Barney’s Blog no 7 - Water Leaks From Plastic Pipes

Water leak detection preoccupies all water companies who are regularly interrogated by the regulator as to progress in reducing water leakage. Water companies and their contractors are well organised to find leakage on mains and are generally successful at reducing mains leakage.


As costs of water increases and customers have to pay for it, the water companies are happy to inform the customer that they have a leak on their service pipe, and then "give them a ticket" to get it repaired – So far, not many of the water companies actually find and fix leaks between the street and the building, but usually pass the job the job on to a contractor who invoices the customer, who may or may not have insurance cover for the find and repair process.

Insurance companies generally provide this cover as part of a household policy, now the water companies themselves invite an extra payment to repair pipes and drains in the event of leaks or blockages.

Soon the water companies themselves will be obliged to take more responsibility for the service pipes to buildings, and the implications are that water charges will need to rise to cover those costs.

Nearly all recent installations of service pipes are non-metallic, blue HDPE plastic pipe being the popular material, or alkethylene. Older service pipes are usually metallic, lead, copper and galvanised steel are used, and it is relatively easy to find a leak on metallic pipe because the water escaping through the hole in the pipe wall creates a distinctive leak noise that travels through the pipe wall and the ground.

Traditional methods use a metal listening stick and the experienced ear of the water company's technicians. More modern approach is to use ground microphones, and some versions of the equipment have a noise capture system to allow the strength of the leak noise to be evaluated at different points on the surface, along the pipe run, to aid pinpointing the leak. But any listening device, traditional or recent, needs to have an audible leak noise; generally the noise is easy to find on metallic pipes, but leaks from plastic pipes are often impossible to hear on the surface.

So is there a method to find leaks on plastic service pipes if there is no leak noise?

In fact there has been an alternative method for more than thirty five years, using a tracer gas mixture, put into the pipe and detected where it rises to the surface above the leak. The Sensistor Company in Sweden invented a sensor to detect Hydrogen, and improved and perfected the methodology for finding leaks.


The tracer gas

It is a good idea to use Hydrogen for a tracer gas. H2 is a very small light molecule and it tends to disperse rapidly from its source, and of course because its very light, it rises quickly. Hence its use in early airships; but everyone remembers the Hindenburg disaster. Not a problem with the gas mixture that is used to trace leaks, it uses 95% Nitrogen and 5% Hydrogen, and this mixture is inert and can not be ignited. The Hydrogen.Nitrogen trace gas method has Drinking Water Inspectorate approval for use in Potable Water. Nitrogen makes up about 78% of the atmosphere, and the Nitrogen in the tracer gas acts as a carrier for the Hydrogen. And the overall mixture is relatively cheap to produce.
Some other trace gas systems use Sulphur Hexafluoride SF6 or Helium. Helium is much more difficult to produce, and SF6 is a far heavier molecule, slow to rise to the surface, and also very poisonous.


Injecting the tracer gas

Depending on the task, the tracer gas can be fed into the pipe either when the water is under pressure, or empty.
Using suitable adapters the tracer gas can be injected through a garden tap, fire hydrant, or a meter housing after the meter is removed

If the pipe is in use, then output pressure from the gas bottle regulator needs to be set to about 1-1.5 Bar above the water pressure. This ensures that the gas enters the water stream and dissolves, like CO2 in a fizzy drink. Once the gas is dissolved in the water, the water looks cloudy/white, and this is a useful and simple way to check that the gas has arrived at the desired point of the pipe past the leak, and of course gas will be present and detectable as it escapes from the water pipe.
It is helpful to use the water in the pipe as a means to carry the dissolved gas towards the leak position, and it requires less gas to be used compared to an empty pipe.


Detecting the tracer gas

Providing the pipe route is known, detecting the gas requires the operator to test the ground at say 0.5M intervals along the pipe route. The operator needs to be aware that the gas takes some time to come to the surface, depending on the pipe depth and the nature of the ground between pipe and surface. The gas takes 15 or 20 minutes to rise through loose ground, and maybe 10 minutes longer if the ground is very compacted, or very wet and compacted or the pipe is very deep. If there are large impervious rocks in the backfill, the gas will go around them. In concrete paths and roads, or block drives, the gas will generally rise through cracks before rising directly through the concrete.


