1. Office of Research and Development National Risk Management Research Laboratory, Air Pollution Prevention and Control Division Photo image area measures 2” H x 6.93” W and can be masked by a collage strip of one, two or three images. The photo image area is located 3.19” from left and 3.81” from top of page. Each image used in collage should be reduced or cropped to a maximum of 2” high, stroked with a 1.5 pt white frame and positioned edge-to-edge with accompanying images. Near-field production pad assessment with GMAP-REQ EDF Synthesis Workshop on Methane Leakage October 17-18, 2012, Boulder Colorado E. D. Thoma, Bill Squier – EPA / ORD / National Risk Management Research Laboratory, 109 TW Alexander Drive, Durham NC. 27711 NRMRL Fugitive and Area Source Group Source and Fenceline Measurements Methods and Technology Development

2. 1 Motivation for development of a mobile, near-field emission assessment approach Number of oil and gas production facilities is increasing –Impact of VOC emissions to ozone attainment is uncertain –GHG emissions estimates (fugitives and tanks) can be improved Proximity of potential sources to populations is increasing –Growing need to understand HAP emission potential Limited measurement data, can be difficult to estimate emissions –Many source types and engineering configurations –Significant variability in maintenance states and product composition This presentation describes a new inspection tool called GMAP-REQ –Locates leaks by drive-by inspection –Allows for rapid, offsite source assessment –Provides complimentary information to on-site and top down techniques

3. 2 Fugitive and area sources EPA’s National Risk Managment Research Laboratory Agricultural operations, industrial fugitives, coal mining, waste water, oil and gas production, landfills, and more spatially variable temporally variable large spatial extent

4. 3 Fugitive and area source measurements (a growing number of near-field tools) We are fortunate to have an increasing number source assessment approaches –tracer correlation, dial lidar, OTM-10 flux plane, SOF, bLs, eddy covar., mp-fenceline The best tool to use depends on the source characteristics and objectives – Landfill measurements by tracer correlation: whole facility emissions, a snapshot in time –Landfill measurements by eddy covariance: cover types, a long term view Oil and gas production has some unique source characteristics –Many sources over large areas, variable maintenance and product composition GMAP-REQ helps provide information on some aspects of O&G emissions – Strengths (finding emissions, off-site assessment, rapid deployment) – Weaknesses (limited range of use conditions, not as accurate as on-site approaches)

5. Many types of emissions complicates assessment FLIR Video File

6. Source: Microsoft Bing Maps (© Microsoft Corporation Pictometry Bird’s Eye © 2010 Pictometry International Corp ) 5 As the separation distances of potential sources to populations decrease, the need for periodic inspection increases

7. Off-site assessment with GMAP-REQ ( Geospatial Measurement of Air Pollution – Remote Emissions Quantification) driving path wind direction 6 Spike in CH 4 indicates emission CH 4 Position vehicle in the plume Acquire CH 4 and wind data for 20 minutes Pull a 30 second canister sample for VOC information

8. 3D sonic anemometer 1.4 litter canister placement Auto-north met station High-res GPS Quad Sampling Port In the truck: High-precision CH 4 Instrument (critical component) Batteries, control system, IR camera, rangefinder 7 GMAP REQ measurement equipment

9. 8 Estimating emissions with GMAP-REQ (two methods: PSG and bLs) Filter off-axis info and fit CH 4 vs. wind dir., determine local background Point Source Gaussian (PSG) – Use distance and atmospheric stability to find expected σ y,σ z (lookup) – Perform simple 2-D integration (q = 2π·σ y ·σ z ·u·c) Backwards Lagrangian Stochastic (bLs) – Use distance,CH 4, and 3D sonic data in model WindTrax 2.0 Estimate VOC emissions by canister ratio approach with CH 4

10. 9 CH 4 release-recovery experiments (testing the GMAP REQ approach) 15 m -150m Wind Release methane gas from a variety of scenarios (0.6 g/s ) – Free release, obstructions, and simulated tanks – Different ground surfaces and distances – A range of atmospheric conditions and wind speeds – Three different GAMP systems

