satellite observations of no 2 and methane over u s oil
play

Satellite observations of NO 2 and methane over U.S. oil and gas - PDF document

7/6/2020 Satellite observations of NO 2 and methane over U.S. oil and gas production areas Barbara Dix 1 , Joep de Bruin 1,2 , Esther Roosenbrand 1,2 , Tim Vlemmix 3 , Colby Francoeur 1,4 , Alan Gorchov-Negron 5 , Brian McDonald 1,4 , Mikhail


  1. 7/6/2020 Satellite observations of NO 2 and methane over U.S. oil and gas production areas Barbara Dix 1 , Joep de Bruin 1,2 , Esther Roosenbrand 1,2 , Tim Vlemmix 3 , Colby Francoeur 1,4 , Alan Gorchov-Negron 5 , Brian McDonald 1,4 , Mikhail Zhizhin 1,4,6 , Christopher Elvidge 6 , Pepijn Veefkind 2,3 , Pieternel Levelt 2,3 , Joost de Gouw 1 1. Motivation and background 1 University of Colorado 2 University of Delft, the Netherlands 2. Satellite data used in this work 3 KNMI, the Netherlands 3. Emissions: observation and source attribution 4 NOAA 4. Methane monitoring metrics 5 University of Michigan 5. Summary and outlook 6 Colorado School of Mines Acknowledgements: Ronald van der A, Henk Eskes, Bud Pope NASA ACMAP program, Colorado Energy Research Collaboratory NOAA Cooperative Institute Agreement, Rocky Mountain Institute 1 Motivation and background: U.S. oil and natural gas production … … is at an all-time high. … is was at an all-time high. COVID-19 economic impact Data: Baker Hughes Production and drilling activities vary on different time scales. 1

  2. 7/6/2020 1 Motivation and background: U.S. oil and natural gas production https://www.washingtonpost.com/graphics/national/united-states-of-oil/ 2019: U.S. is the largest oil producer worldwide. 1 Motivation and background: Atmospheric impacts What are the atmospheric impacts of oil and gas production? • NOx: nitrogen oxide = NO (nitric oxide) + NO 2 (nitrogen dioxide) CH 4 NOx combustion  on-site motors/transportation Source: VOCs Impacts: toxic (NO 2 ), air pollution  ozone/particle formation Lifetime: ~ hours • VOCs: Volatile Organic Compounds/hydrocarbons Sources: oil/gas/fracking chemicals Impacts: harmful to toxic, air pollution  ozone/particle formation Lifetime: ~ hours - days • CH 4 : methane Sources: infrastructure leaks, venting, flaring Impact: greenhouse gas (warming potential ~ 25 times CO 2 ) consumerenergyalliance.org Lifetime: ~ 9 years  NOx Air quality  CH 4 Climate 2

  3. 7/6/2020 1 Motivation and background: NOx 0 1 several (but mostly in Dallas) 4 3 3 0 ● EPA sites with NO 2 and O 3 sensors Surface air quality monitoring in O&G production regions is very limited. 1 Motivation and background: Methane Why are levels of methane rising globally? What is the contribution of O&G system emissions? www.eurekalert.orgmultimediapub128882.php 3

  4. 7/6/2020 1 Motivation and background: Methane in the U.S. * * Some studies suggest EPA emissions are underestimated 72% of O&G system methane emissions come from production. www.epa.gov 2 Satellite data used in this work Tropospheric Monitoring Instrument Ozone Monitoring Instrument (TROPOMI) (OMI) ESA Sentinel 5P satellite NASA AURA satellite October 2017 - present July 2004 - present 24 x 13 km 2 7 x 3.5/7 km 2 Resolution: Resolution: UV-vis channel: NO 2 , HCHO UV-vis + SWIR channel: NO 2 , HCHO, CH 4 Global coverage: 1 day Global coverage: 1 day 4

  5. 7/6/2020 2 Satellite data used in this work How does a remote sensing satellite work? Top of Atmosphere NADIR MEASUREMENTS air air CH 4 NO 2 air SATELLITE NO 2 NO 2 air SUN air backscattered CH 4 CH 4 CH 4 Solar Radiation air air CH 4 air CH 4 air NO 2 NO 2 Surface EARTH Column density measurement in units of: molecules per cm 2 or • • volume mixing ratio, e.g., ppb (parts per billion) Satellite observations of emissions are affected by surface reflectance and clouds. 2 Satellite data used in this work What does a remote sensing measurement mean? Total column density = Top of Atmosphere air air Emission + NO 2 air CH 4 NO 2 CH 4 Background + air NO 2 CH 4 CH 4 CH 4 air air air Atmospheric chemistry + air CH 4 CH 4 NO 2 NO 2 + OH  HNO 3 air Atmospheric transport air CH 4 CH 4 NO 2 CH 4 NO 2 air Surface Satellite measurements of emissions are affected by background concentrations and atmospheric chemistry and transport. 5

  6. 7/6/2020 2 Satellite data used in this work NO 2 and CH 4 over O&G production areas can be observed from space. OMI NO 2 2018 QA4ECV version 1.1; www.temis.nl TROPOMI CH 4 TROPOMI 2018/05- NO 2 2020/04 2018/05- 2020/04 Level 2, offline; https://scihub.copernicus.eu/ Level 2, offline; https://scihub.copernicus.eu/ 3 Emissions: NOx source attribution 2016: Bakken Denver- Julesberg 2017: San Juan Permian Eagle Ford 2018: Dix et al., GRL , 2020 NO 2 serves as proxy for NOx emissions. 6

