As perdas elétricas, que no Brasil giram em torno de 14,4%, são prejudiciais ao desempenho técnico e financeiro das concessionárias de energia elétrica, principalmente em um cenário de uma eminente crise energética e alta regulação. A mitigação das perdas elétricas está diretamente relacionada com sua correta estimação. Para operar um sistema de energia elétrica, é de fundamental importância definir a correta modelagem dos elementos do sistema. As metodologias de estimação das perdas de energia, para sistemas de distribuição, vigentes não levam em conta possíveis erros na correta modelagem das cargas conectadas. Desta forma, este trabalho tem o objetivo de analisar a influência causada pela utilização dos modelos de carga, na estimação das perdas elétricas em sistemas de distribuição. Esta análise abrange as metodologias de fluxo de carga backward-forward sweep por soma de corrente, por soma de potência e Newton-Raphson. A perda de energia é calculada pela diferença entre a energia injetada no sistema, menos a energia entregue. O estudo de caso é realizado em um sistema de distribuição teste de 13 barras da IEEE. No caso proposto, são realizados cálculos das perdas de energia para o sistema de distribuição, considerando diferentes modelos de carga. Desta maneira, o trabalho estimou a diferença no cálculo das perdas para cada tipo de modelo de carga, em relação a perdas calculadas com o padrão original das cargas. Outro ponto analisado foi o desempenho das metodologias de fluxo de carga, frente à alteração dos modelos de carga. Os resultados demonstram que a alteração dos modelos de carga influência a estimação das perdas elétricas nos sistemas de distribuição, e o desempenho dos fluxos de carga. / Electrical losses, which in Brazil are around 14.4%, are harmful to the technical and financial performance of electric utilities, especially in a scenario of an imminent energy crisis and high regulation. Mitigation of electrical losses is directly related to its correct estimation. To operate an electric power system, it is of fundamental importance to define the correct model of the system elements. The methodologies for estimating energy losses, for the existing distribution systems, do not take into account possible errors in the correct model of connected loads. Thus, this study aims to examine the influence caused by the use of different load models, in the estimation of electrical losses in distribution systems. This analysis covers the backward-forward sweep load flow methodologies by the sum of current, by the sum of power and Newton-Raphson. The energy loss is calculated by the difference between the energy injected into the system, minus the energy delivered. The case study is performed on the IEEE 13 Node Test Feeder. In the proposed case, calculations of energy losses in the distribution system are performed considering different load models. Therefore, the study has estimated the difference in the calculation of energy loss for each type of load model, for the losses calculated with the original pattern of loads. Another point discussed is the performance of load flow methodologies, related to the change of load models. The results have shown that the change in load models influence the estimation of electrical losses in distribution systems and in the performance of load flows.

As perdas elétricas, que no Brasil giram em torno de 14,4%, são prejudiciais ao desempenho técnico e financeiro das concessionárias de energia elétrica, principalmente em um cenário de uma eminente crise energética e alta regulação. A mitigação das perdas elétricas está diretamente relacionada com sua correta estimação. Para operar um sistema de energia elétrica, é de fundamental importância definir a correta modelagem dos elementos do sistema. As metodologias de estimação das perdas de energia, para sistemas de distribuição, vigentes não levam em conta possíveis erros na correta modelagem das cargas conectadas. Desta forma, este trabalho tem o objetivo de analisar a influência causada pela utilização dos modelos de carga, na estimação das perdas elétricas em sistemas de distribuição. Esta análise abrange as metodologias de fluxo de carga backward-forward sweep por soma de corrente, por soma de potência e Newton-Raphson. A perda de energia é calculada pela diferença entre a energia injetada no sistema, menos a energia entregue. O estudo de caso é realizado em um sistema de distribuição teste de 13 barras da IEEE. No caso proposto, são realizados cálculos das perdas de energia para o sistema de distribuição, considerando diferentes modelos de carga. Desta maneira, o trabalho estimou a diferença no cálculo das perdas para cada tipo de modelo de carga, em relação a perdas calculadas com o padrão original das cargas. Outro ponto analisado foi o desempenho das metodologias de fluxo de carga, frente à alteração dos modelos de carga. Os resultados demonstram que a alteração dos modelos de carga influência a estimação das perdas elétricas nos sistemas de distribuição, e o desempenho dos fluxos de carga. / Electrical losses, which in Brazil are around 14.4%, are harmful to the technical and financial performance of electric utilities, especially in a scenario of an imminent energy crisis and high regulation. Mitigation of electrical losses is directly related to its correct estimation. To operate an electric power system, it is of fundamental importance to define the correct model of the system elements. The methodologies for estimating energy losses, for the existing distribution systems, do not take into account possible errors in the correct model of connected loads. Thus, this study aims to examine the influence caused by the use of different load models, in the estimation of electrical losses in distribution systems. This analysis covers the backward-forward sweep load flow methodologies by the sum of current, by the sum of power and Newton-Raphson. The energy loss is calculated by the difference between the energy injected into the system, minus the energy delivered. The case study is performed on the IEEE 13 Node Test Feeder. In the proposed case, calculations of energy losses in the distribution system are performed considering different load models. Therefore, the study has estimated the difference in the calculation of energy loss for each type of load model, for the losses calculated with the original pattern of loads. Another point discussed is the performance of load flow methodologies, related to the change of load models. The results have shown that the change in load models influence the estimation of electrical losses in distribution systems and in the performance of load flows.

