The calibration and deployment of a low-cost methane sensor

Since 1850 the atmospheric mixing ratio of methane (CH₄), a potent greenhouse gas, has doubled. This increase is directly linked to an escalation in emissions from anthropogenic sources. An inexpensive means to identify and monitor CH₄ emission sources and evaluate the efficacy of mitigation strategies is essential. However, sourcing reliable, low-cost, easy-to-calibrate sensors that are fit for purpose is challenging. A recent study showed that CH₄ mixing ratio data from a low-power, low-cost CH₄ sensor (Figaro TGS2600) agreed well with CH₄ mixing ratios measured by a high precision sensor at mixing ratios between 1.85 ppm and 2 ppm. To investigate, as a proof of concept, if this low-cost sensor could be used to measure typical ambient CH₄ mixing ratios, we operated a TGS2600 in conjunction with a Los Gatos Ultra-portable Greenhouse Gas Analyzer (UGGA) in controlled laboratory conditions. We then explored the sensor's long-term reliability by deploying the TGS2600 near an onshore gas terminal to calculate emissions from May to July 2018. Our initial studies showed that previously published linear algorithms could not convert TGS2600 output to CH₄ mixing ratios measured by the UGGA. However, we derived a non-linear empirical relationship that could be used to reliably convert the output of a TGS2600 unit to CH₄ mixing ratios over a range of 1.85–5.85 ppm that agree to a high-precision instrument output to ±0.01 ppm. Our study showed that the TGS2600 could be used to continuously measure variability in CH₄ mixing ratios from 1.82 to 5.40 ppm for three months downwind of the gas terminal. Using a simplified Gaussian Plume approach, these mixing ratios correspond to an emission flux range of 0–238 g CH₄ s⁻¹, with average emission of 9.6 g CH₄ s⁻¹ from the currently active North Terminal and 1.6 g CH₄ s⁻¹ from the decommissioned South Terminal. Our work here demonstrates the feasibility of utilizing a low-cost sensor to detect methane leakage at concentrations close to ambient background levels, as long as the device is routinely calibrated with an accurate reference instrument. Having a widely deployed network of such low-cost CH₄ sensors would allow improved identification, monitoring and mitigation of a variety of CH₄ emissions.