Machine learning and the future of water quality monitoring

Advancements in artificial intelligence are reshaping water quality monitoring, with machine learning (ML) playing a pivotal role in optimizing treatment processes. In a recent issue of Water Innovations, Carollo’s Andrew Salverson and Kyle Thompson, along with Dr. Kate Newhart of Oregon State University, explored how ML-powered soft sensors are revolutionizing potable water reuse systems.

Improving Water Treatment with AI-Powered Soft Sensors

Traditional water quality monitoring relies on physical sensors and laboratory testing, but machine learning introduces a new approach. The authors discuss how soft sensors, intelligent algorithms that predict water quality based on existing operational data, offer a faster and more cost-effective way to optimize treatment. Carollo has applied these innovations in real-world settings, demonstrating their potential to enhance efficiency and sustainability.

Predicting Total Organic Carbon in Virginia

The article talks about the Hampton Roads Sanitation District’s (HRSD) SWIFT Research Center, where Carollo developed an ML-powered soft sensor to predict total organic carbon (TOC) levels, a critical parameter in carbon-based reuse systems. Using historical data from HRSD’s demonstration facility, the team created a boosted trees model that significantly improved TOC prediction accuracy. This advancement allows for more precise ozone dosing, leading to potential energy savings and more effective water treatment.

Reducing Energy Use in California’s Potable Reuse

Carollo also applied machine learning at Las Virgenes Municipal Water District’s (LVMWD) Pure Water Demonstration Facility to address challenges with N-nitrosodimethylamine (NDMA), a disinfection byproduct. Dr. Kate Newhart developed a predictive model using data from Orange County Water District’s Groundwater Replenishment System, which demonstrated the potential to reduce UV energy consumption by up to 26%. They emphasized how this model is now being refined with expanded data collection at LVMWD, potentially leading to even greater operational efficiencies.

The Future of AI in Water Quality Monitoring

These case studies highlight the growing role of machine learning in optimizing water treatment. They highlight that by leveraging AI, utilities can improve monitoring accuracy, reduce costs, and enhance sustainability. As more facilities adopt ML-powered solutions, the industry moves closer to a future where real-time data drives smarter and more efficient water management.

To learn more about these groundbreaking applications of AI in water treatment, read the full article in Water Innovations.