Underground sewers that safely manage wastewater are essential to modern cities. However, since they are buried underground and difficult to access, their condition can be hard to monitor. Traditional camera inspections are costly and time-consuming, but we need regular inspections so aging pipes do not continue to deteriorate until they reach a breaking point. When damaged pipes leak, they can pose serious risks to public health by contaminating surrounding water, while more severe failures can even damage infrastructure above them, causing sinkholes in city streets.
To help prevent future disasters, researchers at Tohoku University developed a new method to more efficiently predict the risk of severe corrosion in underground sewer pipes. The researchers measured DNA from sulfate-reducing bacteria found in the sewer wastewater, which can contribute to pipe corrosion. Among the bacteria studied, Desulfobulbus showed the strongest relationship with pipe deterioration and was the most important microbial predictor in the model.
“We were surprised by how useful Desulfobulbus, a type of sulfate-reducing bacteria, was for predicting pipe corrosion,” said Professor Daisuke Sano. “Among all the biological factors we tested, it was the strongest predictor, second only to pipe age.”
Then, using ecological modeling, they estimated where these bacteria were more likely to occur and used machine learning to identify pipe sections that may be at higher risk of severe corrosion.
To test the method, the team collected wastewater samples from 40 manholes across a residential sewer network and measured DNA from sulfate-reducing bacteria. They combined these measurements with information on pipe slope and upstream building area to map bacterial habitat potential across the network. The resulting habitat estimates were then combined with pipe age, diameter, and slope in a Random Forest machine-learning model. The model was tested against inspection records from 3,498 pipe sections. Adding the microbial information improved its ability to identify severely corroded pipes.
“By combining wastewater DNA with ecological modeling, we converted invisible biological signals into clear geographic maps,” said Sano “This can help municipalities prioritize inspections more efficiently and identify high-risk areas before costly failures occur.”
Mapping where corrosion-related bacteria are likely to occur can help identify pipes that may need closer attention. This could allow cities to focus limited inspection resources on higher-risk areas and address problems before serious failures occur. After all, when it comes to the infrastructure hidden beneath our cities, out of sight should not mean out of mind.
- Publication Details:
Title: Wastewater environmental DNA-based mapping of sulfate-reducing bacteria for network-scale sewer corrosion screening using species distribution modeling
Authors: Junming Zhang, Kanki Watanabe, Wakana Oishi, Daijiro Mizutani, Mohan Amarasiri, Daisuke Sano
Journal: Journal of Environmental Engineering
