DRAFT - Do not cite or quote
DRAFT EPA Review of Available Documents for
Developing Proposed Emissions Factors for
Flares, Tanks, and Wastewater Treatment
Systems
-------
DRAFT - Do not cite or quote
This page intentionally left blank.
11
-------
DRAFT - Do not cite or quote
DRAFT EPA Review of Available Documents for Developing Proposed Emissions Factors
for Flares, Tanks, and Wastewater Treatment Systems
Coniracl No i;iM)-l I-OK4
Work Assignment \o 2-12
I S I ji\iron menUil Protection Agency
Office of Air Ouulity Planning and Standards
Sector Policies and Programs Division
Research Triangle Park, North Carolina 27711
August 2014
in
-------
DRAFT - Do not cite or quote
Disclaimer
This report has been reviewed by the Sector Policies and Programs Division of the Office of Air
Quality Planning and Standards, Office of Air and Radiation, EPA, and approved for publication.
Mention of trade names or commercial products is not intended to constitute endorsement or
recommendation for use.
iv
-------
DRAFT - Do not cite or quote
Table of Contents
Page
Section 1 Summary 1
Section 2 Review Studies 4
2.1 Office of Inspector General Review 4
2.2 Shine Review 8
Section 3 DIAL Studies 9
3.1 Alberta DIAL Study 11
3.2 Houston Area DIAL Study 14
3.3 Houston Ship Channel/Shell Deer Park DIAL Study 19
Section 4 Passive FTIR Studies 24
4.1 2010 TCEQ Flare Study 24
4.2 Marathon Petroleum Flare Tests . 25
4.3 Flint Hills Flare Test 26
4.4 Additional Flare Tests 26
Section 5 Tank I-missions Tests ... 28
5.1 Global Companies 1.1.(' South Portland. Maine 28
5.2 Spraguc Opeiatinu Resource LLC- Seaisport, Maine 29
Section (•> Proposed Conclusions ... 31
Section 7 References... 33
Appendix A 38
v
-------
DRAFT - Do not cite or quote
List of Figures
Figure 3-1. Schematic of DIAL System Measuring Emissions (from Chambers and Stosher,
2006) 9
Figure 3-2. Comparison of DIAL and AP-42 Emission Estimates in CONCAWE 90-hour
Study (Smithers et al., 1995) 12
List of Tables
Table 1-1. Scientific Studies Cited in the May 2013 Complaint 1
Table 2-1. TCEQ's Recommended Emission Factors lor I-'lures (TCEQ, 2013) 7
Table 3-1. Summary of Emissions Comparison from the Houston Area DIAL Study 16
Table 3-2. Summary of Emissions Comparison Included in the Shell DIAI. Study Report.. 21
vi
-------
DRAFT - Do not cite or quote
Section 1
Summary
On May 1, 2013, Air Alliance Houston, Community In-Power and Development
Association, Inc. (CIDA), Louisiana Bucket Brigade, and Texas Environmental Justice
Advocacy Services (TEJAS) (collectively, "Plaintiffs") filed a lawsuit against the U.S.
Environmental Protection Agency (EPA) alleging that the I - IV\ had failed to review and, if
necessary, revise emissions factors at least once every three years as required in Section 130 of
the Clean Air Act (CAA). Air Alliance Houston, el al. \ McCarthy. \o. l:13-cv-00621-KBJ
(D.D.C.). In the complaint, which is included as Appendix A to this report, the Plaintiffs sought
to compel the EPA to expeditiously complete a re\ iew of the volatile organic compounds (VOC)
emissions factors for industrial flares ("flares"), liquid storage tanks ("tanks"), and wastewater
collection, treatment and storage systems ("wastewater treatment systems"), and, if necessary,
revise these factors. In Paragraphs 35 and 3f> of their complaint, the Plaintiffs cited several
reports and studies that allegedly show VOC emissions from lltires, tanks, and wastewater
treatment systems "can he se\ ertil orders of magnitude higher than AP-42 emissions factor
estimates." Complaint. • 3o The reports cited in the complaint are listed in Table 1-1 in the
order in which they were cited
Table l-l. Scientific Studies Cited in the Msiv 2013 Complaint
UiT
No.
Cited DoiuniiiH
Studj
1 > pe
1
I'aliick \1illmaii. I'rank \l;iiiiiisk\. Kevin Good, and Bill Nelson. Office of Inspector General.
Evaluation Report: El'. 1 ('an Improve Emission Factors Development and Management, Report
No. 20(i(i-|,-o() I" (2u()(i) \\ ;iil;ihle at: htto://www.era.eov/oie/reoorts/2006/20060322-2006-P-
00017 pdl"
Review
2
Brenda Shine, EPA SPPI) Memorandum to EPA Docket No. EPA-HQ-OAR-2003-0146 dated
July 27, 2007. Potential Low Bias of Reported VOC Emissions from the Petroleum Refining
Industry. Available at htlD://www.reeulations.eov/. search for EPA-HO-OAR-2003-0146-0010.
Review
3
David T. Allen and Vincent M. Torres, Univ. of Tex. at Austin, Ctr. Energy & Envtl. Res.,
TCEQ 2010 Flare Study Final Report (2011), available at
www.tcea.texas.eov/assets/OTiblic/imi3lementation/air/rules/Flare/2010flarestudv/2010-flare-
Passive
FTIR
s t ud v -11 na 1 - rc do rt. od f
4
David Randall & Jeff Coburn, EPA, EPA 453/R-10-002, Critical Review of DIAL Emission Test
Data for BP Petroleum Refinery in Texas City, Texas, at ES-2 tbl. 1 (2010), [BP Texas City, TX]
available at
www.eoa.sov/airtoxics/br) dial review reoort 12-3-10.t>df
DIAL
1
-------
DRAFT - Do not cite or quote
UcT
No.
( ik'd Dociimcnl
Sluclj
1 > PC-
5
Loren Raun & Dan W. Hoyt, Bur. Pollution Control & Prevention, City of Houston,
Measurement and Analysis of Benzene and VOC Emissions in the Houston Ship Channel Area
and Select Major Stationary Sources Using DIAL (Differential Absorption Light Detection and
Ranging) Technology to Support Ambient HAP Concentrations Reductions in the Community
(DIAL Project) (2011), [Shell Deer Park, TX] available at
www. ereenhoustontx. eov/dial20110720.odf
DIAL.