11. GMAP REQ release and recovery experiments (PSG results only)

12. Site A

13. Site B open thief hatch

14. GMAP REQ Comparison of two sites (A and B) from a recent campaign Thief Hatch Closed

15. GMAP REQ draft results 10/05/12 (PSG only, values off scale) N = 97N = 82 N = 22 N = 179 Draft 100512 - do not cite

16. 15 Summary and next steps GMAP-REQ can be useful inspection tool to complement on-site measurements for the oil and gas sector Preliminary results provide interesting comparisons with direct emissions measurements and inventory estimates Data analysis continues – Development of QA checks and comparisons with CFD modeling – Visualization and infrared camera databases GMAP REQ method development activities continue – New user interface software with source location indicators – Expand to UV detection for BTEX (facility LDAR applications)

17. 16 Backup Slides

18. 17 Remote Emission Measurements OTM 10, bLS, DIAL, Tracer Correlation  0.2 km Source  1 km < 0.1 km 0.1 < 1 km 0.1 <  0.5 km OTM 10 bLs DIAL Tracer Correlation (  1 km)

19. PSG:bLs combined emission estimate results for CH 4 release experiments (N=27) Release rate band (horizontal lines) Comparison of PSG and bLs results for release and field data (N=321) CH 4 release-recovery experiments and PSG to bLs model comparisons 18 PSG bLs Average Overall Average

20. Example methane time series 19 Figure 2: Example of (a) 10 Hz CH 4 concentration, (b) 10 Hz wind direction with 10second moving average, (c) 20-mintue time average concentration (ppm) vs. wind direction with Gaussian fit. (a) (b) (c) From AWMA Control # 2012-A-21-AWMA June 19, 2012 San Antonio, TX

21. GMAP REQ plume (over background) and background comparisons (N=480)

22. 21 GMAP REQ Field Data CH 4 Emissions ColoradoFort Worth TXWyoming N = 104N = 87N = 103 (11.9, 14.2)*(10.3, 20.6)*(8.4, 10.3, 11.1, 19.0)* *off scale  mean  median Analysis based on GMAP REQ DA analytical and QA filter criteria as of February 2012 From AWMA presentation Control # 2012-A-21- AWMA June 19, 2012 San Antonio, TX

23. 22 GMAP REQ Field Data VOC Emissions (14.5, 7.3, 3.5, 3.5)* (8.9, 7.2, 4.8, 3.3)* N = 52N = 59N = 75 *off scale  mean  median 0.17 ColoradoFort Worth TXWyoming Analysis based on GMAP REQ DA analytical and QA filter criteria as of February 2012 From AWMA presentation Control # 2012-A-21- AWMA June 19, 2012 San Antonio, TX

24. 23 Each vertical column is one canister sample (N= 60) VOC to Methane Ratio (Greeley CO)

25. 24 Each vertical column is one canister sample (N= 60) VOC to Methane Ratio (Fort Worth TX)

26. 25 Dry Gas Condensate and Gas Oil EPA and City of Forth Worth studies approximate measurement areas EPA (some wet gas) City of Forth Worth (mostly dry gas) From production data From AWMA presentation Control # 2012-A-21- AWMA June 19, 2012 San Antonio, TX

27. 26 GMAP REQ compared to on-site measurements VOC Emissions Dry gas sites Condensate oil tanks ColoradoTexas N = 23N = 24N = 59 *off scale  mean  median N = 52 N = 167 (14.5, 7.3, 3.5, 3.5)*(6.4, 4.4)* 0.17 Analysis based on GMAP REQ DA analytical and QA filter criteria as of February 2012 From AWMA presentation Control # 2012-A-21- AWMA June 19, 2012 San Antonio, TX

