Multivariate analysis of the pressure variation in intermittent water supply systems and the impact ondemand satisfaction
Abstract In intermittent drinking water distribution systems, large volumes of the water are wasted due to leaks in the distribution networks. Similarly, user service is not always satisfied in the time required to fill the storage, nor with sufficient pressure. Hence the importance of this study. Measuring the variability of pressure in the distribution network and determining the factors that influence the definition of a sufficient minimum hours of service, is a first step to change to a continuous service 24/7, in order to minimize the volumes of lost water and meet demand. 347 pressure sensors were placed in a network to detect changes in pressure and obtain data for three years. This study presents a new approach to determine the operating policy of the operating agency that provides the service intermittently. Two objectives are pursued: pressure variability – to minimize Leaks – and define the minimum hours of service. The analysis was performed using multivariate statistical techniques, including Principal Component Analysis, Correlation Matrix and ANOVA's, to explore the association between objectives. The results obtained show that the pressure distribution has a Gaussian behavior and that the hours of service has a Poisson distribution.