Final Report: AGCS Sensor for Gas Leak Detection

EPA Contract Number: 68D02058
Title: AGCS Sensor for Gas Leak Detection
Investigators: Nelson, Loren D.
Small Business: OPHIR Corporation
EPA Contact: Manager, SBIR Program
Phase: II
Project Period: June 1, 2002 through June 1, 2004
Project Amount: $224,999
RFA: Small Business Innovation Research (SBIR) - Phase II (2002) Recipients Lists
Research Category: Ecological Indicators/Assessment/Restoration , SBIR - Monitoring , Small Business Innovation Research (SBIR)


This Phase II research project covers all efforts performed under the original Basic Effort as well as the Option Effort. The Phase II Basic Effort included six tasks: (1) transceiver development, (2) transceiver evaluation for moderate and long ranges, (3) completion of the U.S. Department of Energy (DOE) CO2 sequestration field test, (4) completion of a ground-based transmission pipeline test for the Williston Basin Interstate Pipeline Company, (5) a system performance review and upgrade, and (6) completion of the Phase II Final Report. The Option Effort included four tasks: (1) explore sensor optical improvements, (2) perform long-path, ground-based sensor field demonstrations, (3) perform field test data analyses, and (4) prepare a field test report summarizing the Option Effort field tests and data analyses.

Summary/Accomplishments (Outputs/Outcomes):

At this time, all technical work under both the Basic and Option Efforts has been completed. The Phase I sensor was modified to support a colocated transmitter and receiver (a “transceiver”) and the detection of natural gas was completed. An extensive effort was undertaken to identify a thermal source to support long-range and airborne detection of methane and ethane. The most powerful source, an arc lamp, was purchased and installed in the transceiver. This source, however, did not provide sufficient power for moderate- or long-range trace gas monitoring in the transceiver geometry while using natural surfaces as the remotely located reflector. To achieve long-range (greater than 1 mile) and airborne-pipeline monitoring using natural surfaces, the Ophir Corporation will be required to use sources at shorter wavelengths than were developed for this sensor. At the conclusion of the transceiver tests, Ophir completed a field test in Hobbs, NM, over the Pearl Queen Oil Field for the DOE, National Energy Technology Laboratory (NETL). Two tests were completed in support of DOE's CO2 Sequestration Program, with a test being performed both before and after the injection of CO2 into the oil and gas reservoir. These tests demonstrated that Ophir's Active Gas Correlation Spectrometer (AGCS) technology can effectively monitor large surface areas for the emission of natural gas (methane and ethane). Fence-line distances of 305 m (1,000 ft) were used. Next, the system successfully was demonstrated in the field over an operational pipeline. The pipeline was owned and operated by the Williston Basin Interstate (WBI) Pipeline Company. The test was completed in Glendive, MT, and fence-line optical pathlengths up to 527 m (1,729 feet) were used. This demonstration test proved that Ophir’s technology could effectively detect Class 1, 2, and 3 leaks from a compromised, operational transmission pipeline. System upgrades and improvements were completed throughout the period of performance. These included: improved detectors, superior analog-to-digital converter hardware, a filter wheel system to enable methane and ethane to be monitored at an 8 Hz rate, and software improvements.

Under the Option Effort, Ophir improved the sensor's thermal stability in high-temperature test environments. In addition, an improved optical detector was incorporated. Next, two very successful field tests were completed. The first Option Effort test was performed over an operational natural gas pipeline owned and operated by El Paso Energy. This test included the very first detection of actual leaks from operational natural gas pipelines. Ophir's gas concentration data showed good agreement with measurements taken with Flame Ionization Detectors (FIDs), given the differences in their sampling location and method. Finally, a very thorough test was performed to demonstrate the system's ability to detect methane and ethane emitted from a natural crude oil seep near Santa Barbara, CA. This test was cofunded by the Alyeska Pipeline Company and included the very first demonstration of detecting "oil on water" using an atmospheric remote sensing technology.


A summary of Phase II Basic and Option Effort findings include:


• The Ophir/U.S. Environmental Protection Agency (EPA) Phase II prototype sensor was deployed successfully in: (1) two tests (DOE/NETL) over an oil and gas reservoir in Hobbs, NM; (2) three test sites (WBI Interstate Pipeline Company) over operational pipeline in Glendive, MT; (3) three actual leak sites (El Paso Energy) on an operational pipeline near St. David, AZ; and (4) two test setups (Alyeska Pipeline Company) in the Santa Barbara Channel natural oil and gas seeps off Coal Oil Point, CA.