As perdas elétricas, que no Brasil giram em torno de 14,4%, são prejudiciais ao desempenho técnico e financeiro das concessionárias de energia elétrica, principalmente em um cenário de uma eminente crise energética e alta regulação. A mitigação das perdas elétricas está diretamente relacionada com sua correta estimação. Para operar um sistema de energia elétrica, é de fundamental importância definir a correta modelagem dos elementos do sistema. As metodologias de estimação das perdas de energia, para sistemas de distribuição, vigentes não levam em conta possíveis erros na correta modelagem das cargas conectadas. Desta forma, este trabalho tem o objetivo de analisar a influência causada pela utilização dos modelos de carga, na estimação das perdas elétricas em sistemas de distribuição. Esta análise abrange as metodologias de fluxo de carga backward-forward sweep por soma de corrente, por soma de potência e Newton-Raphson. A perda de energia é calculada pela diferença entre a energia injetada no sistema, menos a energia entregue. O estudo de caso é realizado em um sistema de distribuição teste de 13 barras da IEEE. No caso proposto, são realizados cálculos das perdas de energia para o sistema de distribuição, considerando diferentes modelos de carga. Desta maneira, o trabalho estimou a diferença no cálculo das perdas para cada tipo de modelo de carga, em relação a perdas calculadas com o padrão original das cargas. Outro ponto analisado foi o desempenho das metodologias de fluxo de carga, frente à alteração dos modelos de carga. Os resultados demonstram que a alteração dos modelos de carga influência a estimação das perdas elétricas nos sistemas de distribuição, e o desempenho dos fluxos de carga. / Electrical losses, which in Brazil are around 14.4%, are harmful to the technical and financial performance of electric utilities, especially in a scenario of an imminent energy crisis and high regulation. Mitigation of electrical losses is directly related to its correct estimation. To operate an electric power system, it is of fundamental importance to define the correct model of the system elements. The methodologies for estimating energy losses, for the existing distribution systems, do not take into account possible errors in the correct model of connected loads. Thus, this study aims to examine the influence caused by the use of different load models, in the estimation of electrical losses in distribution systems. This analysis covers the backward-forward sweep load flow methodologies by the sum of current, by the sum of power and Newton-Raphson. The energy loss is calculated by the difference between the energy injected into the system, minus the energy delivered. The case study is performed on the IEEE 13 Node Test Feeder. In the proposed case, calculations of energy losses in the distribution system are performed considering different load models. Therefore, the study has estimated the difference in the calculation of energy loss for each type of load model, for the losses calculated with the original pattern of loads. Another point discussed is the performance of load flow methodologies, related to the change of load models. The results have shown that the change in load models influence the estimation of electrical losses in distribution systems and in the performance of load flows.

