Análise do padrão de pressões e de correlações empíricas para caracterização de vazamentos na rede de distribuição de água de Fortaleza/CE

• Ano de defesa: 2019
• Autor(a) principal: Sousa, José Raimundo de
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Palavras-chave em Português: Recursos hídricos; Análise de séries temporais; Água - Desperdício; Água - Distribuição; Leakages; Water distribution network; Empirical correlations
• Link de acesso: http://www.repositorio.ufc.br/handle/riufc/47383

At a time when natural resources are increasingly scarce, concern about Efficient use of water is increasing. In the state of Ceará, where water supply in the recent years is well below historical average levels, the situation is no different. The leaks in distribution grids are regarded as one of the major villains of water loss and that is why they are receiving special attention. The objective of this paper was to analyze and develop characterization of the variables involved in the water distribution system and generate information to assist in the control of leaks with common data to be obtained for possible for its application by the companies that manage the supply. Were Pressure variation patterns were identified in 17 sensors installed by the city network Fortaleza / CE, evaluating the stationarity of the series, the autocorrelation and the analysis spectral. In a second step, we analyzed the relationships of HDI, network length, consumption and population compared to leakage data during 2018, as well as their characterization. It was observed that the pressure series follow a daily variation cycle for each sensor, having a great correlation with the measurements recorded over multiple 24-hour periods, with indications that during at dawn are the highest pressures, being the time most susceptible to break in the network. The analysis of the leaks shows that its more than 76 thousand occurrences are distributed throughout throughout the year, and there is no period that can be determined as operating. network, which makes it difficult to use detection methods. The empirical correlations show that among all variables tested the number of links has the highest with the leaks, with a coefficient of determination of 0.88. Along with this result is A map of the regions that the company should focus on to combat the leaks is presented, which could exceed 100,000 by 2045 if no intervention is made.