6
Marathon Petroleum Co., LLC, Performance Test of a Steam-Assisted Flare with Passive FTIR
(May 2010), available at
www.tcea.texas.eov/assets/t>ublic/imt)lementation/air/rules/Flare/2010flarestudv/mt>c-txc.t>df
Passive
FTIR
7
Marathon Petroleum Co., LLC, Performance Test of a Steam-Assisted Elevated Flare with
Passive FTIR -Detroit (2010), available at
www.tcea.texas.sov/assets/i3ublic/imi3lementation/air/rules l;hiio/2010riarestudv/mi3c-detroit.i3df
Passive
FTIR
8
Flint Hills resources Port Arthur, LLC, PFTIR Test of Steam- Issisied Elevated Flares-Port
Arthur (2011), available at
www.era.eov/comi3liance/resources/OTiblications/civil/i3rosin)ms/c;i;i ix>it;in 1 iur-rei3ort.i3df
Passive
FTIR
9
Allan Chambers & Mel Strosher, Alberta Research ( oimcil. Inc., Refinery Ijemonstration of
Optical Technologies for Measurement of Fugitive Emissions and for Leak Deled ion (2006),
available at
www.environmentalintesritv.org/i3df/i3ublicauoiit. 1 ill' A.tt D Tmal Unset.odl
DIAL
EPA entered into a consent decree ^ ilh the Plaintiffs lo settle l lie lawsuit. Under the terms of the
consent decree, by August 19, 2014, ]-l\\ will rc\ icw and cither propose revisions to the VOC
emission factors for llares. tanks and wastewater treatment systems under CAA section 130, or
propose a determination under C.\.\ section I 3<) that re\ ision of these emission factors is not
necessary. By December N. 2<)|4. l-PAwill issue linal revisions to the VOC emission factors
for flares, tanks and wastewater treatment systems, or issue a final determination that revision of
these emission factors lor Hares is not necessary N\\ will post each proposed revision or
determination (or combination thereof), and each final revision or determination (or combination
thereof), on its AP-42 website on the dates indicated above.
There are three different types of cited reports listed in Table 1-1:
1. "Review" studies summarize information from a number of other studies.
2. "DIAL" studies contain differential infrared absorption LIDAR [light detection and
ranging] (DIAL) measurement data and may include comparisons with emissions
inventory estimates. These studies generally contain measurement data for a variety of
different emissions sources at a refinery.
3. "Passive FTIR" studies contain passive Fourier transform infrared (FTIR) spectroscopy
measurement data collected to assess the performance (combustion efficiency) of refinery
flares.
2
-------
DRAFT - Do not cite or quote
EPA reviewed each of the cited reports to determine if the data provided in these reports could
be used to support emissions factor development (or to support no revisions to certain emissions
factors). Each report reviewed is briefly described and discussed in subsequent chapters of this
report. Section 2 describes the "Review" studies, Section 3 describes the "DIAL" studies, and
Section 4 describes the "Passive FTIR" studies. Additionally, EPA identified and reviewed two
passive FTIR studies not listed in the complaint. These studies are discussed along with the cited
passive FTIR in Section 4 of this report. Furthermore, a scries of lests were conducted to
quantify storage tank emissions at two facilities in the northeastern United States. The results of
these tests are discussed in Section 5 of this report. We pro\ ide our conclusions regarding all of
these reports in Section 6.
Based on our review, we are proposing thai, while the cited reports provide valuable
information regarding the potential emissions from petroleum refinery and oilier process units,
the data from these studies (with the exception of the passi\ e IIIR studies and one DIAL study)
cannot be used for emissions factor de\ elopmenl due to (I) the lack of operational data by which
to normalize the emissions rates, (2) the lad thai many of the measurements do not isolate one
particular emissions source and 'or (3) the lacl thai the studies did not attempt to characterize the
range of normal operating conditions Howe\ er. the I-PA is proposing that it is possible to form
emissions factors for certain pollutants for industrial llares using DIAL and passive FTIR data.
Sections 3 2 and 4 discuss this in greater detail
3
-------
DRAFT - Do not cite or quote
Section 2
Review Studies
Plaintiffs cited two review studies to support the allegation that "[t]he EPA has
acknowledged, and scientific studies show, that the AP-42 emissions factors for flares, tanks, and
wastewater treatment systems significantly underestimate VOC emissions from these processes."
Complaint, ]} 35.
2.1 Office of Inspector General Review
The Office of Inspector General (OI(i) Re\ iew (Ref No. 1 in Table 1-1) ("OIG Review")
"sought to determine whether the air emissions factors used by the Environmental Protection
Agency (EPA) are of acceptable quality for making key em ironmental decisions, and whether
EPA's process for developing, impro\ ing. and rating emissions factors is sufficient to meet
users' needs." The report provides the following recommendations:
• Develop emissions factors guidance that addresses the development and appropriate
use of emissions factors for non-inventory purposes
• Establish a rating system that provides the quantitative range of uncertainty for
emissions factors lor both imemory and non-inventory purposes.
• Work with industry. State and local agencies, and others to leverage available
resources for meeting increasing demands for new factors.
• 1-slaMish a workgroup to de\ elop a comprehensive strategic plan for the Emissions
I 'actors Program and ensure that requested resources are used to achieve program
goals
The OIG Review notes that EPA has increased the number of rated emissions factors
included in AP-42 from 8,838 in 1996 to 17,110 in 2004, but it also notes that the majority of
AP-42 emissions factors still have a below average (D) or poor (E) rating factor. In 1996, 56
percent of the AP-42 emissions factors were rated D or E, while 62 percent of the emissions
factors were rated D or E in 2004. Thus, while the report notes that the EPA has made progress
in developing rated emissions factors, the quality of the emissions factors remains low.
4
-------
DRAFT - Do not cite or quote
The OIG Review also highlights the point that AP-42 emissions factors are intended for use
in wide-area emissions inventories and should be used only as a last resort when developing site-
specific emissions estimates. The OIG Review described the misuse of AP-42 emissions factors
in situations such as setting permit emissions limits and estimating a facility's potential to emit.
The introduction to AP-42 (EPA, 1995) includes the following recommendations and limitations
on the use of emissions factors:
• "Data from source-specific emission tests or continuous emission monitors are usually
preferred for estimating a source's emissions because those data provide the best
representation of the tested source's emissions ""
• "Emission factors in AP-42 are neither NW-rccommcnded emission limits (e. g., best
available control technology or BACT. or lowest achievable emission rate or LAER) nor
standards (e. g., National Emission Standard lor I lazardous Air Pollutants orNESHAP,
or New Source Performance Standards or \SPS) I se of these factors as source-specific
permit limits and/or as emission regulation compliance determinations is not
recommended by EPA. Because emission factors essentially represent an average of a
range of emission rates, approximately hall" of the subject sources will have emission
rates greater than the emission factor and the other hall" will have emission rates less than
the factor. As such, a permit limit using an AP-42 emission factor would result in half of
the sources being in noncompliance "
• "ll"iepresentati\ e source-specific data cannot be obtained, emissions information from
equipment -s endors. particularly emission performance guarantees or actual test data from
similar equipment, is a better source of information for permitting decisions than an AP-
42 emission factor. When such information is not available, use of emission factors may
be necessary as a last resort. Whenever factors are used, one should be aware of their
limitations in accurately representing a particular facility, and the risks of using emission
factors in such situations should be evaluated against the costs of further testing or
analyses."
The OIG Review highlighted three industries for which emissions factors were
considered unacceptable for the decisions being made, one of which was petroleum refineries.
