Operators who fund dehydrator methane controls without knowing which component emits most can spend on the smaller source first. The Environmental Protection Agency (EPA), the federal environmental regulator, runs the Greenhouse Gas Reporting Program (GHGRP), which requires large emitters to report greenhouse gas data [1]. Its oil and gas rule, Subpart W of 40 CFR Part 98, the program’s federal regulation, covers petroleum and natural gas systems [2]. In September 2025, EPA proposed suspending Subpart W reporting obligations until reporting year 2034 (RY2034) [3].

An interim final rule of October 14, 2026 moved the reporting year 2025 (RY2025) deadline from October 30, 2026, to March 10, 2027 [4]. The reconsideration itself remains a proposal, not a final rule [4]. OGMP 2.0, the Oil & Gas Methane Partnership, is the United Nations Environment Programme’s flagship methane programme and asks members for measurement-based reporting [5].

A glycol pump circulates triethylene glycol (TEG), the dewatering liquid most gas dehydrators use to pull water out of natural gas [6]. A gas-assisted glycol pump runs on high-pressure gas, and that assist gas joins the glycol and is emitted when the glycol is regenerated [6]. Subpart W Method 1 reports still vent and, if present, flash tank methane separately, plus pump type, but no line isolates pump drive gas [2].

A 2026 Colorado State University study by Mdigo and colleagues modeled gas-assisted pumps at about 90% of dehydrator emissions in six uncontrolled units [7]. The model was developed as a modular component of MAES, the Mechanistic Air Emissions Simulator [7]. It models normal operation plus equipment failures, and returns a probability range for a facility’s emissions [8]. Below, for engineers and emissions managers choosing dehydrator controls, the post covers the 90% finding, Subpart W reporting and MAES modeling, then control spending.

Gas-Assisted Pumps Made Up About 90% of Modeled Dehydrator Emissions

Across the six modeled units with uncontrolled gas-assisted pumps, pump emissions were 90.25% of total dehydrator emissions and 63.10% of site-level emissions [7]. The abstract reports one aggregate over six modeled units, drawn from 154 operator-reported units, not a per-unit distribution [7]. The figure applies to six units with gas-assisted pumps whose flash tanks were uncontrolled [7].

Six modeled dehydrators with uncontrolled gas-assisted pumps: pumps 90.25% of dehydrator and 63.10% of site emissions; derived remainder: 6.8% other dehydrator, 30.1% rest of site.

The chart’s grey ‘Other dehydrator’ slice is non-pump dehydrator emissions; Subpart W Method 1 takes pump type as an input but reports no pump share [2].

Nine-tenths from one component is not a marginal finding. For units with gas-assisted pumps and uncontrolled flash tanks, electric pumps or routing flash gas to fuel or compression are likely the highest-impact investments.

Interested in building a Measurement-Informed Inventory for your operations? Contact us to learn about our MAES-based estimation services.

Subpart W Reports Dehydrator Methane by Vent, Not by Pump

Subpart W Method 1 reports each dehydrator’s methane separately for the still vent and, where a flash tank is used, the flash tank vent [2]. No reported field isolates the glycol pump’s drive-gas share [2].

Method 1 simulations include the gas-assist pump; an EPA guide says a gas-assisted pump sends about three times an electric pump’s gas to the regenerator. Reducing the glycol circulation rate lowers the gas the pump draws [9]. A flash tank separator recovers about 90 percent of the methane from the rich glycol, when its gas is routed to fuel or compressor suction [9].

Replacing the gas-assisted pump with an electric pump removes the driving gas altogether [10]. Subpart W’s reported vent totals do not show which option targets the larger emission share.

The Model Runs as a ProMax-Trained Surrogate Inside TetraSoft MAES

Inside TetraSoft MAES, the dehydrator model uses second-order polynomial regressions trained on ProMax simulations, so stream estimates respond to glycol circulation ratio, pressure and temperature [7]. Against ProMax output, the outlet-stream regressions reached a root mean square error (RMSE) of 0.002 or lower, a fit to simulation, not field data [7].

The MAES TEG module takes glycol circulation ratio and wet gas pressure and temperature as inputs, so rerunning it with seasonal values updates its estimate. Conditions change through the year; Subpart W Method 1 lets reporters use periodic simulations to capture them in the annual estimate [2].

Contactor pressure matters for pump emissions: one Kimray energy-exchange pump model uses 1.7 scf of gas per gallon at 300 psig, 8.3 at 1500 psig [11]. Higher glycol circulation means more stroke cycles per unit time for a gas-assisted pump. Pump gas use per gallon falls as absorber pressure drops, but net pump methane also depends on the circulation rate needed.

Component-Level Modeling Changes How Operators Allocate Control Investment

Operators of TEG dehydrators under 0.4 MMscf/d using Subpart W Method 2 default factors get no component split from them, so may misdirect spending. In those six units, controls that miss pump gas address under a tenth of modeled dehydrator emissions.