Load models play an essential role in many studies, including calculating voltage drops and technical losses in distribution systems, for distributed generator (DG) integration planning, and in tariff analysis and design models. The Herman-Beta transform used in the low voltage network modelling studies in South Africa is based on loads modelled as Beta probability density functions. Recently, the transform was extended to make it useful also for probabilistic load flow modelling in medium voltage (MV) networks with non-unity power factor loads and DGs. The electricity supply industry in South Africa has transformed and saw an increased penetration of Independent Power Producers as a result of the government encouraged the renewable independent power procurement programme (REIPPP). There has also been a steady decrease in the costs of procuring power from renewable energy sources, mainly from photovoltaic (PV) systems. South Africa also saw significant tariff increases in the recent past. These have resulted in both new load patterns and uncertainties in the power systems inputs required for network planning and tariff development. Other factors affecting loads and renewable energy output include weather, location and economic factors. Load models are essential for technical and tariff studies. Long term and short term planning models in both technical and tariff modelling require information about the usage behaviour of customers. Planning cannot be separated from the financial impact and tariffs in general. The literature review indicated that planning has the objective of designing a network for optimal usage, thus minimising the costs and deferring investment where possible. Load patterns have been recognised to represent the usage behaviours of customers better and these behaviours influence the planning parameters. There have been studies by numerous researchers to extract parameters from the load profiles for load flow modelling and simulation purposes. The same challenge exists for South Africa, where there has been progress made on the development of LV models, and the same is not replicated in the MV network space. The derivation of load models primarily involves the classification of loads, identifying and estimating the parameters of loads, and assigning load profiles to different loads for studies. Customer measurements are an essential input in load model development and load estimation. Identification of parameters is one of the areas where research is ongoing since there is no global consensus on which attributes best describe customer load profiles. In this study, a proposition on how the parameters for technical and tariff analysis models should be defined was made. The use of 24-hour load profiles to classify calendar days into typical days was also suggested. The availability of measurements data made it possible to develop load models for MV and conduct a study on actual customer data. The customers' measurements data, made it possible to identify the parameters and develop load models that could be used for technical and tariff analysis and conduct a pilot study to evaluate the load models. This study proposes a load model that can be used to model typical days and to model customer loads. The load models proposed here uses the k-means clustering algorithm as the basis for classification. The load models enable the classification of loads and assignment of load profiles accordingly. The results of this study indicated that load parameter models could be extracted from the customer measurements, for technical and tariff studies in distribution networks. It has also been possible to identify and determine the parameters from the load profiles and proposed a process for developing a load model for technical, economic and tariff analysis. The results also indicate that of the five identified parameters, the most significant parameters that affected the clustering results were the load factor, average power and the normalised peak usage parameter when the results of each of the factors were compared on an individual basis. The study also revealed improvements to the clustering results when all the parameters identified in this study were combined and a PCAbased clustering algorithm was used. Finally, the results indicate that the loads in the different economic activitybased classifications do not necessarily have similar shapes although they belong to the same cluster. The modelling process developed in this study may be implemented by utilities for determining load parameter models for MV feeders when measurements are available. The process may also be used to guide future data collection.

Esse trabalho investiga o impacto do comportamento transitório e em regime permanente da carga sobre a estabilidade de tensão do sistema elétrico. Para isso, utiliza-se uma modelagem detalhada da rede elétrica, capaz de representar os principais eventos inerentes aos fenômenos de instabilidade e colapso de tensão. A simulação numérica do sistema algébrico-diferencial resultante é realizada utilizando-se o solver DASSLC. Ao final desse trabalho, faz-se uma análise da resposta dos diferentes modelos de carga sobre a estabilidade de tensão do sistema de 14 barras do IEEE. / This work investigates the impact of transient and steady state load behavior on power systems voltage stability. In order to do this, a detailed electric power system model is used to reproduce the main aspects of voltage instability and collapse phenomena. The numerical solution of the resulting non-linear differential-algebraic equations is carried out by using the DASSLC solver. An analysis of different load models behaviour for some voltage instability situations is presented for IEEE 14 bus system.

