In advanced water metering infrastructure (AMI), leak detection plays a pivotal role in managing non-revenue water and promoting conservation. Two primary approaches are employed in leak detection systems: local detection at the water meter itself and system-wide analytics performed at the headend or utility control centre. While both serve the same ultimate goal, identifying continuous, abnormal consumption indicative of a leak, their methodologies, capabilities, and implementation implications differ significantly.
Meter-Based Leak Detection
Modern ultrasonic smart meters, such as Kamstrup or Badger, are equipped with internal logic and high-resolution sensors capable of detecting minute flow variations. These meters continuously monitor flow at set intervals (e.g., every 60 minutes) and apply algorithms locally to identify usage patterns inconsistent with typical household behaviour. For example, a meter may flag a leak if water is flowing continuously over 24 hours without interruption, a scenario unlikely in most residential settings.
The primary advantage of meter-based detection is speed and granularity. Since the logic resides at the edge, the meter can identify a potential leak in near-real time (60-minute intervals) and trigger alerts immediately.
In Canada’s colder climates, where seasonal freezing may lead to pipe bursts, early detection at the meter can reduce property damage.
Furthermore, meters with onboard leak detection logic can store event flags and transmit this data via the AMI network during scheduled communication windows, reducing data transmission volume while maintaining actionable intelligence.
Headend-Based Leak Detection
By contrast, headend systems (Badger, Neptune, Master Meter, and Sensus), typically housed within utility data centres or cloud-hosted platforms, analyze consumption data aggregated from thousands of meters. Using historical trends, machine learning algorithms, and customer-specific consumption profiles, the headend can detect anomalies such as increasing base flow, nighttime usage spikes, or seasonal variation discrepancies.
This centralized approach benefits from system-wide visibility. It can correlate customer water usage with environmental variables (e.g., weather data, billing cycles, district metered areas) to identify widespread issues such as zone-level leaks, unauthorized use, or meter tampering. Unlike local meter logic, headend detection can look at broader patterns and perform cross-customer comparisons, yielding more advanced predictive insights.
These advanced insights can be correlated to identify leaks most precisely in water mains and service lines. This approach is far more utility focused versus the in-meter approach which is more end user focused. Most utilities are more interested in their own infrastructure instead of the infrastructure owned beyond the meter in the customer’s premises.
However, headend systems can be more data-intensive and typically require frequent interval reads and robust AMI infrastructure. In rural Canadian municipalities with low meter density or limited connectivity, headend detection may not provide timely results, making it better suited for urban environments with dense populations and mature data systems.
Pressure Monitoring
The integration of pressure monitoring within water meters introduces a powerful new dimension to leak detection.
By continuously measuring pressure fluctuations alongside flow data, smart meters can detect anomalies such as pressure drops from burst pipes, slow declines from hidden leaks, or pressure transients that may precede infrastructure failure.
When combined with acoustic and flow-based detection, pressure sensing enhances diagnostic accuracy and enables utilities to distinguish between customer-side issues and system-wide events.
In Canadian systems, where freeze–thaw cycles and aging infrastructure pose ongoing challenges, embedded or temporary pressure monitoring offers utilities a more proactive and data-rich approach to identifying and responding to leaks in real time.
System-wide pressure monitor offers far greater insights into leak detection compared to edge leak detection solutions alone. However, when they are used in concert to each other, the value is dramatically enhanced.
Acoustic Leak Detection in DMAs
To enhance leak detection beyond meter-level flow data, many utilities are deploying acoustic monitoring systems within DMAs. These technologies use sound wave propagation to identify and pinpoint leaks along water mains and service lines; often before surface signs appear.
One example is Itron’s MLOG system, which integrates a permanently installed acoustic sensor with or nearby a meter endpoint. The sensor listens for vibrations in the pipe, and when abnormal acoustic patterns are detected, such as those caused by a pinhole leak or corrosion failure, alerts are generated and sent via the AMI network to the utility headend for review.
Similarly, acoustic leak sensors installed on fire hydrants, such as the Echologics sensors, allow for non-invasive monitoring of transmission and distribution mains. These sensors continuously record and analyse acoustic signatures in the pipeline. When correlated across multiple points in a DMA, they can triangulate the leak location with high precision. This approach is particularly valuable in dense urban environments where excavation is costly and disruptive.
Acoustic detection is well suited for municipalities with aging infrastructure, especially where pipe material, soil conditions, and historical leak patterns warrant proactive surveillance.
In Canada, where frost heave, pipe corrosion, and long winter seasons contribute to main breaks, the early detection offered by acoustic systems reduces water loss and mitigates the risk of catastrophic failure.
