Paola Prado, McGill University: Temporal variability of methane emissions from non-producing oil and gas wells in Western Canada
Paola Prado, Mary Kang, Civil Engineering, McGill University
Abandoned oil and gas (AOG) wells are one of the most uncertain sources of anthropogenic methane emissions in Canada and around the world. It is estimated that Canada is home to ~400,000 of these wells, which can act as subsurface leakage pathways for methane, and may cause the contamination of groundwater, soil, and the atmosphere. There currently exists only one published study with direct methane emission measurements from AOG wells in Alberta and Saskatchewan, the provinces housing 87% of all abandoned oil and gas wells in the country. Moreover, most of these measurements only provide a snapshot of the emission rates, and the extent to which the rates can vary with time remain poorly understood. We aim to further characterize this emission source by carrying out direct measurements from at least wells in Alberta, Saskatchewan, and British Columbia, including at wells we measured previously. These measurements will be conducted using static chamber methodology. Beyond improving emission estimates, there is opportunity in informing environmental policy and examining site reclamation and remediation strategies. Therefore, our study will also target a portion of specific well sites to examine the viability of conversion of these abandoned wells into geothermal energy wells.
Nishant Narayanan, University of Waterloo
Nishant S. Narayanan1, Paule Lapeyre1, Martin Chamberland2, Marc-Andre Albert2, Jim Cormack3, Kyle J. Daun1
1 Department of Mechanical and Mechatronics Engineering, University of Waterloo, Canada
2 Telops, Canada
3 Lidar Services Inc., Calgary Canada
Operators and regulators require accurate techniques that can detect fugitive methane emissions. Particularly in view of emerging methane emissions regulations, it is also increasingly important to also quantify these emissions. Airborne technologies are attractive as they can scan large areas and multiple sites at once and access locations that are otherwise difficult to reach by ground.
Current commercial airborne systems use the absorption of ground-reflected, near-infrared (NIR) radiation to generate column density maps for methane. The incident light may come from a passive source (e.g., GHGSat’s SatD system) or using a tunable diode laser (e.g., Bridger’s GML system). An alternative is to use long-wave infrared (LWIR) hyperspectral imaging, in which methane column densities are inferred by thermal radiation emitted by the ground and absorbed by the gas, or thermal radiation emitted by the gas directly. This technique does not require incident irradiation from above and is therefore robust to ground cover that may be highly absorbing.
We evaluate the performance of this technique through a controlled release field trial campaign carried out at Carbon Management Canada’ Newall County facility. The Telops Hyper-Cam Airborne Mini was used operating between 760 cm-1 to 1365 cm-1. The results highlight the capability of the system to detect methane emissions, and particularly the accuracy of a tool used to predict the minimum detection limit for a given set of environmental conditions. The estimated methane emissions underestimate the true releases, highlighting the need for further development of the measurement model.
Khali El Hachem, St. Francis Xavier: Conjunctive Analysis of Measurements of Soil Emissions and Surface Casing Vent Flows from Oil and Gas Wells in Alberta
Khalil El Hachem1, 2, Sarah Kennedy1, Meghan Flood1, Mark Argento1, Scott A. Mundle3, David Risk1
1 Department of Earth and Environmental Sciences St. Francis Xavier University, Antigonish, Nova Scotia, Canada
2 Department of Civil Engineering, McGill University, Montreal, Quebec, Canada
3 Department of Chemistry and Biochemistry, University of Windsor, Windsor, Ontario, Canada
Oil and gas wells can lose their integrity leading to emissions from surface casing vents and surrounding soils. Measurements of methane emissions from soils surrounding oil and gas wells have been limited, and estimates of oil and gas well soil emissions in Canada’s greenhouse gas inventory are based on measurements collected in the 1990’s. Here, to obtain a sample representative of the different well operational statuses, we conduct and present 1241 measurements of soil emissions near 15 active, 28 unplugged abandoned, 10 plugged abandoned and 7 other oil and gas wells in Alberta and British Columbia and analyze this data with detections of methane from 41 surface casing vents. Relative to distance from the well we find that more than 85% of soil emissions occur within 3 meters from the well. We find that plugged and unplugged abandoned wells have higher soil emissions compared to active and other wells, where point level emissions reach upper limits of 1126.06, 1012.87, 22.68, and 0.16 mg CH4/hour/m2 for plugged abandoned, unplugged abandoned, active, and other wells, respectively. Of the 41 wells where we tested surface casing vents, we find that 38 wells (92%) have methane detections above background. Our data and analysis provide valuable insights into soil emissions near oil and gas wells and can be used to inform Canada’s greenhouse gas inventory. Moreover, our analysis of detections of surface casing vent flows in conjunction to soil emissions allow for a more comprehensive understanding of well leakage.
Daniel Blackmore, University of Waterloo: A Bayesian Approach to Truck-based Methane Emissions Estimates Uncertainties
Daniel Blackmore1, Jean-Pierre Hickey1, Augustine Wigle2, Paule Lapeyre1 Kyle J. Daun1
1 Department of Mechanical and Mechatronics Engineering, University of Waterloo, Canada
2 Department of Statistics and Actuarial Science, University of Waterloo, Canada
Canada’s oil and gas industry is committed to the regulatory requirements of a 75% methane emissions reduction compared to 2012 levels by 2030. To achieve this, a variety of different technologies have been developed and employed to quantify methane emissions. Understanding the uncertainties associated with these emissions estimates is crucial in the application of regulations, and in knowing what actually occurs with respect to emissions and the environment. Understanding uncertainty can help explain discrepancies between reported emissions inventories, and recent measurements made with these technologies. Key to uncertainty is the nature of the error of models these technologies employ to obtain their emissions estimates.
This poster presents the application of a Bayesian statistical framework to understand the uncertainty of emission estimates provided by a truck-based tunable diode laser absorption spectroscopy technology. Concentration data from a controlled release study in April 2022 is compared to the predictions of a Gaussian plume model to investigate the model error. A large eddy simulation of similar release conditions is used for further model error investigation. The results are used within a Bayesian framework, where the model error is key in shaping the likelihood function. This framework is applied to data from a second controlled release campaign from September 2022, to obtain posterior distributions of emission rates. Uncertainties associated with emissions estimates can be quantified using credibility intervals on the posterior distributions. Additionally, a deeper understanding of the model error is obtained, including the shape and form of the likelihood function when using this method.