Neste trabalho é apresentada uma abordagem de controle de demanda para consumidores residenciais de baixa tensão, visando melhoria da eficiência energética em Smart Grids. Inicialmente, um emulador de cargas elétricas residenciais é modelado com base na literatura. O emulador é composto pelo modelo de um reservatório de aquecimento de água (boiler), o modelo de um aparelho de ar condicionado e também modelos de consumo de iluminação, televisores e uma geladeira. Utilizando o software Matlab foi realizada a implementação e simulação do emulador. Os principais algoritmos de controle de demanda são investigados, a fim de verificar o seu desempenho quando aplicados ao conjunto de cargas residenciais. Esses algoritmos normalmente realizam o controle de demanda a partir de um sistema de prioridades. Ainda, a partir dessa analise demostra-se que estes algoritmos consideram níveis de conforto do usuário, porém não permitem o acionamento de duas ou mais cargas em um mesmo período caso a demanda da residência ultrapasse um limite predeterminado. Portanto, propõem-se um algoritmo de controle de demanda adaptativo que utiliza o método de busca Rosenbrock, com o objetivo de sobrepujar tais limitações. O procedimento proposto realiza a operação das cargas residenciais de forma gradual considerando níveis de prioridade e parâmetros de conforto dos usuários. Demonstra-se através de simulações e experimentos que através do método proposto é possível realizar a ativação de diversas cargas concorrentemente, desde que respeitados os níveis de conforto e de demanda. Para obtenção dos resultados experimentais o controlador de demanda foi implementado em um sistema embarcado e testado com o emulador de cargas elétricas residenciais implementado em uma arquitetura HIL (Hardware-in-the-loop). Analisando os resultados, observou-se que o consumo de energia foi o mesmo para todos os cenários simulados sendo que a demanda se manteve abaixo dos limites parametrizados. Porém com o limitador de demanda ativo, se obteve uma redução de até 52% no tempo de aquecimento da água utilizando o controlador de demanda adaptativo, dessa forma o desconforto dos usuários pode ser minimizado. / This work presents an approach to control demand for residential low voltage consumers, aiming to improve energy efficiency in Smart Grids. Initially, an emulator of residential electric loads is modeled based on the literature. The emulator consists of a reservoir for water heating model, the model of an air conditioner and also models of consumption for lighting, televisions and a refrigerator. The implementation and simulation were performed using software Matlab. The demand control algorithms are investigated in order to verify its performance when applied to the set of residential loads. These algorithms typically perform control demand from a system of priorities. Still, from this analysis it demonstrates that these algorithms consider levels of user comfort, but do not allow the drive of two or more loads in the same period of residence if the demand exceeds the limit. Therefore, we propose a control algorithm that uses Rosenbrock search of demand adaptive method, aiming to overcome these limitations. The proposed procedure performs the operation of residential loads gradually considering priority levels and parameters of comfort of users. It is shown through simulations and experiments using the proposed method can perform the activation of several concurrently loads, provided they comply with the limits of comfort and demand. To obtain the experimental results demand the controller was implemented in an embedded system and tested with the emulator residential electrical loads implemented in a HIL (Hardware-in-theloop) architecture. Analyzing the results, it was observed that the power consumption is the same for all scenarios simulated and demand remained below parametric limits. But with the demand limiter active, we obtained a reduction of up to 52% in heat water using the demand controller adaptive, so the discomfort of the users can be minimized.

