April 2015
Inventory of U.S. Greenhouse Gas Emissions and Sinks: Potential Revisions to
Pneumatic Controller Emissions Estimate (Production Segment)
Overview of Current Inventory Method
EPA calculates emissions for pneumatic controllers in the production segment for natural gas and
petroleum systems using a national estimate for counts of pneumatic controllers, potential emission
factors, and Natural Gas STAR data on reductions.
Activity Data
To calculate national emissions for these sources for the GHG Inventory, a set of industry activity data
drivers was developed and is used to update activity data, as a time series of activity data are not
available for this source. Pneumatic controllers are estimated each year by applying a regional factor for
the number of pneumatic controllers per well (GRI/EPA 1996) to annual regional data on gas well
population. The EPA-GRI study collected equipment data from 28 companies that included information
on almost 7,000 pneumatic devices from over 11,000 gas wells and over 3,000 oil wells. The EPA-GRI
study was focused on all methane sources across the oil and gas production industry, and therefore
included wells without pneumatic devices. These factors range from 0.5 to 1.6 pneumatic controllers per
well. For the petroleum production segment, pneumatic controllers are estimated each year by
applying a factor for the number of pneumatic controllers per heater/treater (4), and pneumatic
controller per battery without a heater/treater (3). The activity counts for pneumatic controllers in the
GHG Inventory include emergency shutdown devices.
Potential Methane Factors
The basis for the GHG Inventory's potential methane emission factors for pneumatic controllers in the
natural gas and petroleum production industry segment is the 1996 GRI/EPA report. The factor for
natural gas systems represents a mix of the average emissions from continuous bleed and intermittent
natural gas-driven pneumatic controllers in the 1996 GRI/EPA report. The region-specific factors are
developed using the GRI/EPA factor and regional gas composition data. For petroleum systems, it was
then assumed that 65% of pneumatic controllers in the petroleum production segment are low bleed
pneumatic controllers, and 35% of controllers are high bleed. The GRI/EPA factors are applied to these
populations.
Emission Reductions
The calculated potential emissions estimates are reduced using data on voluntary emission reductions
reported by industry partners to the Natural Gas STAR Program. The reductions undergo quality
assurance and quality control checks to identify errors, inconsistencies, or irregular data before being
incorporated into the GHG Inventory.
This approach of calculating potential CH4 emissions and then applying reductions data to calculate net
emissions is used to ensure an accurate time series that reflects real emission trends. Key data on
emissions from many sources in the Inventory are from data collected in 1992 (EPA/GRI 1996). Since the
time of this study, practices and technologies have changed. While the study still represents best
available data for some emission sources, using these emission factors alone to represent actual
emissions without adjusting for emissions controls would in many cases overestimate emissions. As
updated emission factors reflecting changing practices are not available for most sources, the 1992
emission factors continue to be used for many sources for all years of the Inventory, but they are
1

-------
April 2015
considered to be potential emissions factors, representing what emissions would be if practices and
technologies had not changed over time.
In previous Inventories, all production segment reductions related to pneumatic controllers reported to
Natural Gas STAR were assigned to the natural gas systems category. Since some portion of these
reductions would be more appropriately assigned to the petroleum systems category, in the public
review draft of the 2015 Inventory, the total reductions were apportioned by the potential emissions for
pneumatic devices in petroleum versus natural gas. This update resulted in an increase in natural gas
CH4 emissions (increase of around 5 MMT C02e from the previous inventory estimate for 2012) and a
corresponding decrease in petroleum systems CH4 emissions. See table 1 below.
Table 1. Year 2012 Pneumatic Controller Emissions in Previous (2014) Inventory and in 2015 Final
Inventory for the Production Segment of Petroleum and Petroleum Systems.