The Role of Traditional Acoustic Equipment in Modern Leak Detection 
Handheld hydrophones, ground microphones, and portable correlators are foundational tools for water leak technicians. These systems detect and amplify the sound generated by pressurized water escaping from pipes. Leak noise travels along the pipe and through surrounding materials, and trained technicians can interpret these acoustic signals to locate the leak’s origin with impressive accuracy.
Other technologies, such as HighSense Solutions and PipeMinder, are both popular analysis technologies as well. Here is a brief description highlighting HighSense Solutions and PipeMinder in the context of water network pressure monitoring:
HighSense Solutions, Inc., based in Canada, specializes in smart utility monitoring and leak detection systems designed to minimize non‑revenue water in municipal and commercial settings. Their offerings include devices such as the HSS‑Sense Pro, which integrates with water meters and utility infrastructure to detect abnormal pressure patterns and leaks in real time.
Meanwhile, PipeMinder, developed by Syrinix (now under Badger Meter’s umbrella), delivers high-resolution pressure and flow monitoring for distribution and transmission mains. With sampling rates up to 128 samples/sec, up to five years of battery life, and cellular connectivity, the PipeMinder‑ONE logger paired with the RADAR analytics platform provides immediate alerts for pressure transients, leak and burst events, and integrates water quality data for deeper diagnostics.
Both solutions offer advanced granular monitoring capabilities and alerting; HighSense targets decentralized, meter‑level monitoring in Canadian settings, while PipeMinder delivers network‑wide, high‑frequency pressure and flow surveillance with deep analytics.
Despite the rise of permanent, integrated leak detection systems like in-meter, headend systems, Itron’s MLOG and fire hydrant-mounted acoustic sensors, handheld systems offer critical tactical advantages:
Verification of AMI or acoustic alerts: A sensor may detect a potential leak, but utilities often rely on portable hydrophones or ground microphones to confirm and localize the issue before excavation.
Detection in non-instrumented areas: In rural or low-density parts of Canada where AMI penetration is limited or acoustic nodes are not economically viable, walk-around surveys with portable equipment are the most practical option.
Mobile response to emergent issues: After a customer reports unusual usage or pressure loss, field crews equipped with correlators can investigate in real-time, without waiting for scheduled data transmissions from the AMI network.
Mobile Equipment Overview and Capabilities
Hydrophones and Ground Microphones These tools detect leak sounds transmitted through soil or pavement. Operators listen through headphones and use sound filters to isolate leak characteristics. Advanced models provide visual spectrum analysis, enhancing accuracy in noisy urban environments or during freezing weather when acoustic transmission may be dampened by frost.
Leak Noise Correlators These portable systems use two or more sensors placed on access points (e.g., hydrants or valves) to triangulate the leak location based on sound arrival times. Correlators are especially useful on metallic or PVC mains where leak noise propagates well and the distance between access points is known.
Tracer Gas and Dye Testing Though less common, in cases where acoustic detection fails, such as on plastic pipes or in saturated soil, tracer gas systems and fluoroscopically visible dyes may be deployed. These are particularly relevant in locations with high ambient noise, such as near railways or industrial sites.
Integrating Traditional and Digital Systems
A well-designed utility leak detection strategy does not rely solely on automation. Instead, it integrates:
AMI/Smart Meter Alerts: For initial detection of customer-side or service line issues.
Headend Analytics: For identifying persistent usage patterns, anomalies, or area-wide losses.
Fixed Acoustic Sensors (e.g., Itron MLOG, hydrant sensors): For continuous monitoring and early detection of leaks in DMAs.
Walk-Around Surveys with Handheld Devices: For detailed confirmation, mapping, and follow-up investigations.
Utilities often build a tiered response model where suspected leaks from AMI or acoustic data are prioritised for field crew inspection using hydrophones and correlators. In municipalities like Toronto, Vancouver, or Edmonton, this layered model reduces unnecessary excavations, accelerates response times, and improves public trust by showing proactive management.
Canadian Considerations 
In Canada, where winter frost, freeze–thaw cycles, and expansive soil types can obscure or dampen acoustic signals, manual verification using ground microphones remains critical. In northern or remote communities without full AMI coverage, traditional equipment is the only viable method of leak detection. Furthermore, Indigenous, and rural utilities may benefit from mobile leak detection trailers equipped with correlators, allowing rapid deployment to high-risk areas.