For petroleum refineries, the report cites a Texas 2000 Air Quality Study (TNRCC, 2000)
("TRNCC Study") that compared ambient measurements of VOC concentrations with those
5
-------
DRAFT - Do not cite or quote
projected by emissions inventory estimates. According to the OIG, the TNRCC Study found that
VOC emissions were under-reported, primarily due to under-reporting for flares, process vents,
cooling towers, and process equipment leaks. Based on the TNRCC Study, Texas revised its
emissions estimation guidelines and thereby improved its emissions inventory. Based on the
improved emissions inventory, Texas revised its State Implementation Plan to include goals for
VOC emissions reductions and to relax the previous goal for NOx emissions reductions. In this
example, AP-42 emissions factors were being used for a wide-area emissions inventory, so their
application was not necessarily incorrect, but reliance solely on AP-42 emissions factors rather
than more detailed site-specific information resulted in an inaccurate assessment of what was
needed in order to meet ozone air quality standards The OIG re\ ieu noted that Philadelphia and
California's Bay Area Air Quality Management District found similar issues with their
inventories. Philadelphia then adopted the guidelines developed by the Texas Commission on
Environmental Quality (TCEQ) to impro\ e their emissions in\ entory and the Bay Area Air
Quality Management District issued a new rule | Regulation 12 Rule 11, which required
monitoring flare vent gas flow and composition | to obtain more accurate emissions data.
As the OTCi Re\ ieu does not contain any emissions measurement data or recommended
emissions factors, the current TCI-O emissions in\entory guidelines (TCEQ, 2013) ("TCEQ
Guidelines") were re\ ieued to better understand the emissions inventory improvements cited in
the OIG re\ ieu The TCI-O Guidelines contained improved instructions on how to estimate
emissions and u hen and how to use AP-42 emissions factors (including instructions for using the
TANKS model). but in most cases the TCEQ Guidelines did not provide revised emissions
factors. I or example, the TCI-O Guidelines for estimating emissions from cooling towers is to
use measured VOC concentrations and water recirculation rates to estimate cooling tower
emissions II" VOC concentration data in the cooling water are not available, then the TCEQ
Guidelines specify that the uncontrolled AP-42 emissions factor should be used. Based on these
guidelines, the controlled AP-42 emissions factor should never be used. However, in the
absence of these guidelines, most refineries used the controlled AP-42 emissions factor for
cooling towers because they considered monitoring of chemical additive rates or other operating
parameters to warrant the use of the controlled cooling towers AP-42 emissions factor (Lucas,
2007).
6
-------
DRAFT - Do not cite or quote
The one exception to the TCEQ Guidelines not providing revised emissions factors is the
table of emissions factors forNOx and CO from flares. TCEQ's revised emissions factors are
based on the results of the historical flare study data (circa 1983 and 1985) with which the AP-42
emissions factors were developed (listed as reference 1 of AP-42 Section 13.5). As seen in Table
2-1, the TCEQ Guidelines subcategorized the emissions factors by flare type (steam-assisted
versus air-assisted or unassisted) and by heat content of the flare vent gas ("high Btu" versus
"low Btu"). The TCEQ Guidelines also provide instructions to use data with high time
resolution in order to apply the correct emissions factor As lliciv were limited data within each
subcategory, and these data are quite old, it is question a hie u hclhcr the subcategorization of
these data are statistically justified. Furthermore, as discussed in detail in Section 4 of this
document, significantly more data by which to c\ aluate the NOx and CO llare emissions factors
in AP-42 is now available.
Table 2-1. TCEQ's Recommended Emission Factors for Flares (TCEQ, 2013)
Contaminant
Assist
Type
Waste Gas Stream
Net Heating Value'7'"
Emission Factor
NOx
Steam
High Btu
0.0485 Ib/MMBtu
Low Btu
0.068 lb/MMBtu
Air or
Unassisted
High Btu
0.138 lb/MMBtu
Low Btu
0.0641 Ib/MMBtu
CO
Steam
High Btu
0.3503 lb/MMBtu
Low Btu
0.3465 lb/MMBtu
Air or
Unassisted
High Btu
0.2.755 lb/MMBtu
Low Btu
0.5496 lb/MMBtu
" High Btu: > 1000 Btii/scf
6 Low Btu: 192-1000 Btii/scf
Based on our review or me TCEQ Guidelines, the emissions inventory improvements
made by Texas were not based on improvements to or replacement of AP-42 emissions factors.
The emissions inventory improvements were accomplished through better instructions on how to
determine site-specific emissions estimates based on available monitoring data along with
specific instructions on how and when to apply the AP-42 emissions factors when site-specific
7
-------
DRAFT - Do not cite or quote
monitoring data are not available. These guidelines effectively implement the recommendations
and limitations on the use of AP-42 emissions factors as stated in the AP-42 introduction.
2.2 Shine Review
The Shine Review (Ref No. 2 in Table 1-1) summarized the findings from several
studies, including the DIAL Alberta study (Ref. No. 9 in Table 1-1), the TNRCC Study and the
OIG Review, that indicated that the refinery emissions in\ enlories inay be under-estimating
VOC emissions. Key points made in the Shine Review include
• Emissions inventories generally do not include emissions that occur during process
upsets, malfunctions, startups or shutdowns
• Unusual or unexpected emissions, such as leaks in heat exchange systems that cause
high emissions from cooling towers or lank roof landings, are often omitted from
emissions inventories.
• Improper characterization of process operating parameters used in emissions
estimation methodologies can significantly impact the estimated air emissions.
The Shine Re\ iew did not critique or criticize AP-42 emissions factors or AP-42
emissions estimation methodologies, hut rather the scope and rigorousness of compiled
emissions inventories The Shine Re\ iew also did not present any new emissions factors or data
by which new emissions factors could he de\ eloped In efforts to improve refinery emissions
inventories to address the issues identified in the Shine review, the EPA funded the development
of the I .mission Estimation I'roiocol for Petroleum Refineries (RTI, 2011) ("Refinery Protocol")
to provide guidance on preferred means to develop site-specific air emissions estimates,
including methods lor estimating emissions during process upsets, malfunctions, startups or
shutdowns. Similar to the TCI-O (fuidelines and in accord with the AP-42 discussion on the use
of AP-42 emissions factors, the Refinery Protocol describes a preference for continuous
emissions measurement data or site-specific test data, with AP-42 emissions factors used as a last
resort. The EPA required petroleum refineries to submit detailed, site-specific emissions
inventories following the methodologies outlined in the Refinery Protocol in order to develop an
improved emissions inventory for its sector rulemakings, including the risk and technology
review (EPA, 2011).
8
-------
DRAFT - Do not cite or quote
Section 3
DIAL Studies
DIAL is a laser-based measurement method for determining pollutant concentration
profiles in the ambient air. The DIAL monitoring system has been used in a variety of studies to
measure emissions from petroleum refinery and petrochemical sources. The DIAL studies cited
in the complaint are report Ref Nos. 4, 5 and 9 in Table 1-1.
The DIAL monitoring system is typically situated downwind of an emissions source and
the laser beam is shot across the emissions plume and pivoted to form a 2-dimensional vertical
scan plane (see Figure 3.1). Small portions of the light from the laser are backscattered due to
particles and molecules in the measurement path. This backscattered light is collected through a
telescope system adjacent to the laser and measured via a sensitive light detector. The timing of
the received light provides a measure of the distance of the emissions plume.
