Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
9mapping projects. However, UIT believes that with appropri- ate incorporation of new ideas, an S-wave system tailored to deep utilities can be developed. Modification of field tech- niques, measurement system hardware, and data processing methodologies could produce a system that is more field effi- cient for this purpose, less time-consuming, less costly, and most important, more effective at finding the deep targets. This system requires development testing because previous research on seismic properties of soils at the scale needed for utility mapping has been limited. UIT used the results from the R01B project in which initial measurements of seismic properties of soils were made to optimize receiver specifications. The R01B project developed an S-wave seismic system for mapping dense, shallow net- works of utilities. This project intended to use similar tech- niques and develop instrumentation and analysis techniques that could be complementary to the R01B project. The technology used accelerometers as receivers to improve higher-frequency data quality. As stated above, the old system was unable to collect high-frequency data because of the geo- phone receivers used. The signal sources for this project were electronically driven impact sources of two different sizes. The larger source was used to collect lower-frequency data for large-scale velocity analysis, and the smaller source was used to collect shallow velocity data, as well as reflection data, for imaging the target. Having a more complete knowledge of velocity over the depth range of interest allows accurate cal- culation of target location and depth. The methodology chosen for this part of the project was an attempt to find a technology, unlike GPR, that will image a small target over a depth range of 0 to 50 ft. In some cases, it is possible to use GPR for finding deep utilities, but often GPR waves do not have the necessary depth of penetration. GPR signals are subject to attenuation, especially in clay soils, that vastly decreases depth of penetration. Also, for deep util- ities, lower-frequency GPR signals are necessary, and anten- nas to operate at the lower frequencies are often no longer Phase 2 dealt with the execution of the planning performed during Phase 1. Three distinct activities were carried out dur- ing Phase 2. Innovation prototypes were constructed and ini- tial testing was carried out. On the basis of the results of bench testing and limited field tests, some of the technologies were eliminated from field testing. A series of presentations and documents were prepared to inform the end-user com- munity of the efforts during this project. Finally, in-service testing was performed with two of the candidate technologies on a DOT site. This testing was witnessed by a SUE firm to provide independent feedback. Innovation Prototypes Development Seismic Reflection Locator UIT was the primary subcontractor for the seismic reflection approach that was part of the development of prototypes (Task 4). The proposed research focused on imaging deeply buried utilities with shear wave (S-wave) seismic techniques. The research team believed the best targets were larger, deeper pipes, but testing was necessary to determine the detectable limits of size and depth. The team proposed using standard common midpoint stacking of S-wave seismic data as typically generated by seismic surveys. The cross section in Figure 3.1 shows a hyperbolic diffraction associated with a 54-in. sewer pipe at 38-ft depth. Hyperbolic diffractions are created when a survey traverses a linear or point target, similar to the signa- tures obtained with GPR surveys. The pipe image may not be as clear as desired, but the example shows the potential for seismic technology in an area where GPR can image only the upper 1 to 2 ft. This pipe signature is diffuse because the seismic survey used lower-frequency hardware and standard imaging techniques normally used for deeper geologic mapping. The techniques require intense field effort and time-consuming data process- ing, making such surveys probably too costly for infrastructure C h a P t e r 3 Technology Development
10 available because of FCC regulation of the frequency bands and power levels (lowering the frequency requires raising the power to maintain the same performance). Very low-frequency GPR antennas are not shielded because of their size, allowing noise from outside sources and reflections from surface objects to contaminate the data. In searching for a measurement tech- nique to serve the projectâs purpose, the seismic techniques appear to be good candidates. As seen in Figure 3.2, seismic techniques have shown technical success in mapping a target that is of interest in this project. Seismic Analysis Techniques Several possible seismic techniques are available. P-wave refraction and reflection, S-wave refraction and reflection, and surface wave techniques are all currently used for various purposes in the industry. In choosing the best one for the proposed project, it was important to keep the depth range and target size in mind, but the research team also considered detailed technical aspects of the measurements, such as wave- length, capability