration of the real operating changes of the individual household appliances, variation of customer behavior, effect of line impedances, cancellation and attenuation effects, etc. This thesis presents the development of a time-series stochastic model of the low-order harmonic emission of aggregate residential customers based on a top-down measurement-based approach. The model represents the daily variation of the harmonic magnitudes and phase angles. Besides, the model includes the representation of the harmonic unbalances, which is of great importance for the proper analysis of harmonic propagation in medium-voltage networks. The model is parametrized for German networks, but the methodology can be applied to find the models of other regions or countries.

Harmonic propagation studies of public distribution networks require accurate models of aggregate residential customers (groups of customers) that simulate the harmonic emission of the multitude of household appliances in the network. Most of the present models were developed with the component-based approach, where models of individual household appliances are combined to build the model of multiple customers. This approach requires high amount of input data, like models of individual household appliances and detail information of customer behavior and device composition, which is usually not easy to acquire. However, with the increasing number of PQ-analyzers in the networks, the measurement-based approach is now more and more considered for the modeling of aggregate customers. The measurement-based approach uses measurements of the network in combination with top-down methodologies to obtain models of the aggregate customers. Compared to the component-based approach it has several advantages, like inherent consideration of the real operating changes of the individual household appliances, variation of customer behavior, effect of line impedances, cancellation and attenuation effects, etc. This thesis presents the development of a time-series stochastic model of the low-order harmonic emission of aggregate residential customers based on a top-down measurement-based approach. The model represents the daily variation of the harmonic magnitudes and phase angles. Besides, the model includes the representation of the harmonic unbalances, which is of great importance for the proper analysis of harmonic propagation in medium-voltage networks. The model is parametrized for German networks, but the methodology can be applied to find the models of other regions or countries.

Bei Erdschlüssen in Mittelspannungsnetzen mit Resonanz-Sternpunkterdung wird der einpolige Fehlerstrom, genannt Erdschluss-Reststrom, oft von den Oberschwingungsanteilen dominiert. Dadurch wird in manchen Netzen die betrieblich festgelegte Obergrenze für den Erdschluss-Reststrom, auf deren Grundlage die Erdungsanlagen ausgelegt wurden, überschritten. Für eine Bewertung hinsichtlich des Personenschutz muss geprüft werden, ob dies eine Überschreitung der zulässigen Berührungsspannung nach DIN EN 50522 zur Folge hat. Dazu wird die erdpotenzialanhebende Wirkung der Harmonischen im Erdschluss-Reststrom für repräsentative Anordnungen untersucht. Dies wird mittels eines Modells realisiert und durch Messergebnisse verifiziert. Hierbei werden jeweils Aussagen zur charakteristischen Stromverteilung im Erdungssystem, den wirksamen Erdungsimpedanzen sowie zur resultierenden Erdungsspannung getroffen. Wesentliche Einflussfaktoren werden benannt und eine vom Fehlerort abhängige Risikobewertung durchgeführt. Im Ergebnis kann eine Vielzahl potenzieller Fehlerorte als grundsätzlich unkritisch eingestuft und somit umfangreiche Untersuchungen sowie Fehlinvestitionen vermieden werden.