New Ultrasonics Systems Locate Leaks Beneath Pavement

by James Hall
James Hall and Associates, Mission Viejo, CA

One of the most useful technologies I have utilized to detect leaks in underground lines is ultrasound.  Ultrasonic detectors being sensitive to a specific range of ultrasound, between 20 and 100 kilohertz, sense the ultrasonic emissions produced by a leak.  A problem with early ultrasound systems was that they were not particularly helpful in finding leaks beneath concrete and asphalt, or even 'in pipes or pressurized cables buried at depths of a few feet beneath bare dirt.  But, improved techniques and new technology have changed that.

Ultrasounds, by definition, are beyond the limits of normal hearing.  It takes a sophisticated detector to transpose ultrasonic signals to the range of human hearing.  Ultrasonic detectors are lightweight, versatile, easy to use, and relatively inexpensive.  Some, such as the Ultraprobe 2000, are battery operated and contained in a hand-held "pistol" housing.  Some have a frequency adjustment dial to provide tuning capability.  With these devices, one "hears" the ultrasounds caused by leakage through earphones, and can gauge their intensity by the deflections of the analog meter.   With these new ultrasonic detection systems, I am amazed at how deep I can go underground to detect these leaks.

Pinpoints Leak Beneath Asphalt at 5-Foot Depth
One of my most interesting recent challenges was at the University of Southern California (USC), Los Angeles, where I found a leak 5 1/2 feet beneath an asphalt-topped parking lot and inside a steel conduit.  The leak had developed at a coupling in a 2-inch copper condensate line that was in a 16-inch conduit together with a 4-inch steam line.

When I went to the college that morning to meet with the men who had laid the line, they admitted they were unable to find the leak, although they had been searching for several days!

I asked them to introduce air at 80 psi into the line, and shut off the 4-inch steam main, to avoid creating ambient noise that might interfere with my picking up the ultrasonic signal. I looked over the ground thinking the leak was at the further manhole and did not find it.  I got about 20 feet away from the manhole, the distance between each manhole being 235 feet.  I was sure that, between the air leaking out and the sound of the water emerging from the leak side, I should be able to pick up the leak.

I used the Ultraprobe in the contact mode with a 31-inch extension rod supplied with the instrument and an optional 8-foot interconnect cable.  By probing the asphalt, touching the surface with the metal waveguide, moving from one spot to another, and tuning the frequency between 20 and 25 kHz, I picked up the sound and narrowed it to 22 kHz. After a few minutes search, I observed a minute deflection of my ultrasonic scanner's meter, which along with the volume of ultrasound in the headset indicated a leak.   It had a low level hissing quality, much as one would expect from a leak behind a wall or under a floor.

Fortunately, because of the high frequency short wave nature of the ultrasounds, I could probe to within a foot in each direction.   Once I heard the sound, I knew I could hit it dead on. I took a can of paint, sprayed a point on the asphalt, and extended it several feet so as to not lose it when the workers came in with a backhoe.  We opened up the asphalt, got down to the 16-inch conduit, which we opened with an acetylene torch, and there it was - a 2-inch coupling that had split.

Leak Found in 30 Minutes
We had found the leak within 30 minutes and pinpointed it to within six inches of its site.  It was that precise.

Before this, the greatest depth at which I had detected leaks was 2 1/2 feet under a concrete platform used to test the performance of leak detection devices.  In that case, I found the leak in a 1/2-inch copper line with 5 psi of air.

I also found leaks for a leak detection company whose owner, having seen ultrasonic detectors at a trade show, knew that some day he would have to employ ultrasonics to find leaks under swimming pools.  That day came, and he called me.  I pinpointed the leak under 18 inches of pool decking.  He bought two units on the spot.

The first step in leak detection is to have enough air pressure introduced into the line to create turbulence at the leak site.  Using earphones, one can pick up the sound and narrow in on the white noise emanating from the leak site. Flowing through a restricted orifice under pressure, a gas passes from a pressurized laminar flow to low pressure turbulent flow.  This turbulence generates a broad spectrum of sound, usually referred to as "white noise," which has strong ultrasonic components.  Because the ultrasounds are loudest at the leak site, a sharp rise in ultrasound during an ultrasonic scan is usually a clear indication of leakage.

I suggest using the frequency tuner in the 20 to 40 kHz range.  Taking that one controller and running it between 20 and 40 kilohertz lets one narrow in on the leak.  Most leaks I've found were in the 20 to 40 kHz range.

Several procedures can be incorporated to improve your chances of success. Since an ultrasonic detector senses a turbulent flow, it is necessary to try to produce this effect at the leak site in non-pressurized lines or cables by injecting some gas pressure into the line.  In some cases, isolating sections of a line increases the ease of detecting the turbulent flow.

If this is not possible in a line carrying liquids, adding a gas to the liquid will help. Even though sections cannot be closed off, gas pressure adds strength to the leak signal at the leak site.

If there is reason to suspect a slow leak rate or if it is difficult to utilize gas injection, one must try to get the detector closer to the leak source with long metal rods driven into the ground near the pipe. Touch these rods with the ultrasonic detector's contact probe and listen for a slight rushing sound. The coupling of the rods with the earth will increase the instrument's ability to sense the sound of the flow as the fluid leaks from the pipe. At times it may be difficult to drive rods all the way down. For these situations, spikes or large nails can be driven into the surface and then touched with the contact probe. These procedures are most effective when the instrument is tuned to about 20 kHz.