to generate the frequencies needed, attenu- ation in the types of soils expected, and data acquisition system parameters. Separately, the team considered data pro- cessing, imaging, and interpretation matters as well. Refraction and surface wave techniques are commonly employed to measure depth to layers such as water table, bedrock, and geologic layers of interest. They tend not to be effective at detecting, locating, or mapping small linear tar- gets such as utilities, thus leaving P-wave reflection or S-wave reflection as candidates. Choosing between these two is com- plicated. However, it has been shown that shorter wavelengths are important for detecting smaller targets. Thus, the team wanted to keep the wavelength as short as possible. Wave- length is a function of the velocity at which the wave travels through the soil. Typically, the smallest target that can reliably be detected by a wave will have a diameter equal to about one- quarter of the wavelength. Figure 3.3 shows quarter wavelengths of three kinds of waves in soils the research team would likely encounter: GPR (for comparison), S-waves, and P-waves. The values in the figure are computed using parameters that represent âgoodâ soil conditions for each technique. In suitable soils, GPR has a very good detection capability. The task is to find a tech- nique that can do as well in other soils not suitable for GPR. Seismic waves tend to be more suited for soils that are wet and clay rich, so they are a good complement to GPR. The figure demonstrates that the overall S-waves mimic the wavelengths of GPR well under the assumed conditions. Because it is generally easier to generate source waves that are lower in frequency (similar to common GPR wave- lengths), the team decided to use S-waves instead of P-waves. It is necessary to generate and propagate S-waves in the 200- to 700-Hz range to have 10- to 2-in. target diameter detection capability. Based on experience, the team could readily gener- ate waves in this frequency range. Achieving a detection capa- bility of as small as 2-in. diameter with P-waves requires generating frequencies of at least 3,000 Hz, which tends to be difficult, and they deteriorate quickly with distance. The original approach to designing the S-wave sources was based on mathematical modeling of the soils-to-source interaction. From the R01B project, a large database of seismic-wave velocities in soil was amassed to populate the model. The performance of the model was not adequate. Seismic components were redesigned and tested empiri- cally. One round of functional testing was completed in the Houston area. Figure 3.1. Stacked S-wave return data. Figure 3.2. Seismic reflection concept.
11 It became clear during the course of the project that no innovation prototype for seismic reflection technology would be ready within the schedule and budget of the R01C project. This technology was not included in the field-testing task. As of this writing, UIT continues work on this technology out- side of the SHRP 2 projects. Appendix D contains additional technical support infor- mation for the seismic reflection technology. Active Acoustic Locator GTI proposed to develop a portable, active acoustic pipe- location prototype suitable for proof-of-concept testing on natural gas and water mains and sewers/sewer laterals or any open conduit. This work endeavored to use a common technol- ogy platform with the seismic reflection technique for the sake of economy. In the active acoustic technique, a known signal is injected into the medium being carried by the pipe. The known signals are generated by an acoustic driver connected at the service of a natural gas line, a water hammer generated at a hydrant, or an acoustic driver hanging in a sewer manhole. The acoustic wave propagates through the medium in the pipe, not along the pipe wall. However, as the signal travels through the medium inside the pipe, be it liquid or gas, a portion of the signal couples into the pipe wall. Vibrations from the pipe wall propagate to the surface of the ground, where they are detected. In applications where the appropriate signal can be injected, the prototype should provide the operator with the location and depth of the buried utility. The prototype was based on two technologies previously developed and patented by GTI. The first was an active acoustic technique for locating plastic pipe. The second technology was field-portable acoustic hardware developed to pinpoint buried leaks, along with algorithms to eliminate interference. Field crews successfully used this hardware, shown in Figure 3.4, to pinpoint small natural gas and steam leaks in the presence of substantial background noise. Active Acoustic Signal Injection The prototype used two techniques to locate the pipe: ampli- tude and time of flight (TOF). The amplitude technique relies on the signal strength to localize the pipe. Because the injected signal is known, correlation signal-processing techniques can be used to discriminate the pipe-location signal from background noise. The TOF technique uses an acoustic signal composed of a series of bursts of a known waveform, with each burst sepa- rated by silence. This provides a different advantage: the tech- nique is