The recent development of a hand-held prototype acoustic (sonic) locator has shown promise in its ability to help utility personnel locate buried pipes. The operating principle of the locator is to send a sonic impulse into the ground and detect an echo from the pipe. A series of echo readings are taken at several locations, and result in an immediate cross-sectional display showing the pipe or pipes in lateral position.

The locator can find pipe through dirt covered with grass, concrete, or blacktop. The response is essentially the same for plastic or metallic pipes of the same diameter. The locator has been successfully tested in the laboratory and at a few utility locations around the United States. Its development was sponsored by the Gas Technology Institute and the Operations Technology Development (OTD), and the manufacturer has signed the license to commercialize the device.

Research effort

The driving force behind this research effort is the need for technologies for detecting and locating buried underground natural gas pipes. Natural gas is distributed locally and over long distances by pipelines. Pipelines buried a few feet under the ground are subject to leaks and to failures due to accidental damage. The vast majority of new gas distribution pipes are made of polyethylene. It is often very difficult to locate buried pipes, especially those made of plastic where no tracer wire was installed. The damage and injuries from accidental penetration of buried pipelines are very costly; therefore, it is extremely desirable to produce a method for detecting and locating plastic pipes. This system should operate near the surface of the ground, be suitable for most soil conditions, and simple to operate.

The goal here is to present the results of work using sonic (acoustic) waves to detect and identify buried objects, including small-diameter plastic and metal pipe. A portable, hand-held, sonic pulse-echo system was developed and tested. The device detects buried small diameter metallic and non-metallic utility pipes at shallow depths. This technology was evaluated at both local and field locations around the United States. Good acoustic penetration was achieved in practically all types of soil and moisture conditions, and most surface conditions to depths of five feet.

Both GPR and acoustic (sonic) technologies were seriously considered at the beginning of the project. Ground-Penetrating Radar (GPR) is the primary technology that is used to detect uncooperative targets that are underground. Uncooperative targets are those that contain no active transmitted signal. GPR works well in some regions, but in other regions the attenuation of the soil to microwaves is too high to allow the detection of small-diameter pipes. For example, soils that contain high quantities of clay, magnetite, or water, especially salty water, are difficult or impossible to penetrate with GPR at frequencies high enough to detect small-diameter plastic pipe. Figure 1 shows the relative advantages and disadvantages of each technology. Sonic technology was chosen for development because it was more universally applicable. The concept diagram of the sonic system is shown in Figure 2.

Technical background

A sonic pipe detector operates as shown in Figure 3. A transducer transmits a short acoustic impulse into the ground. The acoustic wave reflects from any acoustic discontinuity. The interface between a solid and a gas (inside of a natural gas pipe) provides a practically 100% reflection coefficient. The reflection coefficient of earth and the outside of a utility pipe is lower, and between earth and rock is still lower. A receiving transducer detects the acoustic echo as well as a surface wave that travels directly from the transmitting to the receiving transducer.

The abilities to distinguish small diameter pipes and pipes in close lateral proximity are improved as the acoustic frequency is increased (the wavelength is shortened) but the attenuating power of the earth also increases with frequency. Earth attenuation sets the maximum usable frequency at any particular depth. The attenuation law is:

A=A0 10–kfx/20

Where: A = Acoustic amplitude at distance x with attenuation

A0 = Amplitude ignoring attenuation

f = frequency (KHz)

x = distance into the earth (cm)

k = soil attenuation constant (db/KHz cm).

The published values for attenuation range from 0.1 to 0.9 db/KHz cm depending on the type of soil, moisture level and compaction. The attenuation measurements performed during the project indicate typical values of 0.2 to 0.6 db/KHz cm.

There is a minimum frequency that can be used. This occurs (approximately) when the wavelength of the acoustic wave equals the circumference of the smallest diameter pipe to be detected. The relationship between frequency and acoustic wavelength is:

f ë= v

Where: ë = wavelength (cm)

v = propagation velocity (cm/sec).

The ability of the pipe to reflect the wave back to the receiving transducer is related to the pipe’s acoustic scatter cross-section which is illustrated in Figure 4. The vertical axis is the log of the scatter cross-section, ó, divided by the physical cross-section (ð r2 where r = radius of the pipe). The horizontal axis is the log of the ratio of the circumference of the pipe (2 ð r) over the wavelength. As can be seen the scatter cross-section is close to the physical cross-section if the wavelength is shorter than the circumference of the pipe. If the circumference is smaller than the wavelength, the scatter cross-section drops rapidly. This means that at lower frequencies, the wave diffracts around the pipe without reflection even if the reflection coefficient is 100%. Similar observations are also true for GPR reflections from non-metallic pipes.

Range resolution (the ability to distinguish pipes one behind the other) depends on acoustic bandwidth. Put it another way, the acoustic pulse must be short because

R = v/2 ?f

and, ?f = 1/ô (approximately)

Where: R = resolution (cm) - needed pipe separation to produce distinct echoes

?f = bandwidth (Hz)

ô = pulse duration (sec)

What this all means is that one should use the maximum frequency permitted by the depth of the pipe and soil attenuation. This frequency determines the wavelength and hence the smallest pipe that can be detected. The maximum bandwidth (shortest pulse length) should also be used with the understanding that the higher part of the acoustic spectrum will be attenuated if the operating frequency is also high.