Submitted in fulfillment of the requirement for the Degree of Master of Management Sciences Specialising in Business Administration, Durban University of Technology, Durban, South Africa, 2017. / In the electricity prepayment metering industry the continuous increase in customers’ expectations and technological innovation demand that leading firms in the industry differentiate themselves from the competition by going beyond customers’ present expectations. Thus, organizations which have set their goals on mere customer satisfaction are, from a customers’ perspective, deemed to be of limited value and may subsequently lack the anticipation and preparation essential to meet the demands of the future. Electricity prepayment metering market competition today is forcing organisations to seek the means to gain customer satisfaction, and thus aim to sustain their future in times of economic instability. The study revealed that the most effective way to retain customer satisfaction is consistent service quality that decreases the cost of attracting new customers and raises sales and market share. The quality of service is one of the most effective factors in creating competitive advantages and advance business. Factors that influence customer satisfaction in the electricity prepayment industry were investigated and tested. The study also attempted to identify the critical factors, as identified in the literature, among those investigated: service quality, customer expectations, customer perception and customer retention. Against this background, the study sought to determine the interrelationship between customer satisfaction and the critical factors of customer satisfaction with Conlog Electricity Prepayment Meters in KwaZulu-Natal. The research developed to ascertain subjects’ perceptions of the critical factors of customer satisfaction; determine the interrelationship among the critical factors of customer satisfaction; and explore the relationship between customer satisfaction and the critical factors of customer satisfaction. The research population comprising of Conlog electricity prepayment metering customers in KwaZulu-Natal was selected and interviewed. Questionnaires were distributed through electronic mail and self-administered for data collection and a sample of twelve respondents was selected for interview. Interviews were analysed using conversational analysis and the data collected from the interviews were merged with the questionnaire data, seeking depth as well as breadth. Data collected from respondents was analysed using descriptive and inferential statistical techniques. The tool utilised to analyze data was the Statistical Package for the Social Sciences (SPSS) version 22 (SPSS Inc., Chicago, Illinois, USA). A p value < 0.05 was considered as statistically significant. Conclusions and recommendations were drawn from the literature and the findings of the study. The results of the study revealed that service quality is the most significant critical factor of customer satisfaction. The study recommends that the management of Conlog can use the specific data obtained from the measurement of service quality in their strategies and plans. The study further recommends that Conlog allocates resources to this effect to provide better service to their customers. The results and findings of the study will provide further information concerning customer satisfaction and customer perception in the prepayment metering industry in KwaZulu-Natal. It will also contribute to future industry research, setting the ground for further research in measuring levels of customer satisfaction in the electricity prepayment industry in South Africa. / M

This study estimates the price and income elasticity coefficients of electricity demand in the mining sector of South Africa for the period ranging from April 2006 to March 2019. A time varying parameter (TVP) model with the Kalman filter is applied to monitor the evolution of the elasticity estimates. The TVP model can provide a robust estimation of elasticities and can detect any outliers and structural breaks. The results indicate that income and price elasticity coefficients of electricity demand are lower than unit. The income elasticity of demand has a positive sign and it is statistically significant. This indicates that mining production – used as a proxy for mining income – is a significant determinant of electricity consumption in the mining sector. In its final state income elasticity is estimated at 0.15 per cent. On the contrary, price does not play a significant role in explaining electricity demand. In fact, the price elasticity coefficient was found to be positive which is contrary to normal economic convention. This lack of response is attributed mainly to the mining sector’s inability to respond, rather than an unwillingness to do so. A fixed coefficient model in a form of Ordinary Least Squares (OLS) is used as a benchmark model to estimate average price and income elasticity coefficients for the period. The results of the OLS regression model confirm that price does not play a significant role in explaining electricity consumption in the mining sector. An average price elasticity coefficient of -0.007 has been estimated. Income elasticity was estimated at 0.11 for the period under review. The CUSUM of squares test indicate that parameters of the model are unstable. The Chow test confirms 2009 as a breakpoint in the data series. This means that elasticity coefficients of electricity demand in the mining sector are time variant. Thus the OLS results cannot be relied upon for inference purposes. The Kalman filter results are superior. This study cautions policy makers not to interpret the seeming lack of response to price changes as an indication that further prices increases could be implemented without hampering electricity consumption in the sector. Furthermore, it recommends that the electricity pricing policy should take into account both the negative impacts of rapid price increases and the need to invest in long-term electricity infrastructure in order to improve the security of energy supply. A long term electricity price path should be introduced in order to provide certainty and predictability in the price trajectory. This would allow all sectors of the economy sufficient time and space to make investment and operational decisions that would have the least adverse effects on economic growth and job creation. / Economics / M. Com. (Economics)

MSc (Statistics) / Department of Statistics / Short-term load forecasting in South Africa using machine learning and statistical models is discussed in this study. The research is focused on carrying out a comparative analysis in forecasting hourly electricity demand. This study was carried out using South Africa’s aggregated hourly load data from Eskom. The comparison is carried out in this study using support vector regression (SVR), stochastic gradient boosting (SGB), artificial neural networks (NN) with generalized additive model (GAM) as a benchmark model in forecasting hourly electricity demand. In both modelling frameworks, variable selection is done using least absolute shrinkage and selection operator (Lasso). The SGB model yielded the least root mean square error (RMSE), mean absolute error (MAE) and mean absolute percentage error (MAPE) on testing data. SGB model also yielded the least RMSE, MAE and MAPE on training data. Forecast combination of the models’ forecasts is done using convex combination and quantile regres- sion averaging (QRA). The QRA was found to be the best forecast combination model ibased on the RMSE, MAE and MAPE. / NRF