2012 in 2014
2012 in 2015
Data Element
Inventory
Inventory
Petroleum Systems
# of Pneumatic Controllers
414,797
441,311
Potential Cm (kt)
435
464
Reductions of Cm (kt)
0
246
Net CH4 (kt)
435
218
Natural Gas Systems
# of Pneumatic Controllers
477,606
468,466
Potential CH4 (kt)
1,208
1,185
Reductions of CH4 (kt)
873
628
Net CH4 (kt)
335
557
Natural Gas and
770
775
Petroleum Systems Total


Net CH4 (kt)


Table 2 below shows the average potential methane factors, and average net factors (after Gas STAR
reductions are removed) for pneumatic controllers in the GHG Inventory.
Table 2. Overview of GHG Inventory Emission Factors for Year 2012
Inventory/Factor type
Average Device
Methane Emission
Rates (scf/h)
Basis
2014 GHG Inventory:
Gas Wells, Potential1
Ranging from 14.1 to
16.8 (dependent on
region); averaging 15.0
EF is a weighted average from GRI, based on device
populations observed in EPA-GRI study and regional field
gas compositions.
2014 GHG Inventory:
Gas Wells, Net
Ranging from 3.9 to 4.7
(dependent on region);
averaging 4.2
2014 Inventory method assigns all Gas STAR reductions
for pneumatic devices to gas segment.
2014 GHG Inventory:
Oil Wells
Averaging 6.2
(independent of region)
Used GRI assumption of 65% of pneumatic controllers in
the petroleum production segment are low bleed, and
35% of controllers are high bleed; use of GRI factors
2014 GHG Inventory:
All Wells,
Effective Net
Averaging 5.1
Recognizing that some Gas STAR reductions are likely at
oil wells, this value represents the potential emissions for
oil and gas reduced by the total Gas STAR reductions.
2

-------
April 2015
2015 GHG Inventory:
Averaging 5.3
The 2015 Inventory allocates Gas STAR reductions to both
All Wells,

gas and oil wells
Effective Net


Table 2 footnotes:
1 This factor represents potential emissions. 2014 GHG Inventory methodology subtracts Gas STAR reductions from this
emission source, resulting in total net emissions that are approximately one-third of potential emissions.
Potential Revisions to Pneumatic Controller Estimates
Several recent studies and upcoming data from GHGRP provide new information on pneumatic
controllers. Below is a brief overview of the data and potential uses of the data to update the GHG
Inventory estimates.
Allen et al. 2014
Allen et al. 2014 measured emissions from 377 pneumatic controllers at 65 oil and gas production sites,
operated by 10 participating companies, across 4 regions of the United States.1 Of the three studies
discussed in this section, the Allen et al. study is the most comparable to U.S. national emissions
estimate and it is described in the most detail.
In addition to measuring methane emission from each device, the study collected data on each device's
design, application and operating characteristics. The pneumatic controllers included the full range of
pneumatic controller types used at oil and gas production sites in a range of applications. Device types
included low, high or intermittent bleed pneumatic controllers. Application characteristics included
liquid level, temperature and pressure control.
For each controller, measurements were made over a 15-minute period and included the number of
actuation events, the duration of each event, and the flow rate during each event. This information was
used to calculate the average emission rate from the device over a given time period.
For most measurements, Allen et al. used a gas supply line meter for testing wherever situations
allowed. However, for 44 device tests they were not able to use a supply line meter, and instead
monitored the exhaust port of the actuator with a Hi Flowฎ sampler. The impact of the use of different
methods is discussed in detail in Allen et al.
The average emission factors from Allen et al. were generally in agreement with average emission
factors from the GHG Inventory. However, the study observed great variability in emissions that is not
reflected in the average factors used by the GHG Inventory. In addition, Allen et al. data indicate that
the population of pneumatic controllers may be from 50% to 100% greater than the population assumed
in the GHG Inventory.
Allen et al. 2014 Emission Data
Of the 377 devices measured in the Allen et al. study, 241 devices had no measurable emissions due to
no observed emissions (136 devices) or due to emissions below the measurement threshold of the
instrument (105 devices). Since the measurement period was only 15 minutes, devices which vented
less frequently would typically have undetected emissions. This would be typical of a separator level
1 Allen, D.T., Pacsi, A., Sullivan, D., Zavala-Araiza, D., Harrison, M., Keen, K., Fraser, M., Hill, A.D., Sawyer, R.F., and
Seinfeld, J.H. Methane Emissions from Process Equipment at Natural Gas Production Sites in the United States:
Pneumatic Controllers, Environmental Science & Technology, 10.1021/es5040156. Available online:
http://pubs.acs.org/doi/pdf/10.1021/es5040156
3

