Geologic carbon storage is a key component of any policy roadmap to reach net zero carbon emissions by mid-century. Recent studies have quantified the amount of carbon dioxide that needs to be stored annually, with results generally in the gigatonne per year range in the US. However, field trials over the past two decades have shown that such high storage rates may not be feasible with current technology due to subsurface engineering challenges. We show how properties of the subsurface place limits on storage rates and how field projects have provided evidence of those limits hindering carbon storage efforts. Without urgently addressing these limits, realizing injection rates consistent with net-zero emissions modeling would require drilling activity that would exceed the entire US oil and gas industry. We conclude with recommendations for overcoming these challenges through focused collaborative efforts at information sharing, technology transfer, and supportive policies.

The U.S. federal government undertook a flurry of new actions to monitor and regulate methane emissions beginning in years 2021 and 2022, including the Inflation Reduction Act (IRA). The IRA and Waste Emissions Charge (WEC) represent a historic legislative effort to regulate U.S. methane emissions, yet its early effectiveness remains empirically unverified. Here we estimate changes in U.S. oil and gas methane emissions during 2019--2024 using an inverse model constrained by satellite observations. We find no detectable inflection point in U.S. oil and gas methane emissions in the years following these legislative and regulatory actions. Although methane emissions intensity from the oil and gas sector declined by 9.3--13.6\% during 2019--2024, this trend likely reflects gradual technological, operational, and regulatory developments rather than any abrupt shift associated with federal policy. Furthermore, we identify a divergence in inventory reliability: state-level inventories align with satellite estimates in the highest emitting states with greenhouse gas (GHG) emission targets but underestimate emissions by a factor of about 4 in the highest emitting states without GHG targets. Our results underscore the limitations of transient legislative signals, suggesting that without long-term regulatory durability, the industry lacks the incentive to commit to methane emission reductions.

Natural gas transmissions and storage compressor stations account for the largest share of methane (CH4) emissions in New York State (NYS). Yet, NYS’s CH4 emissions inventory is based on measurements that are a decade old and unlikely to be representative of NYS operations. Here, we present results from a comprehensive, multi-scale aerial CH4 measurement campaign across all NYS transmission and storage compressor stations. We find a skewed emissions distribution, with 20% of stations accounting for 74% of total CH4 emissions. Emissions at engine-driven compressor stations are, on average, 3-4x higher than emissions at turbine-driven compressor stations, thus demonstrating the need for separate emissions factors for engine- and turbine-drive compressor stations. Overall, measurement-informed emissions inventory from midstream transmission and storage compressor stations in NYS are 72% and 69% lower than the current NYS inventory, respectively. We estimate updated emissions factors of 464 [95% CI: 162 – 920] metric ton (MT) CH4/station/yr and 139 [97, 191] MT CH4/station/yr for engine- and turbine-based transmission compressor stations, respectively. Similarly, we estimate an updated emissions factor of 413 [164, 733] MT CH4/station/yr for engine-based storage compressor stations. These updated emissions factors, along with improved activity data, enable effective reconciliation of NYS inventory with measured emissions.

Methane emissions from the oil and gas sector are a tangible target for near-term mitigation of climate warming. Recently, the U.S. congress passed the Pipes Act of 2020, mandating the Pipeline and Hazardous Materials Safety Administration (PHMSA) to require new standards for leak detection and repair for U.S. natural gas pipelines to enhance safety and reduce methane emissions. In this work, we develop the FEAST-Pipeline model, a techno-economic model to assess the impact of PHMSA's proposed regulations on pipeline emissions at distribution, gathering and boosting, and transmission pipelines. Comparisons are made between LDAR programs employing PHMSA's proposed rules and current methods to identify and repair leaks. We evaluate the factors that influence the effectiveness of pipeline LDAR programs including pipeline material, installation age, corrosion rate, detection sensitivity, survey frequency, and repair time frame. Results of our simulation experiments indicate that PHMSA's proposed rules can greatly reduce pipeline emissions across all sectors, timely repair of leaks ensures the efficacy of LDAR programs, and LDAR programs that prioritize leak-prone pipelines can increase the efficiency of emissions mitigation. Methane leaks from natural gas pipelines contribute to climate change. This study models the effectiveness of leak detection and repair programs at reducing methane leaks from natural gas pipelines.

