November 2018
Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990-2017:
Other Updates Under Consideration
In supporting documentation associated with the development of EPA's 2018 Inventory of U.S. Greenhouse Gas
Emissions and Sinks (GHGI), EPA stated plans to consider improvements to the 2019 GHGI including incorporating
newly reported data from EPA's Greenhouse Gas Reporting Program (GHGRP) and assessing options to update
activity data sources and/or methodologies related to onshore oil and gas wells.
In June 2018, EPA released two memos detailing preliminary considerations for such improvements:
•	Inventory of U.S. GHG Emissions and Sinks 1990-2017: Updates Under Consideration for Incorporating
GHGRP Data (June 2018 GHGRP-based Updates memo)1
•	Inventory of U.S. GHG Emissions and Sinks 1990-2017: Updates Under Consideration for Well-Related
Activity Data (June 2018 Well-related Activity Updates memo)2
EPA reviewed stakeholder feedback on the June 2018 memos and held two public meetings (June 6, 2018 webinar
and October 17, 2018 workshop). EPA released two memos in October 2018 that focused on specific GHGRP-
based updates:
•	Inventory of U.S. GHG Emissions and Sinks 1990-2017: Updates Under Consideration for Natural Gas
Gathering and Boosting Emissions (October 2018 G&B Updates memo)3
•	Inventory of U.S. GHG Emissions and Sinks 1990-2017: Updates Under Consideration for Liquefied Natural
Gas Segment Emissions (October 2018 LNG Updates memo)4
This memo summarizes previous analyses from the June 2018 memos and includes additional analyses and
considerations toward updating the emissions and/or activity data for the sources listed in in Table 1 below. This
memo also introduces a new topic under consideration for potential GHGI improvements—accounting for
anomalous leak events that are not likely reflected in current GHGI emission factors (EFs). EPA continues to seek
stakeholder feedback on how to use available data and improve methodologies for the sources discussed in this
memo; refer to Section 4 for specific questions.
Table 1. Topics Covered in this Memo
Topic
Discussed in
Memo Section...
New Information since June 2018 Preliminary
Updates Memos
Incorporating Available GHGRP Data
1

Incorporating GHGRP data for hydraulically
fractured (HF) oil well completions and workovers
1.1
• Summary of stakeholder feedback (Section
1.1.4)
Incorporating GHGRP data for flaring N2O emissions
1.2
• None
Incorporating GHGRP data for transmission pipeline
blowdowns
1.3
•	Analysis of available data (Section 1.3.2)
•	Time series considerations (Section 1.3.3)
•	Preliminary national emissions estimates
(Section 1.3.4, Table 7)
•	Summary of stakeholder feedback
Updating Well-related Activity Data
2

Updating well drilling activity estimates
2.1
• Summary of stakeholder feedback
1	https://www.epa.gov/sites/production/files/2018-06/documents/2019-ghgi-updates-incorporating-ghgrp-data_2018-06-08.pdf
2	https://www.epa.gov/sites/production/files/2018-06/documents/2019-ghgi-updates-well-activity_2018-06-08.pdf
3	https://www.epa.gov/sites/production/files/2018-10/documents/ghgi_2018stakeholders_boosting.pdf
4	https://www.epa.gov/sites/production/files/2018-10/documents/ghgi_2018stakeholders_segment.pdf
Page 1 of 19

-------
November 2018
Updating well completions and workovers activity
estimates
2.2
•	Analysis of reported gas well event GHGRP
data compared to GHGI (Table 8)
•	Summary of stakeholder feedback
Defining oil versus gas well types
2.3
•	Analysis of alternative definition (Figure 1)
•	Summary of stakeholder feedback
Updating heavy versus light crude equipment
service assumptions
2.4
•	Analysis of available EIA data
•	Summary of stakeholder feedback
Identifying HF wells
2.5
None
Anomalous Leak Event Emissions
3
New topic
1 Incorporating Available GHGRP Data
Subpart W of the EPA's GHGRP collects annual activity and emissions data on numerous sources from onshore
natural gas and petroleum systems that meet a reporting threshold of 25,000 metric tons of C02 equivalent (mt
C02e) emissions. Facilities that meet the subpart W reporting threshold have been reporting since reporting year
(RY) 2011; however, HF oil well completions and workover data elements and transmission pipeline blowdowns
were first required to be reported in RY2016. Subpart W activity and emissions data are currently used in the
GHGI to calculate CH4 and C02 emissions for many production, processing, and transmission and storage sources.
GHGRP subparts W and Y (petroleum refining) include reporting of N20 from flaring. The GHGRP calculation
methodologies specify that subpart W reporters must calculate N20 emissions from flares using an EF of 0.0001 kg
N20 per million BTU, and subpart Y reporters using an EF of 0.0003 kg N20 per million BTU. N20 emissions are
also reported to GHGRP for engine exhaust and other combustion sources, combustion emissions from which are
generally included within GHGI estimates from fuel combustion, separate from natural gas and petroleum
systems.
The GHGRP data used in the analyses discussed in this memo are those reported to the EPA as of August 5, 2017.
Appendix A documents the subpart W calculation methodologies for certain sources discussed in this memo. EPA
will assess data for RY2017 as they become available. Stakeholders have suggested additional or alternate uses of
GHGRP data, such as for certain sources using measurement data only. Stakeholders have also suggested
modifications to the reported GHGRP data for use in the GHGI, such as through removal of stakeholder-identified
outliers. In the current GHGI, EPA uses the publicly available GHGRP data set without modification for the GHGI,
to ensure transparency and reproducibility of GHGI estimates. Prior to public release of the GHGRP data, the EPA
has a multi-step data verification process for the data, including automatic checks during data-entry, statistical
analyses on completed reports, and staff review of the reported data. Based on the results of the verification
process, the EPA follows up with facilities to resolve identified potential issues before public release.
1.1 Incorporating GHGRP Data for HF Oil Well Completions and Workovers
As discussed above, HF oil well completions and workovers were newly required to be reported in RY2016. EPA
analyzed the RY2016 subpart W data for this source as an initial step for considering potential updates to the
existing GHGI methodology.
1.1.1 Current GHGI Methodology
In the current GHGI methodology for HF oil well completions, controlled and uncontrolled CH4 EFs were
developed using data analyzed for the 2015 NSPS OOOOa proposal. The current GHGI estimates C02 emissions
using C02 EFs developed by applying a default production segment ratio of C02-to-CH4 gas content. As such, this
approach for does not fully account for C02 emissions from flaring.
The 2018 GHGI activity data time series (counts of HF oil well completions, which is also referenced in calculating
non-HF oil well completions), was developed from analyzing Drillinglnfo data on well-level dates of completion or
first reported production. The existing GHGI methodology also includes assumptions to develop activity factors
Page 2 of 19