& sorption
DIAL tubes
Height
(m)
Wind
Scan
plane
Emissions
Concentration
(mg/m3)
Figure 3-1. Schematic of DIAL System Measuring Emissions
(from Chambers and Stosher, 2006).
9
-------
DRAFT - Do not cite or quote
DIAL measurements can be made specific to one compound, such as benzene, or general
for a class of similar compounds, such as saturated hydrocarbons. Two different wavelengths of
light are pulsed in quick succession: one wavelength that is absorbed strongly by the pollutant of
interest and one of similar wavelength that is not absorbed. The difference in the returned signal
strength between these two light pulses provides a measure of the concentration of the pollutant.
Thus, a unique advantage of the DIAL monitoring system is that it can provide spatially resolved
pollutant concentrations in the 2-dimensional scan plane
Using the DIAL'S measured pollutant concentration profile across the scan plane, and
site-specific wind speed and direction measurements, it is possible to calculate a mass emissions
rate for the pollutant of interest. If the wind speed and di recti on are loo variable during the
measurement scan, there can be high uncertainties in the calculated mass emissions rate.
Therefore, the ability to accurately measure emissions from an emission source or set of sources
using the DIAL instrumentation is dependent on an acceptable wind direction and wind speed
conditions. Furthermore, measurement scans close to large objects, such as tanks, can be
affected by wind speed changes caused by the large objects These wind speed changes can
create inaccuracies in the DTAT. measurements
At a refinery or petrochemical facility, there are typically multiple emission sources in
close proximity. Emissions sources upwind of the DIAL scan plane can contribute to the
measured emissions rate, therefore the DIAL in\estiuator must perform scans upwind of the
source of interest to rule out interfering sources I lo\\e\ er, depending on the configuration of the
plant, the wind direction, and timing of the measurement scans, it is not always possible to
perform upwind scans or to isolate a single emissions source when performing upwind scans. As
such, DIAL measurements can include emissions contributions from several sources, and it can
be difficult to attribute emissions to particular sources, if upwind scans are not done correctly or
if site-specific configurations limit the ability to isolate an emissions source.
A typical measurement scan is about 10-15 minutes in duration, so the method provides
good time resolution. Multiple scans can be completed in succession, as long as the wind
direction remains consistent. If winds are too variable, only periodic scans can be performed.
Measurements made over short time periods may not be representative of average emissions
from the source. Therefore, the most useful DIAL data are those where multiple DIAL scans
were made at different times to account for variable source operating conditions. For purposes
10
-------
DRAFT - Do not cite or quote
of emissions factors review, measurements are only useful when the operating conditions of the
emitting source at the time of the measurements are known.
3.1 Alberta DIAL Study
The Alberta DIAL Study (Ref No. 9 in Table 1-1) was the earliest of the DIAL studies
examined in this report, and it found emissions from product storage tanks, cooling towers, and
delayed coking unit (DCU) process area to be higher than expected based on the emissions
inventory for the facility. Unfortunately, no corresponding process-specific data were available
for the units tested. It is unknown, for example, what products were being stored in the tanks
tested, what type of controls, if any, were on the storage tanks (fixed roof or floating roof), and
whether the tanks were being actively filled dining the test. Due to the lack of process operating
data at the time of the test, it is impossible to de\ clop emissions factors from the study data (i.e.,
no values are available to normalize (he emissions).
The Alberta DIAL Study used the emissions measured during the study to project annual
emissions by assuming the emissions rales measured occurred continuously for 8,760 hours per
year. The extrapolated YOC emissions lor the refinery's storage tanks were projected to be
5,090 tonnes/yr compared to I 53 tonnes yr as reported in Canada's National Pollutant Release
Inventory (NPRI). I lo\\e\ er. as pre\ ionsly noted, measurements made over short time periods
may not he rcpicsentali\ e of a\ erage emissions from a source. There are a variety of reasons
why short-term emissions may he significantly higher than annual average emissions. For
exam pie. the AlhertaDIAI. Stud y was conducted in the summer (at higher than average
temperatures) and during the day (when \ essel loading activity is higher). These conditions
would lead to higher than a\ erage emissions. Given the short-term nature of the DIAL
measurements, it is difficult to conclude that the high short-term emissions rates measured during
this study are truly representative of the long-term average emissions from storage tanks at the
facility. Given the lack of process data, it is also difficult to conclude whether the emissions
inventory estimates were properly determined or if there may have been unusual process
conditions that explain the high short-term emissions. Considering these limitations, the Alberta
DIAL Study is insufficient to support the assertion that the emissions estimation methodology
for storage tanks as presented in AP-42 underestimates the long-term emissions from refinery
storage tanks.
11
-------
DRAFT - Do not cite or quote
Prior to the Alberta DIAL Study, CONCAWE, a division of the European Petroleum
Refiners Association, investigated and compared DIAL measurement data for external floating
roof storage tanks with emissions estimates made using the AP-42 emissions factor methodology
(see Smithers, et al., 1995). This study used a longer (90 hour) measurement period than the
Alberta DIAL Study. Direct process and meteorological data were used to provide hourly input
data for the organic liquid storage tank emissions factor equations developed by the American
Petroleum Institute (API), which at the time were published as API Publication 2571. API
retains the copyright to these equations, but has granted the EPA permission for the
nonexclusive, noncommercial distribution of this material, which is provided in Section 7.1 of
AP-42. Thus, the equations presented in AP-42 are essentially the same as those in the
Addendum to API Publication 2571 (although more recent revisions to the API methods have
also been incorporated into the 2006 update of AP-42 Section 7.1). As seen in Figure 3-1, the
temporal variations in the hourly emissions calculated using the Addendum API 2517 (i.e., AP-
42) equations agree well with the DIAL measurements.
DIAL Addendum API 2517 API 2517
Time (h)
Figure 3-2. Comparison of DIAL and AP-42 Emission Estimates
in CONCAWE 90-hour Study (Smithers et al., 1995).
