An accurate estimation of residential end uses of water is helpful in developing efficient water systems. If not obtainable through direct metering, this information can be gathered by disaggregating and classifying household-level water-use data. However, most automated techniques require fine-resolution data (e.g., 1 s) and end-use parameters which may be unavailable to water utilities. To fill the above gap, this study presents a method for the automated disaggregation and classification of indoor water-use data collected at the 1-min temporal resolution, and by exclusively relying on the end-use parameter values available in the literature. Specifically, the features of each water-use event detected at the household level are compared against the most common event features for the selected end-use category. The results obtained by testing the method with real data collected at 14 households in two different countries (Italy and the Netherlands) confirm its potential in disaggregating and classifying water end-use events with an average accuracy higher than 90% and an average (normalized) root-mean-square lower than 0.06 despite the lack of information about end uses in individual households. This demonstrates that end-use detection is possible even with data whose resolution is closer to that of most commercial water meters.

@en

In the water industry, an accurate estimation of end-use water consumption is helpful for the implementation of efficient water systems and water-saving technologies. This study aimed to explore the characteristics of water consumption at nine households north of Amsterdam (the Netherlands), subjected to water consumption monitoring at high temporal resolution (i.e. 1 s). Overall, 36,297 water-use events monitored over about 447 days were automatically segmented into 44,115 individual events by means of a new rule-based filtering algorithm, and then labelled by expert analysts. A multi-stage analysis was then conducted in order to evaluate daily per capita end-use water consumption, daily end-use profiles, average end-use parameter average, and their statistical distributions. The results achieved provide insight into the features of end-use consumption, confirming that the largest components are typically related to showers/bathtubs, toilets, and washing machines, whereas different end-use parameter distributions can emerge.

@en

Application of smart meters to the residential sector can provide insight into where and when water is used, thereby enabling utilities to achieve an efficient management of water distribution systems. Moreover, detailed information about domestic water use can be obtained by disaggregating smart meter data collected at the household inlet point. In this paper, a rule-based, automated methodology for disaggregating household water-use data into end uses is presented. The methodology is applicable to 1-min temporal resolution data, whose granularity is slightly lower than the one generally used in other methodologies, potentially allowing it to be applied to several contexts in the field of water-use monitoring. The methodology was set up and validated with data collected for 2 months through intrusive monitoring of four households in Bologna, Italy, and represents a pioneering case in which disaggregation performance is directly assessed by the comparison against data collected at each end use. The results obtained showed that the methodology enables household water use to be efficiently disaggregated even if detailed information about end-use features is not available. @en

This paper proposes a short-term water demand forecasting method based on the use of the Markov chain. This method provides estimates of future demands by calculating probabilities that the future demand value will fall within pre-assigned intervals covering the expected total variability. More specifically, two models based on homogeneous and non-homogeneous Markov chains were developed and presented. These models, together with two benchmark models (based on artificial neural network and naïve methods), were applied to three real-life case studies for the purpose of forecasting the respective water demands from 1 to 24 h ahead. The results obtained show that the model based on a homogeneous Markov chain provides more accurate short-term forecasts than the one based on a non-homogeneous Markov chain, which is in line with the artificial neural network model. Both Markov chain models enable probabilistic information regarding the stochastic demand forecast to be easily obtained.

@en