less susceptible to variations in coupling between the sensor and ground. Laboratory instrumentation powered by an inverter was used to verify the TOF technique. Portable hard- ware needs to be developed to apply the technique. In addition, lower frequencies need to be used to increase the depth of detec- tion. The key technical issues to be solved are development of accurate depth-location algorithms. Measurements were made at GTIâs pipe farm facility, and at the Talbotton Road test site in Georgia. The depth of applica- tion was extended by using lower frequencies. The sensor pack- age from the passive leak detection unit was modified for use at a frequency range centered around 300 Hz as opposed to the original 1,000 Hz. Each sensor, its signal-conditioning electron- ics, and the radio are housed in a single unit. The digital signal processor used in the leak pinpointer was replaced with a more energy-efficient model. The goal was a unit with a reasonable form factor and long run times between recharging. An acoustic driver was developed for use in empty pipes. The driver converts the signal from the main unit into sound. A device was also proposed that creates water hammer pulses and is equipped with a sensor to relay the water hammer sig- nal to the prototype central processor; this device was not put Frequency (Seismic-Hz / GPR-MHz) Qu ar te r W av el en gt h (in .) Quarter Wavelengths Figure 3.3. GPR versus seismic quarter wavelengths in soil. Figure 3.4. GTI digital acoustic leak detector.
12 into practice because of time and budgetary constraints. The prototype was programmed to use both the TOF and the amplitude techniques. The main unit was programmed to automatically interpret the results, display them, and give the operator instructions for the next step. Passive Acoustic Detection GTI also believes that this prototype can perform passive detection of the acoustic signatures of various utilities. Pas- sive noises include flow noise generated by natural gas or water flowing through the pipe or acoustic hum generated in electrical cables. Three-phase electrical cables are designed to minimize the electrical field around them. However, inter- actions between the cables generate large acoustic vibrations that can be detected at a distance. Because the frequencies are similar, the passive and active techniques can be implemented with the same sensors, signal-processing hardware, and read- out display. The exception is that the active technique requires injection of a signal. Seismic and Acoustic Compared One of the overall goals of the project was to develop a suite of complementary technologies. The seismic reflection method has the advantage at locations where the buried facility is inaccessible. The active method has some advantages over the seismic reflection method in those instances where there is access. Because the acoustic signal only travels from the piping to the ground surface, it suffers less attenuation. The active signal is injected into a particular pipe; therefore, the pipe is positively identified. The signal originates in the pipe instead of reflecting from its surface, relaxing the requirement that the wavelength must be scaled to the pipe diameter. The active method can use lower frequencies, which suffer less attenuation to achieve extra depth. Active acoustic location of water mains was expected to be particularly effective because of the large signal expected from a water hammer tool and because of the high conduction of sound through water. Likewise, the passive acoustic technique was expected to be effective on live electrical cables. As a result of time constraints, the demonstration system used a laptop rather than an embedded system display to dis- play the pipe location and depth results. This version of the hardware underwent preliminary testing at a site in Manteno, Illinois. The results of the testing were used to make adjust- ments to the prototype preparatory to further field testing. The output of this task was an innovation prototype sys- tem that can locate buried pipelines by injecting an acoustic signal into the medium within the pipe. Depth estimation is available in the active signal injection mode. Depth estima- tion is not supported in the passive mode. Appendix E contains additional technical support informa- tion on the active and passive acoustic locating technology. Scanning Electromagnetic Locator GTI proposed to develop a portable electromagnetic (EM) pipe-location prototype suitable for proof-of-concept testing on metallic natural gas and water mains and sewers/sewer lat- erals. In this EM technique, a rotating EM field is projected into the soil containing metallic pipes. The projected field induces eddy currents in metallic objects, which in turn pro- duce a detectable field. The rotating field is generated by a set of driven coils, so phased as to gradually rotate the driven field through 360° about a central axis. The primary frequency of the EM signal and the rate of the field rotation are indepen- dently adjustable. The concept is shown in Figure 3.5. One or more sensing coils monitor the EM field as it rotates. The sens- ing coil signals are captured, along with the angle of the Figure 3.5. Scanning EM locator concept.
