A Pipe Dream? Tackling Uncertainties in leak detection practices as we transition to a more sustainable future

Becoming involved with experienced scientists changed my outlook of my time at StFX. On my journey toward graduation FluxLab provides a learning environment driven by a diverse, small group of like-minded, motivated researchers with qualifications ranging from high-school diplomas to doctoral degrees. As a young scientist, being guided by experienced researchers and peers with similar interests was incredibly helpful in shaping my project goals in a way that naturally plays to my strengths. Dave Risk is particularly good at quietly guiding everyone’s unique approach to the scientific process. In my time at the lab, I worked on two conceptual design projects focused around minimizing the uncertainty of two methane gas detection techniques, and now I continue methane research in my Master’s studies.

The field of Leak Detection and Repair (LDAR) is an interdisciplinary field that crept into my constantly vacillating research interests. Methane emission research and solutions integrate economics, engineering, and statistics, and engages policymakers, industry and scientists.  When I began work in the lab, Canada had just finalized a new energy sector methane emission reduction policy to limit emissions of this greenhouse gas (GHG), whose global warming potential is 28-34 times more powerful than carbon dioxide on a 100-year horizon [1]. Compared to carbon dioxide, methane is a short-lived GHG. Methane typically lives in the atmosphere for about 12 years, whereas carbon dioxide’s lifetime is closer to 100+ years [2]. This means that short-term efforts to reduce methane emissions can be a powerful tool to reduce the global warming potential of the atmosphere, with the possibility to see results in a 4-8 year window.

Methane emission inventories suggest that focusing methane mitigation efforts on the oil and gas (O&G) sector is worthwhile – a 2020 study shows the O&G sector contributes 44% of Canada’s total methane emissions [3]. A large proportion of these emissions originate from what’s called fugitive emissions. My philosophy professor laughed when I used the term (he judged my poster on the topic at the annual research poster fair) but fugitive emissions, also referred to as leaks, are the unintentional release of hydrocarbons from infrastructure that should not be emitting gasses (as opposed to the controlled and intentional, vented releases) [4].

In 2019 I worked on a project with a goal to develop a field-deployable benchmarking system for Optical Gas Imaging (OGI) minimum detectable leak rate (MDL) that doesn’t directly involve dangerous or environmentally controlled releases of methane. This project involved building a model using field hardware to predict resolution thresholds for instrumentation that integrates image analysis and calibration.

In the spring of 2019, Dave discussed the problem OGI uncertainty presents regulators, researchers, and industry in detecting fugitive methane leaks from oil and gas infrastructure. For example, an energy company hires two OGI operators to detect leaks and one operator finds 4 leaks at one point in the day and the other finds 0 leaks at another time. When this happens, neither the energy company nor the operator can have full understanding of what’s causing this discrepancy in the number of leaks found. Uncertainty in OGI performance is attributed to many factors including environmental conditions, operator experience, survey practices, leak size distributions, and gas composition [5].  The discrepancy in the number of leaks could be associated with any of these factors, none of which correlate to leak rate. This research area is interesting because both energy companies and the atmosphere benefit from more efficient leak detection.  Operators’ consistency to successfully locate and repair leaks is conducive to reducing fugitive emissions, working towards meeting Canada’s 2025 emissions targets, and providing more product for sale and thus economic stability for oil and gas producers [6].

To address the above scenario, I conducted a feasibility experiment to find a safe and readily available proxy gas to replace methane for field benchmarking. Once this proxy gas was established as visible under similar temperature contrast conditions between the methane and the background by the FLIR OGI camera, further experiments were carried out that mimicked weather conditions in the field.

Since OGI cameras define methane leaks on a pixel-by-pixel basis, where leak rate drives both the pixel intensity and number of pixels, we were able to develop a quantitative approach for image analysis using Python that relies on pixel output. For each experimental condition, a pixel intensity threshold was defined, and the pixel number was used to establish a relative response factor for methane and the proxy, so that they could be related. With additional experiments to increase the sample size of data, I believe it will be a generally straightforward process to perform a linear regression analysis to produce a non-parametric correlation coefficient to act as a response factor.

Tackling research problems like these puts researchers at the forefront of a turning point in our country’s energy history policies. As decisions are made about our energy future, informing policies to reduce methane emissions places scientists in a mediary position as society moves towards a more sustainable future. In an era dominated by knowledge advancement and technological development, I believe this area of research allows us to advance policy that drives us ahead of the curve of climate change, and empowers economic growth and reform in the financial market.

My time at FluxLab signalled a positive turning point during my undergraduate studies. I came to understand my university degree was not only a checkpoint leading to a career; my degree prepared me to apply what I learned in the classroom to society. From here, as I begin my Master’s, I wish to explore the question, “Can we develop technologies transferable from academia to industry that are simple, effective, and cheap enough for energy companies to stomach?” I’m looking forward to finding out if there is an answer, and contributing to this pipeline of research supporting a hopeful future.

by Ellen McCole

[1] Natural Gas Vehicles for America (2018). Understanding Global Warming Potential and other Greenhouse Gas Emission Metrics [PDF file]. Retrieved from: http://www.ngvamerica.org/wpcontent/uploads/2018/06/Understanding-Global-Warming-Potential.pdf

[2] EDF (2008). Greenhouse Gasses: How Long Will They Last? [Blog Post]. Retrieved from: http://blogs.edf.org/climate411/2008/02/26/ghg_lifetimes/#:~:text=About%2050%25%20of%20a%20CO,lifetime%3A%2050%2D200%20years.

[3] Matthew R. Johnson, David R. Tyner (2020) A Case Study in Competing Methane Regulations: Will Canada’s and Alberta’s Contrasting Regulations Achieve Equivalent Reductions? [PDF file]. Retrieved from: https://doi.org/10.1525/elementa.403.s1

[4] T. A. Fox, T. E. Barchyn, D. Risk, A. P. Ravikumar, C. H. Hugenholtz. (2019). A review of close-range and screening technologies for mitigating fugitive methane emissions in upstream oil and gas. Available at: https://iopscience.iop.org/article/10.1088/1748-9326/ab0cc3.

[5] A. P. Ravikumar, J. Wang, and A. R. Brandt. Are Optical Gas Imaging Technologies Effective for Methane Leak Detection? American Chemical Society, Available at: https://pubs.acs.org/doi/10.1021/acs.est.6b03906