• The measurement distances (or optical pathlength) varied from as short as 27 m (88 feet) aboard the Clean Seas' oil cleanup vessel to 527 m (1,730 feet) during these tests.

• The system was demonstrated over an actual leaking pipeline, controlled leak points over operational pipeline, over seawater in the region of natural gas and oil seeps, and along a pier near the natural seep region.

• In all cases, ethane was demonstrated to be a much superior trace gas to indicate pipeline leaks, natural hydrocarbon seepage, and to perform integrity monitoring, as compared to methane.

• The SBIR Phase II Basic and Option Efforts included: (1) the first demonstration of simultaneously detecting and quantifying both methane and ethane from controlled leaks from operational pipelines; (2) the first optical remote sensing detection of both methane and ethane emitted from actual pipeline leaks from operational natural gas pipelines; (3) the first demonstration of using atmospheric methane and ethane as a tracer gas to determine the effectiveness of CO2 sequestration sites; and (4) the first demonstration of measuring atmospheric methane and ethane from oil and gas emitted from natural seeps into an ocean environment.

Ground-Based Pipeline Measurements

• For the purposes of pipeline integrity monitoring, the smallest leak rates will be measured when the wind speed was calm. Thus, one effective method for detecting very minute leaks using optical remote sensing hardware was to install ground-based sensors and leave them operating overnight to inspect the pipeline during calm nighttime conditions. Automatic data collection and analyses is important for an operational system and for documentation of the presence or absence of very minute leaks within that pipeline section under test.

• The AGCS sensor agreed well with the FID readings. One test site did not agree well with the AGCS prototype sensor, indicating a very small leak and the FID sensor (a 20 percent lower explosive level measurement value). This discrepancy may be because of the relative location of the two measurement techniques. The AGCS system monitored the air approximately 1 m (3 feet) in the air above the leak, while the FID sensor measured right at the ground surface. Alternately, wind speed may have played a more significant role in diluting the leak during the AGCS test as compared to the FID test.

• Although the FID and AGCS measurements agreed relatively well, further efforts are required to finalize the calibration of the AGCS system. Efforts to explore measurement drift sources and calibration stability should be undertaken.

Shipborne Seep Measurements

• The AGCS system successfully was deployed aboard the Clean Sea's vessel over the natural oil and gas seeps in the Santa Barbara Channel near Goleta, CA. Although the ship-mounted test pathlength was very short (27 m) for the optical system designed for long-path measurements (approximately 300 m), the AGCS system still achieved adequate system sensitivity to demonstrate methane and ethane atmospheric trace gas concentration measurements in this dynamic environment.

• Elevated methane and ethane concentrations were recorded throughout the test region. A high degree of temporal correlation between the methane and ethane concentrations can be observed in the data record. On most occasions, methane and ethane concentrations were seen to increase and decrease together. This suggests that the methane and ethane were well mixed within the atmosphere and likely are from a common emissions source.

• A great deal of variability exists within the recorded seep concentrations. For example, the first entry of the Seep Tents bubble emission region resulted in a measured methane concentration of approximately 45 ppm and an ethane concentration of about 5.5 ppm. Reentering this region at a later time resulted in peak concentration measurements of about 100 ppm for methane and 10 ppm for ethane. Finally, when this region was reexplored 2 hours later, approximately 130 ppm methane and 12 ppm ethane were measured. This variability likely is because of geographic differences between the actual test sites, wind direction, wind speed, wave action, time of day, solar heating, oil and gas seep rates, etc. Methane and ethane gas concentrations generally were higher in the afternoon than in the morning.

• Although variability exists within the data record, the highest gas concentrations always were recorded in regions containing both oil on water slicks and surface bubble emissions. Significant increases in both methane and ethane always were recorded when entering regions containing both oil on water and bubbles.

• Regions where there were no surface bubble emissions—just oil on water—still contained significant methane and ethane atmospheric trace gas concentrations. Oil on water (with no bubbles), methane, and ethane concentrations varied from about 4 ppm methane and 0.7 ppm ethane (at 10.35 hours), to 18 ppm methane and 3 ppm ethane (at about 11.2 hours), to 40 ppm methane and 3-5 ppm ethane (at 12.7 hours), dropping to 20 ppm methane and 1 ppm ethane over an old and weathered oil layer (at 12.85 hours), to 30 ppm methane and 4 ppm ethane (at about 12.3 hours). These various measurements were taken at different locations within the Seep Tents to Coal Oil Point Seeps regions.

