Pipeline & Gas Journal
April issue 2000:

For 15 years I worked at Gas Research Institute (GRI) in safety research, attempting to develop improvements in distribution operations such as leak surveying, leak pinpointing, pipe location (as it relates to promptness of leak location and repair) and odorant monitoring. During that period, I witnessed the evolution of operations, like leak surveying, from vegetation surveying to the recently commercialized Optical Methane Detector (OMD). In the future, other new and advanced technologies, such as laser remote sensing, will produce even greater benefits for utilities. In what follows, I will describe how these advanced technologies were assessed at GRI and what will be their future potential for application to the safety operations of local distribution companies (LDCs).
Projecting Benefits From Candidate Technologies

A candidate concept could gain approval for funding at GRI as a safety R&D project with convincing evidence that it benefits to LDCs included significant reductions in:

• the dangers of gas leakage without increasing existing survey or repair costs;
• or, costs of a given operation without sacrificing the achievable level of safety;
• or, ideally, both of the above.

In addition, this candidate technology required the approval of GRIs Distribution Project Advisor Group (DPAG), a group of industry advisers.  Finally, assuming initial project tests would go well, there had to be reasonable expectation of a commercializer in the wings in order to deliver the benefits of this technology to the industry.(1) This was doubly challenging since LDCs, with its 1,000 plus private and publicly owned utilities, constitute a small market in terms of number of sales.

To maximize the likelihood of success, and weed out would be losers, we developed for each state-of-the-art (SOA) safety operation a kind of template, or model, of performance parameters against which candidate technologies could be compared. These parameters included the economic, operational and technical factors that characterize each SOA operation. Thus, the benefits of a candidate concept could be estimated by a comparison of its projected performance parameters with those of the SOA.

Leak Surveying SOA Technologies
The flame ionization unit (FID), with its modest cost, good sensitivity and ease of use, has been the gas detector of choice throughout the industry. A table of its performance parameters is shown in Table 1. Despite its almost universal usage, the FID possesses several drawbacks. The FID is a point-sampling detector which can only find a gas leak if its air sampling probe is brought to the immediate area of the gas leak plume. This places a premium on experienced surveyors who are alert to local ground cover conditions and the possible occurrence of unlikely leak site locations. SOA leak surveying, then, is a time-consuming, labor-intensive operation as suggested in Table 1. Industry-wide efficiencies may likely be further impacted by the widespread need for downsizing and retrenchment in recent years which has taken a toll of older and more experienced workers.
For vehicle leak surveys, vehicle speed is limited by the finite sampling time that must be spent at each ground level unit area to intake a detectable level of gas. Leaks carried by the wind in the wrong direction may go undetected.   Thus, average coverages for walking or mobile techniques are reported to be in the range of 5-10 miles per day.

Leak surveying with the FID is subject to false alarms, both positive and negative. The FID is a non-selective detector, responding to all combustible, or oxidizable substances in the air.  This means that false positives from sewer or swamp gases or miscellaneous surface chemicals (such as fertilizer materials) are possible. Thus, it is a common survey practice when potential leaks are detected by foot or vehicle to make a bar hole at the site of the FID response and, by means of a combustible gas indicator (CGI) to sample air in the barhole (another time consuming operation). A high percent reading from the CGI is generally accepted as a confirmation of escaping underground gas. In addition, false negatives may likely occur, even with a dedicated and experienced leak surveyor.

Significant improvements in SOA leak surveying could be achieved with a technology possessing the following desirable features:

1. allows the survey to be performed at greater speed (i.e. with greater productivity);
2. provides a selective response for natural gas (i.e., no false positive alarms);
3. improves the search ability of the surveyor (i.e., no false negative alarms).

Advanced Technologies
Optical Methane Detector: Heath Consultants, Inc. launched the commercialization of the Optical Methane Detector (OMD) for leak surveying by vehicle. I was fortunate enough to become involved in the technology development of the OMD when, in the early 1990s, I attended a presentation of scientists at the Westinghouse Science & Technology Center (WSTC) at the request of Dr. Irv Solomon, then chief scientist for GRI. In laboratory studies, WSTC had developed an electrooptic technique to detect methane based on its unique absorption of infrared light at certain wavelengths. This technique appeared to possess a number of attractive features including: 1) good sensitivity and selectivity for methane; 2) potential for increased speed of surveys by vehicle, 3) hardware that was relatively inexpensive and with no moving parts.
The power requirements of the WSTC approach could easily be satisfied by a survey vehicle; but, for a walking surveyor, the added weight from a battery precluded initial development of a portable OMD detector. The initial OMD configuration was envisioned atop the survey van with its infrared light source pointed at the street ahead of the vehicle. However, initial discussions revealed that, in order to achieve useful range for the light source, the electrooptics would become rather complex and uncertain in design.

Somewhat disillusioned, I suggested we consider a simpler, short-range configuration in which the system was mounted on the front bumper and the light source shined from one side of the detector to the other. As the vehicle moved forward and the plume of street level gas leakage passed into the light beam, attenuation of the infrared light by methane would be detected. Initial demonstrations by WSTC with a fieldable breadboard were successful. The instant response of the detector permitted significantly increased vehicle speed.  Productivity increases of at least 20-50% were predicted. Successful development of the OMD followed, with GRI funding initial stages at WSTC, followed by further efforts at Carnegie Mellon Research Institute,(2), and culminating with the co-funding and commercialization by Heath Consultants.