Problem leaks

I was asked to help to locate a leak in a pedestrianised shopping area. There were 14 small shops, all fed from the same main meter. The managing agent had determined that the leak was on the main feed, because the main meter continued to indicate flow when they shut off all the individual stop-taps to the shops.
The main pipe route was uncertain, and it was also not clear if it was a metallic or plastic pipe or a mixture of both.
To add to the complication the ground material was perpetually wet as these shops were built on what had formerly been the riverbank of this town, but to make it easier the whole of the area was paved using square concrete paving blocks, and the gaps between them should allow the gas to emerge.

Correlation produced some points of interest, but there was no satisfactory leak sound on the surface. This suggested that we were dealing with mostly plastic pipe, so tracer gas was the solution.
We arranged to inject the tracer gas via a 15mm pipe close to the end of the 60M pipe run, after shutting off the supply feeding the shopping area at the incoming water meter. We let the gas fill the pipe and run for about 30 minutes, but strangely, no gas appeared on the surface at that time. After 40 minutes the gas was detected and the leak marked.

For this trace gas detection we were using both the Sensistor equipment, and also a new H2 trace gas detector from Seba KMT.
This is a newly developed version of the HL5000 ground microphone. The main amplifier has new hardware to display the results of the H2 trace gas detection, when used with the new sensor for detection of H2, and also has its usual variety of sensors to act as ground microphones.

We left site and the digging crew arrived a few days later to dig on our marked place. They dug down towards the pipe that they assumed was blue HDPE; but they did not find their pipe. They found 2 cables, power and telecom, an old rusty water pipe and a clay ware drain duct that they expected to be the drainage or rainwater system. But no blue pipe. They were not happy.

So we returned to the site a few days later, injected the gas as before and obtained the same position result as before. Pete and Chris the excavators were giving me a hard time unhappy at not finding the leak last time in the same place. They removed some of the paving tiles to get a more precise position of the trace gas, and there was gas detected at a different position, but still the greatest response in original position. They dug down to the original position found the cables, old pipe and clay ware duct/drain. At the second position they found another cable and drain; neither position showed blue plastic pipe. They looked me up and down, and measuring me against the size of the hole..

The reason for the gas spreading was clear as the excavation progressed; a large sheet of plastic damp course material was laying in the backfill, so obviously the gas spread out around it. But still no blue plastic pipe was revealed. Next we tested the walls of the trench and found a higher concentration of gas in one area. As one of them excavated the Seba HL5000H2 went off the scale, but right above the clay ware duct/drain. So they extended the excavation along the duct and with the HL5000 on minimum sensitivity it was clear that the source was closely. Another fork full of backfill and one of them heard the gas escaping from the pipe, as they got to the end of the duct they found the blue pipe, about 30 cm from our original marking, and probably deflected by the plastic material. A little further along the pipe they found the cause of the leak – a coupling between two pipe lengths that had a split insert. And I never thought that excavation could be so exciting!

Now the reasons for the difficult location was clear; obviously the water from the leak ran away along the duct that housed the supply pipe, and much of the gas went the same way. Rest of the gas ascended but was deflected by the plastic damp course material in the backfill.

So the leak was found and repaired and the client was happy.

New two way solutions for finding leaks on plastic service pipes.

Tracer gas has several advantages to locate hard to hear leaks. One of the reasons why this sort of leak from plastic pipes does not create noise is that a water leak into saturated ground makes little noise, and there is no noise from the water escaping through the hole in the pipe wall.
However gas leaving the pipe with water into saturated ground forms into bubbles and a bubbling noise can be heard on the surface along with the hiss of the gas, using one of the ground microphone attachments for the Seba HL5000H2.

Because the hissing and bubbling noise will be present immediately the gas escapes from the pipe, it will help to pinpoint the leak even before the gas reaches the surface.
Using the new HL5000H2 the operator can pinpoint and confirm the leak when trace gas is used by 2 methods, acoustic as well as using the H2 gas sensor. In both methods, acoustic and using the trace gas, the noise capture histogram and the gas concentration value histogram can be most helpful to pinpoint the leak. Of course, good leak pinpointing means more accurate leak location and less costs of reinstatement.


For any more information on this subject please email