• Wind direction was observed to be a consistent and important factor in the measured methane and ethane concentrations.

Shipborne Measurements of Potential Diesel Exhaust Contamination

• The diesel exhaust tests did not show conclusively whether diesel exhaust would result in a false alarm leak declaration. All diesel exhaust tests were performed in a region already indicating the presence of methane and ethane. Although 360 turn and up/down wind tests were performed, no clear trend was observed to indicate whether diesel exhaust is a potential contaminant. In one event, during which diesel exhaust was swept into the sensor measurement volume, the methane concentration was seen to increase (from approximately 9 to 10 ppm), while ethane could not be measured because of solar saturation effects. As the exhaust plume was removed from the measurement volume, however, methane and ethane concentrations were observed to rise. The installation geometry solar saturation problem, combined with wind direction and the variable methane and ethane concentrations as a result of the close proximity of the seeps, precluded a concrete assessment of diesel exhaust contamination.

• The diesel exhaust contamination test should be repeated in a region far removed from significant sources of methane and ethane, and in an installation geometry where solar reflection contamination does not occur.

• Post-test research on the composition and chemicals associated with diesel exhaust indicated that light hydrocarbons (such as methane and ethane) are not considered major or minor exhaust components. Major gaseous components included CO2, CO, NO2 and SO2. Minor hydrocarbon components primarily included large aromatic hydrocarbons and smaller substituted hydrocarbons, not methane or ethane.

Long-Path Pier Measurements of Atmospheric Methane and Ethane

• Long-path (about 460 m or 1,510 feet) measurements were completed successfully along the Venoco pier. Atmospheric background levels of methane (about 1.5 to 2.5 ppm) were detected. The methane concentration was seen to decrease from the morning to the early afternoon, and then increased in the late afternoon.

• Minute traces of ethane (50 to 60 ppm[m]) were detected from about 12:20 p.m. (12.3 hours) to 1:00 p.m. (13.0 hours) and from approximately 2:10 p.m. (14.2 hours) to about 2:40 p.m. (14.65 hours). If this gas were distributed evenly over the entire 460 m optical pathlength, the average ethane concentration would be approximately 120 ppb.

• The weather data indicated that the wind was from the south during all ethane detection periods. Also, the wind speed was sufficiently high and of a long enough temporal duration to transport atmospheric gases from a source as far away as about 10 miles. Thus, the meteorological data suggest that it is possible that the ethane source was from an oil seep within the Santa Barbara Channel.

• The Venoco Pier data demonstrated that ethane is a significantly better indicator of small hydrocarbon concentrations than methane, because of the relatively large natural background concentration of methane in the free atmosphere.

Pier-Based Measurements of Gasoline Vapors

• Gasoline vapors readily are detected with the AGCS line (with a surface area of 1 ft2) placed approximately 12 inches below the optical sampling beam. The ambient atmospheric temperature (likely similar to the gasoline temperature) was approximately 17.8°C (64°F) at this time.

• Ethane is a significantly better indicator of gasoline than is methane. No apparent increase in methane was seen during the gasoline exposure test.

Future Sensor and Field Test Improvements

• The sensor, in its installation geometry aboard the Clean Sea's vessel, experienced a solar saturation effect when the boat turned to a specific range of headings. This effect was not encountered during previous field tests, even when the receiver was oriented toward the sun. Ophir believes that this effect is a result of the canvas reflecting (and solar absorbing and reradiating) surfaces located near the sensor's line of sight on the ship. These surfaces could redirect considerably more solar energy into the receiver than would a typical ground-based test geometry.

• Future test geometries using the EPA prototype AGCS sensor should avoid solid reflective surfaces near the receiver line of sight. Also, future AGCS sensor configurations, developed for routine field deployment, should include improved solar blocking filters.

Journal Articles on this Report : 1 Displayed | Download in RIS Format

Other project views: All 7 publications 1 publications in selected types All 1 journal articles
Type Citation Project Document Sources
Journal Article Spaeth LG, O'Brien MJ. An additional tool for integrity monitoring. Pipeline and Gas Journal. 68D02058 (Final)
not available

Supplemental Keywords:

Active Gas Correlation Spectrometer, AGCS, transceiver, natural gas, trace gas, sensor, gas leak, pipeline leak detection and monitoring, optical remote sensing, optical pathlength, flame ionization detector, FID, oil, diesel exhaust, hydrocarbons, gasoline, methane and ethane detection, small business, SBIR., Scientific Discipline