Equipment

All data was taken with a personal computer with auxiliary digital-to-analog (D/A) converters to create the pulses from a digital vector, and analog-to-digital (A/D) converters to digitize the echo data. The initial prototype (Figure 5) used a desktop computer. The hand-held portable system (Figure 6) used a board level computer with custom-designed auxiliary circuits. Eighteen-bit converters were used in the prototype with gain control circuits, but the hand-held used 24-bit converters and all gain functions were controlled with software. Matlab was used for all signal processing.

Results

Initially, sonic piezoelectric transmitting and receiving transducers were developed to determine if pipe detection was feasible using sonic technology. Experiments were performed on buried pipes in top soil, sand and clay. Three problems were discovered: the transducers were not sensitive enough, it was difficult to couple the sound into the earth, and there was a strong surface wave that traveled directly from the transmitting transducer to the receiving transducer. Despite these problems the feasibility of detecting buried pipes was demonstrated by using a high voltage pulse amplifier and integrating hundreds of pulses to improve the signal-to-noise ratio. These high voltage circuits and long integration times were unsuitable for a portable instrument. These early systems used an array of several transmitting and receiving transducers to cover a larger area. It was discovered that getting all of the transducers coupled at the same time, in general, was not an easy problem to solve.

Parallel experiments were also conducted with GPR. The GPR easily penetrated sand (as does acoustic waves) and penetrates topsoil. However, the GPR signal penetrated into the clay only a few inches, but the sound wave detected pipes from three to five feet deep with the use of acoustic technology. It was on this basis that work continued on only the sonic system. The surface wave could be separated from pipe echoes by judicial separation of the transmitting and receiving transducers and using a pulse that did not ring. Generally, reducing the ringing also reduces the sensitivity of the transducer which was already too low to begin with the development. Hence, the transducer development became a focus of the earlier effort. Ultimately magneto-restrictive transmitting transducers produced enough transmitting power at low drive voltages and suitably mounted accelerometers provided wideband receiving elements for the prototype acoustic locator.

Portable systems

The project planned development included two portable systems for the locating device. The first prototype (Figure 5) allowed experiments to be done in the laboratory and outside, but was not hand-held. These transducers can be evaluated in the artificial environment of buried pipes and also on public utilities on the street. The second system (Figure 6) was both portable and hand-held with batteries that lasted several hours of street operations to locate buried pipes. This second portable system was sent to several utility sites around the country for evaluation of the technology.

Prototype

During this prototype development phase, echoes were shown on a trace that displays echo amplitude vs. time (A-trace). Based on the modeling of the system and experimental data, the relation between sound propagation velocities, operating frequencies and minimum detectable pipe diameters was determined. This exercise also concluded the importance of operating at the highest frequency that soil attenuation permits if small diameter pipes are to be detected. As the transducers moved over the pipe, the echo from the pipe increased in amplitude and then decreased as the transducers moved away. Other echoes registered were related to the base of the concrete foundation.

Hand-held prototype

A portable, hand-held prototype was built and taken to several utility sites around the United States. These tests were done in conjunction with utility companies in their routine work. The field trials were conducted on dirt, asphalt, compacted rock, and grass in residential and industrial areas.

The hand-held prototype data were compared with available data from utilities and was concluded that practically pipes were detected at most locations accurately. Note that no actual digs were performed to expose pipes to confirm the results. Pipes were indicated in some locations where the available data shows that there were no pipes. This was attributed to abandoned utility pipes as confirmed by utilities from the system maps.

In some locations there were echoes from a fixed depth below the surface underneath a road. These were from the interface between the road base and native soil. These were distinguished from a pipe because the pipe echo would disappear as the measurement point was moved laterally from the pipe but road bed echoes remained as the measurement locations were moved in all directions. The pipe echo showed as a second echo beyond the road base echo.

There were intermittent results on grass and crushed rock due to inconsistent coupling of the transducers to the surface. On grass, the transducers could be repositioned and pressed into the soil to obtain the pipe echo. Coupling to rocks and other rough, solid surfaces remain a problem. Pipe echoes were obtained through rocks, but only with difficulty. Other issues with the prototype were related to the signal processing time (10 to 15 seconds) and an expert to interpret the data.

As a result of this experience, input from the project sponsors, and the commercial partner, several refinements have been made and currently included in a commercial emulator of the device. The processing time has been reduced to about one second. One additional receiving element has been incorporated to allow automatic, rather than operator interpretation, to distinguish between surface wave phenomenon and echoes. This technique also permits the automatic rejection of some other nuisance reflections. Also, a sensor has been placed in the transmitting transducer that automatically determines coupling contact with the surface. This sensing element automatically senses coupling issues and alerts the operator. The operator can either address the coupling issue or move the transducer a few inches, and recollect the data. In addition, the device can be configured for two surface conditions – one for concrete and blacktop, and other for grass and dirt – with two sets of optimized transducers. These improvements are being evaluated and plan to be incorporated in the first production prototypes.

Conclusions

A hand-held acoustic system has been developed and successfully tested at several utility sites. The signal analysis has been dramatically improved so that data can be easily interpreted by the locator crew. A commercially available sonic pipe detection system based on this technology is planned to be introduced next year. n

Acknowledgment

Based on a paper presented at the American Gas Association Operations Conference held in Phoenix, Arizona.