Smithers et al, (1995) determined that the DIAL measurement was 10% greater than
predicted using the Addendum to API 2517, but noted that there were times when measured
12
-------
DRAFT - Do not cite or quote
emissions were higher than predicted by the AP-42 emissions equations. During hours 2 through
6 and hours 14-16, there were periods of high wind gusts. The deck fitting correlations were
developed for average wind speeds from 2 to 15 miles per hour (mph) and API 2517 (as well as
AP-42) specifically notes that these factors only apply when the average ambient wind speed is
below 15 mph. It appears that application of the deck fitting correlations during times when
wind speeds exceed 15 mph will result in an underestimate of the deck fitting emissions. The
increase in the emissions during hours 55 to 64 appeared to be correlated with filling of the tank
from half full to full during hours 55 to 62. While the AIM2 emissions methodology approach
failed to predict this increase in emissions, Smithers el ill. (11^5) emphasized the overall
difference between the measured and calculated emissions over the duration of the 90 hour
testing period was only 10 percent. Thus, the CONCAWE study appears to support the
assertion that the AP-42 emissions methodology can provide an accurate estimate of emissions
from storage tanks in situations where detailed, sile-speeilie data are used in the calculations,
The question was then posed is hourly modeling needed to obtain accurate annual
average emissions estimates, or can annual a\ erage conditions he used to accurately estimate
annual average emissions0 To answer this question. Cohui n and Icenhour (2008) developed an
external executable program to run the TANKS model (which implements the AP-42 emissions
methodology) numerous times and sa\ e the results This allowed input of hourly meteorological
data and post-processing of the results to simulate the annual emissions that would be estimated
if the TANKS model allowed for hourly input data, Emissions estimates were developed for a
generic storage \ essel storing a single fluid (gasoline RVP-7) and using meteorological data for
Houston, Texas They round that the annual emissions estimated for the model tank when using
hourly input data agreed within 2d percent with the annual emissions estimated using annual
average inputs. This study also found that, for this case, the highest hourly emissions rate
(summer, high winds and day time high temperatures) was a factor of 5 to 10 times higher than
the annual average emissions rate. Consequently, it is reasonable to conclude that long-term
emissions rate extrapolations from DIAL measurements conducted over a very short time period
could be up to 10 times higher than the reported annual average emissions. However, it is also
possible that high-emitting tanks measured by DIAL are defective and in need of repair.
Although the Alberta DIAL Study measured emissions exceeded the annual average
emissions inventory by a factor of 30, there is not enough data to determine the cause of the
13
-------
DRAFT - Do not cite or quote
discrepancy. Because no information is provided on the calculations for the emissions inventory,
it is difficult to know if these calculations were performed in accordance with actual site
conditions or in keeping with the recommended methodologies in AP-42. Likewise, it is
unknown if the tanks measured may store different liquids over the course of the year, which
could cause greater uncertainties in extrapolating annual emissions from short-term emissions
measured for a single stored fluid. Based on these considerations, there is no direct evidence
from the Alberta DIAL Study that the equations in AP-42. when correctly applied, systematically
underestimates VOC emissions from storage tanks, and. hasctl 011 the lack of process operating
data at the time of the test, it appears that no emissions factors lor tanks can be created from the
data in the Alberta study.
3.2 Houston Area DIAL Study
The Houston Area DIAL Study (Robinson el ill . 2<)0S) measured VOC emissions from
multiple petrochemical industry sites in the I louston area during ltie summer of 2007, including a
refinery in the Houston area (BP Texas City Refinery Y The I louston Area DIAL Study at the BP
Texas City Refinery included measurements lor \ arious storage tanks, the DCU process area, the
wastewater treatment system, and se\eral llares A critical rc\ie\\ of the VOC measurements
performed at the BP Texas City Refinery was conducted hy Randall and Coburn (Ref No. 4 in
Table I -1) to compare the DIAI. emissions estimates lor sources at the BP Texas City Refinery
with refined emissions estimates bused 011 AP-42 emissions methodologies, taking into account
the process operating characteristics that existed during the time of the measurements.
Specifically, storage tank emission estimates were developed using the TANKS model,
emissions from the wastewater treatment system were developed using WATER9, and emissions
estimates for llares were de\ eloped assuming various destruction efficiencies. For the critical
review, representatives from the BP Texas City Refinery provided detailed information about
tank contents, loading rates, wastewater flows, flare flow and composition. The refinery did not
do any special sampling and analysis of storage tank or wastewater contents during the
measurement period, but unlike the Alberta DIAL study, there was a significant amount of
process data (e.g., tank contents, temperature, daily loading rates) that could be used for
emissions model input. Table 3-1 provides a summary of the comparison of the measured DIAL
emissions versus the air emissions estimates based on AP-42 methodologies. The column
14
-------
DRAFT - Do not cite or quote
"Estimated emissions using standard estimating procedures with actual conditions at the time of
the DIAL test, lb/hr" are the emissions estimates developed by Randall and Coburn (2010). The
column labeled "2007 EIQ average ozone season emission rates, lb/hr" are the average ozone
season emissions rates reported to TCEQ by the BP Texas City Refinery in their emissions
inventory.
There are over 100 storage tanks at the BP Texas City Refinery, but the Houston Area DIAL
Study included measurement scans for only 14 tanks (Robinson, el al, 2008). In their Critical
Review, Randall and Coburn (2010) identified 26 storage tanks lhal they opined may have
contributed to the measured emissions attributed to the 14 tanks specified in the Houston Area
DIAL Study. For the BP Texas City Refinery storage tanks, the reported measured emissions
were generally higher than the emissions estimated using the AP-42 emissions methodology
when considering only the tanks included in the I lonston Area DIAL Study (In Table 3-1, the
lower value in the range of estimated emissions is based on only the tanks reported to be
measured in the Houston Area DIAL Study ) However, u hen emissions were projected for all
tanks upwind of the scan plane (the upper value in the range of estimated emissions in Table 3-
1), the estimated emissions often agreed well with the DIAL measurements, although the extent
to which the upwind tanks contributed to the emissions is not well understood. The possibility
that some of the target tanks were deleeti\ e and emitting more than AP-42 estimates cannot be
ruled out Additionally, in some eases, non-storage tank emissions sources appeared to the
authors to contribute to higher than estimated emissions for certain storage tanks. For example,
Tanks I <>52. I <>53. and I < >55 (crude oil storing tanks) were downwind of the wastewater
treatment system, and it appeared to the authors that the high emissions attributed to these tanks
were likely caused In the emissions from the upwind wastewater treatment system. It is also
possible that one or more of the tanks had defective floating roofs, or that the product factors for
crude oil storage used in the AP-42 methodology, which specifically reduce the estimated
emissions from crude oil storage to account for weathering, are incorrect. The largest
discrepancy noted between measured and modeled emissions is for Tanks 53 and 55. Tank 53
was being actively filled during the time of the DIAL measurements. Emissions associated with
working losses during loading events are generally much higher than breathing losses during
non-loading periods. This likely contributed to the higher than expected measured emissions. It
15
-------
DRAFT - Do not cite or quote
Table 3-1. Summary of Emissions Comparison from the Houston Area DIAL Study
Source
Source Description
Compound
Average
DIAL flux,
lb/hra
Estimated emissions using
standard estimating procedures
with actual conditions at the time
of the DIAL test, Ib.'hr
2007 EIQ average
ozone season
emission rates. Ib/hr"
Tanks 1020,1021,
1024, and 1025
EFRC tanks storing crude oil
voc
6.4s
1.3-1.9®
2.6 - 3.5e
Tanks 1052,1053,
and 1055
EFR tanks stoning crude oil
voc
16.3d
1.8 - 2.3e
2.4 - 2.9e
Tanks 501, 502, 503,
and 504
EFR tanks storing light distillates
voc
8.6a
3.0 - 3.96
6.7 - fi.Ge
Tank 43
VFR' tank storing fuel oil #6
voc
2
1.3
0.2
9.3
1.3
0.2
Tanks 60, 63, 11, 12,
18, 42, 61, and 65
VFR and EFR tanks storing
various products
voc
9
0.6 — 9.1*
4.6 - NA*'9
Tanks 54, 55, 56,
and 98
VFR and EFR tanks storing
various products
voc
3.1"
0.3 - 9.7e
1.0 - NAe'3
Tanks 53 and 55
VFR tanks storing diesel fuel
voc
23.8d
4.8 - 5.2*
1.0 -2.0s
F-8 EBU
Activated sludge unit
voc
30
22 - 55"
6.7
API separator
API separator
voc
7
ND1
NA9
Wastewater vents
Vents from collection system
voc
9
ND
NR1
Flare #6
Ground flare
voc
13
17*
40
Temporary flare
Temporary flare
voc
6
100 - 300*
196
ULC flare
Ultracracker flare
voc
192
3 - 25*
28.3
Coker Unit C
Coker
voc
18
ND
NR
Coker Unit C
Coker while cutting coke
Benzene
1.8
ND
NR
a The tabulated values typically represent the average of calculated fluxes for several scans.
