Sensors capable of quantifying methane concentration and discriminating between possible sources are needed for natural gas leak detection where multiple spatially overlapping sources including wetlands and agriculture may be present. We report on the fabrication by an additive manufacturing process of a four electrode La0.87Sr0.13CrO3, Indium Tin Oxide (In2O3 90 wt%, SnO2 10 wt%), Au, Pt mixed potential electrochemical sensor using yttria-stabilized zirconia (YSZ) as a solid electrolyte to natural gas detection. Artificial neural networks (ANNs) are used to automatically decode the possible source and concentration of methane. The ANNs trained on sensor data are capable of correctly discriminating between three sources of methane emissions from simulated mixtures of emissions from cattle, wetlands, or natural gas with >98% accuracy. Quantification error for methane in mixtures of CH4 in air, CH4 + NH3 in air, and simulated natural gas is less than 1.5% ppm when a two-temperature dataset is employed.
Mixed-potential electrochemical sensor arrays consisting of indium tin oxide (ITO), La0.87Sr0.13CrO3, Au, and Pt electrodes can detect the leaks from natural gas infrastructure. Algorithms are needed to correctly identify natural gas sources from background natural and anthropogenic sources such as wetlands or agriculture. We report for the first time a comparison of several machine learning methods for mixture identification in the context of natural gas emissions monitoring by mixed potential sensor arrays. Random forest, artificial neural network, and nearest neighbor methods successfully classified air mixtures containing only CH4, two types of natural gas simulants, and CH4 + NH3 with > 98% identification accuracy. The model complexity of these methods were optimized and the degree of robustness against overfitting was determined. Finally, these methods are benchmarked on both desktop PC and single-board computer hardware to simulate their application in a portable internet-of-things sensor package context. The combined results show that the random forest method is the preferred method for mixture identification with its high accuracy (>98%), robustness against overfitting with increasing model complexity, and < 0.1 ms training time and < 10 ms inference time on single-board computer hardware.
The US Environmental Protection Agency estimates emissions from US natural gas drilling and processing amounted to 197 million metric tons of CO2 equivalent in 2019.1 Research by Marchese et al. suggests that CH4 emissions could be up to three times higher.2 Current natural gas emissions detection methods consist mainly of active optical monitoring, typically with long path IR sensors.3 These monitoring methods capable of ppb level accuracy are expensive, fragile, and require accurately calibrated mirrors. Mixed potential electrochemical sensors (MPES) are robust, low-cost, and selective sensors, making them appropriate option for methane emission monitoring.4 Previously we have reported on a four electrode MPES with 3 mol% YSZ electrolyte fabricated with direct write 3D printing5, here we investigate substrate and electrode material effects on the sensor response. The response of a four electrode La0.87Sr0.13CrO3, Indium Tin Oxide (In2O3 90 wt%, SnO2 10 wt%), Au, Pt mixed potential electrochemical sensor to methane and simulated natural gas at operating temperatures of 450 – 600 °C was studied. Sensors were fabricated on substrates of yttria stabilized zirconia, ceria stabilized zirconia, and magnesia stabilized zirconia manufactured by direct write extrusion 3D printing of aqueous pastes. These materials are selected to maximize the mixed potential difference by better isolation of the sensing electrodes. Data is presented with a solid electrolyte of 3 mol % yttria stabilized zirconia, and various other solid ceramic electrolytes will be studied. The effect of sensor response with respect to the ionic conductivity of substrate material is studied. The sensor limit of detection for CH4, natural gas, CH4/NH3, and CH4/H2 mixes will be reported. We will also train artificial neural networks to quantify the concentration of CH4 and other subcomponents given a subset of signals taken from the sensors operated at different temperatures. Acknowledgements This project was supported by US Department of Energy Office of Fossil Energy and Carbon Management through award DE-FE0031864. References (1) US EPA, O. Estimates of Methane Emissions by Segment in the United States https://www.epa.gov/natural-gas-star-program/estimates-methane-emissions-segment-united-states (accessed 2021 -12 -13). (2) Marchese, A. J.; Vaughn, T. L.; Zimmerle, D. J.; Martinez, D. M.; Williams, L. L.; Robinson, A. L.; Mitchell, A. L.; Subramanian, R.; Tkacik, D. S.; Roscioli, J. R.; Herndon, S. C. Methane Emissions from United States Natural Gas Gathering and Processing. Environ. Sci. Technol. 2015, 49 (17), 10718–10727. https://doi.org/10.1021/acs.est.5b02275. (3) Aldhafeeri, T.; Tran, M.-K.; Vrolyk, R.; Pope, M.; Fowler, M. A Review of Methane Gas Detection Sensors: Recent Developments and Future Perspectives. Inventions 2020, 5 (3), 28. https://doi.org/10.3390/inventions5030028. (4) Garzon, F. H.; Mukundan, R.; Brosha, E. L. Solid-State Mixed Potential Gas Sensors: Theory, Experiments and Challenges. Solid State Ion. 2000, 136, 633–638. (5) Halley, S.; Tsui, L.; Garzon, F. Combined Mixed Potential Electrochemical Sensors and Artificial Neural Networks for the Quantificationand Identification of Methane in Natural Gas Emissions Monitoring. J. Electrochem. Soc. 2021, 168 (9), 097506. https://doi.org/10.1149/1945-7111/ac2465. Figure 1. Four electrode MPES response at 600 °C to CH4 in simulated air, low ethane simulated natural gas in air, and high ethane simulated natural gas in air, from left to right, respectively. Top row – ceria stabilized zirconia substrate. Bottom row – magnesia stabilized zirconia substrate. Figure 1
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.