AMI + Analytics curb water loss and boost the bottom line

Somewhere in the U.S. and Canada, a water main broke today. That’s a safe estimate because together, the U.S. and Canada see a whopping 260,000 main breaks every year, an average of 712 breaks daily. This results in an 11% average water loss, as H₂0 that’s been treated for resale seeps back into the ground or flows down streets rather than hitting customer taps.

The Utah State University research that crunched those numbers also revealed that 20% of water lines in these North American countries are beyond their useful lives and need to be replaced.

Those are the pipes waiting to burst, to force emergency repairs, open up sinkholes, increase utility overtime expenses, and cause countless hours of inconvenience. Distribution system leaks create what utilities call real losses, while water that gets consumed but not paid for constitutes what’s called apparent losses. Both types of losses hit a utility’s bottom line, but AMI data, plus analytics can help utilities cork some of the losses they see from a variety of causes.

Let’s get real

One of the best ways to tackle distribution system leaks is via acoustic leak detection, which uses data loggers to listen for the sound of water escaping from a hole or crack in underground pipes. Just as air rushing out of a leaking balloon can make a sound, so do leaks in pipes, and those sounds carry along the length of the pipe itself.

The sounds differ by pipe composition and pipe size, and powerful analytic engines leverage machine learning to teach algorithms to factor these things in. Then the system uses data from multiple loggers to correlate where the leak is, providing utility crews with precise locations so that they can take action before the leaks grow into more serious problems.

Gutermann AG, a Switzerland-based company, is a global leader in acoustic leak detection. With some 60 years of experience creating leak-detection technology, the company has an extensive proprietary database of manually classified real-life sound samples. It is this data Gutermann used to train its own artificial intelligence model used to detect and locate underground leaks.

This is why Aclara chose the Gutermann system to integrate into its AMI solution for utilities that want an acoustic leak detection solution. Together, Aclara and Gutermann technologies give utilities peerless leak detection capabilities combined with superior communications technology for a top-tier, highly reliable way to monitor the integrity of water systems continuously.

When the City of Elmhurst, Ill., implemented an Aclara AMI system, it added the Gutermann leak detection to its AMI network. Before getting advanced metering, Elmhurst estimated it was losing 800,000 gallons a day between water main leaks and under-registering older meters. These 292 million gallons of non-revenue water equaled some 20% of the city's annual water purchase, or $1.4 million. Three-quarters of the loss came from distribution leaks.

Acoustic leak detection resulted in a $1 million reduction in purchased water cost in 2019. The city also saved more than $500,000 in fixed network cost by having meters and leak detection on the already-installed Aclara RF network.

Learn more about acoustic leak detection in Elmhurst. Download the case study.

How analytics exposes apparent losses

According to AWWA, there are three causes of apparent loss in water distribution:

1. Unauthorized consumption

There’s another word for this: theft. However, data collected by AMI can be analyzed to identify suspicious consumption patterns and detect potential theft. In addition, the AMI system can send the utility a tamper alert if a customer tries to disconnect or route around a meter.

2. Data handling errors

Meter reading, billing, and archiving errors contribute to apparent water losses by reducing the amount of revenue the utility collects. AMI solutions, which read meters automatically, reduce the errors that occur when data is entered manually into a handheld computer and, ultimately, the billing system.

3. Inaccurate meters
Meters not registering properly, either because they are failing or improperly sized, are a source of apparent water loss that is difficult to quantify without analytics. Short of pulling meters out and bench-testing them, it’s hard to determine if the meters are dying or decaying, which often happens with age. When meters are incorrectly sized for the type of service needed, they also register water usage incorrectly. The City of Elmhurst estimated that 5% of its water loss was due to aged meters that had started under-registering.

Role of Analytics

Water utilities can use a variety of analytics approaches to recover top-line revenue. One way is to analyze data to identify dying meters that are under-registering water usage. These meters are often not found until they actually fail.

Even meter-replacement programs are unlikely to remediate under-registering meters because utilities often conduct such programs on regular schedules and don’t differentiate between good and bad meters. This means that utilities replace good meters along with bad ones at the utility’s expense. Analytics programs can pinpoint bad or incorrectly sized meters, allowing utilities to find faulty meters faster and take a more surgical approach to meter replacement.

To properly identify dying or decaying meters, apparent water loss detection software analyzes historic AMI data and then ongoing monthly water-use data. This establishes a baseline for each meter, allowing irregularities from the pattern to show up through proprietary algorithms and ongoing analyses.

More Ways to Save

Analytics can also ferret out other types of revenue and customer service improvements. The intelligence derived from analyzing payment patterns, for instance, can help utilities calculate the risk of specific customers defaulting on payments. This analysis enables utilities to develop proactive outreach to at-risk customers and create special programs designed to reduce the number of cut-offs for delinquent payments.

Check out our analytics white paper and discover more ways to leverage AMI to stem real and apparent water losses.

This blog was originally published in 2020 and updated in July, 2025