-------
April 2015
controller at a well producing little condensate, a plunger lift valve actuated once a day, or an
emergency shutdown valve that is rarely actuated. To estimate the high-end impact of correcting the
emission data for infrequently actuating devices, the study assumed that all of these infrequently
actuating devices were actuated at frequency intervals just greater than the 15 minute test period. This
high-end assumption increases the estimated emission factor by 11%. A statistical analysis by Allen et al.
on the actuation patterns of the measured emissions indicated that the average emission factor would
likely be increased by 2% to 6% if the test period had been extended indefinitely. Allen et al. did not
adjust their emission factor of 4.9 scf CH4/hr for this likely error due to its small value relative to other
uncertainties.
There were 49 emergency shutdown pneumatic devices (ESD) in the Allen et al. study of 377 pneumatic
controllers (13%). Of the 49 ESD, 31 had no detectable emissions and 5 had well-defined actuations
during the 15 minute test period. Allen et al. also noted that the measured emissions for some of the
ESD included device leaks rather than actuations. Because some studies of pneumatics exclude ESD from
their analysis, Allen et al. considered the impact excluding of ESD on the pneumatic controller emission
factor. When Allen et al. excluded ESD data from the emission tests, the resulting emission factor for
non-ESD pneumatics was 5.5 scf CH4/hr.
The Allen et al. study included 26 measurements on pneumatic controllers at sites that identify as oil
production sites, out of their 377 measurements. Allen et al. treated the data from oil and gas wells as a
single data set due to variability in the definitions of oil and gas wells vary, largely depending on gas to
oil production ratios.
The study found that 19% of controllers were responsible for 95% of emissions.
Allen et al. 2014 Activity Data
The Allen et al. study observed 2.7 controllers per well surveyed. Only sites with venting pneumatic
controllers were visited for the study. However, the study notes that some sites use non-venting and/or
non-pneumatic controllers, and that therefore 2.7 controllers per well represents an upper bound. To
develop rough estimates of national emissions, in one scenario Allen et al. assumed that 75% of all
controllers were pneumatic-style controllers, of which 75% would be actively venting methane. This
assumption results in an estimated average of 1.5 venting pneumatic controllers per well.
Potential Applicability of Allen et al. 2014 to Revising GHG Inventory Emission Factors
Emission factors calculated with the Allen et al. data set are compared to the GHG Inventory emission
factors in Table 3. The first column of values presents the 2014 GHG Inventory factors for potential
emissions from gas wells in the year 2012. The second column presents the 2014 GHG Inventory
emission factors for net emissions from wells when adjusted for voluntary reductions provided by
companies converting high bleed devices to lower bleed devices in 2012. The voluntary reductions were
only applied to gas wells in the 2014 GHG Inventory. The same emission factor was applied to all oil
wells, regardless of region. The third column presents the 2015 GHG Inventory net emission factors for
all wells. The 2015 Inventory distributed the voluntary emission reductions to both gas and oil wells
based on estimated population of devices for each well type. The last column presents the net emission
factor developed by the Allen et al. study for gas wells and oil wells, combined. Allen et al. noted that
there was close agreement between their net emission factor and the 2014 GHG Inventory net factor
when compared at the national level.
4

-------
April 2015
A review of the GRI-EPA 1996 study which provided the activity data used in the GHG Inventory
indicates that ESD valves were included in their survey of pneumatic devices. Therefore the comparison
below includes ESD values in the Allen et al. emission factors.
Table 3. Comparison of GHG Inventory Emission Factors for 2012 to Allen et al. Emission Factors
(scf CH4/device-hour


2014 GHG




Inventory
2014 GHG
2015 GHG
Allen et al.