b In their 2007 emissions inventory, BP reported average ozone season emissions in Ib/d; these values were divided by 24 to estimate the tabulated average
hourly emission rates.
c EFR means external fixed roof tank.
d The results for storage tanks typically are summarized for a group of tanks because the DIAL scans typically could not isolate individual tanks. All scans along
the same path and covering the same range were grouped, and the calculated fluxes for the scans in a group were averaged. The averages for all groups of
16
-------
DRAFT - Do not cite or quote
scans that apply to a group of tanks were then averaged to obtain the tabulated flux. Note that some groups of scans captured emissions from all of the listed
tanks, while other groups of scans were downwind of only some of the listed tanks.
e Emissions were estimated for all tanks that appeared to be upwind of a group of scans, and the estimates for the individual tanks were summed. For each group
of scans, the total emissions were estimated by summing the applicable individual tank emissions estimates. The upper end of the tabulated range represents the
average of these sums. The lower end of the range represents the average emissions assuming only the tank(s) to which NPL attributed emissions were upwind
of the scans.
f VFR means vertical fixed roof tank.
g The upper end of the range could not be determined because some of the tanks could not be found in the 2007 emissions inventory. The specific API separator
of interest also could not be identified in the inventory.
h The low end of the range is based on pollutant properties used in modeling by BP, the measured benzene concentration, and annual average concentrations for
other pollutants. The high end of the range is based on using the default pollutant properties in WATER9, the measured benzene concentration, and an
assumption that all other pollutant concentrations at the time of DIAL testing were higher than average by the same percentage as benzene.
' ND means not determined.
1 NR means not reported in the annual inventory.
k Estimated emission rates are based on assumed 98 percent destruction of C3+ hydrocarbons in flare gas at the time of DIAL testing. A range is presented for the
temporary flare and ultracracker flare because the flow and composition of the flare gas varied significantly during the DIAL test period.
17
-------
DRAFT - Do not cite or quote
appears that the AP-42 emissions methodologies can reasonably predict the emissions from
storage vessels, assuming the storage vessels are not defective and the correct inputs are used.
The wastewater treatment system had measured emissions rates during the Houston Area
DIAL Study higher than projected by the refinery in its annual emissions inventory. The refinery
typically runs two activated sludge units in parallel. During the time of the DIAL measurements,
one of the activated sludge units was down for maintenance and all of the wastewater flow was
managed in the one operating activated sludge unit. The unusually high VOC emissions
measured during the test were caused by the reduced residence lime in the operating activated
sludge tank due to all of the wastewater flow being processed in the single operating activated
sludge tank. When the single activated sludge unil was modeled based on the flow and
concentrations during the time of the DIAL measurements, the modeled emissions estimates
(AP-42 emissions equations as implemented in WATF.R9 emissions model) of 22 to 55 lbs
VOC/hr agreed well with the DIAL emissions measurements of 30 lbs VOC/hr. Therefore, it
appears that the AP-42 emissions equations were accurate lor estimating the emissions from the
activated sludge system for the defined operating conditions
The Houston Area DTAT. Study was one of the first studies that documented high
emissions from a Hare due to poor destruction efficiency DI AI. measurement scans were
performed for three flu res at the refinery the ultracracker (I l.(') flare, Flare No. 6, and a
temporary flare for many of the measurement scans, two Hares were upwind of the scan plane,
butDIAI.s ability to identify and map the location of the emissions plume indicated that the
bulk of the emissions were coming from the I I.C flu re Based on vent gas flow rates, vent gas
composition data, and DIAI. measurement data, the temporary flare appeared to have a
destruction efficiency of approximately 99.9 percent, and Flare No. 6 appeared to achieve a
destruction efficiency of approximately 98 percent. However, the ULC flare appeared to achieve
a destruction efficiency of only 5<) to 80 percent. The ULC flare had a high steam to vent gas
flow and did not have a visible flame, which suggested the poor combustion efficiency was
likely caused by over-steaming of the flare. This finding led to a number of passive FTIR studies
to evaluate flare performance, particularly for steam-assisted flares. These studies are described
in greater detail in Section 4 of this document. The DIAL measurements along with the process
data provided by BP for Flare 6 appear to be sufficient for use in calculating a VOC emissions
factor; however, the Houston Area DIAL Study data from the ULC and temporary flares does
18
-------
DRAFT - Do not cite or quote
not appear to be useful in developing flare emissions factors for several reasons. First, most of
the scan planes for the ULC flare also had the temporary flare as an upwind contributing source
and it is not appropriate to assess a combined emissions factor for the two flares. Second, there
were only three DIAL scans where emissions measurements included only the temporary flare;
these three scans represent less than one hour of measurement data and do not appear to be
sufficient for use in developing an emissions factor. Third, the operation of the ULC flare is not
expected to be representative of a properly operated flare The data suggest that the flare was
oversteamed. Also, the net heating value of the fuels in the combustion zone (considering the
steam addition rate) for the ULC flare were often below 2<)o IJriiish thermal units per standard
cubic feet (Btu/scf). The EPA has recognized that this combustion zone heating value is too low,
and such low values will not be allowed for petroleum refinery flares assuming the newly
proposed refinery standards are finalized as proposed ( see 70 FR 36880)
Because the BP Texas City DTAT. measurements often included a number of upwind
emissions sources and the tests are conducted over a limited operating range (temperatures/wind
speeds), there appears to be no direct means by u liich to use the emissions measurements made
during this study to de\ clop new emissions factors for storage tanks and wastewater treatment
systems. Additionally, the comparison of the l)l.\l. measurement data with emissions model
estimates using site specific data suggests that the AP-42 emissions equations for both storage
tanks and wastew ater treatment systems appears to pro\ ide a reasonably accurate means for
estimating emissions from these sources The l)l.\l. YOC measurement data for Flare 6 are
available and have been incorporated into the emissions factor development process (EPA,
2014). The DIM. VOC measurement data for the ULC and temporary flares do not appear to be
appropriate for use in emissions I actor development because most of the DIAL scans include
emissions contributions from both flares, and it unclear what emissions are associated with each
flare. It is necessary to know the emissions attributed to a particular emissions source in order to
use data in emissions factor development.
