Submitted by Automa??o e Estat?stica (sst@bczm.ufrn.br) on 2017-04-03T20:52:32Z No. of bitstreams: 1 AdrianaCavalcanteMarques_DISSERT.pdf: 6652160 bytes, checksum: f4c1dfa84af7460823c1ea1643624042 (MD5) / Approved for entry into archive by Arlan Eloi Leite Silva (eloihistoriador@yahoo.com.br) on 2017-04-06T21:19:34Z (GMT) No. of bitstreams: 1 AdrianaCavalcanteMarques_DISSERT.pdf: 6652160 bytes, checksum: f4c1dfa84af7460823c1ea1643624042 (MD5) / Made available in DSpace on 2017-04-06T21:19:34Z (GMT). No. of bitstreams: 1 AdrianaCavalcanteMarques_DISSERT.pdf: 6652160 bytes, checksum: f4c1dfa84af7460823c1ea1643624042 (MD5) Previous issue date: 2016-06-27 / O setor el?trico brasileiro tem apresentado diferentes mudan?as, desde a desverticaliza??o das empresas at? a cria??o do ?rg?o regulador (Aneel - Ag?ncia Nacional de Energia El?trica) respons?vel pela regula??o econ?mica dos segmentos de transmiss?o e distribui??o. Contudo, a Aneel aplica as pol?ticas de tarifa??o separadamente entre os segmentos que comp?em o sistema el?trico, n?o considerando as rela??es existentes na cadeia de suprimentos. O presente trabalho tem como objetivo propor um novo modelo de an?lise de efici?ncia do sistema el?trico brasileiro a fim de tornar a revis?o tarif?ria melhor ajustada para a realidade do cen?rio atual. Os segmentos do setor (gera??o, transmiss?o e distribui??o) foram fundamentados por meio de um conjunto de vari?veis recorrentes na literatura para retratar os processos realizados no sistema el?trico, no ?mbito nacional e internacional. A partir disso, o modelo conceitual considera 12 vari?veis distribu?das em tr?s dimens?es: custo, produtividade e qualidade. A amostra contempla 94 empresas, dentre as quais segmentam: 16 empresas de gera??o, 28 empresas transmissoras de energia e 50 empresas distribuidoras. As Unidades Tomadoras de Decis?o (Decision Making Units - DMU) foram classificadas como os caminhos com as diferentes empresas da gera??o at? a distribui??o de energia, de forma que a amostra totalizou em 362 DMUs. Na condu??o da pesquisa, elaborou-se um modelo para avalia??o do sistema el?trico com a utiliza??o da An?lise Envolt?ria de Dados (DEA ? Data Envelopment Analysis), com o prop?sito de identificar quantitativamente o n?vel de efici?ncia da rede. De modo a poder representar a complexidade do sistema el?trico, o modelo proposto utiliza o DEA em redes (Network Data Envelopment Analysis ? NDEA) com tr?s est?gios, cada um simulando os segmentos do setor, para identificar os impactos entre os processos e as causas de inefici?ncia. Esses est?gios da rede foram avaliados para verificar a performance das empresas em cada segmento e validar o modelo de retornos constantes de escala, por meio dos modelos cl?ssicos CCR em compara??o com o NDRS, utilizado pela Aneel. A partir dessa compara??o entre os modelos cl?ssicos, validou-se o uso do modelo de retornos constantes de escala para as an?lises do setor. Os resultados do modelo NDEA apontam para a adequa??o do m?todo proposto realizando a an?lise agregada do sistema el?trico, ao apresentar resultados mais discriminat?rios que os modelos cl?ssicos. Obteve-se que nenhum caminho da rede observado possui as tr?s empresas constituintes eficientes em sua totalidade. Contudo, foi poss?vel identificar em quais liga??es as empresas devem focar seus investimentos para alcan?ar melhores desempenhos e quais redes ter como benchmarks, contribuindo para a tomada de decis?o gerencial e o planejamento eficiente das empresas. Por fim, identificou-se os fatores ex?genos que influenciam na efici?ncia, como a regi?o geogr?fica e porte, sendo poss?vel ent?o, obter diagn?sticos para uma tarifa??o mais adequada ? realidade do setor, beneficiando tanto os consumidores, quanto as empresas e o governo. / The Brazilian electricity sector has shown different changes, since unbundling of companies to the creation of the regulatory agency (Aneel - National Electric Energy Agency), responsible for the economic regulation of the transmission and distribution segments. However, Aneel applies charging policies separately between the segments that make up the electrical system, not considering the relationship in the supply chain. In this context, this paper aims to propose a new model of the Brazilian electric system efficiency analysis in order to make the best-adjusted tariff review with the system reality. The segments of the industry (generation, transmission and distribution) were founded through a set of recurring variables in the literature to portray the processes carried out in the electrical system at the national and international levels. The conceptual model includes initially 12 variables divided into three dimensions: cost, productivity and quality. The sample includes 94 companies, consisting of in 16 generation companies, 28 electricity transmission companies and 50 distribution companies. The Decision Making Units (DMU) were considered the paths from generation to distribution, so that the sample contains 362 DMUs in total. In conducting the research analysis, a model was developed to evaluate the electrical system using the data envelopment analysis (DEA) to quantitatively identify the network efficiency level. To be able to represent the complexity of the electrical system, a unique grouping model was proposed and analyzed by Network DEA's model (NDEA) in three stages, each one simulating a particular segment of the industry, in order to identify the impacts between the processes and the causes of inefficiency. These stages of the network were evaluated to verify the performance of the companies in each segment and validate the model of constant returns to scale, through the classic models CCR compared to the NDRS used by Aneel. From this comparison between the classic models, the use of constant returns to scale model for sector analysis was validated. The results of the NDEA model point to the adequacy of the proposed method by performing the electrical system aggregate analysis, by presenting more discriminatory results than classic models. Furthermore, none of the paths have efficient companies in all of its three stages. However, it was possible to identify which links those companies should focus their investments to achieve better performance and which networks are its benchmarks, contributing to the decision-making and the efficient planning of companies. Finally, we identified the exogenous factors that influence efficiency, such as geography and size, thereby making possible to obtain a more appropriate tariffing regarding the reality of the sector, benefiting consumers, companies and the government.