Esse trabalho investiga o impacto do comportamento transitório e em regime permanente da carga sobre a estabilidade de tensão do sistema elétrico. Para isso, utiliza-se uma modelagem detalhada da rede elétrica, capaz de representar os principais eventos inerentes aos fenômenos de instabilidade e colapso de tensão. A simulação numérica do sistema algébrico-diferencial resultante é realizada utilizando-se o solver DASSLC. Ao final desse trabalho, faz-se uma análise da resposta dos diferentes modelos de carga sobre a estabilidade de tensão do sistema de 14 barras do IEEE. / This work investigates the impact of transient and steady state load behavior on power systems voltage stability. In order to do this, a detailed electric power system model is used to reproduce the main aspects of voltage instability and collapse phenomena. The numerical solution of the resulting non-linear differential-algebraic equations is carried out by using the DASSLC solver. An analysis of different load models behaviour for some voltage instability situations is presented for IEEE 14 bus system.

The main aim of the present study is to evaluate the safety level of Turkish type precast prestressed concrete bridge girders designed according to American Association of State Highway and Transportation Officials Load and Resistance Factor Design (AASHTO LRFD) based on reliability theory. Span lengths varying from 25 m to 40 m are considered. Two types of design truck loading models are taken into account: H30S24-current design live load of Turkey and HL93-design live load model of AASHTO LRFD. The statistical parameters of both load and resistance components are estimated from local data and published data in the literature. The bias factors and coefficient of variation of live load are estimated by extrapolation of cumulative distribution functions of maximum span moments of truck survey data (Axle Weight Studies) that is gathered from the Division of Transportation and Cost Studies of the General Directorate of Highways of Turkey. The uncertainties associated with C40 class concrete and prestressing strands are evaluated by the test data of local manufacturers. The girders are designed according to the requirements of both Service III and Strength I limit states. The required number of strands is calculated and compared. Increasing research in the field of bridge evaluation based on structural reliability justifies the consideration of reliability index as the primary measure of safety of bridges. The reliability indexes are calculated by different methods for both Strength I and Service III limit states. The reliability level of typical girders of Turkey is compared with those of others countries. Different load and resistance factors are intended to achieve the selected target reliability levels. For the studied cases, a set of load factors corresponding to different levels of reliability index is suggested for the two models of truck design loads. Analysis with Turkish type truck models results in higher reliability index compared to the USA type truck model for the investigated span lengths

En 2010, ERDF (Electricité Réseau Distribution France) a entamé la mise en place du projet « Linky » dont l'objectif est d'installer 35 millions de compteurs intelligents en France. Ces compteurs permettront de collecter les données de consommation en « temps réel », avec lesquelles des modèles de charge plus précis pourront être envisagés. Dans ce contexte, cette thèse définit deux objectifs: la définition de modèles prédictifs de charge pour la conduite et la conception de modèles d'estimation de charge pour la planification. En ce qui concerne la conduite, nous avons développés deux modèles. Le premier exploite le formalisme mathématique des séries chronologiques ; le second est basé sur un réseau de neurones. Les deux modèles cherchent à prévoir la charge des jours « J+1 » et « J+2 » à partir des informations collectées jusqu'au jour « J ». Le modèle « série chronologique » repose sur les propriétés temporelles des courbes de charge. Ainsi on découpe la courbe de charge en trois parties : la tendance, la périodicité et le résidu. Les premiers deux sont déterministes et indépendamment développés en deux modèles : le modèle de tendance et le modèle de cyclicité. La somme de la prévision de ces deux modèles est la prévision finale. Le résidu quant à lui capture les phénomènes aléatoires que présente la courbe de charge. Le modèle de prédiction ainsi développé s'aide de nombreux outils statistiques (e.g., test de stationnarité, test ANOVA, analyse spectrale, entres autres) pour garantir son bon fonctionnement. Enfin, modèle « série chronologique » prend en compte plusieurs facteurs qui expliquent la variation dans la courbe