Capturing leaked methane can be a win for both firms and the environment. However, leakage volume uncertainty can be a barrier inhibiting leak repair. We study an experiment at oil and gas production sites that randomized whether site operators were informed of methane leakage volumes. At sites with high baseline leakage, we estimate a negative but imprecise effect of information on endline emissions. But at sites with zero measured leakage, giving firms information about methane leakage increased emissions at endline. Our results suggest that giving firms news of low leakage disincentivizes maintenance effort, thereby increasing the likelihood of future leaks.

Growth in liquefied natural gas (LNG) demand has been accompanied by debates about its greenhouse gas (GHG) emissions impact. Studies have shown that activity-based inventory accounting methods in the oil and gas sector significantly underestimates methane emissions. As a result, target-based policies and voluntary initiatives are moving towards adopting direct measurements to assess the GHG emissions intensity (EI) of LNG supply chains. Yet, most supply chain assessments of LNG do not incorporate measurement data. In this work, we develop a probabilistic, geospatial, and measurement informed life cycle assessment model to estimate the GHG EI of global LNG supply chains covering over 90% of LNG trade. We find a ~5x range in supply chain GHG EI from 8.6 g CO2e/MJ in Qatar to over 39 g CO2e/MJ in Algeria. Overall, our work suggests supply chain-weighted GHG EI of global LNG trade is underestimated by up to 31% compared to prior estimates that did not incorporate measurements. Probabilistic models of GHG EI of LNG exhibit a heavy tailed distribution, revealing the importance of low likelihood but high emitting supply chains. Incorporating direct measurements in supply chain assessments is necessary to avoid underestimation from activity-based inventories and increase confidence in target-based policies to address methane.

Hydrogen is of interest for decarbonizing hard-to-abate sectors because it does not produce carbon dioxide when combusted. However, hydrogen has indirect warming effects. Here we conducted a life cycle assessment of electrolysis and steam methane reforming to assess their emissions while considering hydrogen's indirect warming effects. We find that the primary factors influencing life cycle climate impacts are the production method and related feedstock emissions rather than the hydrogen leakage and indirect warming potential. A comparison between fossil fuel-based and hydrogen-based steel production and heavy-duty transportation showed a reduction in emissions of 800 to more than 1400 kg carbon dioxide equivalent per tonne of steel and 0.1 to 0.17 kg carbon dioxide equivalent per tonne-km of cargo. While any hydrogen production pathway reduces greenhouse gas emissions for steel, this is not the case for heavy-duty transportation. Therefore, we recommend a sector-specific approach in prioritizing application areas for hydrogen.

Landfills contribute 18% of all anthropogenic methane emissions in the US. While prior work has shown that official inventories underestimate landfill methane emissions compared to measurements, no systematic study of the spatio-temporal variation in landfill emissions exist. In this work, we report on results from quarterly, source-resolved, aerial methane measurements across 46 landfills in the Appalachian basin. While total landfill emissions were found to be consistent across all surveys, we observe significant variability in emissions from individual landfills, with mean annual emissions of 2532 metric tons per year (tpy). Despite observing no seasonal patterns, local precipitation events up to a week before surveys resulted in approximately 29% higher emissions. Landfills with renewable natural gas facilities emitted an average of 5011 tpy, 63% higher than the mean emission rate of landfills without any gas-to-energy projects. In addition, we find that methane emissions can be empirically predicted based on the volume of bulk waste disposed, providing an independent method to validate field measurements. We estimate a mean measurement-informed landfill gas collection efficiency of 53%, significantly lower than the mean reported value of 81%. Our findings underscore the need to update reporting methods that account for measured data to improve accuracy of landfill methane emissions.

