Análise das perdas de água em um sistema público de abastecimento: enfoque na influência da classe metrológica e do tempo de uso dos hidrômetros

Abstract

In Brazil, the average revenue losses of water utilities is around 36%. This value, considered very high, is due to water losses, that corresponds not only to the real losses (leaks in pipes, reservoir overflow, etc.) but also correspond to the so-called apparent losses (illegal service connection, meter under-registration, reading errors, etc.), which can represent more than 50% of total water losses. This study aimed to analyze the water losses in a public supply system, evaluating the influence of metrological class and usage time of meters on apparent losses caused by under-registration. This research was divided into three stages: (i) quantification and analysis of water losses in a condominium located in Itabira-MG, with 83 active connections, by performing daily meters readings of class B meters, that were already installed in the condominium, and posteriorly of the class C meters, which replaced the firsts, in order to compare their performance; (ii) calculation of return time of the investment in meters replacements, aiming to assess the economic viability of this action; and (iii) performing tests in a laboratory scale in 12 of the 83 class B meters of the condominium to determine their indication errors and verify the existence of a possible correlation between this parameter and the usage time of these meters. The condominium average water loss reached 18,9% and 10,4% for micro-measurement performed with class B and C meters, respectively. It was proven that the replacement of class B meters by the class C meters significantly reduced apparent water losses of the condominium, at the 1% significance level. The required time to recover the invested capital for the meters replacement was estimated to occur in 5,6 months, and this investment was considered economically viable. Could not detect evidence of correlation between the indication errors and usage time of meters, at the 5% significance level.