Spatiotemporal Variation of Point Source Emissions in Permian Basin and its Implication on Basin Level Inventory Building and Mitigation

Introduction

Methane emissions from oil and gas sector have high spatial and temporal variability. Previous studies with limited sample size and very few (if any) numbers of repeated measurements have often been used to estimate basin and national scale emissions, leading to high uncertainties to the inventory estimation. This study investigates methane emission data on point sources collected in repeated surveys of Permian Basinover 2019–2023¹, focusing on the spatiotemporal variability of emissions and the sampling strategy for developing a relatively accurate annual emission inventory for large oil gas regions with limited number of surveys that capture the temporal variation and a fraction of the facilities that can cover the spatial variability.

Methodology

We have developed a multi-faceted approach reorganizing repeated flight overpass data into surveys to estimate source persistence and construct emission events characterized by unique combinations of rates, duration (in terms of number of surveys), and frequency (number of repeated events). Using multiple time series construction approaches including discrete event simulations, we model the annual variability of point source emissions at multiple scales: sub-basin (25, 100, 400 km²) and basin (4000 km²), and use random sub-sampling to determine minimum coverage criteria to capture annual emission variability accurately. Our criteria are based on two parameters—the number of surveys conducted and % facilities covered in the basin.

 

Results and Discussion

 Our findings highlight that considerable temporal variability in methane emissions can occur, especially for areas of sub-basin scale (≈100–400 km²). The emission rates of detected facility-scale sources vary significantly across years (average Coefficient of Variation, COV ≈ 2.3). Our results also reveal that emission events are short-lived (≥75% events lasting ≤ 2 surveys), occur infrequently (average persistence ≤ 33%), yet significantly contribute to total detected emissions (~70%). Nevertheless, these intermittent facility-scale emissions translate to significantly lower variability across sub-basin scale surveys (COV ≈ [0.2–1.3]) and basin-scale surveys (COV ≈ 0.5).

We also study the spatial and temporal variability of emission estimates for sub-basin and basin-scale surveys. At the basin scale, even a single survey sufficiently captures the annual emission variability (bias ≤ 10%) and the emission distribution across different sub-basins (cosine similarity of 0.8–1.0). However, at a sub-basin scale, even estimates based on 25 surveys can be biased (bias ≈ 20%). We find low correlations (Spearman R < 0.5) of time series patterns of sub-basin-sub-basin and sub-basin-basin-scale areal emissions, pointing to the diverse emission patterns within the basin.

 

Concluding Remarks

Our findings on spatial (areal coverage) and temporal (number of surveys) considerations can help build annual emission measurement strategies for oil and gas production regions.

Reference

• Cusworth, et al., 2022. Strong methane point sources contribute a disproportionate fraction of total emissions across multiple basins in the United States, PNAS, 119 (38) e2202338119, https://doi.org/10.1073/pnas.2202338119