Ethane Detection: At my request, WSTC made a brief study to determine whether or not the OMD approach could be extended to the detection of ethane. The results were positive.(2) Since the presence of ethane is almost unique to natural gas, the response of the “OED” would be specific to natural gas and capable of eliminating all false positives. In fact, a combined “OMED” is possible without greatly changing the hardware.(2) However, since most natural gas typically contains only 1-3% ethane, a detection limit on the order of 25-50 ppb or less would be desirable.

Laser Remote Sensing Detector (LRS Detector): Like the OMD, a laser-based leak detector is based on the unique absorption of certain infrared wavelengths by methane. In principle, the surveyor would employ a handheld system, which he operated like a flashlight. Standing at a distance, he could scan large areas rapidly, as he searched for leaks. Customer service lines, with front or side meter sets, could be surveyed from the curb or sidewalk. Inaccessible areas such as complex aboveground pipes could easily be surveyed. The result would be a huge increase in productivity. It can be seen that laser remote sensing overcomes the chief defect of the FID. It may take time, but laser-based technology represents the next generation leak survey detector. Others in the gas industry recognized this long ago.(3)  

How it Works: Using an infrared laser, at a wavelength uniquely absorbed by methane, a surveyor would shine the LRS detector at distant areas to be surveyed. As the laser light shined upon the ground, it would be scattered in all directions. A portion of the backscattered light would be intercepted by a small antenna on the surveyor’s LRS unit. Should the laser beam intercept a plume of natural gas escaping from the ground, some of the laser light would be absorbed by methane causing a reduction in the intensity of the backscattered radiation received at the LRS unit. This change in signal level could be announced by an audio alarm whose frequency and intensity would vary with the level of attenuation of the backscattered radiation. One can envision at least three configurations:
1. A truly standalone handheld LRS unit employed by a surveyor on foot;

2. A handheld system with power and data processing capabilities supplied to the surveyor by a cable from the survey van;

3. A van-mounted LRS operating from a rooftop turret for survey of street and service lines.
This technology is not yet ready. Power requirements, for example, make options 2 and 3 the more likely first generation product. Other technical problems, such as background methane, can be resolved. Continued advances in laser technology are bringing the LRS closer to reality.
Gas Leak Imaging: Several years ago, when infrared imaging at a specific wavelength first appeared, I obtained funding at GRI for an imaging study where the goal was to examine, for the first time in a video format, the appearance and movement of real world gas leaks, both indoors and outdoors.(4) Further advances in infrared detector technology have advanced this technology to the point where application as leak survey tool has become a reality. This concept represents the ultimate leak surveying tool.(5)

How it Works:  Instead of a single sensor in the detector, a rectangular array of microsensors are employed.(6) Optics in the detector would fill the sensor array with an image of the backscattered infrared light. Outputs from each microsensor would be converted to pixels in a combined video image. Gas leak plumes would appear as blackened wisps emanating from the ground. As the surveyor scanned the area, an auxiliary audio alarm would alert him that a leak had been detected. He would then locate the specific site with the help of the real time video image. If a user-friendly system could be developed, guidance of the surveyor by video image would represent the ultimate in quality of information.

Recognizing the enormous potential benefits of these approaches, GRI continues to support the development of these technologies. The sensor arrays represent a relatively young technology whose materials and manufacture are still evolving in development. Presently, they are expensive. However, as applications of these devices continue to expand, the technology will mature and it may be expected that the price will drop significantly. In the end, higher capital costs of the LRS detector will be more than offset by improvements in productivity.

Automated Leak Detection: Portable or handheld PCs, in concert with global positioning satellite technology and wireless communication, can be exploited to send a record of each leak site to a dispatch center, establishing a paperless record-keeping and leak management system. This technology is here and applicable. Cost savings for clerical record-keeping alone can be significant. P&GJ

FOOTNOTES

1. Final funding approval for any new start project in GRI’s Distribution Operations Business Unit depended upon its ranking with other new start projects, together with the level of available funds from its annual budget.

2. Tom Henningsen, H. Lessure, T. Oblak, “Development Of The Optical Methane Detector,” Topical Report to GRI, Report Number 96/0477 (February 1994 - September 1996).

3. A.C. Eberle, “Hand Held Remote Sensing Methane Detector For Leakage Survey,” AGA Operating Section Proceedings, p. 608 (1984).

4. T.G. McRae, L. L. Altpeter, Jr., “Natural Gas Leak Dispersion Studies Using Infrared Gas Imaging,” Presentation and Proceedings of the 1992 International Gas Research Conference.  Government Institutes, Inc., Rockville, MD, 1993.

5. This is true for aboveground application.  Actually, the ultimate system will reside within the pipes as a set of autonomous monitors; but this is far in the future.

6. e.g., a 32 x 32 or 256 x 256 array of detectors.