MAES is a mechanistic emissions model that generates probability distributions of expected emissions for each equipment component [12]. TetraSoft offers MAES by subscription through the MAES Platform [8].

The TEG dehydrator module is one modular component of MAES, which estimates emissions from the component level up across its equipment types. Colorado’s Oil and Natural Gas Annual Emission Inventory Reporting (ONGAEIR) is the state portal where operators file annual emissions inventories by June 30 [13]. Colorado operators may find component-level estimates useful when reviewing the dehydration-unit entries in their ONGAEIR submissions.

Frequently Asked Questions

What is a gas-assisted glycol pump?

A gas-assisted glycol pump uses high-pressure natural gas to drive glycol circulation through the dehydration system [10]. The spent driving gas leaves the pump with the rich glycol at low pressure [10]. Without a flash tank, it passes to the reboiler, where the entrained gas separates and water boils out, and that mixture vents from the still [10]. Electric pumps eliminate the driver gas stream, though gas absorbed in the glycol still vents from the regenerator [6].

Why do gas-assisted glycol pumps dominate TEG dehydrator methane?

Without a flash tank, entrained methane, including pump gas, vents from the regenerator; a flash tank routed to fuel or compression recovers about 90 percent [9]. In the study’s six uncontrolled units, non-pump emissions, which vary with glycol rate, pressure and temperature, were under 10% of the dehydrator total.

Does the 90% pump share apply to dehydrators with electric pumps?

The study reports the 90.25% pump share for six modeled units with gas-driven pumps and uncontrolled flash tanks [7]. Facilities that have replaced gas-assisted pumps with electric pumps eliminate that emission pathway [10]. For those facilities, still gas and flash gas become the dominant dehydrator sources, and the MAES TEG module models those components separately.

How does component-level modeling complement GHGRP Subpart W dehydrator reporting?

A MAES-based inventory can complement Subpart W by estimating the glycol pump’s share, which Subpart W does not report, for control planning. MAES estimates the glycol pump, still vent and flash tank contributions separately, rather than one unit-level total. EPA’s proposal to suspend Subpart W reporting until RY2034 remained unfinalized as of November 3, 2026; see GHGRP Subpart W Deadline Moved to March 10, 2027.

References

  1. US EPA, Greenhouse Gas Reporting Program (GHGRP) web page. https://www.epa.gov/ghgreporting
  2. US EPA, 40 CFR Part 98, Subpart W (Petroleum and Natural Gas Systems), Greenhouse Gas Reporting Program. https://www.epa.gov/ghgreporting/subpart-w-petroleum-and-natural-gas-systems
  3. US EPA, Reconsideration of the Greenhouse Gas Reporting Program, proposed rule, 90 FR 44591, September 16, 2025. https://www.govinfo.gov/content/pkg/FR-2025-09-16/html/2025-17923.htm
  4. US EPA, Extending the Reporting Deadline Under the Greenhouse Gas Reporting Rule for Reporting Year 2025 to March 10, 2027, interim final rule, 91 FR 65112, October 14, 2026. https://www.govinfo.gov/content/pkg/FR-2026-10-14/html/2026-20982.htm
  5. United Nations Environment Programme, Oil & Gas Methane Partnership 2.0 (OGMP 2.0). https://ogmpartnership.org/
  6. US EPA Natural Gas STAR Program, Electric Glycol Circulation Pumps. https://www.epa.gov/natural-gas-star-program/electric-glycol-circulation-pumps
  7. Mdigo et al., Mechanistic Modeling of TEG Dehydrator Emissions in Oil and Gas Industry, Fuels 7(2), 21, 2026, DOI:10.3390/fuels7020021. https://doi.org/10.3390/fuels7020021
  8. TetraSoft, MAES Platform, https://tetrasoftco.com/maes-platform/maes-landing.html
  9. US EPA Natural Gas STAR Program, Lessons Learned: Optimize Glycol Circulation and Install Flash Tank Separators in Glycol Dehydrators. https://19january2017snapshot.epa.gov/sites/production/files/2016-06/documents/ll_flashtanks3.pdf
  10. US EPA Natural Gas STAR Program, Lessons Learned: Replacing Gas-Assisted Glycol Pumps with Electric Pumps. https://19january2017snapshot.epa.gov/sites/production/files/2016-06/documents/ll_glycol_pumps3.pdf
  11. Kimray, Energy Exchange Glycol Pumps, catalog page 10:10.1. https://kimray.com/sites/default/files/import/documents/catalogpages/10_10.1.pdf
  12. Santos et al., Using Measurement-Informed Inventory to Assess Emissions in the Denver-Julesburg Basin, ACS ES&T Air 2, 1598-1611, 2025, DOI:10.1021/acsestair.5c00089. https://doi.org/10.1021/acsestair.5c00089
  13. Colorado Department of Public Health and Environment, Oil and Natural Gas Annual Emission Inventory Reporting. https://cdphe.colorado.gov/ongaeir