-------
November 2018
(AFs) for apportioning total counts into control categories. In 2008, Colorado and Wyoming adopted regulations
that require reduced emission completions (RECs); the current GHGI assumes that 7% of completions are RECs
with 95% control efficiency, from 2008 forward.
For workovers, the current GHGI methodology estimates emissions from all oil well workovers without
distinguishing HF from non-HF, using an EF developed for conventional wells and an assumption that 7.5% of all
oil wells are worked over in each year.
1.1.2 Analysis of Available Data
EPA analyzed the RY2016 subpart W data for HF oil well completions and workovers to consider updating the
existing GHGI methodology, which estimates emissions from HF oil well completions based on historical
rulemaking data and does not include a specific emissions estimate for HF oil well workovers (as discussed in
Section 4.1). The new subpart W data allow development of separate GHGI emissions estimates for HF
completions and workovers, in parallel control categories that exist for HF gas well events (reflecting
combinations of REC use, venting, and flaring).5
Additionally, as summarized in Section 4.1, the current GHGI HF oil well completion C02 EF is calculated by
applying an associated gas C02-to-CH4 content ratio, which does not account for C02 conversion during
hydrocarbon combustion. This current methodological limitation would be obviated by using subpart W data to
directly calculate CH4 and C02 EFs, parallel to the current methodology for HF gas well events.
This section documents development of EFs and activity data for HF oil well completions and workovers according
to the general methodology used in the current GHGI for HF gas well completions and workovers. The 2018 Well-
related Activity Data companion memo details considerations for potentially improving the approach to
estimating national total activity data for all completions and workovers (e.g., Drillinglnfo query methodology,
workover rate assumptions).
Table 2 below shows EFs calculated using RY2016 subpart W data for HF oil well completions and workovers for
each event type/control category, compared to current GHGI EFs.
Table 3 shows AFs for each event type/control category.
Table 2. Emission Factors Calculated from Subpart W Compared to Current GHGI, for Year 2016
Event Type
Control
Category
CH4 EF (mt/event)
CO; EF (mt/event)
2018 GHGI
Subpart W
2018 GHGI
Subpart W
Non-REC
Vent
6.76
36.0
0.38
0.8
Flare
1.1
248.8
REC
Vent
0.34
1.3
0.02
0.1
Flare
2.6
287.1
5 The GHGI methodology for HF gas well completions and workovers incorporates GHGRP data. For HF gas well completions and workovers,
EFs are developed from reporting year-specific GHGRP subpart W data (2011 through 2016), with year 2011 EFs applied for earlier time
series years. The EFs are developed for four control categories: non-REC/vented; non-REC/flared; REC/vented; and REC/flared. The total
counts of HF completions are developed from Drillinglnfo data for years prior to 2011, and GHGRP data are used for year 2011 forward (as
the directly reported counts are higher than Drillinglnfo-based estimates). The counts are apportioned into control categories based on
year-specific GHGRP data for 2011-2016; for years 1990-2000, it is assumed all events are non-REC, and 10% of events flare; interpolation
is used to develop AFs in intermediate years. For HF gas well workovers, it is assumed that 1% of the count of existing HF gas wells in a
given year (estimated from analyzing Drillinglnfo data) are worked over.
Page 3 of 19

-------
November 2018
Table 3. Activity Factors Calculated from Subpart W Compared to Current GHGI, for Year 2016
Event Type
Control
Category
HF Completions
HF Workovers
Subpart W
2018 GHGIJ
Subpart W
# of Events
% of total
# of Events
% of total
# of Events
% of total
Non-REC
Vent
111
3%
11,567
93%
35
11%
Flare
542
13%
16
5%
REC
Vent
1,345
33%
871
7%
186
56%
Flare
2,061
51%
93
28%
Total

4,059
100%
12,438
100%
330
100%
a - For years 2008 forward, the current GHGI assumes 7% of HF oil well completions are controlled via REC due to state-
specific regulations. The current GHGI does not include specific estimates for HF oil well workovers.
To develop national total activity data for HF oil well completions, EPA analyzed counts derived from the
Drillinglnfo data set compared to reported counts. For HF gas well completions, counts reported under GHGRP
exceed Drillinglnfo-based estimates, so are assumed to represent national coverage and used directly as national
total activity in the GHGI. For HF oil well completions, this is not the case; Drillinglnfo-based counts exceed
reported counts. Therefore, to develop the preliminary national emissions estimates presented in Section 1.1.4,
Drillinglnfo-based activity data are used in conjunction with the EFs and AFs in Table 2 and
Table 3, respectively.
Workover data are not contained within EPA's Drillinglnfo analysis data set, so an assumption of 1% annual
workover rate is applied for HF gas wells in the current GHGI. In each year of the time series, 1% of existing HF
wells (estimated from the Drillinglnfo data set) are assumed to undergo workovers. For HF gas wells, this
approach results in national total activity data that exceed HF workover counts reported under subpart W. For the
preliminary national emissions estimates presented in Section 4.4, EPA applies the same assumption to HF oil
wells to calculate national total workover activity. Similar to HF gas wells, this approach results in national total
activity data that exceed HF oil well workover counts reported under subpart W.
As stated above, the 2018 Well-related Activity Data companion memo details considerations for potentially
improving the approach to estimating national total activity data for all completions and workovers in the GHGI,
which might include refining the Drillinglnfo query methodology and/or further incorporating subpart W data. For
example, the 2018 Well-related Activity Data memo estimates that within the RY2015-2016 subpart W data for
gas wells, an overall workover rate is 5-6% in recent years (compared to the current GHGI assumption of 4.35% for
non-HF gas wells and 1% for HF gas wells).
1.1.3 Regional Variability and Time Series Considerations
For HF oil well completions and workovers, this memo presents preliminary emissions estimates (see Section 4.4)
according to the existing GHGI methodology to develop estimates for HF gas well events; EFs and AFs are
calculated at the national level. EPA seeks stakeholder feedback on whether a region-specific approach should be
considered for these sources.
To develop the time series AFs for HF oil well completions and workovers based generally on the existing
methodology for gas well events, and incorporating current control assumptions for HF oil well events, the
following assumptions could be applied:
•	For years 1990-2007, all completions and workovers are non-REC, and 10% of events flare.
•	For the first year in which subpart W data are available, 2016, control fractions across the four categories
are developed directly from reported subpart W data.
Page 4 of 19

