Until now, large-scale pipeline condition assessment has been difficult and expensive. In this article, we look at a fast, cost-effective and non-invasive way of giving your pipes a check-up: Water hammer waves.

Water hammer waves: A cost-effective, non-invasive way to check your pipes

The Transient Research Group (part of the Intelligent Water Decisions group) at the University of Adelaide has recently developed a technology to screen the wall condition of water pipes in a non-invasive and cost-effective way. It is achieved by the smart use of controlled hydraulic transient pressure waves, which are also known as “water hammer” waves.

Water hammer waves are pressure waves travelling along a pressurised pipe. Water hammer waves can be generated by a sudden change in the flow, eg a pump failure or a fast valve closure.

In some metallic pipes, wave speed can be up to 1400 m/s. For a century, water hammer was regarded as dangerous and harmful to water distribution systems. The focus of both the industry and academia was on how to prevent the occurrence of water hammer waves. But, controlled water hammer waves, with relatively small amplitude, can be used as tool to detect defects along a pipeline. It’s a non-invasive and cost-effective condition assessment method.

How water hammer-based pipe condition assessment works

The process of water hammer-based pipe condition assessment is akin to the use of sonar waves to detect remote objects.

A typical testing configuration is illustrated in Figure 1. Controlled incident pressure waves, with a typical size of 5 m to 10 m in pressure head, are generated by closing a specialised side-discharge valve. That valve is installed in a pipe through an existing connection point, such as a fire hydrant or an air valve. If the pipe is uniform and in good condition, the pressure wave will propagate along the pipe without any reflections.

Any physical changes along the pipe, such as a short section with a thinner wall due to corrosion, or the spalling of the cement mortar lining, will induce wave reflections. The wave reflections will propagate in the direction opposite to the incident wave. They can can be measured by pressure transducers installed at multiple locations along the pipe. Once again, fire hydrants or air valves are used for the connection to avoid any excavation.

An example of the wave reflection induced by a pipe section with a thinner wall is given in Figure 2.

Safely assess kilometres of pipe at once

Specialised data analysis algorithms are used to interpret the pressure data collected, and to reconstruct an image of the pipe wall condition by estimating the remaining wall thickness. The method is not at all destructive for the pipe, and no physical items are inserted into the pipe. The technique is totally non-invasive. And because the high wave speed enables kilometres of pipe to be assessed in a single test, therefore making the technique cost-effective.

This figure demonstrates the typical configuration used in the field for assessing pipeline conditions using transient pressure waves. It demonstrates the pressure transducers and wave generators at external points like hydrants, along the pipe. The incident pressure waves go in each direction down the pipe from the transient wave generator, and the pressure is picked up along the pipeline. Reflected waves come back towards the transient wave generator, where there is also a pressure transducer. The figure shows that the waves are consistent until the interior of the pipe changes. That change causes waves to ‘reflect’ and travel back towards the generator and pressure transducer.

Figure 1. Typical configuration used in the field for pipeline condition assessment using transient pressure waves.

 This figure illustrates the wave reflection induced by a pipe section with a thinner wall. On the horizontal axis is a time scale. On the vertical axis is a pressure scale. A steady state shows a lower pressure travelling in a straight line. A ‘step wave’ is shown by a sharp rise, and on the graph it looks like a step. Along a steady state wave, a reflection appears like a dip: The line drops slightly, travels along and goes back up. This sectioned reduction in pressure shows that the pipe has a damaged or thinner wall than the rest of the pipeline.

Figure 2. Illustrative example of the wave reflection induced by a pipe section with a thinner wall thickness

The needs of different pipelines require different types of analysis

Various data analysis algorithms have been developed by the Transient Research Group to suit different pipeline condition assessment needs.

Inverse transient analysis

An inverse transient analysis technique has been developed to achieve detailed and relatively high-resolution pipe wall condition assessment [1].

It is an iterative model calibration process. It aims to find a numerical pipe model that produces the same pressure responses as the real pressure measurements. Optimisation algorithms are used to search the optimal parameters (eg wall thicknesses) for the pipe model.

Direct wave analysis

A time-domain direct wave analysis technique focuses on the major pressure wave reflections and therefore the major deterioration in the pipe [2-4].

The size of a wave reflection is used to estimate the wall thickness of the corresponding deteriorated section, and the arrival time of the reflection is used to determine its location. This technique is relatively fast and easy to implement, and the spatial resolution can be scaled to satisfy the different needs of the clients.

