Pressure management is the basic step of reducing water losses from water supply systems (WSSs). The reduction of direct water losses is reliably achieved by reducing pressure in the WSSs. There is also a slight decrease in water consumption in connected properties. Nevertheless, consumption is also affected by other factors, the quantification of which is not trivial. However, there is still a lack of much relevant information to enter into this analysis and subsequent decision making. This article focuses on water consumption and its prediction, using regression models designed for an experiment regarding an administrative building in the Czech Republic (CZ). The variables considered are pressure and climatological factors (temperature and humidity). The effects of these variables on the consumption are separately evaluated, subsequently multidimensional models are discussed with the common inclusion of selected combinations of predictors. Separate evaluation results in a value of the N3
coefficient, according to the FAVAD concept used for prediction of changes in water consumption related to pressure. The statistical inference is based on the maximum likelihood method. The proposed regression models are tested to evaluate their suitability, particularly, the models are compared using a cross-validation procedure. The significance tests for parameters and model reduction are based on asymptotic properties of the likelihood ratio statistics. Pressure is confirmed in each regression model as a significant variable.
This is an open access article distributed under the Creative Commons Attribution License
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited