An evaluation of airborne mass balance and tracer correlation approaches to estimate site-level CH₄ emissions from LNG facilities using CO₂ as a tracer of opportunity

Abstract. Accurate and representative quantification of methane (CH₄) emissions from individual oil and gas facilities is crucial to improve our knowledge of CH₄ sources, improve the reliability of emissions reporting and help facilitate mitigation opportunities. Liquefied Natural Gas (LNG) terminals primarily emit carbon dioxide (CO₂) but are also potentially large CH₄ sources in the gas supply chain. Here, in work supported by the United Nations Environment Program International Methane Emissions Observatory (UNEP’s IMEO), we evaluate two airborne measurement approaches to quantify CH₄ emissions from four LNG facilities. The first approach applies a downwind mass balance method to quantify emissions of both CH₄ and CO₂. Since operator-reported CO₂ emissions are relatively well-established, we evaluate the method’s performance by comparing to operator-reported CO₂ values. Using this approach, we show that an individual facility CO₂ mass balance quantification has a mean relative difference to operator reporting of ±20 %, with no significant mean bias across all estimates. The second approach uses measured CH₄:CO₂ mole fraction ratios as an alternative method for estimating site-level CH₄ emissions. Using this tracer correlation approach, we show that uncertainties in the ratio of CH₄:CO₂ from a single day can be below ±10 % at the 95 % confidence level. Due to uncertainty in the CO₂ emission rate, the resulting mean 2σ uncertainty on CH₄ emissions is ±30 %. CH₄:CO₂ ratios are found to vary with height, with larger variation at lower altitudes (<250 m). Representative sampling across both horizontal and vertical space is needed to enhance the accuracy of the tracer correlation approach for individual emission estimates. We find that when repeated over multiple days and different atmospheric conditions, ratio measurements below 250 m provide a median estimate that is within 5 % of the mass balance and multi-height ratio methods. Whilst the CO₂:CH₄ emission ratios derived from measurements could be applied to estimate CH₄ emissions over longer timeframes, the degree of representativeness will depend on the variability of the ratio over time. Our results indicate that the tracer correlation approach using CO₂ as a tracer of opportunity can be used to efficiently estimate CH₄ emissions from LNG facilities with a low level of uncertainty.