Potential
Inventory Net
Inventory Net
All well types,
Region/Well Type
Emissions
Emissions
Emissions
Net Emissions
Gas Wells
Northeast
15.5
4.3
8.5
1.7
Mid Continent
15.1
4.2
8.3
4.9
Southwest
14.7
4.1
8.1
Rocky Mountain
14.1
3.9
7.9
0.7
West Coast
16.8
4.7
9.3
Gulf Coast
16.1
4.5
8.9
10.6
Gas Wells Average
15.0
4.2
8.2
4.9
Oil Wells Average
6.2
6.2
3.4
-
National Average

5.1
5.3
4.9
Table 3 footnotes:
1	Attributing all Gas STAR reductions to gas wells in the natural gas systems source category, none to oil wells in the petroleum
systems source category.
2	Apportioning Gas STAR reductions to both gas and oil wells according to their respective well populations.
There are a number of ways the Allen et al. study results could be used to update emission factors in the
GHG Inventory. A key finding of the study is that a small fraction of controllers emit a large fraction of
emissions. It may be possible to stratify the Inventory activity data in a way that would allow calculation
of emissions for the high-emitting population separately from emissions from the lower emitting
population. EPA is seeking comment on how to quantify this population throughout the time series.
Please see Request for Stakeholder Feedback.
The Allen et al. emission data could be organized into device type (i.e., high bleed, intermittent, and low
bleed) to develop emission factors based on device type. However, almost half of the devices in the
sample could not be classified by the company. Similarly, the Allen et al. data could be organized by
application of the controller, such as on separator level control or on emergency shutdown devices
(ESD). However, EPA is unaware of a source for activity data on the national populations of pneumatic
device by application.
There is a very high degree of regional variation in the Allen et al. study emission factors (a factor of 10,
as seen in Table 3 above) which differs from the regional variation observed in the EPA-GRI study.
Potential Applicability of Allen etal. 2014 to Revising GHG Inventory Activity Data
The activity factors (number of pneumatic devices per well) observed and calculated in the Allen et al.
study are compared to the activity factors used in the 2014 GHG Inventory for the year 2012 in Table 4.
Table 4. Comparison of 2014 GHG Inventory Activity Factors for 2012 to Allen et al.
(average number of pneumatic controllers per well)
5