-------
November 2018
• For intermediate years, 2008-2015, control fractions are developed through linear interpolation.
This produces AFs across the time series that are generally consistent with the existing GHGI assumption that oil
well RECs are introduced beginning in year 2008, during which 7% of completions and workovers are REC, and
10% of both REC and non-REC events flare. EPA seeks feedback on the assumptions above used to develop these
control category AFs.
To apply EFs across the time series, EPA would apply year-specific EFs for GHGRP years, and EFs from the earliest
GHGRP year to all prior years, consistent with the approach for HF gas well events. For the 2019 GHGI, this
approach means that EFs calculated from RY2016 data would be applied for years 1990-2016, and RY2017 data
would be used to develop EFs for year 2017.
1.1.4 Preliminary National Emissions Estimates
Table 4 below shows national total activity data and CH4 emissions for select time series years based on the
updates under consideration described above.
Table 4. Preliminary National Activity and Emissions Estimates for HF Oil Well Completions and Workovers,
Select Years
Data Element
1990
2000
2005
2010
2015
2016
HF oil well completions (#)
3,075
2,246
4,594
8,188
12,438
12,438
Non-REC/Vent (%)
90%
90%
90%
61%
12%
3%
Non-REC/Flare (%)
10%
10%
10%
11%
13%
13%
REC/Vent(%)
0%
0%
0%
11%
29%
33%
REC/Flare (%)
0%
0%
0%
17%
45%
51%
HF oil well workovers (#)
846
848
947
1,235
1,916
1,884
Non-REC/Vent (%)
90%
90%
90%
64%
19%
11%
Non-REC/Flare (%)
10%
10%
10%
8%
5%
5%
REC/Vent(%)
0%
0%
0%
19%
50%
56%
REC/Flare (%)
0%
0%
0%
9%
25%
28%
Total CH4 emissions (kt)
128
101
180
222
95
46
2018 GHGI CH4 emissions (kt)a
21
15
31
52
79
79
Total CO2 emissions (kt)
100
79
142
688
2,179
2,402
2018 GHGI CO2 emissions (kt)a
1
1
2
3
4
4
a - Does not include estimate for workovers. The 2018 GHGI does not specifically estimate emissions from HF oil well workovers; the
estimate for all (non-HF and HF) oil well workovers is negligible compared to the magnitude of other estimates shown in this table (<0.1 kt
across the time series).
Stakeholder feedback generally supported incorporating GHGRP data for this source, using an approach
consistent with that currently used for HF gas well events. Stakeholders expressed interest in developing separate
EFs for completions versus workovers (in the current approach for both gas and oil well events, both types of
events are combined). EPA reviewed available data and notes that due to the approach wherein year-specific EFs
are calculated for each of four control categories, the number of data points in some categories is relatively small.
Therefore, EPA intends to implement the existing approach in the 2019 GHGI, wherein HF oil well completions
and workovers data are combined to calculate EFs (while activity for each event type is separately estimated).
1.2 Flaring N2O Emissions Updates Under Consideration
The current GHGI does not estimate N20 emissions for natural gas and petroleum systems. However, with recent
updates that use GHGRP data to estimate CH4 and C02 flaring emissions, the EPA is considering updates to
incorporate N20 emissions for the same flaring sources. The EPA would apply the existing source-specific
methodology for using GHGRP CH4 data to develop N20 EFs.
Page 5 of 19

-------
November 2018
For purposes of presenting preliminary national total flaring N20 emission estimates, EPA calculated a ratio of the
GHGRP reported N20 emissions to C02 emissions and then multiplied the N20-to-C02 ratio by the 2018 GHGI C02
emissions, for each emission source. Table 5 presents reported GHGRP N20 and C02 flaring emissions, the
calculated N20-to-C02 ratio, 2018 GHGI C02 emissions, and the resulting scaled N20 emissions, for RY2016. This
table focuses on sources that currently use a GHGRP-based methodology in the GHGI, but also includes reference
GHGRP data for sources in this memo where updates are being considered.
Table 5. Preliminary National N20 Emissions Estimates for Flaring Sources in Natural Gas and Petroleum
Systems, Year 2016


GHGRP Flaring


Estimated
Emission Source
GHGRP N20
C02
Ratio of
2018 GHGI
National
(as reported)3
(as reported)
N20:C02
CO2
Total N2O

(mt)
(mt)
(xl00,000)
(mt)
(mt)
Natural Gas & Petroleum Production
Tank Flaring
9.3
4,966,089
-
8,510,234
16.7
NG: Large Condensate Tanks w/Flares
1.0
1,063,935
0.1
1,172,292
1.0
NG: Small Condensate Tanks w/Flares
+
31,800
0.1
35,039
+
Petro: Large Oil Tanks w/Flares
8.2
3,859,139
0.2
7,281,742
15.6
Petro: Small Oil Tanks w/Flares
+
11,215
0.1
21,161
+
Associated Gas
21.6
7,312,187
-
9,102,967
26.9
Petro: Associated Gas Flaring
21.6
7,312,187
0.3
9,102,967
26.9
NG: Flared Gas Well Completions and
Workovers
2.1
135,343
-
186,054
2.3
HF Completions - Non-REC with Flaring
+
8,872
0.2
8,710
+
HF Completions - REC with Flaring
2.1
110,800
1.9
110,998
2.1
Non-HF Completions - flared
+
1,876
0.2
16,407
+
HF Workovers - Non-REC with Flaring
+
279
0.4
10,669
+
HF Workovers - REC with Flaring
+
1,582
0.2
33,436
0.1
Non-HF Workovers - flared
0
11,933
0
5,836
0
Petro: Flared Oil Well HF Completions
and Workovers
18.2
757,150
-
4,382
+
HF Completions - Non-REC with Flaring
0.3
136,782
0.2
4,365a
+
HF Completions - REC with Flaring
17.9
618,126
2.9
16a
+
HF Workovers - Non-REC with Flaring
+
2,024
0.1
NEa
+
HF Workovers - REC with Flaring
0
218
0
NEa
0
Miscellaneous Production Flaring
7.7
2,633,587
-
3,583,254
10.4
NG
3.3
991,718
0.3
1,128,617
3.8
Petro
4.4
1,641,869
0.3
2,454,637
6.6
Well Testing
+
13,800
-
34,803
0.1
NG
0
220
0
323
0
Petro
+
13,580
0.2
34,481
0.1
Gathering and Boosting
25.9
5,930,105
-
225,373
1.0
Gathering and Boosting Stations
25.9
5,930,105b
0.4
225,373ab
1.0
Offshore Production
10.9
457,617
-
-
-
Offshore Flaring
10.9
457,617
2.4
368,840°
10.9°
Natural Gas Processing
Flare Stacks
10.4
3,621,791
0.3
5,404,328
15.5
Transmission and Storage
Transmission Station Flare Stacks
+
25,116
0.05
88,409
+
Storage Station Flare Stacks
+
2,343
0.2
15,307
+
Page 6 of 19

-------
November 2018


GHGRP Flaring


Estimated
Emission Source
GHGRP N20
C02
Ratio of
2018 GHGI
National
(as reported)3
(as reported)
N20:C02
CO2
Total N2O

