Evidence of Emissions from Oil and Gas Drilling Operations in Northeastern
Colorado
NOAA GMD: Gaby Petron, Steve Montzka, Ben Miller, Adam Hirsch*,
Anna Karion, Colm Sweeney, Jon Kofler, Arlyn Andrews, Ed
Dlugokencky, Laura Patrick, Pieter Tans & NOAA Carbon Cycle Group
*now at the National Renewable Energy Laboratory, Golden, CO
NOAA CSD: Greg Frost, Michael TrainerNOAA PSD: David Welsh, Dan Wolfe
AcknowledgmentsWestern Regional Air Partnership: Tom Moore
Environ: Amnon Bar-Ilan and John Grant
0
2
4
6
8
10
12
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
0
2
4
6
8
10
12
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
Methane emissions (Tg) from natural gas systems EPA US GHG inventory 2010 vs 2011
Field production ProcessingTransmission and Storage Distribution
2010 EPA 2011 EPA
2009 Methane emission estimates: TOTAL 32.7 Tg, Natural gas systems 10.5 Tg, Enteric fermentation 6.6 Tg, Landfills 5.6 Tg, Coal mining 3.4 Tg, Manure management 2.4, Petroleum Systems 1.5 Tg
x2
Northern Colorado Front Range ozone non attainment area: Estimated 40% of total VOC in the region due to oil and gas operations in Denver Julesburg Basin (DJB)
Green River Basin, WY: Very high winter time ozone in natural gas field (Schnell et al., Nature, 2009)
Uinta Basin, UT: hourly ozone in natural gas field up to 155 to 159 ppb last winter
Examples of air pollution from O&G operations in Western States
Jonah, Feb 2008
Pinedale smog
Outline• Oil and gas operations emission signature in
Colorado Northern Front Range• Source estimates & • Comparison with inventories• Conclusions
NOAA Cooperative Tower Long Term Measurement and Sampling Network
Boulder Atmospheric Observatory, Erie, Weld county (NOAA PSD)
SGP
NWF
o 300 meter tall towero 30 sec- Meteorological Datao Daily discrete air samples from 300 meter level since Aug 2007Tower team: Arlyn Andrews, Jonathan Kofler, Jonathan Williams
Denver Julesburg Basin
Denver
Non attainment area for summer time ozone
12,000 gas wells in Weld County~ 900 new ones in 2006
Discrete Air Samples AnalysesNOAA Boulder Lab
o MAGICC System (Carbon Cycle Group): CO2, CH4, N2O, CO, H2, SF6
CH4 repeatability error: 1.2 ppb
o GC/MS System (HATS group): C3H8, nC4H10, iC5H12, nC5H12, C2H2,
C6H6, CFCs, HFCs, PFCs… Most species: total uncertainty <5% n-butane and C2H2: 10-15%
o High precision long term well calibrated measurements
standards
air samples
analyzers
http://www.esrl.noaa.gov/gmd/ccgg/ & http://www.esrl.noaa.gov/gmd/hats/
Logistics: Molly Heller, Chris Carparelli, Jack Higgs,…Analysis: GMD: Tom Conway, Andy Crotwell, Ed Dlugokencky, Pat Lang, Paul Novelli, Kelly StrokerHATS: Ben Miller, Carolina Siso, Steve Montzka
MAGICC
Air samples collected at the BAO and SGP* have a strong alkane signature. Both sites are in major oil and gas production regions.
* SGP is a NOAA aircraft site in Northern Oklahoma. Samples collected below 650 meters were used for this analysis. GMD aircraft program leader: Colm Sweeney.
BAO: Distinct Alkane Signature
NOAA Cooperative Tall Tower Measurement and Sampling Network
SGP
Median summer mixing ratio at 7 NOAA TowersMidday data only (June-Aug 2007-2009)
NWF
BAO
0
100
200
300
400
500
CH4 ppb -1800
C3H8 ppt/10
nC4H10 ppt/4
iC5H12 ppt nC5H12 ppt
Med
ian
mix
ing
ratio
STR, CA WGC, CANWF, CO - Night BAO, COWKT, TX SGP, OKLEF, WI
Alkane source in N&E wind sector: Denver Julesburg Basin
North & East
South
West
Boulder Atmospheric Observatory- 30-sec met data at 3 vertical levels (NOAA PSD)- daily* midday air sample collection from 300 meter level and analysis in NOAA GMD labs
Air samples from the North and East wind sector have the strongest alkane signature
(all year round), suggesting this is where the alkane source is likely located.
0
200
400
600
800
1000
CH4-1800ppb
C3H8 ppt/10
nC4H10ppt /5
iC5H12ppt
nC5H12ppt
N&E Summer N&E WinterS Winter W Winter
Median mixing ratios (N>30) Aug 2007-April 2010
Mobile Lab intensive sampling in Front Range
Picarro Sensor: CO2, CH4, H2O (Colm Sweeney, Anna Karion)
Ozone 2B Analyzer (Laura Patrick)
GPS
Denver Julesburg Basin, O&G wells
1700
1800
1900
2000
2100
2200
11.5 12 12.5 13 13.5 14 14.5 15 15.5 16 16.5Local Time on 7/31/2008
0
100
200
300
400
500
600
PFP 3
PFP 4,5PFP 7,8
PFP 6
PFP 9,10,11
stop at BAO
PFP 2
Mobile Platform to sample close to
sources
July 31, 2008
Targeted flask sampling on most drives: total of 88 flasks collected over
June-July 2008
Landfill (up to 2.4ppm)
Waste water treatment plant (up to 3.2 ppm)
CH4
O3
CO2
Boulder
Regional alkanes enhancements vs point CH4 sources
Denver Julesburg Basin, O&G wells
July 31, 2008Landfill (up to 2.4ppm)
Waste water treatment plant (up to 3.2 ppm)
CH4
O3
CO2
Same alkane signature at BAO and in Mobile Lab samples
y=0.50xr2=0.98
y=0.48xr2=1
The alkanes are strongly correlated in BAO N&E wind sector samples and in Mobile Lab samples.
The alkanes come from the same source located in NE part of the Front Range.
BAO N&E summer Mobile Lab
Slope: 0.10Slope: 0.11
Multi species analysis: Separation of various methane sources
1750
1850
1950
2050
2150
0 5000 10000 15000 20000 25000 30000C H t
Background methane
Oil and gas contribution
Feedlots contribution
Propane ppt
Met
han
e p
pb
Samples collected with the Mobile Lab close to feedlots, a landfill, and a waste water treatment plant are above the oil and gas methane-to-propane correlation line.
July 14, 2008Sample collection: Lloyd Miller, William Kolodzey (no Picarro)
subset
Outline• Oil and gas operations emission signature in
Colorado Northern Front Range• Source estimates & Comparison with inventories• Conclusions
Western Region Air Partnership Oil and Gas Total VOC Inventory
0
10
20
30
40
50
2006
flash
ing
2006
ventin
g
2006
other
2010
flash
ing
2010
ventin
g
2010
other
VOC
em
issi
ons
Gg/
yrOther Categories
Compressor Engines*
Venting - recompletions
Venting - initialcompletionsVenting - blowdowns
Pneumatic pumps
Pneumatic devices
Unpermitted Fugitives
Permitted Fugitives*
Glycol Dehydrator*
Small condensate tanks
Large condensate tanks*
State regulated sources in the Front Range O3 non attainment area (NAA)
2006 based on industry data and reported emissions for permitted sources
2010 projected
Emission estimates for Front Range NAA
Amnon Bar Ilan et al. [2008]
Upstream and midstream operations
only
Speciation profiles of raw natural gas and condensate tanks flash emissions
0% 20% 40% 60% 80% 100%
CO2
N2
methane C1
ethane C2
propane C3
i-butane iC4
n-butane nC4
i-pentane iC5
n-pentane nC5
C6+ heavies
Venting GWA
Venting WRAP
Flashing WRAP
0.00% 0.05% 0.10% 0.15%
benzene
toluene
ethylbenzene
xylenes
2,2,4-trimethylpentane
Used to derive speciated emissions for fugitive/vented emissions (raw gas) and flashing emissions from condensate tanks.