[pt] Este trabalho apresenta um método de Monitoramento Não Intrusivo de Carga de Aparelhos elétricos (do inglês Non-Intrusive Appliance Load Monitoring – NIALM) supervisionado, usando técnicas de análise de estados estacionários, para desagregação do consumo elétrico residencial a partir de uma única medição, sem a necessidade de instalação de medidores individuais nos dispositivos. A metodologia proposta divide o problema em duas etapas. Inicialmente, há um pré-processamento para identificação e desagregação dos aparelhos que apresentam comportamento periódico, modelados a partir da estimação dos parâmetros da série de Fourier. Na etapa seguinte, os resultados obtidos são combinados a um modelo de otimização linear-inteiro misto para desagregação dos equipamentos não-periódicos, buscando minimizar a diferença entre a curva de carga total lida e a soma das curvas de carga desagregadas por dispositivo. Uma aplicação didática é realizada para validação do método proposto com dados reais e, por fim, é apresentada uma análise de viabilidade econômica da migração para a tarifa branca aplicada no Brasil. Os resultados mostram que, ao utilizar dessa metodologia, é possível que o usuário avalie se há ou não vantagem em deslocar parte do seu consumo de energia para fora do horário de ponta para obter benefício na sua fatura de energia elétrica. / [en] This work presents a supervised Non-Intrusive Appliance Load Monitoring (NILM) method, or energy disaggregation, for residential consumption, which aims to decompose the aggregate energy consumption data collected from a single measurement point into device-level consumption estimation using steady state analysis techniques with no need to install individual meters on appliances. The proposed methodology considers two steps to face the problem. Firstly, periodical appliances are modeled from the estimation of Fourier series parameters and extracted from the total power measured. Secondly, the results obtained are combined with a Mixed Integer Linear Programming proposed to disaggregate the remaining appliances, which minimize the difference between the total aggregated load and the sum of the estimated load curves per appliance. A study case is performed with a real case to validate the proposed method and indicates that the model can be useful for practical applications, such as helping evaluate the possibility of the consumers changing the modality of their tariff contract from the conventional tariff to the new Brazilian modality called white tariff.

The forecasting of Eskom transmission electrical network demands is a complex task. The lack of historical data on some of the network components complicates this task even further. In this dissertation a model is suggested which will address all the requirements of the transmission system expansion engineers in terms of future loads and market trends. Suggestions are made with respect to ways of overcoming the lack of historical data, especially on the point loads, which is a key factor in modelling the electrical networks. A brief overview of the transmission electrical network layout is included to provide a better understanding of what is required from such a forecast. Lastly, some theory on multiple regression, neural networks and qualitative forecasting techniques is included, which will be of value for further model developments. / Computing / M. Sc. (Operations Research)

CAPES / Nesta Tese de Doutorado propõe-se um modelo de consumo de energia para redes acústicas subaquáticas que leva em consideração as especificidades do ambiente subaquático, como a comunicação utilizando ondas acústicas, a dependência da largura de banda do canal com a perda de percurso, que varia tanto com a distância quanto com a frequência, e o ruído. O desvanecimento, resultado do multipercurso, normalmente modelado pelas distribuições Rayleigh e Rice nas transmissões terrestres, é modelado pela distribuição K, que melhor representa a severidade do ambiente subaquático. O modelo considera uma rede subaquática linear de múltiplos saltos e a possibilidade de retransmissões para calcular a energia total consumida para cada bit de informação transmitido com sucesso entre a fonte e o destino. A fim de obter o menor consumo de energia, a SNR e a frequência de operação também são otimizadas, sendo considerado o uso de códigos convolucionais, cuja taxa ótima que leva ao menor consumo é determinada. Uma análise teórica é desenvolvida para cenários com e sem limitação de atraso. No primeiro caso avalia-se o consumo de energia quando retransmissões não são permitidas ou devem ser limitadas e, portanto, uma FER residual deve ser tolerada. No segundo caso infinitas retransmissões são permitidas até que um pacote seja recebido sem erros. Para ambos cenários o número ótimo de saltos que minimiza o consumo de energia é determinado, e na sequência o impacto do número de tentativas de transmissão é considerado. Resultados numéricos são apresentados, mostrando que o esquema de múltiplos saltos é mais eficiente em termos de consumo de energia que a transmissão direta. Além disso, os resultados mostram que um número pequeno de tentativas de transmissão é suficiente para alcançar uma redução considerável no consumo de energia em redes de múltiplos saltos, limitando o atraso médio por pacote transmitido, o que é muito interessante em aplicações reais. / In this Thesis, an energy consumption model for underwater acoustic networks is proposed. The model takes into account the specificities of the underwater environment, such as the use of acoustic waves for communication, dependence of the underwater acoustic channel bandwidth with the path loss, which varies with both the distance and frequency, and noise. The fading, usually modeled by Rayleigh and Rice distributions on terrestrial communications, in this work, is modeled by the K distribution, which best represents the severity of the underwater environment. The model considers a linear multi-hop underwater network and the possibility of retransmissions to calculate the total energy consumed for each bit of information successfully transmitted between the source and the destination. In order to obtain the minimum energy, the SNR and the operating frequency are also optimized. The use of convolutional codes is considered and the optimal code rate, which leads to the minimum energy consumption, is determined. A theoretical analysis was developed for two scenarios: delay constrained and delay unconstrained networks, indicating the optimal number of hops which minimizes energy consumption. Next, the impact of the number of transmission trials was considered. A numerical analysis was also performed for both the scenarios. The numerical results validate the theoretical analysis, showing that the multi-hop scheme is more efficient in terms of energy consumption when compared to direct transmission. Furthermore, the results show that a small number of transmission trials is sufficient to achieve a considerable reduction in energy consumption in multi-hop networks, limiting the average delay per packet transmitted, which is very interesting in real applications.