(mt)
(mt)
(xl00,000)
(mt)
(mt)
LNG Storage Station Flare Stacks
+
2,506
_d
NE
+d
LNG Import/Export Station Flare Stacks
0.2
97,940
_d
NE
0.2d
Petroleum Refining
Flare Stacks
36.0
3,604,229
1.0
3,604,229
36.0
NE - Not estimated
+ Does not exceed 0.05 mt
a - Current GHGI does not rely on subpart W data for this source, and 2018 GHGI estimated C02 emissions shown in this table do not fully
account for combustion. Using C02 emissions estimates developed under the draft subpart W-based approaches discussed in the June 2018
GHGRP-based Updates memo, national N20 emissions would be approximately 53 mt for flared oil well HF completions and workovers and
28 mt for G&B station flaring.
b - C02 includes vented and fugitive sources, in addition to flared sources.
c - Current GHGI does not rely on subpart W data for this source. As the GHGRP reported C02 emissions exceed the current GHGI estimate,
the as-reported GHGRP N20 emissions are shown.
d - Current GHGI does not estimate flaring C02 from these sources. Therefore, as-reported GHGRP N20 emissions are shown as surrogate
for national estimates.
1.3 Transmission Pipeline Blowdowns Updates Under Consideration
As discussed above, transmission pipeline blowdowns were newly required to be reported in RY2016. EPA
analyzed the RY2016 subpart W data for this source as an initial step for considering potential updates to the
existing GHGI methodology.
1.3.1	Current GHGI Methodology
The current GHGI shows emissions from transmission pipeline blowdowns as "pipeline venting for routine
maintenance and upsets." Emissions are calculated using a CH4 EF from GRI/EPA 1996 and annual transmission
pipeline miles from the U.S. Department of Transportation's Pipeline and Hazardous Materials Safety
Administration (PHMSA). C02 emissions are calculated from the CH4 EF and a default downstream gas profile of
93.4% CH4 and 1.0% C02.
1.3.2	Analysis of Available Data
A subpart W transmission pipeline facility must report blowdown emissions from natural gas transmission
pipelines (as defined at 40 CFR 98.238) that it owns and operates. EPA calculated a transmission pipeline
blowdown EF from the subpart W data by summing the reported emissions and dividing by the reported
transmission pipeline miles. Table 6 shows the calculated subpart W EF compared to the current GHGI EF. Note,
the subpart W RY2016 data reflect approximately 50% of the total transmission pipeline mileage estimated in the
current GHGI for year 2016 (147,000 of 300,000 miles).
Table 6. Emission Factors (mt/pipeline mile) Calculated from Subpart W Compared to Current GHGI, for Year
2016
Data Source
CH4
CO2
2018 GHGI
0.6
0.01
Subpart W
1.2
0.02
EPA compared subpart W reporters of this source to a PHMSA data set that includes transmission pipeline
mileage by company, state and type (i.e., interstate or intrastate). Based on a preliminary analysis, approximately
71% of the national total interstate transmission pipeline miles and 11% of intrastate transmission pipeline miles
were reported to subpart W in RY2016. EPA may further consider this distinction and calculate unique EFs for
interstate and intrastate pipelines, which would be paired with activity data specific to each type.
Page 7 of 19

-------
November 2018
1.3.3	Time Series Considerations
Preliminary stakeholder feedback suggested that the reported subpart W data for this source might not be
representative of national emissions in recent years (e.g., facilities meeting the reporting threshold might
generally have larger-diameter pipelines leading to higher emissions per mile). As discussed above, EPA is
considering whether sufficient data are available to develop separate EFs and accompanying activity data for
interstate versus intrastate pipelines as a potential means to recognize differences in the reporting versus
nonreporting population. Establishing subcategories might offer improved representativeness, but is dependent
on identifying an accurate method for categorizing subpart W data (interstate versus instrastate designation is not
directly reported but might be inferred through analysis of external data sets such as published by PHMSA) and
the statistical characterization of reported data points in each category (e.g., whether the data set size and
variance support establishing separate EFs).
EPA continues to seek stakeholder feedback on how subpart W data for this source might be used to inform time
series trends, including specific new questions in Section 4 regarding the establishment of subcategories.
1.3.4	Preliminary National Emissions Estimates
Table 7 below shows national total emissions for select time series years based on one approach EPA is
considering for the 2019 GHGI public review draft: using existing GHGI activity (pipeline miles) for all years;
existing EFs for years 1990 through 1992; newly calculated EFs from subpart W data starting in year 2016; and
linear interpolation to calculate EFs in intermediate years. EPA notes that multiple approaches are under
consideration and seeks stakeholder feedback on the most appropriate approach.
Note, the current GHGI methodology involves first calculating potential emissions for this source (i.e., EF
multiplied by activity), then subtracting voluntary reductions (i.e., quantity of emissions reduced as reported to
EPA's GasSTAR program) to calculate net emissions. EPA seeks stakeholder feedback on whether and how to
reflect voluntary reductions in the update under consideration (e.g. does the updated approach reflect net
emissions, and therefore an adjustment to include Gas STAR would double count reductions?).
Table 7. Preliminary National Emissions Estimates for Transmission Pipeline Blowdowns, Select Years
Data Element
1990
2000
2005
2010
2015
2016
2018 GHGI
# Pipeline miles
291,925
298,957
300,468
304,803
300,376
300,645
Cm EF (mt/mile)
0.61
0.61
0.61
0.61
0.61
0.61
CO2 EF (mt/mile)
0.018
0.018
0.018
0.018
0.018
0.018
Potential total CH4 emissions (kt)
178
182
183
186
184
183
Net total CH4 emissions (kt)
178
149
58
58
124
124
Total CO2 emissions (kt)
5
5
5
5
5
5
Update under consideration
CH4 EF (mt/mile)
0.61
0.81
0.93
1.05
1.17
1.20
CO2 EF (mt/mile)
0.018
0.020
0.021
0.022
0.024
0.024
Total CH4 emissions (kt)
178
241
279
320
353
360
Total CO2 emissions (kt)
5
6
6
7
7
7
EPA continues to seek stakeholder feedback on how subpart W data for this source might be used to update the
time series estimates in the GHGI, including specific new questions in Section 4 regarding EFs calculated from
subpart W data.
Page 8 of 19

-------
November 2018
2 Updating Well-related Activity Data
Appendix B provides a table that shows all well-related activity data over the time series as used in the 2018
GHGI, as well as data elements from updates under consideration, as discussed below. Note that several existing
GHGI methodologies documented e.g., developing counts of active wells) rely on EPA's analysis of Drillinglnfo's
subscription-based digital Dl Desktop raw data feed6; this data set is referred to throughout this memo as
"Drillinglnfo data."
2,1 Well Drilling
The U.S. Department of Energy's Energy Information Administration (DOE/EIA) no longer maintains the Monthly
Energy Review well drilling activity data set that was used to develop well drilling activity inputs in past GHGIs
(most recent estimates cover through 2010), so the GHGI requires a new data source for the entire time series, or
at least from 2011 forward.
As described in the June 2018 Well-related Activity Updates memo, EPA did not identify another data set
published by EIA that provides complete well drilling activity over the GHGI time series. EPA developed a
methodology for querying the Drillinglnfo data set to develop estimates of wells drilled and presented preliminary
results in the June 2018 Well-related Activity Updates memo. The Drillinglnfo-based approach is:
•	Count all wells drilled in time series year N with:
o Onshore location -and-
o Spud date within year N -or- spud date not reported, but date of first production within year
N+l
•	Apportion counts between oil, gas, and dry production types
o Dry or temporarily inactive (TIA) wells drilled: spud date within year N, but no production is
reported in year N+l
o Gas wells drilled: GOR in year N+l >100 mcf/bbl
o Oil wells drilled: GOR in year N+l < 100 mcf/bbl
•	Apportion dry/TIA counts to natural gas and petroleum systems according to the year-specific split
between gas wells drilled and oil wells drilled.
Stakeholder feedback on the Drillinglnfo-based approach noted that preliminary estimates appeared reasonable
for incorporation into the GHGI.
For a Drillinglnfo-based methodology, EPA sought feedback on a few additional considerations. First, whether this
approach adequately accounts for dry wells (which are spud but do not achieve reportable production levels).
Based on preliminary analysis of results, this approach does appear to sufficiently represent dry wells at
comparable levels to the EIA data set; on average, dry wells (not including TIA wells which report production after
year N+l) contribute 10% of total wells drilled. Second, whether this approach is overly inclusive of wells that may
not be drilled for oil and gas production purposes but are present in the Drillinglnfo data set. Second, how to
account for time series coverage issues. For states without recently released data, EPA plans to develop a
surrogate methodology wherein an early year's data are assigned to recent years to fill state-level data gaps,
similar to the existing approach for counts of active wells. Additionally, total wells drilled in most recent time
series year cannot be fully estimated by the current approach (i.e., GOR in year N+l is not available); EPA plans to
implement a surrogate approach of using the previous year's estimate.
Appendix B shows well drilling counts across the time series in the 2018 GHGI and preliminary estimates by the
Drillinglnfo data analysis approach described above.
6 https://info.drillinginfo.com/products/di-plus/
Page 9 of 19