The WRAP inventory used average emission profile.
We used the entire documented range.
Venting WRAP: average of industry data for DJB (company proprietary data)Venting Greater Wattenberg Area Study: natural gas samples from 77 wells in DJB (2006, COGCC)Flashing WRAP: EPA TANK model output for 16 condensate tanks in DJB (2002, CDPHE)
Not included in GWA study
Atmospheric Molar Ratios versus Bottom-up Inventory Emission Ratios
0.0
0.5
1.0
1.5
2.0
C3/C1
nC4/C1
nC4/C3
iC5/C
3
nC5/C3
iC5/n
C4
nC5/nC4
iC5/n
C5
mol
ar ra
tios
Flashing GWA raw gas WRAP F+V
BAO N&E Mobile Lab
Very good agreement.
Vertical bars show :•min and max values for flashing and fugitive emissions
•2 sigma for observed atmospheric molar ratios
C1: methane, C3: propane, nC4: n-butane, i or nC5: i or n-pentane, C2: ethane*
0
50
100
150
200
C1 C2 C3 iC4 nC4 iC5 nC5
Gg/
yr
Bottom-upBAOMobile Lab
Emission estimates comparisonThe bottom-up propane source estimate is used to derive top-down* emissions for all other species based on observed atmospheric ratios
~ 50% discrepancy between bottom-up and top-down estimates for methane.
Error bars show min/max range
Methane source= 1.6 to 6% of Weld County natural gas production in 2008
Used to derive top-down* estimates for other measured species
Fraction of total natural gas withdrawal vented to the atmosphere
These estimates are still highly uncertain.
Percentage0 1 2 3 4 5 6
2010 EPA
2011 EPA
Conventional gas
Shale gas
Weld Co/BAO
Weld Co/Mobile Lab
Conventional gas and Shale gas estimates from Howarthet al. [2011]:% of methane produced over the lifecycle of a well.
US average(no error bars)
Estimates for Weld County
Conclusions• High-quality multi-species
measurements from the NOAA GMD Tower network provide unique information on regional sources of GHG and air pollutants.
• Oil and Gas operations in the Northern Front Range have a regional impact on air composition.
• Bottom-up emission inventories for oil and gas operations are quite uncertain.
• This was a study of opportunity to look at methane variability in the region.
• Next possible steps: – Aircraft and Mobile Lab
mapping of the region to derive absolute fluxes
– Active chemistry study
Natural Gas Production in the US since 1936
0
5
10
15
20
25
30
1930 1940 1950 1960 1970 1980 1990 2000 2010Year
Trill
ion
cubi
c fe
et
from Gas Wells from Oil Wellsfrom Shale Gas from Coalbed WellsGross Withdrawal
Alaska13%
Texas29%
Louisiana6%Gulf of Mexico
9%
New Mexico6%
Oklahoma7%
Wyoming10%
Colorado6%
Other states14%
2009 natural gas production in the US = 19.5% of world production2010 US Total Production = 26.8 Trillion cubic feet (Tcf)
From shale gas=3.4 From Coalbed= 2.02010 US Consumption = 24.1 Tcf
2009 breakdown of production by state
Natural gas gross withdrawal by state
0
5
10
15
20
25
30
1967 1977 1987 1997 2007Year
Trill
ion
Cub
ic F
eet
Total Other States AlaskaTexas LouisianaGulf of Mexico New MexicoOklahoma Wyoming
Alaska13%
Texas29%
Louisiana6%Gulf of Mexico
9%
New Mexico6%
Oklahoma7%
Wyoming10%
Colorado6%
Other states14%
2009 Breakdown
Other states…
0
1
2
3
4
5
6
1967 1977 1987 1997 2007
Year
Trill
ion
Cub
ic F
eet
Colorado Alabama
Arizona Arkansas
California Florida
Illinois Indiana
Kansas Kentucky
Maryland Michigan
Mississippi Missouri
Montana Nebraska
Nevada New York
North Dakota Ohio
Oregon Pennsylvania
South Dakota Tennessee
Utah Virginia
West Virginia