-------
April 2015

2014 GHG
2014 GHG
Allen et al. Survey:


Inventory
Inventory
Wells with Venting
Allen et al. Field
Region
Gas Wells
Oil Wells
Pneumatics
Survey: All Wells1
Northeast-Appalachian
0.5
-
2.3
1.3
Mid Continent
1.6
1.9
1.1
Southwest
1.3
Rocky Mountain
1.5
3.6
2.0
West Coast
1.0
Gulf Coast
0.7
3.3
1.9
National Average
1.0
0.8
2.7
1.5
Table 4 footnotes:
1 As described above, Allen et al. only visited wells with venting pneumatic controllers (while not all wells use pneumatic
controllers, and not all controllers actively vent) and found that these wells had, on average, 2.7 controllers per well. Allen et al.
provided a rough estimate that 75% of 75% of the observed per-well controller count to estimate the controller count for the
entire national well population. Column 5 presents regional activity data based on this estimate.
Potential Impact of Allen et al. 2014 on GHG Inventory Estimates
Table 5 below, based on data from the Allen et al. study, presents several approaches for estimating
national emissions from pneumatic devices at natural gas production sites using the Allen et al. study
data for year 2012. The first scenario in the table uses the activity data from in the 2014 GHG Inventory
and applies the average factor from Allen et al. The second scenario in the table uses the rough estimate
from Allen et al. that 75% of wells have 2.7 pneumatic controllers per well (the count per well
observed in the Allen et al. study). The third scenario is based on the Allen et al. study conclusion that a
national emission estimate would need to account for the fraction of controllers that were not
pneumatic and/or not actively venting. They suggested that possibly 75% of all controllers were
pneumatic-style controllers, of which 75% would be actively venting methane. This assumption results in
an estimated average of 1.5 venting pneumatic controllers per well, which is then multiplied by the Allen
et al. emission factor of 4.9 scf/h and the GHG Inventory well count of 470,913 wells. The fourth
scenario is based on the public review draft of the 2015 GHG Inventory which uses the same basic
approach as used in the 2014 Inventory (first scenario) but with a minor adjustment to the number of
devices and the reallocation of some Gas STAR reductions from the natural gas production sector to the
oil production sector.
Table 5. 2012 National Methane Emission Estimates for Natural Gas Production Sites, Various
Scenarios
Activity Factor
Emission Factor
National Methane
Emission Estimate
Number of controllers in EPA 2014
GHG Inventory, 477,606 controllers
Average emission rate for all
controllers measured in this work, 4.9 scf
methane/h
394 kt/y
20.5 bcf/y
Number of wells in EPA 2014 GHG
Inventory, 470,913 wells, with 75% of the
wells assumed to have 2.7 pneumatic
controllers per well, as observed in this
work
Average emission rate for all
controllers measured in this work, 4.9 scf
methane/h
786 kt/y
40.9 bcf/y
6

-------
April 2015
Number of wells in EPA 2014 GHG
Inventory, 470,913 wells, with 75% of the
wells assumed to have 2.7 pneumatic
controllers per well, and assuming only
75% of the pneumatic controllers vent
methane as suggested by some study
participants.
Average emission rate for all
controllers measured in this work, 4.9 scf
methane/h
600 kt/y
31.2 bcf/y
Number of controllers in EPA 2015 (draft)
GHG Inventory, 468,303 controllers
Average emission rate for natural gas
controllers in the Inventory, 8.2 scf
methane/h
653 kt/y
34.0 bcf/y
Oklahoma Independent Petroleum Association
Oklahoma Independent Petroleum Association (OIPA)2 recently conducted a study quantifying methane
emissions from 680 pneumatic controllers across Oklahoma. The study collected data for each controller
and calculated emissions using an engineering equation.
The calculation approach determined controller emissions as the sum of the controller, tubing, and
actuator emissions as a result of actuation plus any continuous bleed and seepage emissions. The
calculation did not include any unintended leaks.
OIPA observed an average of 3.6 pneumatic controllers per well, including 15 wells that did not have
controllers. See Figure 1, below.
The study grouped controllers based on age (new (first production in 2000 or later) or old (first
production 1999 or earlier)) and production (oil or gas), and observed an average controller count per
site 2.2 times higher for new sites than for old sites, which OIPA noted was due to the increased number
of process units at newer sites.
The study calculated that a small fraction of sources (3.5%) were responsible for a large fraction of
emissions (73%).
2 Oklahoma Independent Petroleum Association, Pneumatic Controller Emissions from a Sample of 172 Production
Facilities, November 2014. Available online: http://www.oipa.com/page images/1418911081.pdf
7