-------
November 2018
2.2 Well Completions and Workovers
As described above, the current methodology for well completion and workover event counts involves a mix of
Drillinglnfo data analyses, GHGRP data analyses, and historical assumptions.
EPA has conducted a preliminary analysis to assess how completion counts and workover rates reflected in the
subpart W data compare to current GHGI assumptions. Regarding workover rates, the subpart W data reporting
structure and requirements limit the level of detail for such an analysis, due to: (1) there is not a specific reporting
element indicating whether reported wells are HF or non-HF; and (2) non-HF oil well workovers are not reported.
EPA therefore analyzed the number of completion and workover events and the overall gas well workover rate, as
summarized in Table 8.
In response to stakeholder feedback, EPA developed the following considerations based on this analysis and the
estimation that approximately 70% of gas wells in the U.S. are covered by subpart W reporting:
•	Non-HF gas well completion counts might be over-estimated in recent years of the GHGI.
•	Non-HF gas well workover counts might be under-estimated in recent years of the GHGI.
•	HF gas well workover counts might be over-estimated in recent years of the GHGI.
Table 8. Comparison of Gas Well Event Activity Data and Calculated Workover Rate
Activity Data Element
Current GHGI
GHGRP As-reported
Basis
2015
2016
2015
2016
# Non-HF gas well completions
Scaled from 400 events/year in 1992
786
770
108
88
# Non-HF gas well workovers
4.35% of active non-HF gas wells
7,549
7,315
18,031
14,957
# HF gas well workovers
1% of active HF gas wells
2,521
2,487
263
103
Overall gas well workover rate
4.35% of active non-HF gas wells;
1% of active HF gas wells
6%a
5%a
a - Calculated using number of reported gas well workover events (with and without HF) divided by the number of producing wells
at the end of the calendar year (reported under 40 CFR 98.236(aa), found in the "sub-basin characterization" table).
Appendix B shows all categories of completion and workover counts across the entire time series in the 2018
GHGI. EPA seeks stakeholder feedback on improving the current methodology in the 2019 GHGI; see Section 4 for
specific stakeholder feedback requests.
2.3 Definition of Oil versus Gas Well
The current methodology estimates the count of active gas wells in a given year as all wells in the Drillinglnfo data
set with a GOR > 100 mcf/bbl in that year, and active oil wells as those with GOR < 100 mcf/bbl. By this definition,
oil wells include associated gas wells.
Other data sets (e.g., those published by EIA) use different GOR thresholds for defining oil versus gas wells and
might have different underlying assumptions regarding whether associated gas wells are a subset of oil wells. A
value of 6 mcf/bbl is another common definition threshold, based on the oil and gas energy equivalence factor (6
mcf gas provides roughly the same amount of energy as 1 bbl oil equivalent (BOE); BOE is commonly used in
financial statements to combine oil and gas production into a single measure).
EPA reviewed available data reported under GHGRP subpart W to evaluate how reported GOR values compare to
the current GHGI methodology and consider whether the current production type delineation threshold of 100
mcf/bbl is appropriate. Per subpart W, oil wells are defined as producing from an oil formation, not defined by a
specific GOR threshold.
Table 9 below summarizes reported GOR data based on EPA's review of subpart W RY2015-2016 summary data
(Envirofacts table EF_W_FACILITY_OVERVIEW, containing data for over 400,000 wells reported to the EPA as of
Page 10 of 19

-------
November 2018
August 5, 2017). For most oil wells reported under subpart W (73%), the sub-basin level average GOR falls within
the current GHGI definition (<100 mcf/bbl); while a significant fraction (27%) have higher average GORs. At a
lower delineation threshold (e.g., 6 mcf/bbl), an even higher fraction of subpart W oil wells (roughly 50%) would
be considered gas wells. Therefore, based on this analysis, EPA finds support for the current approach to
delineating oil versus gas wells, but seeks stakeholder feedback (see Section 4 for specific requests on this topic).
Table 9. 2018 GHGI Well-Related Activity Data Summary

# Oil Wells with Specified GOR (mcf/bbl)
Reporting Year
GOR <10
10< GOR <100
100< GOR <1,000
GOR >1,000
2015
123,446
28,104
11,674
48,091
[58%]
[13%]
[6%]
[23%]
2016
117,538
33,346
21,068
32,781
[57%]
[16%]
[10%]
[16%]
Combined
240,984
61,450
32,742
80,872
[58%]
[15%]
[8%]
[19%]
In response to recent stakeholder feedback, EPA calculated oil and gas well counts across the time series by two
different GOR delineation thresholds (100 mcf/bbl and 6 mcf/bbl). The results of this sensitivity analysis are
shown in Figure 1.
100 mcf/bbl: Active - Gas
100 mcf/bbl: Active - Oil
—	— — 6 mcf/bbl: Active - Gas
—	— — 6 mcf/bbl: Active - Oil
700,000
200,000
100,000
600,000
577,515
534,302
-j
500,000
400,000
300,000
462,218
Figure 1. Active Oil and Gas Well Counts by Two Different GOR Delineation Thresholds
2.4 Heavy versus Light Crude Equipment Service
Heavy crude is defined as oil with lower than 20° API gravity. The GHGI uses separate EFs and activity data for
wellheads, separators, and headers in heavy versus light crude service. Currently, the total counts of wellheads
and headers are split into heavy versus light crude categories using an assumed split between heavy crude wells
Page 11 of 19

-------
November 2018
(7.05% of all oil wells) and light crude wells (92.95%). As documented in the 1999 EPA/ICF report7, this
assumption, and its extension to wellhead and header activity, was developed for a 1995 base year; this split is
currently applied to calculate equipment counts in all time series years. The split between heavy and light crude
separators in base year 1995 is also documented in the 1999 EPA/ICF report: EPA assumed 90.1% of separators
are in light crude service, and 9.9% in heavy crude service, and applied this split to all time series years. EPA has
identified multiple data sources that might facilitate improvement to the current methodology by allowing
calculation of the heavy versus light crude equipment splits in recent years.
GHGRP subpart W collects average API gravity data associated with production in all oil sub-basins. Based on an
analysis of RY2015 and RY2016 data, 19% and 18%, respectively, of oil wells reporting to GHGRP produce heavy
crude. This value is higher than the current assumption of 7.05% of wells producing heavy crude (and subsequent
assumption that 7.05% of wellheads and headers are in heavy crude service). To follow the existing methodology
which assumes that per well equipment counts are the same for heavy crude and light crude wells (with the
exception of separators, where on average over the time series each heavy crude well has 0.47 separators while
each light crude well has 0.32 separators), EPA might analyze subpart W data specifically for facilities that produce
heavy crude versus light crude. For this approach, however, only a subset of onshore production facility data can
be analyzed—those with either all heavy crude sub-basin formation types or all light crude sub-basin formation
types—since equipment counts (e.g., separators) are reported at a basin level. Table 10 summarizes the data
availability and preliminary estimates of separator activity factors based on this approach. EPA might use subpart
W data to update the equipment count splits in recent years and reflect updated industry trends.
Table 10. Subpart W Equipment Counts3

Count of


Separators/
Number of Data


Separators
Count of Wellheads
Wellhead
Points (Facilities)

Data Set
2015
2016
2015
2016
2015
2016
2015
2016
Notes/Methodology
All onshore
oil prod








Counts from records classified
76,690
86,890
213,380
221,326
0.36
0.39
319
315
in Table R.4 as "Crude oil








production equipment"15









Counts from records in Table
Heavy








R.4 - from facilities that
crude-only
1,818
345
40,894
40,063
0.04
0.01
14
14
produce only heavy crude (all
facilities








sub-basins are oil with API
gravity <20 in Table AA.l.ii)









Counts from records in Table
Light crude-








R.4 - from facilities that
only
22,153
23,048
54,098
51,428
0.41
0.45
102
103
produce only light crude (all
facilities








sub-basins are oil with API
gravity >20 in Table AA.l.ii)
a - Data reported as of August 5, 2017.
b - For this approach, data from all facilities reporting presence of crude oil production equipment for equipment leak
calculations can be used (ignoring the reported sub-basin formation type(s)).
EPA also reviewed the methodology documented in the 1999 Radian report which was the basis for the 1999
EPA/ICF report estimates. The 1999 Radian report methodology analyzed state-level reported heavy oil
production as a fraction of total oil production, then applied that fraction to state-level oil well counts to estimate
heavy oil well counts in each state, and finally summed heavy oil well counts to estimate the national population
fraction. This approach does not facilitate development of a heavy versus light split for equipment other than
wellheads (e.g., a specific split for separators as in the current methodology); additionally, inherent in this
7 Estimates of Methane Emissions from the U.S. Oil Industry (Draft Report). Prepared by ICF International. Office of Air and Radiation, U.S.
Environmental Protection Agency. October 1999.
Page 12 of 19

-------
November 2018
approach is an assumption that heavy and light crude wellheads have the same average production rates. EPA
recently reviewed EIA state-level production and API gravity data and estimated that 4-5% of crude produced in
the lower 48 states in recent years is heavy crude. Due to the assumptions and limitations of this analysis (i.e.,
using production split between heavy and light as surrogate for wellhead split), EPA is focusing further efforts
toward how subpart W data, such as that summarized in Table 10, might be used to update the GHGI for recent
time series years.
EPA seeks stakeholder feedback on the data sources and approaches described above, or other methodologies to
consider for improving this aspect of the oil production segment major equipment activity estimates; see Section
4 for specific stakeholder feedback requests on this topic.
2,5 Identification of HF Wells
Appendix B shows current GHGI estimates of HF gas and oil well counts. There are limited public data estimating
national total counts comparable to the GHGI; Appendix B shows year 2016 estimates from ElA's Today in Energy
website.
EPA seeks stakeholder feedback on data sources and methodologies that might be used to update the current
approach for identifying HF gas and oil wells within the Drillinglnfo data set. EPA might retain the current
assumption that all horizontally drilled wells are hydraulically fractured but update the methodology for
identifying wells that do not report horizontal drill type but would be expected to be hydraulically fractured based
on location in an unconventional formation. EPA is considering reviewing subpart W sub-basin-level (county-level)
data to assess whether a crosswalk of location and HF indication might be constructed from or verified using
reporting data, in order to estimate total national HF well counts (at least for recent time series years).
si Anomalous Leak Events
In recent GHGIs, EPA incorporated an emissions estimate for the Aliso Canyon gas leak during years 2015 and
2016 in the storage well category. EPA used the California Air Resources Board (CARB) published estimate of the
methane release from the leak.8
EPA seeks stakeholder feedback on existing data sources or suggested methodologies for identifying similar
events across natural gas and petroleum systems, with emissions beyond what is likely accounted for in GHGI EFs.
4 Requests fc keholder Feedback
HF Oil Well Completions and Workovers (Section 1.1)
1.	EPA seeks feedback on the national representativeness of subpart W-based HF oil well completion and
workover emissions factors (emissions per event) and activity factors (i.e., allocation of total event counts
across four control categories).
2.	EPA seeks feedback on how to consider regional and temporal variability for HF oil well completions and
workovers.
3.	EPA seeks stakeholder feedback on the methodology and assumptions for allocating events into the four
control categories across the time series (i.e., control category AFs, as detailed in Section 4.2). Specifically,
for years 1990-2007, it is assumed all events are non-REC, and 10% of events flare; in contrast, the GHGI
methodology for HF gas well event AFs assumes that RECs are introduced earlier, in year 2000.
4.	Historical analyses for HF gas well events data (RY2011-2015) included all HF well event data reported,
and therefore might have included reported data from HF oil well events if any reporters reported data
from these activities in those years. Should EPA revisit these historical EFs (e.g., discard from the EF data
8 https://www.epa.gov/sites/production/files/2017-04/documents/2017_aliso_canyon_estimate.pdf
Page 13 of 19

-------
November 2018
set any events seemingly conducted at oil wells? develop factors specific to oil well events prior to
RY2016?)?
N2Q Emissions (Section 1.2)
5.	EPA seeks feedback on updating the GHGI to include N20 from flaring, based on GHGRP data.
a. Specifically, EPA seeks feedback on time series calculations—such as whether calculated ratios of
N20:C02 from GHGRP from 2016 (or several years') data could be applied to each year of the time
series from 1990-2017.
6.	EPA seeks feedback on other available data sources for N20 emissions.
Transmission Pipeline Blowdowns (Section 1.3)
7.	EPA seeks feedback on the general use of subpart W data to update the current GHGI methodology for
this source.
8.	Are the EFs calculated from RY2016 subpart W data (shown in Table 6) nationally
representative,specifically:
a.	Should EFs calculated from subpart W data be applied to the non-reporting population
(approximately 50% of national total pipeline mileage) or is there a more appropriate approach
for estimating emissions from this population?
b.	Do available subpart W data support development of subcategories that might improve EF
representativeness—for example interstate versus intrastate EFs?
9.	EPA seeks feedback on time series calculations, specifically
a.	Whether GHGI EFs be retained for early time series years or if subpart W EFs should be applied for
all years.
b.	Whether year-specific EF(s) should be developed from subpart W data for RY2016 forward. Based
on preliminary analyses, EPA notes significant variation in calculated EFs in the available years of
subpart W data. Specifically, the average RY2016 CH4 EF was calculated to be 1.2 mt/mile,
compared to 0.6 mt/mile using recently published RY2017 data. EPA seeks stakeholder feedback
on whether these differences reflect industry trends, or if an approach such as combining early
reporting years' data is appropriate.
10.	EPA seeks feedback on whether and how to include GasSTAR reductions in an updated methodology for
this source.
Well Drilling (Section 2.1)
11.	EPA seeks stakeholder feedback on data available to improve annual national activity estimates for well
drilling across the time series. EPA seeks feedback on the most appropriate data source for updating the
GHGI, whether it is highlighted in Section 2.1 (i.e., EIA or Drillinglnfo) or another source that should be
considered.
12.	EPA seeks feedback on how to ensure time series consistency, given the current EIA data source provides
annual activity through 2010. Should a new data source cover only recent years, or be used to entirely
replace current estimates?
13.	As EPA further considers developing annual well drilling activity estimates from Drillinglnfo data, feedback
is sought on the specific methodology and other considerations outlined in Section 2.1. Specifically:
a.	Whether this approach adequately accounts for dry wells (which are spud but do not achieve
reportable production levels). Based on preliminary analysis of results, this approach does appear
to sufficiently represent dry wells at comparable levels to the EIA data set; on average, dry wells
(not including TIA wells which report production after year N+l) contribute 10% of total wells
drilled.
b.	Whether this approach is overly inclusive of wells that may not be drilled for oil and gas
production purposes but are present in the Drillinglnfo data set (e.g., wells drilled specifically for
Page 14 of 19

