Detecção de vazamentos em redes de abastecimento de água através da utilização de séries de pressão e controle estatístico

• Ano de defesa: 2019
• Autor(a) principal: Ximenes Júnior, Celso Lira
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Palavras-chave em Português: Recursos hídricos; Abastecimento de água; Sistemas de informação geográfica; Análise de séries temporais; Sistemas de suporte de decisão; Statistical control; Leakage
• Link de acesso: http://www.repositorio.ufc.br/handle/riufc/48023

The management of the water source in the urban areas of Brazilian cities manages the water and sewage companies faces daily challenges. Increasing or maintaining supply at satisfactory and stable levels for ever-increasing infrastructures to the disorderly growth, coupled with climate change that brings the long drought phases and incompatible with the historical version, impute to the Supply Systems. Exceeded by increased efficiency in water transport processes. In this scenario, stock management becomes one of the main actions to be invested by sanitation companies, especially with regard to active leak control. Given the fact, the present study aims to develop a base method for real-time flow detection for water distribution networks. For this, it is necessary to use the free Geographic Information System (GIS), linked to the process of statistical control of the press series of the city of Fortaleza. The adoption of GIS has enabled a georeferenced platform organization, generating a database with reliable information. The data obtained and treated were applied the statistical control method, in particular the calculation of mean and exponentially weighted moving variance (MVMEP), in series of pressure signals through monitoring of the supply network. The data was related to other methods of the data to the software of Epanet, but did not return satisfactory sitting. Thus, detection values of around 60% were found when compared to the reports of leakage occurrence. The average detection time was 2.66 days after the leak removal event. The same was addressed in applying the method of extending maintenance devices at the macrosystem scale. As the maximums have reached 40%, just pointing to the fact that the supply zones are not totally free and there is no microsystem anymore. Other findings could be determined, such as an inversion of the relationship between detection time and number of existing connections, which are otherwiseindependent of the degree of affinity between macrosystem and microsystem devices. Therefore, the proposed method for the detection of voids proved to be satisfactory, given its own capacity for expression, the robustness features in force, the features of an operational nature, such as reading errors in measurements, subjectivity without passing on information, isolation of paving zones, faults and maneuvers there is no unrecorded margin.