• Both inefficient combustion and unlit flares contribute to substantial methane emissions that greatly exceed standard estimates for flaring.
• Mitigation efforts that address either combustion efficiency or unlit flares (such as operational practices on flare maintenance), or reducing the usage of flares altogether (with alternatives such as re-injection or small-scale gas capture technology) would provide significant methane emission benefits.
• DRE values are skewed. Destruction removal efficiency (DRE) characterizes how well a flare’s combustion destroys CH₄ in the waste gas. The majority of flares function close to expected performance, with DRE values near 98%. However, across all basins, a relatively modest number of poorly performing flares (with DRE values as low as 60%) were observed to cause a significant drop in average performance.
• Unlit flares directly vent unburned gas to the atmosphere due to the flame being extinguished or never properly ignited, resulting in an additional impact on the flaring CH₄ budget. Given, based on observations, that ~3-5% of flares are unlit, the relative contribution of both poorly combusting and unlit flares to total CH₄ flaring emissions is similar.
• Observed DRE combined with rate of unlit flares results in effective flare efficiencies that are considerably lower than the expected 98% across all three basins. The average observed DRE across the three regions of study is 95.2% and the average total effective DRE after accounting for unlit flares is 91.1%. These emissions estimates are ~5x larger than if assuming 98% DRE for all flares quantified by VIIRS and no occurrences of unlit flares (0.10 Tg CH₄/year). This indicates that flaring activities are a much larger part of the CH₄ O&G footprint than previously estimated.

  Inefficient and Unlit Natural Gas Flares Both Emit Large Quantities of Methane (Science, 2022)
  View/Download Fact Sheet (PDF)

• A single NOx emission factor does not capture the range of observed emission factors in the Bakken, Permian, and Eagle Ford. All three basins show a heavy-tail distribution, with a minority of flares emitting the majority of NOx and a nontrivial number of flares producing NOx at least an order of magnitude larger than the EPA value. Bakken and Permian basin averages are 2X-3X the EPA value.
• Extrapolation to basin scale emissions indicates that 20%–30% of flares are responsible for 80% of total basin-wide flaring NOx emissions.
• Overall, NOx emissions from flaring vary widely depending on emission factors and gas composition, challenging the accuracy of bottom-up inventories.
• Basin-specific emission factors and regional gas composition assumptions better capture average NOx flaring performance at the basin scale. However, a basin-specific approach may still miss high emitters in the heavy tail of skewed distributions.
• It is crucial to count very high NOx-emitting flares in any air quality assessment in surrounding regions, as NOx has direct negative health impacts, such as asthma exacerbations.
• Flaring for the studied region does not contribute significantly to the national NOx emissions estimate. However, flares account for a sizeable portion of NOx in O&G production and offer a potential opportunity for reduction.

  Excess Methane Emissions from Shallow Water Platforms Elevate the Carbon Intensity of US Gulf of Mexico Oil and Gas Production (PNAS, 2023)
  View/Download Fact Sheet (PDF)

• The climate impact of offshore production in the U.S. is measured to be double that estimated in inventories (5.7 gCO2e/MJ compared to 2.4 gCO2e/MJ). This difference is primarily due to higher methane emission in Gulf of Mexico shallow waters.
• Offshore U.S. oil and gas production has lower carbon intensity than onshore production.
• Most production (80%) occurs in GOM deep waters, but most emissions (also 80%) occur outside of these areas.
• Offshore carbon intensity varies greatly across the U.S. Gulf of Mexico deep waters show low climate impacts, while the Cook Inlet and Gulf of Mexico shallow waters have the largest carbon intensities.
• The relative importance of CO2 and CH4 emissions differs between regions. For example, in GOM state shallow waters, CH4 emissions contribute significantly more to the carbon intensity than CO2 emissions, whereas CO2 is the dominant source of emissions in Alaska.

  Measurement-Based Carbon Intensity of U.S. Offshore Oil and Gas Production (Environmental Research Letters, 2024)
  View/Download Fact Sheet (PDF)

• Observed methane emissions exceed inventories, elevating the carbon intensity of the basin to over twice the official inventories. Mean observed methane emissions in federal and state waters are 3X and 13X higher than inventories, respectively. Methane emissions are of particular concern because they have greater global warming potential than carbon dioxide, especially in the short term. They should be the first focus of mitigation efforts.
• The carbon intensity of oil and gas production varies widely across the Gulf. Carbon intensity is extraordinarily high in shallow federal and state waters, where production rates are moderate and methane drives the majority of emissions. The observed shallow water carbon intensity far exceeds that of both deep water Gulf of Mexico production and typical global oil production. In contrast, carbon intensity is low in deep waters, where combustion emissions dominate the climate impact and production is high. Observational data of carbon dioxide from this study is generally consistent with inventories, suggesting that combustion is well-represented in the federal inventory.
• Central hub processing facilities are the primary contributor to excessive shallow-water production emissions. High-emission events from these facilities are frequent and can be attributed to cold venting, emissions associated with tanks, and other pieces of equipment.

  In Situ Sampling of NOx Emissions from United States Natural Gas Flares Reveals Heavy-Tail Emission Characteristic (Environ. Sci. Technol., 2024)
  View/Download Fact Sheet (PDF)

• By applying persistence and a newly defined metric, the chance of subsequent detection (CSD), the analysis of four years of repeated airborne sampling data from GOM central hubs reveals the sites that continue to emit methane over extended periods.
• The average CSD for an emitting GOM central hub at any revisit time between one month and four years is 74%. The eight GOM central hubs with the largest emission rates (>750 kg/hr), greatly exceeding the EPA's super emitter threshold of 100 kg/hr, have an average CSD of 96%.
• Forward-Looking Infrared (FLIR) camera imagery has shown cold venting at most of these super emitter sites, which—if it is the standard operating procedure—could explain these facilities’ high emissions and persistence. Modifying infrastructure to burn gas and maintain a consistent flame would support EPA goals and reduce the overall global warming potential of offshore production in the GOM.

  Temporal Variation and Persistence of Methane Emissions from Shallow Water–Oil and Gas Production in the Gulf of Mexico (Environ. Sci. Technol., 2024)
  Infographic (JPG)