Análise de sensibilidade por Teoria de Perturbação Generalizada da frequência de acidente de uma instalação nuclear equipada com um canal de proteção sob envelhecimento

• Ano de defesa: 2019
• Autor(a) principal: Souza, Larissa Fernandes Araujo de
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal do Rio de Janeiro; Brasil; Instituto Alberto Luiz Coimbra de Pós-Graduação e Pesquisa de Engenharia; Programa de Pós-Graduação em Engenharia Nuclear; UFRJ
• Palavras-chave em Português: Análise de sensibilidade; Reatores nucleares; Medidas de segurança; Segurança nuclear; Envelhecimento; CNPQ::ENGENHARIAS::ENGENHARIA NUCLEAR::TECNOLOGIA DOS REATORES::SEGURANCA, LOCALIZACAO E LICENCIAMENTO DE REATORES
• Link de acesso: http://hdl.handle.net/11422/13296

In this work, we present a sensitivity analysis of a protection system of a nuclear plant equipped with a single protection channel, that enters in a wear out state while in the operation period of the plant. This analysis is related to the metric of the plant accident frequency, and considers the variation of parameters of interest that compound the calculation of this metric. By the common method, the calculation of the values for this analysis has a high computational cost considering the useful life of the protection channel, and it is even higher when the calculation considers the wear out period, which makes this task very tiring in terms of time. To work around this inconvenience, it is proposed to use the method of Generalized Perturbation Theory (GPT) applied to the calculation for the accident frequency and its sensitivity analysis for a single protection channel, subject to the parameters of interest, varying its demand rate. The employment of a first order GPT to calculate the accident frequency, with the application of the perturbation concept to vary the demand rate values, showed up fairly effective to systems subject to high demands, however, it did not bring good results to lower values of this parameter. Even so, the efficiency of this method stands out when compared to the common method. It may be noted that the effectiveness of the GPT improves depending on the reference demand value and the perturbed value used, but loses efficiency, when compared with a wider perturbation range.