O presente trabalho aborda a aplicação de códigos fontanais em redes subaquáticas. Tais redes transmitem dados abaixo da água fazendo uso de sinais acústicos e possuem diversas aplicações. No entanto, é sabido que esse tipo de rede é caracterizado por uma baixa velocidade de propagação e largura de banda menor que as redes que operam em meios de transmissão mais conhecidos, tais como a transmissão sem fio via ondas de rádio frequência, resultando num maior atraso na entrega de pacotes. Para tentar minimizar estes atrasos e aumentar a eficiência energética das redes subaquáticas, o trabalho otimizou o sistema de transmissão inserindo um código corretor de erros fontanal no transmissor de mensagens. Dentro desse contexto, foi necessário modelar o consumo energético necessário para a transmissão correta de pacotes de dados em redes subaquáticas utilizando códigos fontanais. Dentre os resultados do trabalho, o mais relevante conclui que o uso dos códigos fontanais é capaz de reduzir em até 30% o consumo de energia quando a distância de transmissão é de 20 km para o caso com a taxa de erro de quadro alvo (FER) de Po = 10^−5, e em ate 25% para a FER alvo de Po = 10^−3. / The present work employs fountain codes in an underwater network, in which data is transmitted using acoustic signals and has many applications. However, underwater networks are usually characterized by low propagation speed and smaller bandwidth than networks that use radio frequency signals, resulting in larger transmission delays. Then, aiming at minimizing the delays and increasing the energy efficiency of underwater networks, the present work employs fountain error-correcting codes at the transmitter. To that end, it was first necessary to model the energy consumption of a success data packet transmission in an underwater network using fountain codes. Our results show that the use of fountain codes is able to reduce up to 30% of energy consumption when the transmission distance is of 20 km for the case with a target frame error rate (FER) of Po = 10^−5 , and 25% for the same distance with a target FER of Po = 10^−3.

The forecasting of Eskom transmission electrical network demands is a complex task. The lack of historical data on some of the network components complicates this task even further. In this dissertation a model is suggested which will address all the requirements of the transmission system expansion engineers in terms of future loads and market trends. Suggestions are made with respect to ways of overcoming the lack of historical data, especially on the point loads, which is a key factor in modelling the electrical networks. A brief overview of the transmission electrical network layout is included to provide a better understanding of what is required from such a forecast. Lastly, some theory on multiple regression, neural networks and qualitative forecasting techniques is included, which will be of value for further model developments. / Computing / M. Sc. (Operations Research)

