Using smart meters data to model customer water consumption and evaluate customer flow patterns

Abstract

ABSTRACT The correct determination of water consumption is key to increasing hydraulic modeling reliability. In this context, Smart Metering (SM) is an important concept. The SM systems can provide consumption at a customer level and in small intervals of time. In this scenario, the work focused on a residential condominium, in Brazil, where an SM system was implemented. The work uses data obtained to model consumers’ water consumption and proposes a system to detect metering failures and internal water loss. Three models were proposed: one uses Autoregressive Integrated Moving Average (ARIMA) theory, another uses the Seasonal Autoregressive Integrated Moving Average (SARIMA), and the last uses multivariate linear regression. The average root mean square error of the first model was 14,73, of the second was 13,50, and of the third was 20,99. A metering and consumption system, able to detect internal leaks and others, was also developed.