We present a low-cost (≈$150) monitoring system for collecting high temporal resolution residential water use data without disrupting the operation of commonly available water meters. This system was designed for installation on top of analog, magnetically driven, positive displacement, residential water meters and can collect data at a variable time resolution interval. The system couples an Arduino Pro microcontroller board, a datalogging shield customized for this specific application, and a magnetometer sensor. The system was developed and calibrated at the Utah Water Research Laboratory and was deployed for testing on five single family residences in Logan and Providence, Utah, for a period of over 1 month. Battery life for the device was estimated to be over 5 weeks with continuous data collection at a 4 s time interval. Data collected using this system, under ideal installation conditions, was within 2% of the volume recorded by the register of the meter on which they were installed. Results from field deployments are presented to demonstrate the accuracy, functionality, and applicability of the system. Results indicate that the device is capable of collecting data at a temporal resolution sufficient for identifying individual water use events and analyzing water use at coarser temporal resolutions. This system is of special interest for water end use studies, future projections of residential water use, water infrastructure design, and for advancing our understanding of water use timing and behavior. The system’s hardware design and software are open source, are available for potential reuse, and can be customized for specific research needs.

<p>Collecting and managing high temporal resolution (< 1 minute) residential water use data is challenging due to cost and technical requirements associated with the volume and velocity of data collected. It is well known that this type of data has potential to expand our knowledge of residential water use, inform future water use predictions, and improve water conservation strategies. However, most studies collecting this type of data have been focused on the practical application of the data (e.g., developing and applying end use disaggregation algorithms) with much less focus on how the data were collected, retrieved, quality controlled, and managed to enable data visualization and analysis. We developed an open-source, modular, generalized cyberinfrastructure system to automate the process from data collection to analysis. The system has three main architectural components: first, the sensors and dataloggers for water use monitoring; second, the data communication, parsing and archival tools; and third, the analyses, visualization and presentations of data produced for different audiences. For the first component, we present a low-cost datalogging device, designed for installation on top of existing, analog, magnetically driven, positive displacement, residential water meters that can collect data at a user configurable time resolution interval. The second component consists of a system developed using existing open-source software technologies that manages the data collected, including services and databasing. The final element includes software tools for retrieving the data that can be integrated with advanced data analytics tools. The system was used in a single family residential water use data collection case study to test the scalability and performance of its functionalities within our design constraints. Testing with a base system configuration, our results show that the system requires approximately six minutes to process a single day of data collected at a four second temporal resolution for 500 properties. Thus, the system proved to be effective beyond the typical number of participants observed in similar studies of residential water use and would scale well beyond this even with the modest system resources we used for testing. All elements of the cyberinfrastructure developed are freely available in open source repositories for re-use.</p>

No abstract

<p>Water end use disaggregation aims to separate household water consumption data collected from a single residential water meter into appliance/fixture-level consumption data. In recent years, the field has rapidly expanded as the value of disaggregated data has been shown for understanding water use behavior, identifying anomalies, and identifying opportunities for conserving water. Several methods have been developed for disaggregating water end uses from high temporal resolution water use data collected using residential smart water meters. However, most existing methods have been incorporated into proprietary software tools and have been tested using datasets that are inaccessible due to privacy issues, with the result being that neither the code nor the data from these studies are available for verification or potential reuse. We describe and demonstrate a new, open source, and reproducible water end use disaggregation and classification tool that builds upon the results of existing smart water metering and end use disaggregation studies. The tool was designed and developed in Python and can be accessed via any current Python programming environment. It was tested on anonymized, high temporal resolution datasets collected from 31 residential dwellings located in the Cities of Logan and Providence, Utah, USA for a period of one month. Results from different meter types and sizes were tested to demonstrate the accuracy and reproducibility of the tool in disaggregating and classifying high temporal resolution data into individual water end use events. Execution of the tool requires approximately one minute for processing one-day of data collected at a four second time interval for one dwelling. The disaggregation tool is open source and can be adapted to specific research needs. The anonymized dataset we used to develop and test the tool is openly available in the HydroShare data repository.</p>

Residential water end-use events (e.g., showers, toilets, faucets, etc.) can be derived from high temporal resolution (<1 min) water metering data. Past studies have collected data at different temporal resolutions (e.g., 4 s, 5 s, or 10 s) without assessing the impact of the temporal aggregation interval on end-use event features (e.g., volume, flowrate, duration) due to the unavailability of data at a sufficient temporal resolution to enable such analyses. We recorded the time between every magnetic pulse generated by a magnetically driven residential water meter’s measurement element (full pulse resolution) using a new, open-source datalogging device and collected data for two residential homes in Utah, USA. We then examined water use events without temporally aggregating data and compared to the same data aggregated at different time intervals to evaluate how temporal resolution of the data affects our ability to identify end-use events, calculate features of individual events, and classify events by end use. Our results show how collecting full pulse resolution data can provide more accurate estimates of event occurrence, timing, and features along with producing more discriminative event features that can only be estimated from full pulse resolution data to make event classification easier and more accurate.