EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Enabling validation of belowground natural gas leak quantification using controlled release experiments

Joelle R. Uribe, G. Venkata Rao, Jui-Hsiang Lo, Kathleen M. Smits

Elem Sci Anth · 11.11.2026 · 14 (1)

DOI: 10.1525/elementa.2026.00034 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Elem Sci Anth, erschienen am 11.11.2026. Verfasst von Joelle R. Uribe und 3 weiteren Autor:innen. Die Arbeit nennt 46 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Interest in methane (CH4) leak quantification from belowground natural gas (NG) pipelines has increased due to emerging regulations mandating routine leak rate quantification. Because leak sources are belowground, emission rates must be inferred from CH4 observations that are strongly influenced by environmental conditions, soil properties, and surface characteristics. A major barrier to improving belowground leak quantification methods is the limited availability of controlled-release datasets suitable for algorithm validation and comparison. This study compiles, curates, and publishes 22 high-fidelity datasets from previously conducted belowground controlled-release experiments and demonstrates their utility to independently evaluate belowground NG leak quantification methods that leverage surface CH4 mole fraction measurements collected during routine field operations. To illustrate their value, 3 publicly available surface-measurement-based leak quantification models, DINA, ESCAPE−1, and Modified ESCAPE−1, were applied to experiments conducted at 4 known release rates (37, 89, 122, and 182 g-CH4 h−1). The models differed considerably in data requirements, with ESCAPE−1 and Modified ESCAPE−1 requiring approximately 18% and 73% more input variables than DINA, respectively. Modified ESCAPE−1 also relied on higher temporal resolution data to capture flow-state dynamics. Model performance varied across leak sizes and environmental conditions, with Modified ESCAPE−1 achieving the lowest overall median absolute percentage error (20%) and reduced bias (−6%) relative to ESCAPE−1 (34.7%, −26.6%) and DINA (58%, −35.7%). Results indicate that incorporating subsurface transport and near-surface atmospheric processes improve precision, particularly at lower emission rates, while increasing data complexity. These findings highlight trade-offs between data burden and estimation performance and demonstrate how model choice and environmental conditions influence inferred leak rates. By addressing critical gaps in data availability and transparency, this work contributes to dataset sharing practices that facilitate integration of individual belowground release experiments, support independent testing of belowground NG leak quantification methods, and inform development of practical approaches compatible with leak investigation workflows.

Zitieren

APA

Uribe, J. R., Rao, G. V., Lo, J., & Smits, K. M. (2026). Enabling validation of belowground natural gas leak quantification using controlled release experiments. Elem Sci Anth, 14(1). https://doi.org/10.1525/elementa.2026.00034

Vancouver

Uribe JR, Rao GV, Lo J, Smits KM. Enabling validation of belowground natural gas leak quantification using controlled release experiments. Elem Sci Anth. 2026;14(1). doi:10.1525/elementa.2026.00034

BibTeX
@article{Uribe2026,
  author = {Uribe, Joelle R. and Rao, G. Venkata and Lo, Jui-Hsiang and Smits, Kathleen M.},
  title = {Enabling validation of belowground natural gas leak quantification using controlled release experiments},
  journal = {Elem Sci Anth},
  year = {2026},
  volume = {14},
  number = {1},
  doi = {10.1525/elementa.2026.00034}
}
Kostenlos starten →