de consommation tels que la température, les cyclicités, le temps, et le type du jour, etc. En ce qui concerne le modèle à base de réseaux de neurones, nous nous focalisons sur les stratégies de sélection de la structure pour un modèle optimal. Les choix des entrées et du nombre de neurones cachés sont effectués à travers les méthodes dites de «régression orthogonale » et de « leave-one-out-virtuel ». Les résultats montrent que la procédure proposée permet de choisir une structure de réseau de neurones qui garantisse une bonne précision de prédiction. En ce qui concerne la planification, un modèle non paramétrique est proposé et comparé avec le modèle actuel « BAGHEERA » d'EDF. Avec l'ouverture du marché d'électricité, la relation entre les fournisseurs, les clients et les distributeurs devient flexible. Les informations qualitatives d'un client particulier telles que sa puissance souscrite, son code d'activité, ses tarifs etc. sont de moins en moins disponibles. L'évolution du modèle BAGHEERA qui dépend ces informations pour classer les clients dans différentes catégories est devenue indispensable. Le modèle non paramétrique est un modèle individuel centré sur le relevé des compteurs. Trois variables de régression non paramétriques : Nadaraya Watson, Local Linear et Local Linear adapted ont été analysées et comparées. Les scénarios de validation montrent que le modèle non paramétrique est plus précis que le modèle « BAGHEERA ». Ces nouveaux modèles ont été conçus et validés sur de vraies données collectées sur le territoire français. / From 2010, ERDF (French DSO) started the “Linky” project. The project aims at installing 35 millions smart meters in France. These smart meters will collect individual client's consumption data in real time and transfer these data to the data center automatically in a certain frequency. These detailed consumption information provided by the smart metering system enables the designs of more accurate load models. On this purpose, two distinctive objectives are defined in this dissertation: the forecasting load models for the operation need and the estimation load models for the planning need. For the operation need, two models are developed, respectively relying on the “time series” and the “neural network” principals. They are both for the objective of predicting the loads in “D+1” and “D+2” days based on the historical information till “D” day. The “time series” model divides the load curve into three components: the trend, the cyclic, and the residual. The first two parts are deterministic, from which two models named the trend model and the cyclic model are made. The sum of the prevision of these two models is the final prediction result. For a better precision, numerous statistical tools are also integrated such that the stationary test, the smoothed periodogram, the ANOVA test and the gliding window estimation, etc. The time series model can extract information from the influence factors such as the time, the temperature, the periodicities and the day type, etc. Being the most popular non linear model and the universal approximator, the neural network load forecasting model is also studied in this dissertation. We focus on the strategy of the structure selection. The work is in collaboration with Prof. Dreyfus (SIGMA lab), a well known expert in the machine learning field. Input selection and model selection are performed by the “orthogonal forward regression” and the “virtual-leave-one-out” algorithms. Results show that the proposed procedure is efficient and guarantees the chosen model a good accuracy on the load forecasting. For the planning, a nonparametric model is designed and compared with the actual model “BAGHEERA” of the French electricity company EDF. With the opening of the electricity market, the relationship among the regulators, suppliers and clients is changing. The qualitative information about a particular client such as his subscribed power, his activity code and his electricity tariffs becomes less and less available. The evolution from the BAGHEERA model to a data-driven model is unavoidable, since the BAGHEERA model depends on these information to attribute every client in the French territory into a pre-defined category. The proposed nonparametric model is individualized and can deal with both temperature sensitive (possessing an electrical heater) and temperature insensitive clients. Three nonparametric regressors are proposed: the Nadaraya Watson, the local linear, and the local linear adapted. The validation studies show that the nonparametric model has a better estimation precision than the BAGHEERA model. These novel models are designed and validated by the real measurements collected in the French distribution network.