-------
November 2018
injection). EPA expects a minor contribution from such wells, since estimates developed from the
Drillinglnfo approach under consideration are comparable with EIA estimates for the overlapping
time frame (1990-2010), as shown in Appendix B. EPA estimates that that injection wells might
contribute approximately 2-4% to total counts; it is difficult to identify such wells in the
Drillinglnfo data set due to the presence of hundreds of reported production types. EPA seeks
feedback on how the Drillinglnfo-based methodology might take this issue into account.
c. How to account for time series coverage issues (due to states with lagging reporting and therefore
not fully represented in the Drillinglnfo data set in recent years).
Well Completions and Workovers (Section 2.2)
14.	EPA seeks stakeholder feedback on how to use available data to improve national activity estimates for
well completion and workover events—specifically, how Drillinglnfo and subpart W data sets might be
used in conjunction, or if one data set should be used to develop estimates and the other to verify
estimates.
Definition of Oil vs. Gas Well (Section 2.3)
15.	EPA seeks stakeholder feedback on whether the current methodology for counting and allocating active
well counts between oil and gas should be updated, and if so, how.
Heavy vs. Light Crude Equipment Service (Section 2.4)
16.	Based on RY2015 and RY2016 subpart W data, 19% and 18%, respectively, of oil wells in that data set
produce heavy crude (API gravity less than 20), compared to the current GHGI basis of 7.05%. This
updated heavy crude fraction is based on reported data for approximately 210,000 active oil wells (out of
approximately 580,000 active oil wells nationwide). EPA seeks stakeholder feedback on incorporating this
updated fraction into the GHGI time series.
a.	Should EPA consider developing geographic-specific (e.g., NEMS region-level) estimates of heavy
crude well fractions?
b.	Should EPA retain the estimate of 7.05% of oil wells producing heavy crude (developed for base
year 1995) for early years of the time series, and interpolate to the updated fraction based on
subpart W data? Or is a different approach more appropriate—for example, where the heavy
crude fraction is more tailored to the specific time period, rather than a set or linearly increasing
value? EPA seeks information on data sources that might offer information to implement a more
tailored approach.
17.	How should EPA use API gravity data in conjunction with equipment count data reported under subpart W
to improve oil production segment major equipment activity estimates? For example, Table 10 above
presents activity factors for separators per oil well developed specifically for heavy and light crude
populations.
a.	Should EPA retain the current approach of extending the heavy/light crude well count split to
wellhead and header activity data (for example, 19% of oil wells produce heavy crude, therefore
19% of headers are in heavy crude service)?
b.	Should EPA retain the current approach of developing specific activity factors for separators in
heavy versus light crude service, as shown in Table 10?
Identification of HF Wells (Section 2.5)
18.	EPA seeks stakeholder feedback on whether it is reasonable to retain the current assumption that all wells
with horizontal drill type according to the Drillinglnfo data set are hydraulically fractured, or if there are
recommendations for improving this assumption.
19.	EPA seeks input on publicly available data sources and methodologies that might be used to identify wells
that do not report horizontal drill type in the Drillinglnfo data set but would be expected to be
Page 15 of 19

-------
November 2018
hydraulically fractured based on location in an unconventional formation (i.e., used to create a new
formation type crosswalk).
a. EPA specifically seeks feedback on how GHGRP subpart W data might be used in this step to
construct or verify such a crosswalk.
Anomalous Leak Events (Section 3)
20.	EPA seeks stakeholder feedback on existing data sources or suggested methodologies for identifying
similar events across natural gas and petroleum systems, with emissions beyond what is likely accounted
for in GHGI EFs.
21.	EPA seeks information from stakeholders on any anomalous leak events that have occurred over the
1990-2017 time series that should be included in the 2019 GHGI, in addition to the Aliso Canyon leak in
years 2015 and 2016.
Page 16 of 19

-------
November 2018
Appendix A, GHGRP Measurement Methodologies from Data Sources Considered for Updates
Emission Source
Measurement and/or Calculation Type
# Sources
Location & Representativeness
EF Calculation Method
Oil Well HF Completions and
Workovers
Emissions calculated for each event, based
on (1) measured actual flowback gas
volumes from the well or (2) calculated
flowback gas volume based on well
parameters (e.g., pressure differentials,
temps).
If flared, then flare control efficiency is
applied.
Emissions data (for 2016) are
available for 4,059 completions and
330 workover events at HF oil wells
Facilities in the U.S. that exceed
25,000 mt C02e reporting
threshold.
For this memo, the EPA used
reported data to calculate,
event/control category specific
(e.g., REC, flare), average EFs
Flare Stacks
CH4 and C02 emissions calculated using: (1)
gas volume sent to the flare, (2)
combustion efficiency (from manufacturer
or assume 98%), fraction of feed gas sent to
an un-lit flare, and (3) gas composition for
C02, CH4, and hydrocarbon constituents.
N20 emissions calculated using amount of
fuel combusted, fuel heating value, and
prescribed EF of 1.0 x 10-4 kg N20/mmBtu
Varies by industry segment
Facilities in the U.S. that exceed
25,000 mt C02e reporting
threshold.
For this memo, to estimate
emissions for each source, EPA
calculated a ratio of the GHGRP
reported N20 emissions to C02
emissions and then multiplied the
N20-to-C02 ratio by the 2018 GHGI
C02 emissions
Transmission Blowdown Vent
Stack
Emissions calculated using:
•	Blowdown volumes, number of
blowdowns, and the ideal gas low
modified for compressibility; or
•	Flow meter to measure emissions for all
equipment associated with a blowdown
event.
Blowdown volumes <50 scf are exempt.
Emissions data (for 2016) are
available from 9,093 blowdowns
(which occurred over 147,187
miles).
Facilities in the U.S. that exceed
25,000 mt C02e reporting
threshold.
For this memo, EPA calculated EFs
as a straight average of all available
data.
Page 17 of 19

