Norway’s Horten Municipality employs Nordic Semiconductor-powered cellular IoT monitoring solution to address drinking water waste from pipe leaks

The solution developed by 7Sense uses Nordic’s nRF9160 low power module to record and relay sensor data to the Cloud enabling a rapid response from engineers if a leak is detected

Horten Municipality in southern Norway, and local smart condition monitoring solution provider, 7Sense, have collaborated on a cellular IoT-based project designed to detect leaks in drinking water networks in the region. The solution is powered by Nordic Semiconductor’s nRF9160 low power module with integrated LTE-M/NB-IoT modem and GNSS, and the project is supported by the Norwegian Institute of Public Health.

According to Norsk Vann, a national association representing Norway’s water industry, Norway has one of the highest water leakage losses in Europe, and as of 2021 the country had an average leakage loss of 30 percent. In several water supplies, 40 percent or more of the purified drinking water is lost due to leaks in the pipe network. The project in Horten Municipality is part of a long-lasting commitment to lower the leakage percentage below the municipality’s individual sustainability goal of 18 percent by 2028. For the last three years it has reported a stable leakage level of 16-17 percent, and this project will help maintain that level or decrease it further by locating small hard-to-find leakages, often found on private pipe connections.

Cellular IoT data transfer to Cloud

To meet the needs of the project, Horten Municipality has worked with 7Sense to deploy a fleet of IoT sensors on its drinking water network that continuously monitor the condition and activity of water pipes using acoustics. High quality microphone heads are magnetically attached to the water pipes or valves to ‘listen’ for activity over a broad spectrum of frequencies that is recorded by the nRF9160-powered sensors. This data is periodically relayed to Horten Municipality’s Cloud-based visualization platform using the cellular IoT connectivity of the Nordic module, from where advanced machine learning (ML) algorithms are used to determine if any anomalies that may occur over time are indicative of a potential leak. If a leak is suspected, Horten Municipality engineers can quickly investigate to limit potential water losses.

“The advantage of continuously monitoring noise levels in the pipes is that the noise is often strongest early in the leakage phase with a clear high pitch,” says Jan Einar Nornes, Project Manager, Smart City Horten. “As the leakage grows the noise levels tend to decrease and it’s harder to distinguish from background noise. The solution was deployed early in 2024, and while it’s still early we have detected at least one leak so far.”

nRF9160 offers superior radio sensitivity

For the system to be effective, enough sensors need to be deployed to give the Cloud platform and the ML algorithms sufficient data to work with. At the same time, because the system is several meters below ground, the superior radio sensitivity and power consumption of the nRF9160 were essential to ensure reliable cellular connectivity and avoid the need to regularly recharge batteries. The solution is powered by a 3.6 V, 14 Ah D-sized lithium primary cell, providing a battery life more than five years thanks in part to the power-saving characteristics of the Nordic module.

“We have worked with Nordic and the nRF9160 since the introduction of the module in 2018,” says Frode Stensaa, CEO, 7Sense Technologies. “The level of integration, Arm processor, NB-IoT modem, and well-known SDK [software development kit]—alongside the radio sensitivity and low power consumption—made the Nordic module a clear choice for the sensor.”

While the current pilot project has been a success, work is ongoing to refine the solution for extended deployment. The final version will include a three-axis accelerometer to monitor the position of the sensor on the pipe to ensure the sensor isn’t unintentionally moved, for example if service personnel accidentally contact it when performing other work, or rust degrades the effectiveness of the magnet. Further firmware and backend algorithm development will also refine the effectiveness of the solution as it is gradually scaled to 1000 plus sensors for Horten Municipality’s mid-size water infrastructure.