Improving Methane Emissions Inventories with Measurements: A Mechanistic Modeling Approach for Midstream Oil and Gas Facilities in the Appalachian Basin

Jacob Mdigo¹, Arthur Santos¹, Daniel Zimmerle¹, Arvind Ravikumar²

Colorado State University,² The University of Texas at Austin

Download Poster PDF

Background

Natural gas (NG) in the U.S. is not only a source of energy but also a transition energy to green energy sources. However, these benefits can be insignificant if emissions from the NG supply chain are not mitigated. With efforts to build more accurate and transparent emission inventories, field measurements can enhance the inventory estimates if correct methodologies are applied. This is often referred to as measurement-informed inventories (MII). Santos et al.2 describe a methodology that uses the mechanistic air emissions simulator3 (MAES) to incorporate measurement data into the inventories. In this study, the methodology was applied to build MII inventories for the Appalachian Methane Initiative4,5 (AMI) midstream facilities.

Reported Sources and Unreported Sources flow into MAES which flows to MII

Motivation:

  • Voluntary Reporting Initiatives – OGMP 2.0
  • Investor and Market Drivers – ESG expectations & EU market requirements
  • Internal Environmental Action Targets – Corporate climate goals
  • Reputation and Transparency in Emissions Reporting

Methodology

Compressor stations, dehydration stations, and gas plants in OH, WV, and PA

a) Study Area for AMI Study Partners: Midstream

Included 168 midstream facilities:

  • 9 Dehydration stations
  • 10 Gas processing plants
  • 149 Compressor stations
MAES (Mechanistic Air Emissions Simulator) logo

b) MAES Simulation Tool, beyond Bottom-Up (BU) Models

  1. Input Data: Equipment information, site configuration, Gas composition
  2. Methods: a) Use of MC methods, Markov Chain Matrices, mechanistic models, and traditional models to estimate emissions
  3. Output: CH4 emissions up to 1-sec resolution at different equipment operational states.
Oil drum icon, spreadsheet icon, MAES logo
Normal CH4 Simulated Emissions and Measurement Data

c) MII Methodology

  1. MAES Inventory Models
    • Replicating reported emissions using MAES
      • Replicates the annual inventories to 1-second resolution.
      • Assumes upset-related emissions are often missed in the inventories.
  2. MAES MII Models
    • a) Detecting unreported emissions (MAES + Aerial Surveys)
    • b) Updating MAES with unreported sources using aerial data

Results

a) MAES Inventory Models vs GHGRP Reported CH4 Emissions (mt)

CH4 Emissions (Metric Tons/Year) with Facilities. Data split into reported vs simulated.
An example of aligning the MAES Inventory Models to GHGRP CH4 reported emissions per facility.

b) Detected Failure Events from Aerial Data by Equipment Type

Equipment# Scanned# FailuresFailure Rate
Flare488680.139
Compressor2864310.011
Facility Piping6682300.344
Tank 113402450.183
Tank 2 (Only Controlled)163110.067
Heater12681820.144

c) Equipment CH4 Emissions contribution in the GHGRP & MII

Total Reported CH4 Emissions (34397.7 mt/year). Tank (0.7% or 244.6 mt/yr), Flare (3.6% or 1254.1 mt/yr) Dehydrator (5.6% or 1916.1 mt/yr) Other (5.1% or 1751.9 mt/yr), Compressor (85.0% or 29230.5 mt/yr). Heater is 0.3 mt/yr.
Total MAES Simulated MII CH4 Emissions (47126.5 mt/year). Tank (11.8% or 5541.0 mt/yr), Separator (0.5% or 249.0 mt/yr), Heater (4.8% or 2244.2 mt/yr), Flare (3.7% or 1737.7 mt/yr), Dehydrator (2.2% or 1037.5 mt/yr), Other (11.1% or 5216.6 mt/yr), Compressor (66.0% or 31100.5 mt/yr).

d) CH4 Emissions by Inventory Comparison

GHGRP (34,093), MAES Inventory Models (-6.2% or 31,968), MAES MII Models (+38.2% or 47,126)

e) Methane Intensity (MI)

Reported MI= Repoted CH4 Emissions/ Reported CH4 Throughput = 34456.5/ 101948986.6 = 0.034%
MAES MII MI = MII CH4 Emissions/ Reported CH4 Throughput = 47104.3/101948986.6 = 0.046%

Conclusions 

  • The MAES inventory estimates were ~6% lower than the reported CH₄ emissions, primarily because the reported data included failure-mode emissions from compressor seals. In contrast, the MAES inventory assumes regular operation and does not simulate such failures. This discrepancy is within the ±15% alignment tolerance recommended by Santos et al.
  • This study shows that abnormal emissions are underrepresented in the reported inventories. In the Appalachian Basin, by incorporating field measurement data, the midstream site emissions are underestimated by approximately 38.2% as shown by the MAES MII models.
  • Further, the MII models indicate that tank and facility piping (i.e., Other) emissions are significantly underestimated in the reported inventories as observed by the field measurement data and as simulated by MAES MII models.
  • The MII MI for the Appalachian Basin was calculated to be 0.046%, slightly higher than the reported MI of 0.034%.
  • This work shows that incorporating modeling with field data helps operators choose aerial survey providers by matching technology and detection limits to the specific goal.

Acknowledgments and Contact Information

This research was funded by the U.S. Department of Energy (DOE) through the Energy Emissions Modeling and Data Lab (EEMDL). It was made possible through the methane monitoring efforts of the Appalachian Methane Initiative. The opinions, findings, conclusions, and recommendations expressed herein are those of the authors and do not necessarily reflect the views of the organizations providing technical input or financial support.


Poster Author:

Jacob Mdigo | Master’s Student | Systems Engineering, Colorado State University | [email protected]


Project PI:

Arthur Santos | Research Scientist | CSU Energy Institute, Colorado State University | [email protected]

 


References

  1. U.S. Energy Information Administration. (2025). U.S. energy facts explained — consumption and production. Independent Statistics and Analysis. Retrieved September 17, 2025, from https://www.eia.gov/energyexplained/us-energy-facts/
  2. Santos, A., Mollel, W., Duggan, G. P., Hodshire, A., Vora, P., & Zimmerle, D. (2025). Using Measurement-Informed Inventory to Assess Emissions in the Denver-Julesburg Basin. ACS ES&T Air, 2(8), 1598-1611.
  3. Mollel, W., Mdigo, J., Santos, A., Vora, P., Duggan, J., & Zimmerle, D. (2024). MAES study sheet guide.
  4. Appalachian Methane Initiative (AMI), Center for Energy and Environmental Systems Analysis, University of Texas at Austin. (2025). Appalachian Methane Initiative (AMI). Retrieved September 17, 2025, from https://www.ceesa.utexas.edu/ami
  5. METEC, Colorado State University. (2025). Appalachian Methane Initiative (AMI). Retrieved September 17, 2025, from https://metec.colostate.edu/appalachian-methane-initiative-ami/
  6. Levi, M. (2013). Climate consequences of natural gas as a bridge fuel. Climatic Change, 118(3-4), 609–623. https://doi.org/10.1007/s10584-012-0658-3
Equipment at the METEC Site