-------
April 2015
Sites	
172 sites {205 wells) visited for data collection
162 sites {190 wells) had natural gas pneumatic controllers
10 sites (15 wells) did not have natural gas pneumatic controllers
Controllers	
680 natural gas pneumatic controllers 659 intermittent vent controllers
77 controller models	21 continuous bleed controllers
Average Controller Coumts	
4.0 pneumatic controllers per site	3.6 pneumatic controllers per well
5.0	pneumatic controllers per new gas site	5.3 pneumatic controllers per new oil site
3.1	pneumatic controllers per old gas site 2.7pneumatic controllers per old oil site
Actuation Frequencies	
538 controllers (79%) had no actuations detected during the observation period and were assigned the default rate
126 controllers (19%) had actuation rates less frequent than the once per 15 minute default rate
16 controllers 12%) had actuation rates more frequent than or eaual to the default rate
Figure 1. Key Observational Results from OIPA Study
Prasino Group Study
The Prasino Group conducted a survey of high bleed devices used in the Fort St. John area of British
Columbia, Canada to determine the average bleed rates from high bleed pneumatic controllers and
pumps3 in field conditions. The study sampled bleed rates using a positive displacement bellows meter
at upstream oil and gas facilities across a variety of producing fields in the area.
The study found that whole gas emissions from high bleed controllers averaged 9.2 scf/h and from high
bleed intermittent controllers averaged 8.7 scf/h. The average of these two type of high bleed devices
was 9.0 scf/h.
The Prasino study did not collect activity data on the population of pneumatic controllers per well.
Summary of Recent Studies
Table 6 summarizes the findings of the above three studies.
Table 6. Summary of Results of Studies Discussed in this Memo
Data Element
Allen et al.
Prasino
OIPA
Pneumatic Controllers per
well
Observed an average of 2.7
controllers per well
measured. Authors suggest
the national average may be
between 1.5 to 2.7
controllers per well.
N/A
Average of 3.6 controllers
per well, with variations
between new and old sites
and oil and gas
Bleed type fraction
N/A
N/A
97% intermittent vent; 3%
continuous bleed
Emission factors
Ranging from 0.7 to 10.6
scf/h (dependent on region);
averaging 4.9 scf/h
9.2 scf/h high bleed; 8.7
scf/h intermittent
controllers. The average of
these two type of high bleed
devices was 9.0 scf/h
Average 1.05 scf whole
gas/hour; 0.40 scf/h
intermittent vent,
21.54 scf/h continuous
bleed
3 The Prasino Group, Final Report- For Determining Bleed Rates for Pneumatic Devices in British Columbia,
December 18, 2013. Available online:
http://www2.gov.bc.ca/gov/DownloadAsset?assetld=lF074ABD990D4EFB8AE555AEB3B8D771&filename=prasino
pneumatic ghg ef final report.pdf
8

-------
April 2015




Observations on distribution
19% of controllers
responsible for 95% of
emissions
N/A
3.5% of controllers
responsible for 73% of
emissions
Greenhouse Gas Reporting Program
EPA's Greenhouse Gas Reporting Program (GHGRP) has collected 3 years of emissions data for
pneumatic controllers at oil and gas production sites. The GHGRP data currently available include total
emissions from controllers by type for each facility. Starting in 2015, data will also be reported to the
GHGRP on the number of controllers, by type, and total number of wells for each facility. This data
could be used to allow EPA to update its assumption on the types and numbers of controllers in use for
the 2016 GHG Inventory.
Table 7. Comparison of GHGRP and GHG Inventory Production Segment Pneumatic Controller CH4 (kt)
Year
GHGRP
2015 Inventory
2011
839
831
2012
879
775
2013
974
760
Regional Surveys
WRAP Survey
The Independent Petroleum Association of Mountain States (IPAMS) conducted a project with WRAP
(Western Regional Air Partnership) "to build on, improve, and address inventory issues." The result
included criteria pollutant emissions for point and area sources associated with the exploration,
production, and gathering operations of oil and gas in the major basins throughout the six-state (CO,
MT, NM, ND, UT, and WY) study region for year 2006. The study collected activity data on pneumatic
controllers and calculated emissions using number of controllers and emission factors.
2011 Oil and Gas Emission Inventory Enhancement Project for CenSARA States
The Central States Air Resource Agencies (CenSARA) identified the need to improve the activity data and
methodologies used by its member states to develop oil and gas area source emissions. ENVIRON and
the Eastern Research Group (ERG) reviewed available data sources and collected basin-level specific
emissions information for calendar year 2011 through industry surveys to operators within the CenSARA
domain. Pneumatic devices were the most significant source of methane emissions in most basins
within the CenSARA region. Emissions from pneumatic controllers were generally calculated using
emission factors and an estimate of pneumatic controllers per well.
Request for Stakeholder Feedback
•	EPA seeks feedback on activity data sources that may be used to estimate the number of
pneumatic controllers per gas well and per oil well, across the Inventory time series (1990
through 2013). EPA seeks feedback on whether certain sub-populations of wells (e.g., stripper
oil wells) warrant sub-population-specific assumptions.
•	EPA seeks feedback on Allen et al. assumptions used to calculate national activity data for
pneumatic controllers and their applicability for the GHG Inventory.
9