Direct wave analysis is ideal as a screening tool for large-scale pipeline condition assessments. It identifies the pipe sections with possible significant deterioration, so they can be assessed in more detail.

Field trials have validated these techniques

The water hammer-based pipe condition assessment techniques as developed by the Transient Research Group have been validated by a number of field trials conducted in Australia. They are applicable to various metallic (with or without cement mortar lining), concrete and asbestos cement pipes.

Figure 3 gives a snapshot of a pressure trace (the red line) measured in an above-ground, mild steel, cement mortar-lined (MSCL) pipe in South Australia. For comparison, the wall thickness of the section of pipe was also measured by the ultrasonic sounding technique (8 circumferential sampling points, 5 meters longitudinal interval in intact sections and down to 1 m interval in deteriorated sections).

It is obvious that there is a strong correlation between the perturbations in the measured pressure traces and the changes in pipe wall thickness. Deteriorated pipe sections were identified and removed from the pipeline system, as shown in the selected photos.

More technical details of this technology can be found in the references, at the end of the article.

This figure shows a chart with two lines, one red and one green, and two sections which break out to show photographs of what the data means. The chart has three series: Wall thickness (mm), Chainage (m), and Dimensionless Pressure. It shows that there is a strong correlation between the perturbations in the measured pressure traces and the changes in pipe wall thickness. Where there is corrosion inside the pipes (shown in the photographs), the graphs show drops in both wall thickness and dimensionless pressure. The chainage series shows us the location of the sections of pipeline that are in poor condition.

Figure 3. Results from a field trial on a mild steel cement mortar lined pipe: measured pressure trace (red line, normalised by the size of the incident pressure wave); average wall thickness as determined by the ultrasonic sounding technique (green line); ultrasonic measurements below 4.2 mm wall thickness (>10% reduction, coloured squares); and selected photos of deteriorated pipe sections (below the main plot).

Improve maintenance and prevent disruptive events

Up to the end of 2015, 67 different systems totalling over 500 km of pipelines had been assessed using this technology. As a result, a number of deteriorated pipe sections and significant defects (such as faulty valves) were detected. The benefits of this early detection include guiding targeted maintenance and preventing disruptive events such as pipe bursts.

In this photo we can see a section of pipeline with a signal generation station. Tthe generator is connected to a side discharge scour valve.

Figure 4. Transient pressure (water hammer) wave generation station.


Water hammer pipeline technology is already available

Even more exciting news is that the novel water hammer-based pipeline condition assessment technology is already available. The company Detection Services Pty Ltd is licensed to deliver tests and analysis on behalf of the University of Adelaide.

Now, it’s time to give your critical water pipes a check-up! For more information please get in touch.

References for more technical details:

[1]          M. L. Stephens, M. F. Lambert, and A. R. Simpson, “Determining the internal wall condition of a water pipeline in the field using an inverse transient model,” Journal of Hydraulic Engineering, vol. 139, pp. 310–324, 2013.

[2]          J. Gong, M. L. Stephens, N. S. Arbon, A. C. Zecchin, M. F. Lambert, and A. R. Simpson, “On-site non-invasive condition assessment for cement mortar-lined metallic pipelines by time-domain fluid transient analysis,” Structural Health Monitoring, vol. 14, pp. 426-438, 2015.

[3]          J. Gong, M. F. Lambert, A. C. Zecchin, A. R. Simpson, N. S. Arbon, and Y.-i. Kim, “Field study on non-invasive and non-destructive condition assessment for asbestos cement pipelines by time-domain fluid transient analysis,” Structural Health Monitoring, vol. 15, pp. 113-124, 2016.

[4]          J. Gong, A. R. Simpson, M. F. Lambert, A. C. Zecchin, Y. Kim, and A. S. Tijsseling, “Detection of distributed deterioration in single pipes using transient reflections,” Journal of Pipeline Systems Engineering and Practice, vol. 4, pp. 32-40, 2013.

Key members in the Transient Research Group

Prof. Angus Simpson angus.simpson@adelaide.edu.au

Prof. Martin lambert martin.lambert@adelaide.edu.au

Dr Aaron Zecchin       aaron.zecchin@adelaide.edu.au

Dr Jinzhe Gong           jinzhe.gong@adelaide.edu.au

Ms Nicole Arbon        nicole.arbon@adelaide.edu.au

Dr Si Tran Nguyen     si.nguyen@adelaide.edu.au