13 rotating EM field. Metallic materials within the range of the instrument disturb the EM field as seen by the sensing coils. The location of the buried metallic facilities would be in cylindrical coordinates referenced to the prototype. This would allow the EM innovation prototype to be used in two distinct modes: deployed on the surface along the path of interest or within a nonmetallic pipe. When used from the surface, the prototype could detect metallic objects in a semicylindrical volume beneath its path, allowing adjacent facilities to be resolved more readily than would be possible with standard EM locators that place a sig- nal on a single facility. EM Prototype Implementation The prototype was based on the metallic joint locator (MJL) previously developed by GTI and successfully used to pinpoint metallic joints, repair clamps, and service tees on natural gas pipelines. The MJL, shown in Figure 3.6, produces a stationary AC magnetic field oriented directly down into the earth. This requires crews to move the MJL over a large area to locate fea- tures or have some prior knowledge of the pipe location. The rotating field prototype allows a wider area to be scanned with each pass. The proposed work was to introduce the rotating magnetic field technique, modulation techniques for the EM field, and signal processing to trace buried pipes. The modula- tion techniques to be evaluated were polyphase sinusoids and pulse or single-phase sinusoids modulated by a polyphase sinusoid. These modulation techniques, in conjunction with multiple drive coils, produce the angular rotation of the EM field axis. The primary EM frequency is limited to 200 kHz or lower, based on prior experience using AC electromagnetic methods. A balance needs to be achieved between the depth of signal penetration and the size of facility detected. The expected result of the active EM technique was a work- ing innovation prototype for detection, lateral location, and depth estimation of metallic buried facilities. A method of tracking the linear motion would be incorporated into the data and their presentation. The proposed means of linear odometry was the combination of INS and GPS technology. Signals sent into the ground, data from the sensors, and the processed signals could be stored for later review. The prototype of the scanning EM locator, shown in Figure 3.7, experienced several difficulties. There were signal strength issues related to crosstalk between the channels. The three-phase emitter coils and the pair of three-phase pickup coils have a degree of mutual coupling. The mutual coupling of the pickups degraded the angle-to-target resolution some- what. This issue was exacerbated by the fact that the pickups were tuned to a specific resonant frequency: the same fre- quency at which the emitter was driven. The resonance pro- vides some additional gain, but the coils go on âringingâ after the field has swept through their position. Simply increasing the output power of the emitter does not improve the range or signal strength. A solution was tested that used nonresonant coils with an additional preamplifier to compensate for the lost reso- nant gain. This improved the situation somewhat, but the best solution was judged to be to migrate to a two-phase design with completely orthogonal coils. This solution would have required substantial reworking of the proto- type hardware and software and could not have been com- pleted within the resources of the current project. GTI believes that the basic technique is viable based on the per- formance of the commercial MJL available from Sensit Technologies. The output of this task was not available during the cur- rent project. The EM innovation prototype that uses a new technique to accurately locate buried metallic utilities and objects requires additional development before field testing Figure 3.6. GTI metallic joint locator. Figure 3.7. Prototype scanning EM locator.