Addressing methane emissions from the oil and gas supply chain has emerged as a key near-term mitigation target. The past decade of research has improved our understanding of methane emissions, with a primary focus on quantifying emissions without describing their underlying causal mechanisms. In this work, we integrate source-specific methane emissions measurement from multiple large-area aerial surveys with source-tracked cause analyses to identify and analyze causal mechanisms that underlie observed emission patterns. Overall, 53% of all observed emissions can be attributed to specific causal categories, with the rest comprising normal operational emissions. While abnormal tank emissions were the most common cause, unloading events exhibited the highest average emission rate. Importantly, we find that large release events are not driven by fundamentally different causal mechanisms than those of small emitters, indicating that escalation due to specific operational conditions, rather than fundamentally distinct causes, drives high-magnitude emissions. In addition, we observe statistically significant quarterly and inter-operator variability in the prevalence of different causal categories, reinforcing the need for adaptive, operator-specific mitigation strategies. These findings support a shift in methane mitigation from generalized leak detection with one-size-fits-all solutions toward risk-targeted, process-informed mitigation.

We compare continuous monitoring systems (CMS) from three different vendors on six operating oil and gas sites in the Appalachian Basin using several months of data. We highlight similarities and differences between the three CMS solutions when deployed in the field and compare their output to concurrent top-down aerial measurements and to site-level bottom-up inventories. Furthermore, we compare vendor-provided emission rate estimates to estimates from an open-source quantification algorithm applied to the raw CMS concentration data. This experimental setup allows us to disentangle the effect of the sensor platform (i.e., sensor type and arrangement) from the quantification algorithm. We find that: 1) localization and quantification estimates rarely agree between the three CMS solutions on short time scales (i.e., 30-minutes), but temporally aggregated emission rate distributions are similar between solutions, 2) differences in emission rate distributions are generally driven by the quantification algorithm, rather than the sensor platform, 3) agreement between CMS and aerial rate estimates varies by CMS solution but is close to parity when CMS estimates are averaged across solutions, and 4) similar sites with similar bottom-up inventories do not necessarily have similar emission characteristics. These results have important implications for developing measurement-informed inventories and for incorporating CMS inferential output into emission mitigation efforts.

Abstract Methane emission distributions are highly skewed, where a small portion of large emitters could contribute over 50% of total emissions. Studies have found that some of the large emitters are intermittent, and they only last as short as a few hours. However, leak detection and repair programs (LDAR) are only conducted periodically. Using the snapshot measurement information to extrapolate the annualized emission inventory brings uncertainties. In this work, we quantified the uncertainties that arise from the survey frequency. This study considers the impact of survey frequency on two major scenarios – without repair programs and including repair programs. The results show that survey frequency has a significant impact on the accuracy of measurement-informed inventory (MII). However, the impact of survey frequency on different scenarios varies.

Continuous Monitoring Systems (CMS) are a promising technology to detect and quantify intermittent and high-volume methane emissions across the oil and gas supply chain. This is particularly salient at midstream compressor stations where the contribution of short duration emission events to total emissions make survey type technologies less suitable to develop accurate measurement informed inventories. In this work, we report on the first concurrent and long-term test of five CMS technologies to detect, localize, and quantify methane emission from two major types of midstream compressor stations found in the US – a turbine-only station and an engine-only station. We find that CMS technologies can distinguish between different operational states of the compressor only under conditions of low background methane emissions. Combining known events at these facilities with in-situ controlled releases, we observe that all CMS technologies generally struggle in identifying short duration or low-emission rate (relative to baseline) events. Critically, we find that positive event detection, based on analysis of underlying methane signals, frequently did not translate to alerts sent to the operators. Deployment of CMS at midstream compressor stations must proceed with caution based on specific applications, site configuration, and the nature of baseline emissions.

