Repository logo

Characterization of methane emissions from gathering compressor stations: final report




Zimmerle, Daniel, author
Bennett, Kristine, author
Vaughn, Timothy, author
Luck, Ben, author
Lauderdale, Terri, author
Keen, Kindal, author
Harrison, Matthew, author
Marchese, Anthony, author
Williams, Laurie, author
Allen, David, author

Journal Title

Journal ISSN

Volume Title


This document is the final report to the U.S. Department of Energy (DOE) for contract DE-FE0029068 awarded to Colorado State University (CSU). CSU and subcontractor AECOM partnered with nine U.S. midstream operators to characterize emissions from natural gas gathering and boosting stations ("gathering stations") – a sector of the natural gas supply chain where few measurements have been made and little data are available for component emissions. Although there is overlap in the classes of equipment on gathering stations with those on production sites or transmission stations that have been measured recently, emissions are likely to differ for functional and operational reasons. Partner companies provided the study team site access to (1) measure methane emissions and (2) collect activity data on equipment and operations. Field measurements were made at 180 facilities in 11 U.S. states during June-November 2017. Measured facilities were sampled from 1705 partner facilities located in 28 American Association of Petroleum Geologists (AAPG) basins. Measurements were made in basins representative of current U.S. production and facilities selected for measurement shared key characteristics in proportion to all partner facilities. The principal deliverable of this study is a set of emission factors for components and major equipment at gathering stations. Leaker and population emission factors were developed for components, and population factors were developed for major equipment. All data was also incorporated into a model to produce a nationally representative estimate of emissions from gathering stations. Emission factors and model results are intended to inform the U.S. Environmental Protection Agency (EPA) greenhouse gas inventory (GHGI). Components were counted on 1002 major equipment units (compressors, dehydrators, separators, tanks, acid gas removal units, and yard piping). Emission measurements were made on 1938 major equipment units. Data from a parallel study performed by GSI Environmental Inc. under the same DOE funding program were also incorporated. The field campaign supported a robust updating of emission factors for fugitive and vented emissions on components and major equipment. In general, the study indicates that study emission factors either agree with, or are larger than, current greenhouse gas reporting program (GHGRP) emission factors for the western U.S. and most GHGI emission factors, and are substantially larger than emission factors used by the GHGRP for the eastern U.S. This study also developed and field-tested two measurement methods to better characterize emissions from unburned methane entrained in compressor engine exhaust (“combustion slip”) and vented and fugitive methane emissions from gas-powered, pneumatically actuated valves and controllers. Emissions from these sources are not well characterized at compressor stations. Long-term, direct measurements of pneumatic controller emissions were made on 72 pneumatic controllers (PCs) at 16 gathering stations; measurements averaged 76 hours in duration. New emission factors could not be developed due to measurement errors with the meters utilized for the study (see Section 4.3), and as a consequence, GHGRP emission factors were utilized to estimate pneumatic controller emissions in station and national estimates. However, the PC emissions data is still useful for a qualitative examination of pneumatic controller emissions [1]. In particular, these long-duration measurements provide insight into PC emissions behaviors that are not reflected in manufacturer’s literature and have not been shown in prior studies. Recorded PC data shows a high degree of variability in operation over the course of hours or days – especially for intermittent vent PCs. Recordings also show an unexpectedly high occurrence of abnormal emission behavior – 25 of 40 intermittent vent controllers show abnormal behavior at some point during the recording, and 5 of 24 were emitting at higher than the low-bleed maximum of 6 scfh. Combustion slip was measured on 102 individual compressor drivers at 51 gathering stations. Results from combustion slip measurements indicate emissions similar to emission factors from EPA AP-42 [2]. Although component fugitive and vented emission factors are higher, current GHGI estimates are based upon whole-station measurements made in a prior field campaign [3] and subsequent national model [4]. Relative to these prior studies, and by extension the GHGI, emissions at stations measured during the field campaign are statistically lower. The current study drew a nationally-representative sample from a larger population of stations than the previous study (1705 stations versus ≈700 stations) while working with a larger group of industry partners (9 partners versus 4 partners), which raises confidence in the current study. While reasons cannot be definitively stated, likely causes of the lower methane emissions in this study are: (1) the previous study measured facilities with substantially higher throughput than the current study (39.5 [0.223 to 382] versus 19.7 [0.068 to 116] MMscfd whole gas); (2) the partner population in the previous study indicated a larger proportion of more complex stations – this study sampled 60% compression-only stations versus 30% in the previous study [3]; (3) the two studies utilized different measurement methods; and, (4) there may have been operational improvements to gathering stations during the intervening four years. To complete national estimates, the study utilized 319 per-basin GHGRP reports for gathering systems in 36 AAPG basins, including 15,895 reported compressors, and counts for other equipment, including gas pneumatic controllers, dehydrators, flares and other equipment. Using GHGRP activity data and data collected in the field campaign, the study estimated 6,108 [5,846 to 6,374] stations nationally, which is statistically higher then the current GHGI estimate of 5,241 stations. However, the study’s national model indicated emissions that are statistically lower than current GHGI estimates for the gathering & boosting sector – 1,484 [1,439 to 1,537] Gg · y−1CH4 versus a GHGI estimate of 1,955.1 Gg · y−1CH4. Reasons for this difference align with those for station emission estimates - updated mix, size and throughput of stations, more complete activity data for stations, better estimates for unmeasured emission sources, including unit and station blowdowns, and possibly improvements in operations at gathering stations since prior studies. Results presented in this report are supported by several supplemental volumes which are cited throughout. Supplemental volumes are further supported by appendices, as cited within. In addition, results will be disseminated in three peer-reviewed publications currently in preparation.


Authors are from: Energy Institute at Colorado State University, Fort Collins, Colorado, USA, AECOM, Austin, Texas, USA, SLR International Corporation, New Braunfels, Texas, USA, Mechanical Engineering, Colorado State University, Fort Collins, Colorado, USA, Fort Lewis College, Durango, Colorado, USA, and University of Texas, Austin, Texas, USA.

Rights Access


national model
emission factors


Associated Publications