Using Meter Readings to Identify Conservation Opportunities: A Low-Cost, No-Code Digital Solution

Water utilities sit on a wealth of customer usage data, yet many utilities struggle to use this powerful asset to affect operations. The sheer amount of information gathered by utilities can be a major hurdle keeping staff from using the data to improve processes and identify opportunities for conservation efforts. As the amount of information available to utilities increases, so too do the technical requirements of utility staff to process and analyze the data in a meaningful, methodologically sound manner. A final barrier is socializing the resulting analysis across utility staff to ensure that findings are acted upon and fully implemented. As water utilities continue to deploy Advanced Metering Infrastructure (AMI), the number of individual measurements grows exponentially. Whereas water meter readings that occur in monthly, bimonthly, or quarterly intervals provide little insight into the true usage patterns of households (except seasonal trends), measurements from AMI - whether hourly or sub-hourly - provide new opportunities to identify usage tendencies that could not be detected previously. State and local governments are developing sustainability plans encompassing every aspect of their operations, including energy efficiency, renewable energy, carbon neutrality, and water management. As water continues to become a scarcer resource and climate events continuously wreak havoc on water supplies globally, water conservation and management are becoming increasingly vital to long-term sustainability efforts. While mass marketing and educational campaigns for all customers may provide some value, targeted efforts typically give higher returns on investment. Going after low-hanging fruit (e.g., residential meters with high usage) is a cost-effective method to reduce water consumption across a utility territory. Identifying the proper customers to target, however, is not straightforward. While hourly consumption data exist, processing and analyzing the information is time-consuming and requires considerable technical skills and resources. A water utility in the western United States sought assistance with processing, cleaning, and analyzing several years of hourly AMI measurements to identify ideal candidates for water conservation marketing. Requirements of the task included a no-code solution that could be used across utility staff without any additional software purchases and minimal training. Tetra Tech developed a low-cost, web-based, point-and-click tool that allows utility staff to easily filter information and set specific thresholds on key water use measurements to identify specific groups of residential meters to target for water conservation education. The interface allows the utility to directly export a list of meters meeting selected criteria to contact for educational purposes and conservation efforts. As an initial step, hourly water use records were transferred to a cloud-computing platform and were cleaned and inspected for completeness. After removing questionable and incomplete measurements, publicly available data on home sizes, property size, and the existence of a pool on the property were added to the working file to provide additional household attributes. Each meter's Homeowner's Association (HOA) was also added to the data, serving as a proxy variable to identify homes of a similar build year and size. Lastly, the latitude and longitude of each property/meter were geocoded from the utility's Customer Information System to allow for straightforward mapping of customers across the utility territory. The resulting data set provided sufficient information to calculate water usage metrics and patterns to identify characteristics of residential meters with relatively high hourly and daily water use. In addition, the hourly measurements provided adequate granularity to identify irrigation system patterns across households, with several distinct watering patterns identified a single hour on Monday/Wednesday/Friday; a single hour on Tuesday/Thursday; a single hour each weekday; or a single hour on Saturday/Sunday. Scheduled irrigation systems contribute disproportionately high amounts to overall meter consumption; pinpointing homes with these systems provided an initial list of residences available for educational campaigns. Extended periods of water use one or more gallons of water used during consecutive hours were identified and tabulated as a percentage of total water use. These often indicate a leak within the home's plumbing system or water delivery, again providing the utility with an easy home to contact for maintenance and water conservation. After fully analyzing and calculating metrics across individual meters, all results were aggregated to the level of individual HOAs. This provided the utility with a holistic approach to target neighborhoods with above-average water use for example, neighborhoods with a high prevalence of pools tended to have higher overall water use. Similarly, an HOA with many homes using scheduled irrigation systems has much higher water use than an HOA with lower irrigation. Before AMI, many of these patterns were simply unidentifiable without significant investment in sub-metering equipment the patterns, especially those with a time-volume aspect (i.e., identifying a scheduled irrigation system that occurs each Monday, Wednesday, and Friday from 9 to 10 AM), would not have been noticeable from monthly or quarterly readings. All of these data and metrics were integrated into a dynamic dashboard consisting of ten separate reports, each focusing on distinct combinations of usage characteristics that allow for analysis and segmentation of two years of hourly readings from approximately 13,000 residential meters. The dashboard maps homes/meters and water usage characteristics for each meter. Filters, sliders, thresholds, and ranked data allow the utility to focus on specific segments of their customers. For example, one report allows the utility to identify all meters with average daily water use above a user-defined threshold or simply show the N meters with the highest average daily water use (where the user defines N, i.e., 50 highest users). Another report allows utility staff to select from rankings across HOAs the N HOAs with the highest average hourly water use, the N HOAs with the highest prevalence of pools, the N highest water usage per square foot. Once the utility is satisfied with its selections, the resulting meters are exportable to a .CSV file for marketing and education. As AMI deployment continues to grow and utilities are tasked with finding solutions to reduce water consumption, leaks, and waste, straightforward and low-cost digital solutions can provide actionable insight to utilities based purely on data.