Voltage Impasse Region (VIR) is a phenomenon in power systemswhose dynamics are describe by a set of Differential AlgebraicEquations (DAE). VIR denotes a state-space area where voltagecausality is lost, i.e. the Jacobian of the algebraic part of DAEis singular. In a Time Domain Simulation (TDS) once system trajectoriesenter VIR, TDS experiences non-convergence of the solution.Then, there is no reason to continue with the simulation. Thisis why it is important to understand the mechanisms that introduceVIR. It is known that VIR appears in relation to static, non-linearload models. However, it remained unknown what the cumulativeeffect of several static, non-linear loads would be.This master thesis has further expanded the concept of VIRby carrying out a structured study on how the load modelling affectsVIR. For this purpose, this thesis proposes a quasi-dynamicmethodology to map VIR in the relative rotor angle space. Themethodology introduces a new discrete index called Voltage ImpasseRegion Flag (VIRflag), which allows to determine if the algebraicequations of DAE are solvable or not and, thus, to locate VIR.A test system is used to test the proposed quasi-dynamic approach.The VIRflag was first used to map VIR for various load combinations.Then, the relationship between TDS non-convergence issuesand the intersection of a trajectory with VIR is examined toverify the proposed methodology.The proposed method has been proved to be efficient in the determinationof VIR regardless of the number of non-linear loads inthe power system. Among the static exponential load models, theConstant Power (CP) load component has been identified as theone with the largest influence on VIR appearance and shape. TheConstant Current (CC) loads induce ”smaller" VIR areas and theConstant Impedance (CI) load can only alter the shape of VIR inthe presence of non-linear load models. / VIR (Voltage Impasse Regions) är ett fenomen i kraftsystem varsdynamiska förlöp beskrivs av differential-algebraiska ekvationer(DAE). VIR betecknar ett område i tillståndsrummet där går förlorad,dvs Jakobianen av den algebraiska delen av DAE är singulärI tidsdomän-simuleringar (TDS) när en trajektoria träffar VIR,konvergerar TDS inte till en lösning. Då finns ingen anledning attfortsätta med simuleringen. Därför är det viktigt att förstå mekanismernasom introducerar VIR. Det är känt att VIR är relateradetill statiska, icke-linjära lastmodeller. Det var dock okänt vadden kumulativa effekten av flera statiska, icke-linjära belastningarskulle vara.Denna uppsats har vidareutvecklat begreppet VIR genom attgenomföra en strukturerad studie om hur lastmodellering påverkarVIR. För detta ändamål föreslår denna avhandling en kvasidynamiskmetod för att kartlägga VIR i det relativa rotorvinkelrummet.Metoden introducerar ett nytt diskret index som heterVoltage Impasse Region Flag (VIRflag), vilket gör det möjligt attbestämma om den algebraiska delen av DAE är lösbar eller inteoch därmed lokalisera VIR. Ett används för att testa det föreslagnakvasi-dynamiska tillvägagångssättet. VIRflag användes först för attkartlägga VIR för olika belastningskombinationer. Därefter granskasförhållandet mellan konvergensproblem i TDS och korsningenmellan en trajektoria och VIR för att verifiera den föreslagna metoden.Den föreslagna metoden har visat sig vara effektiv vid bestämningav VIR, oberoende av antalet icke-linjära belastningar. Bland destatiska exponentiella belastningsmodellerna har konstanteffektlast(CP) haridentifierats som den som har störst inflytande påVIR;s form. Den konstantströmlasten (CC) inducerar mindre"VIRområdenoch konstantimpedanslasten (CI) kan endast ändra formenav VIR i närvaro av icke-linjära belastningsmodeller.

Orientador: José Roberto Sanches Mantovani / Resumo: Neste trabalho propõe-se o desenvolvimento de uma ferramenta computacional para o planejamento da operação de curto prazo de sistemas de distribuição com geração distribuída (GD) considerando uma abordagem probabilística. Uma modelagem sequencial formulada com base na perspectiva das companhias de distribuição (DisCos) é proposta. As decisões operacionais da DisCo são inicialmente otimizadas no estágio de operação day-ahead (DA) e, então, na operação real-time (RT). A operação DA visa maximizar a diferença entre a energia vendida aos consumidores e as compras realizadas no mercado de eletricidade atacadista e da GD, ou seja, os lucros. No estágio RT, busca-se a minimização dos ajustes necessários para acomodar os desvios das quantidades previstas no planejamento DA. Modelos de cargas dependentes de tensão e restrições relacionadas à demanda são explicitamente formulados. A rede é representada através de equações de fluxo de potência AC completo. Propõe-se ainda a incorporação de um mecanismo para precificação nodal de potência reativa. Os modelos resultantes são caracterizados como programas de otimização matemática multiperíodo de grande porte não lineares e não convexos com variáveis contínuas e discretas. Um algoritmo pseudodinâmico baseado na meta-heurística Busca Tabu (BT) é proposto para solução do problema resultante de maneira eficaz, sem linearizações. Os resultados obtidos para alimentadores de distribuição de 69 e 135 barras ilustram a eficiência da metodologia pro... (Resumo completo, clicar acesso eletrônico abaixo) / Doutor