-------
April 2015
•	EPA seeks feedback on an appropriate emission factor structure that may be used in the GHG
Inventory to better account for the high emission rate observed for a very limited number of
emission sources. Are activity data available to support this option?
•	EPA seeks feedback on industry dynamics that should be reflected over the time series—i.e.,
how have emissions or activity data significantly changed over time; and how can EPA apply
updated data across the time series that would reflect such changes?
•	EPA seeks feedback on whether the same or different emission factors and activity factors
should be applied for pneumatic controllers in natural gas and petroleum systems.
•	There is a very high degree of regional variation in the Allen et al. study emission factors (a
factor of 10, as seen in Table 3 above) which is different than the findings of the 1996 EPA-GRI
study. EPA seeks comment on potential application of region-specific EFs versus national EFs in
the GHG Inventory.
•	EPA seeks comment on whether it is appropriate to develop activity data on a regional versus
national basis for the GHG Inventory.
•	The Allen et al. study noted a large number of devices with malfunctions. EPA seeks feedback
on how this can be accounted for within the GHG Inventory, across all years of the time series
(1990-2013).
•	EPA seeks feedback on an approach to use GHGRP data in the GHG Inventory, across all years of
the time series (1990-2013). One comment on the expert review draft inventory suggested
estimating the controller counts in GHGRP and increasing them by 15% to estimate national
controllers. The comment also suggested applying the Allen et al. average emission values to
the population of high and intermittent bleed controllers in GHGRP, and using a value of 6 scf/h
to calculate low bleed emissions.
•	NSPS OOOO regulates new pneumatic controllers. EPA seeks stakeholder feedback on methods
that will allow the GHG Inventory to reflect changes in emissions due to the impacts of this
regulation.
•	EPA seeks comment on the allocation of Gas STAR reductions between the natural gas and
petroleum systems categories. The 2015 public review draft of the Inventory allocates the
reductions based on the potential emissions calculated for each category. Other options could
include a split based on well counts. Future revisions to this category will likely include use a net
calculation approach and will not rely on Gas STAR data.
•	OIPA study noted differences in the number of pneumatic controllers on site for new versus old
well sites. EPA seeks stakeholder feedback on approaches for applying data on new versus old
sites in the GHG Inventory.
•	EPA seeks feedback on data available through the WRAP and CenSARA work and other regional
data collection efforts that could be used to update national emissions estimates.
Preliminary comments EPA has received on this source category include support for the development of
emission factors and activity data on a regional as opposed to a national basis, support for the use of
measurement data for calculating emission factors, and support for a revision of EPA's current approach
to use technology-specific net factors. In addition, commenters support updating activity data for this
source using data from external studies, and/or data from the GHGRP, and noted that these updates
would likely increase the number of pneumatic controllers per well. Commenters support including
malfunctioning devices in emissions calculations, such as through including these emissions in a leaks
category.
10

-------