Addressing methane emissions across the liquefied natural gas (LNG) supply chain is key to reducing climate impacts of LNG. Actions to address methane emissions have emphasized the importance of the use of measurement-informed emissions inventories given the systematic underestimation in official greenhouse gas (GHG) emission inventories. Despite significant progress in field measurements of GHG emissions across the natural gas supply chain, no detailed measurements at US liquefaction terminals are publicly available. In this work, we conduct multiscale, periodic measurements of methane and carbon dioxide emissions at two US LNG terminals over a 16-month campaign. We find that methane emission intensity varied from 0.007% to 0.045%, normalized to methane in LNG production. Carbon dioxide emissions accounted for over 95% of total GHG emissions using 100-year global warming potential (GWP) for methane. Thus, contrary to observations across other natural gas supply chain segments, we find that reported GHG emissions intensity closely matches measurement informed GHG emissions intensity of 0.24-0.27 kg CO2e/kg CH4. In the context of developing LNG supply chain emissions intensity, we conclude that the use of the Greenhouse Gas Reporting Program emissions intensity provides reasonably accurate estimates of total GHG emissions at LNG terminals.

The importance of reducing methane emissions from oil and gas operations as a near-term climate action is widely recognized. Most jurisdictions around the globe using leak detection and repair (LDAR) programs to find and fix methane leaks. In this work, we empirically evaluate the efficacy of LDAR programs using a large-scale, bottom-up, randomized controlled field experiment across ~200 oil and gas sites in Canada. We find that tanks are the single largest source of emissions, contributing to nearly 60% of total emissions. The average number of leaks at treatment sites that underwent repair reduced by ~50% compared to control sites. Although control sites did not see a reduction in the number of leaks, emissions reduced by approximately 36% suggesting potential impact of routine maintenance activities to find and fix large leaks. By tracking tags on leaking equipment over time, we find a high degree of persistence – leaks that are repaired remain fixed in follow-up surveys, while non-repaired leaks remain emitting. We did not observe any significant growth in emission rate for non-repaired leaks, suggesting that any increase in observed leak emissions following LDAR surveys are likely from new leaks. Vent emissions reduced by 38% without a significant reduction in the average number of vents across control and treatment sites, showing the importance of both anomalous vents and temporal variations in vent emissions. Our results show that a focus on equipment and sites that are prone to high emissions such as tanks and oil sites are key to cost-effective mitigation.

Quantitative optical gas imaging (QOGI) system can rapidly quantify leaks detected by optical gas imaging (OGI) cameras across the oil and gas supply chain. A comprehensive evaluation of the QOGI system’s quantification capability is needed for the successful adoption of the technology. This study conducted single-blind experiments to examine the quantification performance of the FLIR QL320 QOGI system under near-field conditions at a pseudo-realistic, outdoor, controlled testing facility that mimics upstream and midstream natural gas operations. The study completed 357 individual measurements across 26 controlled releases and 71 camera positions for release rates between 0.1 kg Ch4/h and 2.9 kg Ch4/h of compressed natural gas (which accounts for more than 90% of typical component-level leaks in several production facilities). The majority (75%) of measurements were within a quantification factor of 3 (quantification error of −67% to 200%) with individual errors between −90% and 831%, which reduced to −79% to +297% when the mean of estimates of the same controlled release from multiple camera positions was considered. Performance improved with increasing release rate, using clear sky as plume background, and at wind speeds ≤1 mph relative to other measurement conditions.