28. 27 GMAP REQ “VOC snapshot measurements” compared to CO condensate tank emissions inventory expressed in g/s. (tanks within 500 m of GMAP measurement, Inv. data provided by Dale Wells, Colorado DPHE) Draft 040212 (14.5, 7.3, 3.5, 3.5)*(3.9, 3.2)*  mean  median In Greeley Colorado, condensate tank emissions are controlled by flares *off scale 0.17 Inv. Uncontrolled: modeled inventory assuming 0% control Capture Efficiency (CE), Inv. CE=75%, RE=83%: State of CO estimate of 75% control CE and 83% Rule Effectiveness (RE), 95% control effectiveness Inv. Reported: Reported inventory assuming 100 % CE, 100% RE and 95% control effectiveness From AWMA presentation Control # 2012-A-21- AWMA June 19, 2012 San Antonio, TX

29. 28 Continuing work on dataset and method Geospatial visualization database Google Earth database Allow viewing of field Data and IR videos Developing Wind QA Chart to help confirm source location and distance Source Wind QA Chart (needs magnetic declination correction) Observation

30. Office of Research and Development National Risk Management Research Laboratory, Air Pollution Prevention and Control Division Photo image area measures 2” H x 6.93” W and can be masked by a collage strip of one, two or three images. The photo image area is located 3.19” from left and 3.81” from top of page. Each image used in collage should be reduced or cropped to a maximum of 2” high, stroked with a 1.5 pt white frame and positioned edge-to-edge with accompanying images. CFD Results of 6-Tank Simulations Case set-up: 6 identical tanks, 12’ dia. 15’ tall, separated by 3’ Domain dimensions: x: -50 m – 200 m; y: -50 m – 50 m; z: 0 m – 100 m Grid size: 0.0125 m – 1.6 m Total cells: 43.2 millions 0 degree wind, U(3m) = 4.5 m/s Emission rate: Case 1: 1 g/s each from 6 tanks Case 2: 6 g/s from one tank 29

31. 30 Method development includes computational simulations to understand flow With good winds, emissions from the tops of the tanks get mixed down by wake Measurements at about 3 m work pretty well in these cases Probe Height

32. Contour and Line Plots: 6 Emission Sources 31

33. Contour and Line Plots: 1 Emission Source 32

34. Contour and Line Plots of Normalized Concentrations: Single Tank 33

35. 34 Basic Data Analysis Approach Estimate CH 4 emissions using concentration and wind data Obtain emission information for other compounds by a ratio of canister to CH 4 data F t = [(C t * F o )/C o ] [M t /M o ] Where: F t =the flux of the target compound (VOC) C t =the measured concentration of the target compound F o =the calculated methane flux C o =the measured methane concentration M t =the molecular weight of the target compound M o =the molecular weight of methane

36. 35 EPA oil and gas measurement team and contractor and vendor acknowledgments Eben Thoma, Bill Squier, Bill Mitchell - EPA Office of Research and Development, National Risk Managment Research Laboratory Dave Olson, Sergey Napelenok, Matt Landis, Karen Oliver - EPA Office of Research and Development, National Risk Managment Research Laboratory Robin Segall, Jason DeWees, Roy Huntley - EPA Office of Air Quality Planning and Standards Adam Eisele, Cindy Beeler, Carl Daly - EPA Region 8 Mike Miller, Michael Morton, Ruben Casso - EPA Region 6 Marta Fuoco, Chad McEvoy, Loretta Lehrman - EPA Region 5 Rich Killian, CarolAnn Gross-Davis, Ron Landy, Howard Schmidt - EPA Region 3 Armando Bustamante, Ken Garing - EPA NEIC M. Shahrooz Amin, Mark Modrak - ARCADIS (Contractor EPA) Chris Lehmann, Joe Ibanez, Buzz Harris, David Ranum - Sage Environmental Consulting Inc. Chris Rella, Eric Crosson - Picarro Inc. Wei Tang, Matt Freeman, Mike Uhl - Lockheed Martin Working with City of Forth Worth TX, States of CO, WY, and TX Disclaimer: Mention of products, companies, or services does not constitute endorsement