Vibration-based Assessment of municipal pipe NETworks for IoT-enabled real-time monitoring and machine learning-aided maintenance planning (VA-NET)

Keeping track of the status of pipe networks for drinking water and sewage is time consuming and costly, partly because they are located underground.

At the same time, parts of these pipe networks are aging and have structural problems such as leakages, sagging and corrosion. This, in turn, leads to operational problems.

A significant problem is that sewer pipes is infiltration from groundwater, precipitation and exfiltration from the drinking water pipes into the sewer pipes.

This so-called infiltration and inflow correspond to the same amount as the wastewater produced in households and other activities and twice the volume that is transported, pumped and purified in Sweden, due to a substandard sewage management system.

This leads to increased electricity, and chemical use for transport and purification and increased volumes of untreated wastewater to lakes and streams.

Poor status of the pipe network can also cause sediment deposits and sewer blockages, which in turn can cause basement flooding.Today´s most widely accepted technique for examining the status of sewer networks is to use a video camera that is manually inserted into the pipe network and films the pipes while an operator simultaneously observes and assesses pipe defects.

This is time consuming, costly and subjective. The goal may be for a medium-sized municipality to come back and film the pipes every ten years.

In smaller municipalities with few resources, the status of the pipes has never been systematically investigated since they were constructed several decades ago.

Therefore, there is a need to investigate and develop other methods to be able to investigate the status of the pipe networks faster and cheaper. One such method could be to make vibration measurements and acoustic analyses of the pipe networks.

An advantage of such a method is that there is no need to go down the pipe network, and faster diagnosis of pipe status. This would make it easier for water and wastewater organizations to work with proactive maintenance.

Acoustic methods have been used to find leakages in drinking water networks and gas pipes also to investigate the status of for example, dam and bridge structures.

However, today vibration methods for assessing the status of sewer networks do not exist.The overall goal of this project is to investigate the potential of using vibration measurements to investigate the status of sewer networks and, by extension, to develop a method that continuously assesses the status of the pipes with real-time measurements and machine learning algorithms.

In this project, a method will be developed to collect and analyze vibration/acoustic data from primarily gravity pipes. The work includes selecting appropriate sensors and frequency ranges and carrying out modeling of the collected data.

Furthermore, the chosen method will be tested on a pipe network that is built up in the lab to identify typical patterns that the signals give for different types of structural wiring defects. Based on the results from the lab, the method will also be tested in the field. There we will make acoustic measurements on pipes that have also been flushed and filmed.

The results from the acoustic measurements will be compared with the results from the filming.

Based on the results of lab and field measurements, recommendations for further development of IoT-based real-time measurements and analysis will be given.This project will be carried out in collaboration between Luleå University of Technology (LTU) and the water and sewerage department at Boden municipality and the water and sewerage companies MittSverige Vatten & Avfall, Tekniska Verken in Linköping and Luleå miljöresurs.

LTU is represented by scientists from Water and Wastewater Technology and Acoustics.

The municipal actors will provide input on which defects should be prioritized to be investigated and contribute with appropriate pipe locations for field measurements. That part includes human resources in the fi