• Agi, Kamil
• Garzon, Fernando H.
• Smith, James
• Ian, Robert
• Halley, Sleight

Abstract

<jats:p>According to the EPA, methane (CH<jats:sub>4</jats:sub>) emissions from oil and gas infrastructure accounted for 211 million metric tons of CO<jats:sub>2</jats:sub> equivalent in 2020 [1]. Actual emissions may exceed this by a factor of three [2]. Current natural gas leak detection technologies largely consist of optical sensors such as IR spectrometers [3]. Optical sensors have high sensitivity, but the high cost and fragility of these sensors limit practical applications and continuous monitoring in the field. Mixed potential electrochemical sensors (MPES) are low cost, robust, selective and sensitive, making them a viable option for continuous natural gas leak detection [4]. While we have previously reported on the development of these sensors for natural gas detection in the laboratory, it is necessary to evaluate how these sensors perform in relevant environments.</jats:p><jats:p>The MPES device consists of La<jats:sub>0.87</jats:sub>Sr<jats:sub>0.13</jats:sub>CrO<jats:sub>3</jats:sub> (LSC), indium tin oxide (ITO, In<jats:sub>2</jats:sub>O<jats:sub>3</jats:sub> 90 wt%, SnO<jats:sub>2</jats:sub> 10 wt%), and Au sensing electrodes with a Pt pseudo-reference electrode, bridged by 3 mol% YSZ solid electrolyte. A low ionic conductivity magnesia stabilized zirconia (MSZ) substrate is used to enhance sensitivity with a demonstrated limit of detection (LOD) of &lt; 40 ppm. The sensor is integrated with an internet of things (IoT) data collection and transmission package developed by SensorComm Technologies.</jats:p><jats:p>Field testing was performed at Colorado State University’s Methane Emission Technology Evaluation Center (METEC). The sensors’ capability of detecting buried pipeline leaks was investigated by varying the leak rate from 7.2 SLPM to 37 SLPM, lateral sensor distance from 0 meters to 3 meters, and vertical distance from 0 meters to 0.28 meters (Figure 1). Machine learning methods were applied to a training dataset collected in the laboratory to quantify the CH4 concentration. These results serve as a first demonstration that a low-cost mixed potential electrochemical sensor system can successfully detect underground pipeline emissions and quantify CH4 concentrations that are in agreement with previously published results [6] collected using more complex and costly methods.</jats:p><jats:p>References:</jats:p><jats:p>[1] O. US EPA, “Estimates of Methane Emissions by Segment in the United States,” Aug. 27, 2018. https://www.epa.gov/natural-gas-star-program/estimates-methane-emissions-segment-united-states (accessed Dec. 08, 2022).</jats:p><jats:p>[2] A. J. Marchese <jats:italic>et al.</jats:italic>, “Methane Emissions from United States Natural Gas Gathering and Processing,” <jats:italic>Environ. Sci. Technol.</jats:italic>, vol. 49, no. 17, pp. 10718–10727, Sep. 2015, doi: 10.1021/acs.est.5b02275.</jats:p><jats:p>[3] T. Aldhafeeri, M.-K. Tran, R. Vrolyk, M. Pope, and M. Fowler, “A Review of Methane Gas Detection Sensors: Recent Developments and Future Perspectives,” <jats:italic>Inventions</jats:italic>, vol. 5, no. 3, Art. no. 3, Sep. 2020, doi: 10.3390/inventions5030028.</jats:p><jats:p>[4] F. H. Garzon, R. Mukundan, and E. L. Brosha, “Solid-state mixed potential gas sensors: theory, experiments and challenges,” <jats:italic>Solid State Ion.</jats:italic>, vol. 136–137, pp. 633–638, Nov. 2000, doi: 10.1016/S0167-2738(00)00348-9.</jats:p><jats:p>[5] S. Halley, L. Tsui, and F. Garzon, “Combined Mixed Potential Electrochemical Sensors and Artificial Neural Networks for the Quantificationand Identification of Methane in Natural Gas Emissions Monitoring,” <jats:italic>J. Electrochem. Soc.</jats:italic>, vol. 168, no. 9, p. 097506, Sep. 2021, doi: 10.1149/1945-7111/ac2465.</jats:p><jats:p>[6] B. A. Ulrich, M. Mitton, E. Lachenmeyer, A. Hecobian, D. Zimmerle, and K. M. Smits, “Natural Gas Emissions from Underground Pipelines and Implications for Leak Detection,” <jats:italic>Environ. Sci. Technol. Lett.</jats:italic>, vol. 6, no. 7, pp. 401–406, Jul. 2019, doi: 10.1021/acs.estlett.9b00291.</jats:p><jats:p><jats:bold>Figure 1:</jats:bold> Sensor response to various heights above a simulated buried pipeline leak on two successive days of testing (a and b), and estimated CH<jats:sub>4</jats:sub> concentrations from sensor data (c).</jats:p><jats:p><jats:inline-formula><jats:inline-graphic xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="2604fig1.jpg" xlink:type="simple" /></jats:inline-formula></jats:p><jats:p>Figure 1</jats:p><jats:p />

Topics

• impedance spectroscopy
• theory
• experiment
• tin
• machine learning
• Indium
• liquid-solid chromatography