Methane emissions from oil and gas operations exhibit skewed distributions. New technologies such as aerial-based leak detection surveys promise cost-effective detection of large emitters. Recent policies such as the proposed US Environmental Protection Agency methane rule that allows the use of new technologies as part of conventional leak detection and repair (LDAR) require demonstration of equivalence with existing optical gas imaging (OGI)-based LDAR programs. In this work, we illustrate the impact of emission size distribution on the equivalency between OGI-based LDAR programs and that of alternative LDAR programs that use site-wide surveys. We find that emission size distribution compiled from aerial measurements across four oil and gas basins include significantly more emitters in the 1 – 10 kg/h and an order of magnitude lower average emission rate for large emitters compared to the emissions distribution in the EPA methane rule. As a result, equivalence between OGI-based surveys and site-wide screening may be achieved at lower site-wide survey frequencies when using technologies with detection threshold below 10 kg/h, compared to the EPA methane rule. However, equivalency cannot be achieved for site-wide screening with a detection threshold of 30 kg/h at any survey frequency because most emitters across most US oil and gas basins exhibit emission rates below 30 kg/h. We find that equivalence is a complex trade-off between the choice of technologies, design of hybrid LDAR programs, and survey frequency that can have more than one unique solution set.

The utilization of greenhouse gas (GHG) life cycle assessments (LCAs) of liquefied natural gas (LNG) has increased over the past decade. In this study, a novel framework for improved supply chain-specific LCAs for GHGs is presented using a gas pathing algorithm aligned with how gas is purchased, sold, and transported within the U.S. Utilizing supply chain emissions and gas purchase data specific to two U.S. liquefaction facilities, we identify 138 distinct gas pathways with GHG emission profiles that can vary by nearly a factor of 6. Reference case GHG intensities are 22-53% lower than prior studies for U.S. LNG delivered to Europe (production through regasification, 100-yr GWP). This study also incorporates recent supply chain measurement data. GHG intensities based on measurement data for U.S. LNG delivered to Europe are 41-52% higher than the reference case (production through regasification 100-yr GWP) and 8-11% higher for production through power generation boundaries (all market destinations, 100-yr GWP) but 20-28% lower than prior estimates employing national or regional nonempirical data. Supply chain-specific LCAs and the integration of emission measurements in LCAs are critical to accurately characterize the differences in GHG emissions from natural gas and LNG supply chains.

Low-carbon hydrogen is considered a key component of global energy system decarbonization strategy. The US Inflation Reduction Act includes incentives in the form of production tax credits for low-carbon hydrogen production, provided the lifecycle greenhouse gas (GHG) emissions intensity (EI) of hydrogen is below 4 kg CO2e/kg H2. Blue hydrogen or hydrogen produced from natural gas coupled with carbon capture and sequestration is one such pathway. In this work, we develop a geospatial, measurement-informed model to estimate supply chain specific lifecycle GHG EI of blue hydrogen produced with natural gas sourced from the Marcellus and Permian shale basins. We find that blue hydrogen production using Permian gas has a lifecycle EI of 7.4 kg CO2e/kg H2, more than twice the EI of hydrogen produced using Marcellus gas of 3.3 kg CO2e/kg H2. We conclude that eligibility for tax credits should therefore be based on lifecycle assessments that are supply chain specific and measurement informed to ensure blue hydrogen projects are truly low carbon.

Growth in US liquefied natural gas (LNG) exports have increased concerns about the climate impacts of methane leakage along LNG supply chains. Current life cycle analysis (LCA) models of US LNG supply chains are based on emissions estimates in national inventories that have been demonstrated to significantly underestimate emissions. In addition, recent top-down measurements of methane emissions exhibit significant sub-national spatial and temporal variation across oil and gas (O&G) basins. In this study, we develop a geospatial, measurement informed LCA model that incorporates recent top-down methane measurements to examine regional differences in greenhouse gas (GHG) emissions intensity of US LNG supply chains for delivery to Europe and Asia. For every megajoule of LNG shipped from the US, the energy allocated GHG emissions intensity of the Permian-UK LNG supply chain is 42% higher compared to the Marcellus-UK LNG supply chain. Disparities in LNG emissions intensity across source basins can be directly attributed to higher measured methane emissions compared to inventory estimates. Developing measurement informed, supply-chain specific lifecycle GHG emissions assessments is critical to enabling a global market for differentiated natural gas.