14 can be performed. The level of effort required to complete this prototype puts it outside the scope of the R01C project. Appendix F contains additional technical support infor- mation on the scanning EM locator technology. Long-Range Radio Frequency Identification Tags VAI proposed to create a long-range pipe-location network based on a public wireless standard, IEEE 1902.1. The tag net- work has the advantage of cost-competitive marker tags that can be placed in both new and existing installations, with detection ranges of up to 40 ft underground, allowing suffi- cient margin that a marker at a depth of 20 ft has a reasonable âaperture of discoveryâ aboveground. The pipe tag trans- forms a buried plastic pipe into a smart pipe that can report location and status. The pipe tag could be built into the pipe, with optional sensors, and small portable readers could harvest key infor- mation directly from the tag, as well as feed into a real-time geo location database. There is some scaling of the device range with the size of the antenna. Tags for near-surface pipes could be quite small, as shown in Figure 3.8. A tag for a deeply buried sanitary or drain line may need to be 6 in. in length. Given the size of these deeper mains, building the device into a standard section still presents a negligible footprint. The final produc- tion versions of the tag would need some means of ensuring their orientation with respect to the facility; the prototypes would not have this. A tag permanently built into a pipe could be aligned with its axis to provide an operator with informa- tion on the direction of the pipe run. A tag installed in an âafter the factâ excavation would need some sort of leveling mecha- nism to provide the best signal strength. IEEE 1902.1, or RuBee, is an international wireless visibil- ity standard. RuBee is not simply radio frequency identifi- cation (RFID) in that it is a two-way, peer-to-peer transceiver protocol, and it can optionally use smart tags with small pro- cessors and sensors. RuBee has been optimized for visibility applications that must work in harsh environments, under- water or underground, and near or on steel and that may also require high security, high human safety, high intrinsic safety, and low electromagnetic interference. RuBee is not like RFID. It is a packet-based protocol that operates at 131 kHz; most of the energy at these frequencies is magnetic. RuBee is the only wireless technology that provides long-range read/write in harsh environments. VAI has developed a smart tag that has a detection range of more than 40 ft. The tag costs less than current passive marker tags and has an expected battery life of more than 20 years, with the possibility of extending that to 50 years. VAI was slated to create two focused prototype tag prod- ucts for this project: ⢠Hardened RuBee marker tags (RMT). These tags are water- proof and explosion proof, meet ANSI/UL 913-88, are designed for an underground life of 50 years, and have a range sufficient for utilities at 20-ft depths. Tags programmed with three-axis location data and, optionally, other relevant information about the buried assets can be placed into a con- struction site trench and near or on top of a pipe or other asset. These tags can be used for new construction, as well as with existing pipes that have been excavated. ⢠Hardened RuBee pipe tag (RPT). RuBee tags were pro- posed to be attached directly onto a pipe or fabricated as part of the pipe, either inside or outside. Tags would have the same basic specification as the RMT tags; however, the form factor would be conformal to the shape of the pipe itself, with a long-term goal of manufacturing the tags as part of the pipe. VAI produced two smart tag interrogation devices for this project: ⢠Commissioning handheld (CH). A short-range (2 ft) read/ write handheld device with barcode reader, RuBee reader and writer, Wi-Fi link, and software capable of adding useful information to the tag, as well as setting data values, such as current GPS coordinates, expected depth, date and time of installation, pipe type, size, content, and other field-critical data. Including field-critical data offers savings on future field localization, since the user would not require access to a database. In times of emergency or even routine localization, local data storage offers many advantages. ⢠Tag localizer (TL). A long-range (5 to 70 ft) read/write and presence-detection portable reader that can locate tags and provide field-critical data from both tag reads and a remote database (Dot Tag server) capable of providing the same details. The TL may also provide real-time localization of the tag based on tag signal and two antennas. The TL may Figure 3.8. Various RuBee tag form factors.
15 be attached to a vehicle or may be hand carried. It will include GPS and GPRS data link. The production of smart tag technology prototypes was delayed by a number of factors. VAI proposed that the perfor- mance of the tags could be improved by reducing the design to a single-chip implementation. The layout and first produc- tion of the new chip was expected to be completed during the second quarter of 2011, but there were multiple delays in the production of the new silicon. The fallback position was for VAI to provide engineering samples for demonstration based on their earlier chipset. This was executed, and VAI produced 20 sample tags with a range in excess of 40 ft, along with two handheld reading devices. Preliminary testing took place in a borehole 20 ft deep. The soil had no impact on the operation of the system. Several samples and readers were delivered to GTI; examples are shown in Figure 3.9. The balance of the hardware was available for Task 5 testing. The outputs of this task were working models of a buried smart tag that have