Training and knowledge retention are equally important. As utilities adopt more AI and cloud-based headend tools, ensuring that field crews maintain strong acoustic interpretation skills is essential. These experienced technicians provide the human expertise needed to validate algorithmic predictions and resolve ambiguous or borderline cases.
Pipe Material Impacts
The effectiveness of acoustic leak detection is significantly influenced by the type of pipe material in a water distribution system.
Metallic pipes such as ductile iron, cast iron, and steel transmit sound efficiently, allowing leak noise to travel long distances; ideal for acoustic sensors, correlators, and hydrant-based monitors.
In contrast, plastic pipes like PVC and HDPE are poor conductors of sound; they absorb and dampen acoustic signals, limiting detection range and reducing accuracy. This poses challenges for utilities using acoustic methods in newer subdivisions or rural areas where plastic mains are common.
Asbestos cement pipes fall in between, with moderate transmission characteristics. In Canadian systems with a mix of pipe types and challenging soil or climate conditions, utilities must adjust sensor spacing, use complementary tools like tracer gas or AMI data, and tailor detection strategies based on pipe material to ensure effective leak identification.
Pit versus Basement Meter Installations for Acoustic Detectors
When an acoustic leak-detecting meter is installed inside the house, its ability to detect distribution-side leaks, such as those on the service line between the curb stop and the dwelling; is limited. Interior placements isolate the meter from external pipe vibrations, making it difficult to capture leak sounds in buried infrastructure.
Conversely, when the meter is located in a pit near the curb, it is acoustically coupled to the service line and can more effectively detect leaks occurring between the main and the property. This positioning enhances early detection of underground leaks and provides more comprehensive system monitoring, especially when integrated with AMI.
For Canadian municipalities, outdoor pit installations offer greater utility-wide value for acoustic monitoring, particularly in areas with aging service lines or frost-susceptible soils.
However, the Pipework material will again have a significant effect on the performance of any acoustic leak detection systems.
Pipework Depth Affects
In colder Canadian climates, deeper pipe burial, often required to prevent freezing, can negatively affect acoustic leak detection performance.
Water mains buried at depths of up to 4 metres (13 feet) and service lines at approximately 2 metres (7 feet), especially when made of PVC, reduce the transmission of leak-generated sound to surface-level sensors.
The increased soil mass, particularly if frozen or water-saturated, attenuates acoustic signals and dampens vibration. This makes it more challenging for both fixed and portable sensors to detect or localise leaks accurately.
To compensate, utilities may need closer sensor spacing, advanced signal processing, or alternative methods like tracer gas in deep or plastic pipe installations.
Technical and Operational Trade-offs
Detection Speed: Meter-based logic offers faster detection and alerting; headend analytics may lag due to data batching and processing delays.
Resolution: Meter-level detection is precise to the individual service connection; headend analysis is better for trend and pattern recognition across multiple connections.
Data Volume: Headend systems may require larger datasets and more frequent reads; meter-based detection transmits fewer, targeted events. Implementation. From experience, one-hour intervals are common for both types of leak detection systems.
Cost: Smart meters with advanced local analytics may carry a higher upfront cost, while headend systems require ongoing software and IT infrastructure investment. The headend system can work equally well with lower cost, Positive Displacement meters making the whole solution far more affordable.
Maintenance and Verification: Field crews can act on confirmed meter alerts with confidence; headend anomalies may require additional validation before dispatching.
“To find a leak is to find value; value in capturing non-revenue water, in service reliability, and in the trust of the people that utilities serve”. – Metercor
Conclusion
Both approaches to leak detection, meter-based and headend-based, are complementary rather than mutually exclusive. A well-designed AMI system in a Canadian municipality should leverage the strengths of both, using local logic for immediate, actionable alerts and headend analytics for strategic planning, billing integrity, and infrastructure management.
While integrated leak detection systems, smart meters, headend analytics, and fixed acoustic sensors, are revolutionising water management in Canada, traditional handheld tools remain vital in confirming, locating, and responding to leaks. These methods offer mobility, precision, and field adaptability that complement high-tech systems. A resilient utility strategy blends innovation with time-tested tools, empowering both technology and technicians in the pursuit of sustainable water stewardship.
By far, the captured lost revenue drives these projects. Non-revenue water losses are quantified and remediated. The new meter ultrasonic technology plays a significant role in capturing low and high flow losses. Most importantly the emerging pressure monitoring feature holds the greatest value in an optimized water distribution system.
As the water utility landscape shifts towards predictive and data-driven operations, integrating both detection layers will be essential in safeguarding water resources, reducing operational costs, and improving customer satisfaction.