  7. 7/6/2020 3 Emissions: NOx source attribution by multivariate regression Permian Basin NO 2 signal caused by: NOx fraction from production NOx fraction from drilling NOx background NO 2 = c 1 ∙ background + c 2 ∙ rig count + c 3 ∙ oil volume Dix et al., GRL , 2020 NOx emissions from production and drilling can be attributed separately. NOx emissions are dominated by those from drilling of new wells. 3 Emissions: NOx source attribution comparison with inventory Comparison with Fuel based Oil and Gas NOx inventory (FOG) Bakken Permian Basin Top-down (satellite) and bottom-up (inventory) source attributions agree. Dominant but declining fraction of NO x emissions come from drilling. Dix et al., GRL, 2020; Francoeur et al., in prep., 2020 Gorchov Negron et al., Environ. Sci. Technol., 2018 7

  8. 7/6/2020 3 Emissions: NOx from flaring Flaring volumes derived from VIIRS satellite night-time images Dix et al., GRL, 2020 Elvidge et al., Energies, 2015 NOx from flaring contributes ~5-10% to total O&G NOx emissions. (NOx from flaring is not (yet) discernable from space.) 3 Emissions: NOx source attribution by spatial correlation Feb 2019 TROPOMI NO 2 Feb 2019 Feb 2019 NO 2 scales with oil and gas production volumes. Esther Roosenbrand, CU Boulder, University of Delft, the Netherlands, master thesis, 2020 Industrial activity data from Enverus (formerly known as DrillingInfo) 8

  9. 7/6/2020 3 Emissions: CH 4 source attribution by spatial correlation Feb 2019 TROPOMI CH 4 Feb 2019 Feb 2019 CH 4 scales with oil and gas production volumes. de Gouw et al., Scientific Reports, 2020 Industrial activity data from Enverus (formerly known as DrillingInfo) 3 Emissions: CH 4 source attribution by seasonal signal TROPOMI CH 4 2018/12 - 2019/03 TROPOMI CH 4 2018/05 - 2020/04 2018/12- 2019/08 Uintah basin Surface CH 4 emissions are “trapped” in temperature inversion during winter. de Gouw et al., Scientific Reports, 2020 9

  10. 7/6/2020 Emissions: CH 4 source attribution by seasonal signal Satellite column measurements are consistent with surface monitoring (Univ. of Utah). de Gouw et al., Scientific Reports, 2020 3 Emissions: CH 4 and NO 2 over the Denver-Julesberg and San Juan basins 2018/05 - 2019/12 Denver-Julesberg San Juan CH 4 NO 2 CH 4 NO 2 gas oil gas oil TROPOMI NO 2 spatial distributions are dominated by urban sprawl. TROPOMI CH 4 spatial distributions show some correlation with O&G production areas. Industrial activity data from Enverus (formerly known as DrillingInfo) 10

  11. 7/6/2020 3 Emissions: CH 4 over the Denver-Julesberg basin: current work Denver-Julesberg 2018/05 - 2020/04 ongoing cooperation with CDPHE  3 research flights in summer 2020 suggested flight tracks Fort Greely Collins Boulder Denver DIA  CH 4 * Research flights will help to link TROPOMI CH 4 observations to local O&G emissions. *  CH 4 = CH 4 – background CH 4 background CH 4 = monthly 10 th percentile  more sensitive to boundary layer excess CH 4 4 CH 4 monitoring metrics: trends in CH 4 and CH 4 vs production slopes How can we monitor the industry’s performance? Slope = CH 4 / prod. unit Note: 10 ppb  0.5% of total column Currently: average CH 4 emissions per oil and gas production ~ constant Reduction in CH 4 emissions:  slope will decrease 11

  12. 7/6/2020 4 CH 4 monitoring metrics: trends in CH 4 – NO 2 ratios 2019/01/31 Slope =  NO 2 /  CH 4 Currently: average  NO 2 /  CH 4 ratios ~ constant Reduction in CH 4 emissions:  ratio will increase de Gouw et al., Scientific Reports, 2020 4 CH 4 monitoring metrics: detection of upset emissions by pixel statistics wind emissions Distributions are determined by emissions and wind.  Upset emissions are captured in 90 th percentile. 12

  13. 7/6/2020 4 CH 4 monitoring metrics: detection of single events 2019/02/26 2019/03/17 2019/03 2019/02 Single day (multi-day?) events get “lost” in monthly averages. 4 CH 4 monitoring metrics: discussion on emissions Natural gas as bridge fuel for a carbon-free future? TROPOMI CH 4 inversion Zhang et al., Science Advances, 2020 Schneising et al., ACPD , 2020 (CH 4 emissions from atmospheric modeling) (CH 4 emissions from mass balance) Permian: Permian: 2.7 ± 0.5 Mt/a 3.16 ± 1.13 Mt/a break-even point based on gas system break-even point based on oil and gas system de Gouw et al., Earth’s Future, 2014 Benefit of natural gas over coal remains only if leakage is below break-even point.  How do we define break-even point? 13

  14. 7/6/2020 5 Summary and outlook Summary • NO x from production and drilling can be attributed separately using NO 2 time series. Top-down and bottom-up source attributions agree on distribution and trend. • Increased columns of NO 2 and CH 4 consistently correlate with oil and gas production. Developed metrics can identify basin-wide trends and local outliers. • Daily global coverage is suitable to observe intermittent industrial activity. • Satellite data can help assess effectiveness of NOx and CH 4 regulations. Outlook • Continued work on metrics development. • Quantifying basin-wide NO 2 and CH 4 emissions and pushing sub-basin scales. 14

Recommend


More recommend