Barwon Water deploys StormHarvester AI to detect sewer blockages before spills

Title

Barwon Water deploys StormHarvester AI to detect sewer blockages before spills

Content

Barwon Water, a water utility serving southwest Victoria in Australia, deployed StormHarvester's AI-powered machine learning platform in 2025 to transition its wastewater network management from reactive to proactive. The project, which began in 2025 and spans multiple locations across Barwon Water's network, focuses on using sewer level sensors combined with ML analysis to generate proactive blockage alerts, with a particular focus on the coastal town of Lorne on Victoria's Great Ocean Road. "Insights from wastewater analytics platforms, such as StormHarvester, have helped us detect and respond to potential sewer blockages earlier, allowing us to investigate and mitigate issues that could have otherwise impacted our customers and the environment," said David Snadden, General Manager, Smart & Sustainable Infrastructure, Barwon Water. The deployment addresses three operational priorities for Barwon Water: detecting developing blockages early before they become incidents, prioritising high-risk and high-consequence areas such as known pollution or flooding hotspots, and ensuring alerts are genuine and actionable enough to deploy operational teams effectively. The platform also provides inflow and infiltration analysis across the monitored network to help target investigations and reduce excess flows, while evidencing compliance with Barwon Water's General Environmental Duty (GED) obligations to the state EPA. StormHarvester's predictive models have already demonstrated tangible impact in three documented cases in Lorne. In the first case, the system detected unexpectedly high sewer levels at a manhole, accounting for rainfall and seasonal conditions, generating an alert for Barwon Water's smart networks team. Jetting teams mobilised days later, located a developing partial blockage, and cleared it before heavy rainfall arrived in the following days. In the second case, an alert enabled proactive scheduling of a line clean ahead of a busy holiday weekend, with crews discovering a fat accumulation about the size of a soccer ball approximately 30 metres from the maintenance hole, along with tree roots and wet wipes; the early intervention prevented a major environmental and customer-impacting spill. The third case saw another developing partial blockage cleared ahead of subsequent heavy rainfall. By detecting abnormal behaviour before high levels or spills occurred, crews were able to intervene and avoid potential incidents and reduce operational disruption during the peak summer period. Lorne was chosen for focused case study analysis because it experiences high wastewater flows during holiday periods and has been historically prone to sewer blockages and overflows, with the conditions heightening consequences during peak tourism periods. Source: StormHarvester case study, April 2026