Constraining Emissions of Methane in Utahs Uintah Basin

Constraining Emissions of Methane in Utah’s Uintah Basin with Ground-based Concentration Observations and a Time-Reversed Lagrangian Transport Model (STILT) Christopher S. Foster (chris. foster@utah. edu)1, Erik Crosman 1, Lacey Holland 2, Derek Mallia 1, Ben Fasoli 1, Ryan Bares 1, John Horel 1, John Lin 1 1 The University of Utah, Department of Atmospheric Sciences, 2 The University of Hawai’i at Manoa, Department of Atmospheric Sciences Motivation and Goals: CH 4 Concentration (ppm) • The Uintah Basin produces approximately 1% of the total U. S. natural gas, but it is suggested to be unique in its large leak rate, ~10% (6. 2 -11. 7%) of the production amount (Karion et al. 2013) • Foster et al. 2017 examined the accuracy of two CH 4 emission inventories within the Uintah Basin and found the NOAA inventory to better characterize the spatial distribution of emissions than the EPA Foster et al. 2017 Observed CH 4 concentration (ppm) over two transects in Sept. and Nov. , 2017 5 4 3 2 Basin CH 4 WFR-STILT Modeling (Foster et al. 2017) Observational Analysis 2015 -2018 1) How do concentrations vary as a function of time of day, time of year, and year to year? Wyoming Utah Mobile Observations: 6 a. b. • FRU excellent background site, remaining below 2 ppm on average except in the winter stagnation periods in basin • HPL generally experiences higher concentrations than other two basin sites • During transitional seasons (spring and fall), HPL and ROO observe higher concentrations than in the summer – this is likely due to increase stability and shorter days/longer nights • In the summer, daytime minimum concentrations return to near background • CSP has a similar diurnal cycle to ROO in the spring and HPL in the winter – possibly due to a combination of meteorology and emissions in region • CH 4 diurnal cycles are impacted by meteorological patterns –boundary-layer height, local and synopticscale flows Colorado Uintah Basin (a) Emission rate of CH 4 (micromole m-2 s-1) within Utah’s Uintah Basin, at 4 km resolution from NOAA. (b) As in (a), except at 0. 1 degree resolution from EPA. Location of gas and oil wells shown as light a. • Foster et al. 2017 analyzed data from April and May, 2015 • There are currently ~ 3 years worth of CH 4 data available at 3 sites within the Uintah Basin: Fruitland (FRU), Roosevelt (ROO), Horsepool) • A 4 th, supplementary site, Castlepeak (CSP), operated between Nov. 2015 and May 2016 • Meteorological data also available at the 4 sites (temperature, wind speed and direction) a. b. Data and Methods: • Foster et al. 2017 study period during SONGEX (Shale Oil and Natural Gas n. EXus) and JAGUAR (Joint Air and Ground Uintah basin Air emissions Reconciliation project) • CH 4 inventory from Ahmadov et al (2015, NOAA) and Maasakkers et al. (2016, EPA) • CH 4 time series analysis April 2015 – present • Missing data is not included in analysis (no interpolation) • Observational analysis conducted on all days with a focus on “quiescent” (light winds) and “non-quiescent” conditions • FRU is considered a background site - due to its location and predominately westerly flow • HPL located to the north of gas wells, ROO is located within oil wells, CSP located within more densely situated gas wells Summary: • NOAA emission inventory reproduced the average CH 4 diurnal cycle within standard error at Horsepool and Castlepeak • CH 4 concentrations at the three basin sites (HPL, ROO, CSP) show daily and seasonal variations consistent with boundary layer evolution and stability • Higher concentrations are observed at night, while near background are observed during the day (slightly elevated above background during day in transitional seasons) • Meteorological conditions, specifically wind speed and direction (in addition to boundary layer evolution), play a key role in observed CH 4 concentrations at the sites Future Work: • Key future research questions: 2. What are the driving factors in the temporal variation of observed CH 4 concentrations? 3. Can an observational CH 4 dataset of this size be used in conjunction with other available data to make claims about the state of CH 4 emissions within an ONG region? Acknowledgements: (a) Average STILT CH 4 contribution (log(ppm)) during the time period 18 April 2015 to 31 May 2015 for Horsepool, UT only for quiescent days. (b) As in a, except for Roosevelt, UT. • Diurnal averages and average contribution include only quiescent days. • At HPL, EPA nighttime average lower than both Ahmadov and observations. • Horsepool contribution shows influence from east-southeast. • Roosevelt contribution shows influence from north-northeast. • Roosevelt unfiltered diurnal average shows spikes at night indicative of local influence. • When CH 4 observations with high relative standard deviation or simultaneous low wind speeds are filtered out, spikes in diurnal average at night are removed. • Simulations perform better at Horsepool than Roosevelt, where nighttime maxes are under simulated by both inventories. • At CSP, NOAA nighttime average closer to observation than EPA, which underestimates nighttime by ~0. 3 ppm. c. Average methane concentration at each hour of the day (MST) during the period 18 April 2015 to 31 May 2015 on quiescent days. Red lines represent observation, the red dashed line represents unfiltered observations, blue lines represent simulation using NOAA inventory, green lines represent EPA inventory. Standard error shaded on each line. (a) Roosevelt. (b) Horsepool. (c) Castlepeak. • The winter is characterized by multi-day episodes with elevated CH 4 concentrations • These periods are the result of cold-air pooling within the basin, which causes pollutants (CH 4 and others) to build up over many days and synoptic scale disturbances clean out the elevated concentrations • The build up of CH 4 from 28 Nov. – 11 Dec. 2015 was gradual (2 ppm to 6+ ppm at all 3 sites) over the roughly 2 week period, while the cleanout was much shorter • Diurnal patterns are still seen superimposed within the gradual build up • FRU also experiences some elevation above background • This study was supported by NOAA Climate Program Office’s Atmospheric Chemistry, Carbon Cycle, and Climate program, award #NA 14 OAR 4310138 • Utah Department of Environmental Quality. • Ben Fasoli and Ryan Bares for his help with the Nerdmobile • Derek Mallia for his help setting up and running the STILT model • John Horel for his help with ceilometer data • Seth Lyman for assistance with observations Selected References: • Ahmadov, R. , and Coauthors, 2015: Understanding high wintertime ozone pollution events in an oil- and natural gas-producing region of the western US. Atmos. Chem. Phys. , doi: 10. 5194/acp-15 -411 -2015. • Karion, A. , and Coauthors, 2013: Methane emissions estimate from airborne measurements over a western United States natural gas field. Geophys. Res. Lett. , 40, 4393– 4397, doi: 10. 1002/grl. 50811. • Lin, J. C. , C. Gerbig, S. C. Wofsy, A. E. Andrews, B. C. Daube, K. J. Davis, and C. A. Grainger, 2003: A near-field tool for simulating the upstream influence of atmospheric observations: The Stochastic Time-Inverted Lagrangian Transport (STILT) model. J. Geophys. Res, 108, doi: 10. 1029/2002 JD 003161. • Maasakkers, J. D. and Coauthors, 2016: Gridded National Inventory of U. S. Methane Emissions. Environ. Sci. Technol. , 23, doi: 10. 1021/acs. est. 6 b 02878.