Monday, November 20, 2017

Detecting leaks in water pipes using sound waves

It doesn’t take an expert to know that pipes are used all over the world to transport water for drinking, washing and cooking, among many other uses. When pipes leak (Figure 1), it causes huge problems for human health and the environment, as well as causing massive economic losses.

Figure 1: Leaking water pipe
Water leakage can lead to contamination of drinking water with viruses and bacteria that can cause illness and disease. Although we may not realise it, water is, in fact, a finite resource that is facing immense pressure globally; therefore, wasting water through leakage is a major unnecessary strain on water supplies. Furthermore, water is an expensive commodity and so unnecessary water losses are costly for water companies, right through to their customers. Identifying leaks quickly and efficiently is therefore paramount if we are to tackle these problems. 

Water companies typically identify leaks using sound waves; sensors are placed along pipes and leaks are detected by recording and analysing the vibro-acoustic emission (VAE), or how sound waves interact with the piping. Vibro-acoustic emissions are effective in locating the leak. Vibro-acoustic emissions could also give data on the leak size and leak flow rate, however no predictive methodology currently exists for this. Such information would be extremely useful for water companies and would allow them to assess and prioritise leakage repairs, so that larger leaks are dealt with urgently.
Figure 2: Accuracy of two sensors in predicting
(a) leak flow and (b) leak area within +/- 10% band
(from Butterfield et al., 2017)

To tackle this problem, STREAM researcher Joseph Butterfield looked at whether it is possible to use vibro-acoustic emissions to predict leak flow rate and leak size. First, an algorithm was devised to use as a predictive model for leak flow rate based on vibro-acoustic emission signals. Next, experiments were conducted on different sized holes and different flow rates in plastic pipes, taking vibro-acoustic emission measurements using two types of sensors. The experiments found that leak flow rate has a strong effect on vibro-acoustic emission signals, with increasing leak flow resulting in increased signal amplitude. The model performed well, showing good accuracy in predicting the leak flow rate regardless of the leak area (Figure 2).

Through this work, it has been demonstrated that vibro-acoustic emissions are a suitable method for predicting leak flow rates in water distribution pipes. Further studies testing different shaped holes and different pipe materials could improve the accuracy of prediction and enable water companies to establish the flow rate and location of leaks. This would, in turn, allow them to prioritise fixing bigger leaks first, which benefits human health, the environment and saves both money and resources.


For full reference:

Butterfield, J.D., Meruane, V., Collins, R.P., Meyers, G., Beck, S.B., 2017. Prediction of leak flow rate in plastic water distribution pipes using vibro-acoustic measurements. Structural Health Monitoring.