Proposta de um Roadmap para avaliação de riscos em projetos De perfuração direcional horizontal

• Ano de defesa: 2016
• Autor(a) principal: Gomes, Mário Batista
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal da Paraíba; Brasil; Engenharia de Produção; Programa de Pós-Graduação em Engenharia de Produção; UFPB
• Palavras-chave em Português: Perfuração horizontal direcional; Lógica fuzzy; Análise de riscos; Horizontal directional drilling; Fuzzy logic; Risk analysis; ENGENHARIAS::ENGENHARIA DE PRODUCAO
• Link de acesso: https://repositorio.ufpb.br/jspui/handle/tede/9921

The relevance of the analysis and risk management in horizontal directional drilling projects are sedimented in the work developed based on tacit knowledge capture experts and builders of experience in order to model this knowledge from conceptual premises, a mathematical model capable to replicate efficiently the conceptual frameworks of the opinion of experts. Historical data of the failure modes that appear in HDD (Horizontal Directional Drilling) projects, allow us to add a simple set of qualitative techniques to manage and anticipate the occurrence of damage that result in financial, economic and social losses, security, productivity or quality. These techniques are easy to implement and eliminate the need for skilled labor to validate the results. The failures of historical data in prior projects, can provide the basic elements for modeling knowledge through a learning tool. For quantitative effect, the tool used was the Fuzzy Inference System (FIS). The variables that indicate the failure modes are taken from the existing knowledge database in the literature and applied to a set of 40 projects. The application of qualitative tools represented a significant reduction of accidents in the research organization. The SIF model (fuzzy inference system) proposed is able to predict based on expert opinion the risk of a project building on 05 key variables. The result of the inference of the quantitative model involves the quantification of the overall risk of a given project, on a scale of 1 to 10. The proposed interface simplifies quantification operation, providing agile the overall risk of the interest of project building on the conceptual indication 05 this risk input variables. The calculated risk be inferred about the risks associated with the project, facilitating the decisionmaking process of managers and builders.