3.3 Houston Ship Channel/Shell Deer Park DIAL Study
The Shell DIAL study (Ref No. 5 in Table 1-1; Raun and Hoyt, 2011) measured VOC
emissions from a combined petroleum refinery and chemical complex (Shell Deer Park facility)
in the Houston Ship Channel area. Like the Houston Area DIAL Study, the Shell DIAL study
19
-------
DRAFT - Do not cite or quote
included DIAL VOC measurements around various storage tanks, the DCU process area, the
wastewater treatment system, and several flares; it also included measurements around the light
olefins processing area. For many of these sources, FTIR measurements were also conducted to
improve compositional characterization of the plume. Estimates based on emissions factors
(presumably based on AP-42 emissions factor methodologies) were also reported. A summary
of the comparison of the emissions measurements and emissions factor estimates from the Shell
DIAL study for some of the measured sources is provided in Table 3-2. (Note: The measured
estimates are labelled "Estimate of 95th Upper Confidence I .imil of the Mean.")
Similar to the Houston Area DIAL Study, there are o\ er I < >0 storage tanks at the Shell
Deer Park refinery and emissions estimates were reported for 19 of these storage tanks. The
emissions measured for the 19 storage tanks cited in the report were higher than the estimated
emissions. In reviewing the scan planes for sped lie storage tank measurements, it appears that
there were often 4 or more tanks upwind of the scan plane (or between "upwind" and
"downwind" scans), but the emissions as reported (see Table 3-2) often include emissions for
only one or two of these tanks. Tt is possible that there could be more emissions sources that are
contributing to the measured emissions than arc being accounted for in the Shell DIAL study
emissions factor estimates, but the data is not conclusi\ e with respect to these tanks.
With respect to the storage tank emissions factor estimates, it appears that some of the
emissions estimates were de\ eloped using time dependent input factors (e.g., there are different
emissions factor estimates lor Southwest Tank AIM 7 for January 15 and January 19), so at least
some site-specific conditions were accounted for in these emissions estimates. However, the
report lacks any real description or documentation regarding how these calculations were made
and the level of site-specific operating data available for these emissions factor estimates. It
could be possible that only w i nd speed and temperature were varied for the differing emissions
factor estimates and data for other site-specific conditions (e.g., whether the tank was being
actively filled, actual composition of tank contents, etc.) were not accounted for in the emissions
factor estimates. The Shell DIAL study does appear to show that most of the tanks at the Shell
Deer Park refinery that were targeted for DIAL measurement scans had higher than expected
emissions during the time of the DIAL test. However, given the lack of process data, it does not
appear possible to develop or revise the emissions factor methodologies for storage tanks based
on the reported Shell DIAL study data.
20
-------
DRAFT - Do not cite or quote
Table 3-2. Summary of Emissions Comparison Included in the Shell DIAL Study Report
Area
Date
Emission
Factor Based
Calculation
(Ibs/hr)
VOC (V)
or
Benzene
(B)
Estimate of the
95th Upper
Confidence Limit
of the Mean
flbs'hrV:
Potential
Underestimation
Multi pliet
A-333
13-Jan
0.43
V
A-330
13-J an
0.45
V
A-332
13-Jan
1.27
V
Total
2.15
20.18
•:
A-3 2 5
15-Jan
0.22
V
Southwest
A-326
15-Jan
0 34
V
T an ks
Total
0,56
13.15
•i i
AP-17
19- Jan
0 48
V
Total
0.46
42.6
O
AP-17
15-Jan
0 25
AP-16
15-Jan
0. ,4
Total
0.39
- .- 1
A-310
1,»" 4
0. >7
v
West Tanks
G-324-R1
1 /' 4
0 26
V
Total
0.43
¦ .T ;
21 -Jan
20.67
CR-3
25-Mar
20.67
Average
20.67
J-327
22-Jan
0.14
J-328
22-Jan
0 s2
J-33! *
22-Jan
4 6?
J-332*
22-Jan
4 6?
Total
0 52
¦ ~ ¦¦ .7
J-327
23-Jan
0 'E
East Tanks
J-328
23-Jan
0 "2
Total
0 27
1 D. U
!- T
J-327
28-Jan
0 ' 1
J-328
28-Jan
0 't
J-33 P
28-Jan
4 6?
J-332"
28-Jan
4 63
Total
Cj ^
25-Jan
6 5
30-Jan
15
Northwest
Wastewater
5-Feb
I I 5
Average
I •
. _
9-Feb
0 0! 9
p
13-Feb
0.2
B
Average
0,11
i .3
bi
* permit limits
** from ProUCL
21
-------
DRAFT - Do not cite or quote
Table 3-2. (Continued)
Area
Date
Emission
Factor Based
Calculation
(ibs/hr)
VOC (V)
or
Benzene
!B)
Estimate of the
35th Upper
Confidence Limit
of the Mean
flbsrtiri"
Potential
Underestimation
Multiplier
East
Wastewater
s-Feb
5 88
Total
5.88
- 3. 2z>
—
Tanks T-
OL913 and T
OL920
T-OL913
8-Feb
1.15
t*
T-OL913
ID-Feb
1.17
P
T-OL913
23-Mar
1.18
B
T-OL920
S-Feb
0.83
C
T-OL92Q
10-peb
0.83
B
T-OL920
23-Mar
0.83
B
Total of Tank
Averages
2.00
1 "i T'*1
ACU BEU
12-Feb
3.4 i
p
15-Feb
3.4 f
p
Average
3.41
1 S.77
26-Mar
2.49
Total
2.49
77.4 o
31
Tanks South
of ACU BEU
D-350
2-Feb
0.03
D
D-351
12-Feb
0.09
B
D-381
15-Feb
03
B
D-352
22-Mar
0.02
-
a 1 o
** ». J
Total
0.44
** from ProUCL
Very high emissions were measured IV0111 the northwest wastewater treatment system. In
late January and early I'ehruary. the northwest wastewater treatment area had measured VOC
emissions that ranged from 4 to So limes higher than those projected based on emissions factors.
According to the Shell refinery representatives, there was a temporary malfunction of a skimmer
on one of the dissoK ed air flotation units, which caused an oil layer to develop on the top of the
tank contents that impacted some of the measurements. The emission measurements taken
during the malfunction are not appropriate for use in developing or revising the emissions factor
methodologies for wastewater treatment tanks. Emissions from the wastewater treatment facility
were also higher than estimates at times when the process was operating normally; however, the
report does not appear to include sufficient process data from which to evaluate the
appropriateness of the modeled emission estimates or the AP-42 emissions factors.
DIAL measurement scans were also conducted for the DCU process area, selected flares,
and the olefins process area. The report did not appear to include emissions factor-based
22
-------
DRAFT - Do not cite or quote
emissions estimates for these sources or operating data by which emissions estimates based on
emissions factors could be determined.