Este trabalho tem por objeto de estudo a correlação entre a utilização per capita de energia elétrica e o Índice de Desenvolvimento Humano (IDH). O objetivo geral da pesquisa é avaliar a compatibilidade das premissas do Plano Nacional de Energia (PNE) brasileiro, em relação à utilização de energia elétrica, com o desafio da elevação do IDH nacional ao nível “muito alto” – o mais elevado conforme a terminologia atual das Nações Unidas. Quanto ao método, a pesquisa é, em relação aos seus objetivos, preponderantemente descritiva, e quantiqualitativa no que se refere ao estudo do problema. Realizou-se tratamento estatístico de diferentes aspectos correlacionados à energia elétrica e ao IDH, de modo a estabelecer padrões significativos e universais dessa correlação, os quais parecem possíveis de serem previstos mediante utilização de fórmula matemática, a qual tornaria então possível avaliar o desenvolvimento humano utilizando a energia elétrica como parâmetro. Verificou-se que a proporção da utilização de energia elétrica é cada vez mais intensa em relação ao total de utilização de energia primária, o que se explica pela absoluta indispensabilidade, motivada pela versatilidade e disponibilidade, da primeira, e definiu-se um valor objetivo mínimo à utilização per capita de energia elétrica, ponderado em função da intensidade energética – no caso, 222 kWh –, necessário ao status de desenvolvimento humano “muito alto”. Definiram-se ainda duas fórmulas matemáticas – a primeira delas possui maior versatilidade descritiva, pois envolve equação logarítmica, enquanto a segunda, simplificada, utiliza função potência – capazes de individualmente estimar, em função do valor de utilização de energia e compondo uma base comparativa global, o IDH correspondente do ente analisado. Ambas são capazes de auxiliar na compreensão, ainda que em cognição preliminar, acerca da situação energética do ente sob análise, podendo ser utilizadas como indicador de possíveis direcionamentos de políticas energéticas que se pretendam voltadas ao desenvolvimento humano. Comprovou-se a acurácia da fórmula logarítmica em termos globais e históricos e, na sequência, verificou-se quão compatíveis são entre si o Plano Nacional de Energia e o IDH “muito alto”, em médio prazo e tendo como cenário os estados brasileiros. Concluiu-se, quanto a este objetivo geral, que a evolução histórica de oferta e demanda de energia elétrica no País é compatível com os números projetados pelo PNE, e que estes são compatíveis com o IDH “muito alto” – para atingir tal nível seriam necessários cerca de 720 TWh anuais, enquanto o PNE projeta, para 2030, 971 TWH e a evolução histórica das últimas duas décadas indica, para o mesmo ano, aproximadamente 830 TWh. O eventual atingimento desse nível de IDH deve, contudo, ser relativizado, pois provavelmente não seria suficiente para tirar o Brasil da atual posição de catching-up em relação aos países desenvolvidos. / This work studies the correlation between the per capita use of electrical energy and the Human Development Index (HDI). The main research objective is to evaluate the compatibility between the Brazilian National Energy Plan’s (PNE) goals, regarding the use of electrical energy, and the challenge of increasing the national HDI towards the “very high” level – the highest one according to the current United Nations terminology. As for the method, the research is, regarding its objectives, predominantly descriptive, and quanti-qualitative in relation to the problem studied. Several statistical aspects correlated to electrical energy and HDI were statistically treated, in such a way as to define significant, universal patterns of this correlation, which seem to be possible to predict through a mathematical formula that would then make it possible to evaluate human development using electrical energy as a parameter. It was noted that the proportion of electrical energy is increasingly bigger in relation to the total amount of primary energy, something explained by the absolute indispensability, through versatility and availability, of the former, and an objective minimum per capita value of electrical energy use – in this case, 222 kwh –, weighted as a function of energy intensity, was defined as necessary to reach a “very high” HDI. Besides it, two formulae were developed, both capable of individually estimate an entity’s HDI from the weighted value of electrical energy use. The first of them is more descriptively versatile, by using a logarithmic equation, while the second one is a simplified version, based on a potency function. Both formulae are capable of helping, yet in a preliminary cognition, to understand the energetic situation of the entity under analysis, possibly serving as decision support for energy policies aiming at human development. The logarithmic formula was checked globally and historically for accuracy, and after that it was used to verify how compatible are the Brazilian Energy Plan and a “very high” HDI, on the medium term and on the framework of the Brazilian states. It was concluded that, regarding this general objective, the historical evolution of national electrical energy supply and demand is compatible with the PNE projected numbers, which, on their turn, are compatible with a “very high” HDI – to reach such a level would require about 720 TWh per year, while the PNE forecasts, to 2030, 971 TWh, and the historical evolution of the last two decades indicates, for the same year, approximately 830 TWh. The possible attainment of this HDI level, however, should be put into perspective, because it probably would not be enough to avoid the catching-up situation Brazil is currently facing in relation to developed countries.