Predictive maintenance of pipe networks - Smart technology for optimised flushing of sewer lines

Abstract

Reoccurring blockages in sewer systems may reduce flow capacity and increase the risk of sewer overflows, resulting in environmental and societal consequences. To minimise the occurrence of such events, regular maintenance flushing of sewer lines with known blockage-related issues can be performed. Converting from a strategy based on pre-scheduled time-intervals to a predictive maintenance strategy could optimise these operations. This study investigates how a predictive maintenance strategy based on water level measurements from sensors mounted in manholes can be designed and implemented. The data analysed in the project was collected from two sensors installed by Sustainable Waste and Water (SWW), in Gothenburg, along with twenty datasets from sensors in Malmö, provided by VA SYD. The proposed method consists of three main steps: data collection, data analysis and warning system design. The results indicate that the proposed method is applicable for various conditions and sewer systems, including Gothenburg. For the studied datasets, the warning system successfully identified increased water levels associated with build-up and blockage events without generating any false alarms. However, longer monitoring periods and data from additional sites are required to fully evaluate the reliability and long-term potential of the method. Especially, regarding pipes which are subjected to blockages consisting of fats, oils and grease. A predictive maintenance strategy could however reduce unnecessary maintenance flushing operations, resulting in lower green house gas emissions and more efficient utilisation of operational personnel and resources.