-------
November 2018
Appendix B. Well-Related Activity Data

1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
Well counts
2018 GHGI
Non-HF gas wells
135,552
141,052
139,823
140,834
142,127
142,245
144,501
144,082
148,078
148,112
144,545
156,621
157,696
162,506
167,595
168,855
175,567
178,970
187,366
186,667
187,098
187,153
182,776
179,300
179,305
173,544
168,151
HF gas wells
62,074
78,538
76,875
82,527
88,242
90,423
99,399
103,820
108,880
113,260
122,364
135,526
142,815
154,003
162,832
179,615
195,683
207,348
227,138
234,576
242,301
248,162
250,614
248,528
252,141
252,107
248,730
Total active gas wells
197,626
219,590
216,698
223,361
230,369
232,668
243,900
247,902
256,958
261,372
266,909
292,147
300,511
316,509
330,427
348,470
371,250
386,318
414,504
421,243
429,399
435,315
433,390
427,828
431,446
425,651
416,881
Non-HF oil wells *
469,317
467,760
454,605
446,499
432,774
423,199
418,579
419,582
401,394
381,938
382,314
379,071
375,274
372,953
372,994
374,960
379,859
380,541
391,513
387,949
389,226
393,598
401,244
404,373
405,284
398,424
373,608
HF oil wells*
84,582
88,843
86,070
87,745
87,088
84,442
86,754
87,650
84,935
82,132
84,785
86,243
86,547
88,223
91,457
94,672
98,627
97,217
112,992
113,657
123,494
135,121
151,260
165,297
184,166
191,593
188,356
Total active oil wells
553,899
556,603
540,675
534,244
519,862
507,641
505,333
507,232
486,329
464,070
467,099
465,314
461,821
461,176
464,451
469,632
478,486
477,758
504,505
501,606
512,720
528,719
552,504
569,670
589,450
590,017
561,964
Total HF wells *
146,656
167,381
162,945
170,272
175,330
174,865
186,153
191,470
193,815
195,392
207,149
221,769
229,362
242,226
254,289
274,287
294,310
304,565
340,130
348,233
365,795
383,283
401,874
413,825
436,307
443,700
437,086
Total active wells
751,525
776,193
757,373
757,605
750,231
740,309
749,233
755,134
743,287
725,442
734,008
757,461
762,332
777,685
794,878
818,102
849,736
864,076
919,009
922,849
942,119
964,034
985,894
997,498
1,020,896
1,015,668
978,845
EIA Today in Energy
Total HF wells
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
670,000
Total active wells
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
NPA
977,000
Drilling event counts
2018 GHGI
Gas wells drilled (incl. fraction
of dry)
15,096
13,066
10,887
13,047
12,232
10,465
11,498
14,473
14,507
14,564
19,863
25,350
20,041
23,582
27,180
31,969
36,536
36,255
35,824
20,266
18,837
18,837
18,837
18,837
18,837
18,837
18,837
Oil wells drilled (incl. fraction
of dry)
17,234
16,796
12,540
11,744
9,587
10,679
11,255
14,105
9,575
5,818
9,424
10,208
7,830
9,251
9,877
12,053
14,893
14,816
18,478
12,537
17,774
17,774
17,774
17,774
17,774
17,774
17,774
Total wells drilled
32,330
29,862
23,427
24,791
21,819
21,144
22,753
28,578
24,082
20,382
29,287
35,558
27,871
32,833
37,057
44,022
51,429
51,071
54,302
32,803
36,611
36,611
36,611
36,611
36,611
36,611
36,611
Preliminary Drillinglnfo Data Analysis (described in Section 2.1)a
Gas
13,417
6,719
7,464
6,794
7,208
7,648
16,608
9,557
9,089
10,507
14,449
17,131
14,390
17,237
20,018
23,060
24,733
24,964
23,600
11,636
12,172
9,206
5,512
4,725
4,385
2,580
1,550
Oil
15,013
11,030
11,854
10,720
9,145
10,210
18,925
12,636
7,746
8,447
10,981
11,232
9,558
11,570
13,498
15,553
16,951
18,820
18,633
12,566
20,697
25,874
29,953
30,422
30,209
14,049
8,246
Dry/TIA
7,971
5,650
5,220
4,641
4,249
3,909
4,228
4,601
3,898
3,601
5,095
6,782
4,875
5,549
6,477
6,872
9,190
7,631
9,232
4,773
5,614
6,271
5,781
5,304
7,037
3,861
1,964
Gas wells drilled (incl. fraction
of dry/TIA)
17,179
8,858
9,481
8,594
9,081
9,322
18,584
11,538
11,193
12,503
17,344
21,227
17,319
20,557
23,886
27,164
30,186
29,315
28,759
13,931
14,251
10,852
6,410
5,438
5,277
3,179
1,861
Oil wells drilled (incl. fraction
of dry/TIA)
19,222
14,541
15,057
13,561
11,521
12,445
21,177
15,256
9,540
10,052
13,181
13,918
11,504
13,799
16,107
18,321
20,688
22,100
22,706
15,044
24,232
30,499
34,836
35,013
36,354
17,311
9,899
Total wells drilled
36,401
23,399
24,538
22,155
20,602
21,767
39,761
26,794
20,733
22,555
30,525
35,145
28,823
34,356
39,993
45,485
50,874
51,415
51,465
28,975
38,483
41,351
41,246
40,451
41,631
20,490
11,760
Completion event counts
2018 GHGI
Gas well non-HF completions
365
405
400
412
425
429
450
458
474
482
493
539
555
584
610
643
685
713
765
778
793
804
800
790
796
786
770
Gas well HF completions
3,769
3,630
2,630
3,425
3,322
3,034
4,057
5,352
4,785
4,583
6,881
8,675
7,536
8,911
10,459
12,866
14,176
14,206
15,223
8,811
8,691
9,749
7,665
7,382
7,141
5,272
3,105
Oil well non-HF completions
9,764
9,644
6,395
5,916
4,742
5,855
6,203
7,671
5,359
3,476
5,844
5,791
4,285
4,618
5,046
6,185
7,369
7,142
8,305
6,117
7,565
3,252
0
0
0
3,315
3,315
Oil well HF completions
3,075
2,944
3,007
2,940
2,606
2,393
2,633
3,535
2,323
1,329
2,246
3,097
2,490
3,511
3,743
4,594
6,016
6,229
8,328
5,073
8,188
12,501
16,335
17,332
19,154
12,438
12,438
Workover event counts
Page 18 of 19

-------
November 2018

1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2018 GHGI
Gas well non-HF workovers
5,897
6,136
6,082
6,126
6,183
6,188
6,286
6,268
6,441
6,443
6,288
6,813
6,860
7,069
7,290
7,345
7,637
7,785
8,150
8,120
8,139
8,141
7,951
7,800
7,800
7,549
7,315
Gas well HF workovers
621
785
769
825
882
904
994
1,038
1,089
1,133
1,224
1,355
1,428
1,540
1,628
1,796
1,957
2,073
2,271
2,346
2,423
2,482
2,506
2,485
2,521
2,521
2,487
Oil well workovers
41,542
41,745
40,551
40,068
38,990
38,073
37,900
38,042
36,475
34,805
35,032
34,899
34,637
34,588
34,834
35,222
35,886
35,832
37,838
37,620
38,454
39,654
41,438
42,725
44,209
44,251
42,147
* Values not published in the 2018 GHGI, but underlie the current estimates of HF oil well completion event counts
N/A - Not applicable
NPA - Not publicly available
NE - Not estimated
a - Results of preliminary approach discussed in Section 2.1. This does not include state-level adjustments to account for lagging reporting that EPA implements for other well-related data elements that are based on Drillinglnfo data. After
adjustments that would be implemented in a final methodology, estimates in recent years (e.g., 2015 forward) will likely increase.
Page 19 of 19

-------