ranges appropriate for utility depths of 20 ft, handheld devices to interact with the tags, and the appropriate system software. This system will allow operators to accurately relocate facilities at a lower cost than current GPS systems. Additionally, the tags provide positive identifi- cation of the buried facility via a unique serial number. Appendix G contains additional technical support infor- mation on the long-range smart RFID technology. Internal Inertial Mapping System On the basis of the required improvements identified in Task 2, an innovation prototype inertial mapping tool for internal deployment in piping was proposed for development. Several areas of research and improvement were identified. These included the following: ⢠âLiveâ insertions. Requiring the utility to be shut down and taken out of service before running inertial mapping tools is a significant limitation. Many lines cannot be shut down, because of the critical nature of their services. Developing a technique that allows the inertial mapping tools to be installed through a âhot tapâ would increase the applica- tion and practicality of using inertial mapping for locating deeply buried utility lines. ⢠Smart tag internal benchmarking. The ranges of inertial mapping tools are limited by cumulative error; the accuracy slowly degrades with distance from the insertion point. Smart tags are an excellent technology for locating deeply buried utility lines, but installing the tags requires access to the utility. If the inertial mapping tools could also be used to install smart tags on the interior of a pipe wall, relocating these facilities in the future would be greatly simplified. Smart tags could be used as a benchmark to enhance the accuracy of inertial mapping. ⢠Small hole insertion. Excavation and restoration costs are a significant factor in determining the practicality of using inertial mapping tools. Developing installation techniques that reduce the size of the excavation, and therefore reduce the total cost of use, would also increase the feasibility of this technology. The original plan was to perform a demonstration of the existing smart probe inertial mapping system in out-of- service facilities. This was to be a demonstration of the iner- tial mapping capabilities only and did not include any of the live insertion aspects. After consideration, however, SHRP 2 and the project team concluded that there was no longer a research component to this particular subtask. Inertial map- ping technology was dropped from the testing program. Appendix H contains technical support information on the use of inertial mapping technology for the internal map- ping of ducts. end-User Outreach and Presentation Over the course of the R01C project, a substantial number of presentations and outreach documents were prepared. The audiences for these materials were varied. Some of the materials were targeted at groups directly involved in the proj- ect, such as the utility advisers and the TETG. Others were presented in public forums, such as the TRB annual meeting, the AASHTO annual ROW and utilities subcommittee meet- ing, and several well-attended utility webinars. The purpose of the public outreach was to keep the community of potential Figure 3.9. VAI long-range tags and handheld reader.
16 end users informed about the progress of R01C, as well as to collect feedback on the direction of the project. User Panel Selection and Interaction A user panel was formed during Phase 1 to serve as an audience and review body for both R01C and R01B that would be repre- sentative of those who would use the tools or products resulting from this work. A set of KPIs was formulated for each of the technologies and reviewed by the user panel (Appendix C). A general information webinar was presented to the user panel in September 2010. A final products summary and presentation was prepared in April 2011 and reviewed by the user panel. TRB Annual Meeting Workshop A workshop for the SHRP 2 R01 projects was presented at the TRB annual meeting in January 2011. It included the follow- ing elements: ⢠An overview giving the motivation for the SHRP 2 projects; ⢠A presentation on R01A (geospatial data repositories); ⢠A presentation on R01B (multisensor platforms for sub- surface 3-D imaging); ⢠A presentation on R01C; and ⢠A presentation on the MTU project. TETG Webinars The first TETG webinar took place in July 2010. This webi- nar acquainted the TETG with the concepts for technology prototypes that were being tested and the proposed sched- ule. Additional TETG webinars were held during the course of the project to keep the TETG apprised of progress and changes that occurred. The last of these was held in Novem- ber 2012. Utility Webinar Presentations Two webinars were given in an open forum for any utility or state agency that wished to attend. These webinars were pro- moted through the TRB website. The purpose of the webinars was to promote the results of the R01A, R01B, and R01C projects and also to get feedback from the potential end users of any tools resulting from this work. Attendees of these webi- nars were able to post questions; the entire list of questions with answers by the appropriate researcher was made avail- able after the webinars. The first utility webinar took place in August 2011 and attracted just over 300 registrants. There were also 31 ques- tions, comments, or suggestions posted during the webinar. A poll of the attendees indicated that over 90% of the attend- ees were satisfied with the content and presentation. The second utility webinar took place in February 2012, with an estimated 186 attendees. The level of satisfaction and ques- tions were comparable to the first webinar. AASHTO Presentations Presentations were given at the AASHTO annual meeting in May 2011 and again in April 2012.