Given the lack of process operating data and occasions where the DIAL scans appeared
to include emissions contributions from additional emissions sources which are not specifically
cited in the study report, it does not appear that the Shell DIAL study can be directly used to
update or revise emissions factors. Additionally, as noted in the Shine review (Ref No. 2 in
Table 1-1), high emissions events that occur during process upsets, malfunctions, startups and
shutdowns are not typically characterized in a facility's emissions inventory and emissions
factors are not generally available or applicable for estimating emissions from these events. The
Shell DIAL study certainly documents that high emissions can occur as a result of such events,
as noted by the high emissions measured at the northwest wastewater treatment area during the
malfunction of the DAF. The AP-42 emissions 1 actors are intended for estimating emissions
during representative normal operating conditions, they are not intended to account for emissions
that occur as a result of process upsets or startup or shutdow n e\ ents.
23
-------
DRAFT - Do not cite or quote
Section 4
Passive FTIR Studies
Following the Houston Area DIAL Study (discussed in detail in Section 3.2), a number of
passive FTIR studies were performed at different refineries to investigate flare performance
under different flare and assist gas flow rates. Many of these studies were cited in the Plaintiffs'
complaint (Ref Nos. 3, 6, 7, and 8 in Table 1-1). Additional flare performance tests were
conducted on one flare at a Shell facility (Shell 201 la and 2<> I 1 h) and on one flare at an INEOS
facility (INEOS 2010a and 2010b). All of these recent 1.1 are performance studies, as well as
historical data on flare performance (see EPA. I lM3 and 1985), were re\ iewed, compiled, and
analyzed. Based on this review, the EPA prepared a report summarizing its findings and
distributed the report for peer review (EPA, 2012) Based on comments recci\ ed during the peer
review process, the data were recompiled and analyzed using increased time resolution (minute-
by-minute data) (see Sertkaya et al . 2d 13) I pon re\ iewing these data, the EPA concluded that
some refinery flares, particularly steam- or air-assisted llares with low flare gas flow rates, were
not achieving 98 percent destruction efficiency Ikvause it is the Agency's position that a well
operated flare should, at minimum, he ahle to achie\ e this destruction efficiency, the EPA has
proposed to amend the Refinery M AC T standards (40 CFR part 63 subparts CC and UUU) to
establish more stringent monitoring requirements and operating limits to ensure flares achieve a
minimum destruction efficiency of 1>K percent (79 FR 36880).
The re\ iew of passi\ e I ""I IR studies summarized above previously focused on
combustion efficiency of flares In further reviewing this data, it appears the passive FTIR
studies provide data that can be used to propose revisions to the existing AP-42 CO and NOx
emissions factors and propose a VOC emissions factor for flares. This section summarizes the
data available in each of the cited passive FTIR studies. We are proposing to revise section 13.5
of AP-42 to incorporate the new emissions factors developed with these data.
4.1 2010 TCEQ Flare Study
The purpose of the 2010 TCEQ flare study project (Ref No. 3 in Table 1-1; Allen and
Torres, 2011) was to conduct field tests to measure flare emissions and collect process and
operational data in a semi-controlled environment to determine the relationship between flare
24
-------
DRAFT - Do not cite or quote
design, operation, vent gas lower heating value (LHV), vent gas flow rate, destruction and
removal efficiency (DRE), and combustion efficiency (CE). The TCEQ's primary objectives, as
stated in the study report, included the following:
Assess the potential impact of vent gas flow rate turndown on flare CE and VOC
DRE.
Assess the potential impact of steam/air assist on flare CE and VOC DRE at various
operating conditions, including low vent gas flow rates
Determine whether flares operating over the range of requirements stated in 40 Code
of Federal Regulations (CFR) § 60.18 achie\ e l lie assumed hydrocarbon DRE of 98
percent at varying vent gas flow rale Uinulou n. assist ratios and vent gas heat content.
Identify and quantify the hydrocarbon species in flare plumes \ isualized with passive
infrared cameras.
Commensurate with the study objectives, the stuck report focused on VOC emissions at
or near the operating conditions w here llare perform a nee deteriorates. Because the study
specifically investigated flare performance u liile \ arying steam flow rates, some of the
emissions measured during these tests are not likely to be representative of normal operating
conditions. However, if the data are limited to those times when the flare was meeting the
recommended combustion efficiency, it may be possible to use the data to develop emissions
factors. Although the study report ( Allen and Torres. 2<)| 1) does not include any data on the CO
andMK emissions, raw data on the emissions concentration measured during this study are
available and have been incorporated into the emissions factor development process (EPA,
2014).
4.2 Marathon Petroleum Flare Tests
Passive FTIR tests were conducted at two different Marathon Petroleum refineries (Ref
No. 6 and 7 in Table 1-1). The main objective of these tests was to better understand the impacts
of steam on the overall performance of the flare CE. As with the TCEQ flare study, many of the
test runs were conducted with varying steam flow rates to determine how varying steam rates
impact CE. Consequently, some of the emissions measured during these tests are not likely to be
representative of normal operating conditions. However, if the data are limited to those times
when the flare was meeting the recommended CE, it may be possible to use the data to develop
25
-------
DRAFT - Do not cite or quote
emissions factors. Although the reports do not include any data on the VOC, CO and NOx
emissions, raw data on the emissions concentrations measured during the Marathon flare tests are
available and has been incorporated into the emissions factor development process (EPA, 2014).
4.3 Flint Hills Flare Test
Passive FTIR tests were conducted on two different flares at the Flint Hills Port Arthur
Refinery (Ref No. 8 in Table 1-1). The overall objecti\ es of the Flint Hills flare test program
were as follows:
Evaluate the impacts of CE over a range of operating scenarios by changing both flare
vent gas composition and steam rales
Evaluate key operating parameters such as steam to vent gas ratio (S/VG) and Net
Heating Value of the Combustion /one (M TYcz) as indicators thai may assist in
maintaining flare operation at high efficiency conditions during day-lo-day operation.
As with the other flare studies. because the objecii\ e of the tests was to evaluate
conditions that affect flare CE, many of the lest runs were conducted with varying steam flow
rates that would likely not be representati \ e of normal operati ng conditions. However, if the
data are limited to Ihose limes u hen the Hare was meeting the recommended CE, it may be
possible to use the data lo de\ clop emissions factors Although the report does not include any
data on the YOC. CO and NOx emissions, raw data on the emissions concentrations measured
during this study are a\ ailaMe and ha\ e been incorporated into the emissions factor development
process (I-PA. 2<>14).
4.4 Additional Mare Tests
In addition to the passi\ e I 'TIR studies cited in the complaint, we are aware of passive
FTIR studies conducted on two other industrial flares (Shell, 201 la and 201 lb; ENIOS, 2010a
and 2010b). These studies were similar to the previous passive FTIR studies in that they were
largely conducted to identify the operating limits where flare performance begins to deteriorate.
As with the other flare studies, because the objective of the tests was to evaluate conditions that
affect flare CE, many of the test runs were conducted with varying steam flow rates that would
likely not be representative of normal operating conditions. However, if the data are limited to
those times when the flare was meeting the recommended CE, it may be possible to use the data
26
-------
DRAFT - Do not cite or quote
to develop emissions factors. Raw data for measured VOC, CO and (for the INEOS study only)
NOx emissions concentrations collected during these studies are available and have been
incorporated into the emissions factor development process (EPA, 2014).
27
-------
DRAFT - Do not cite or quote
Section 5
Tank Emissions Tests
Global Companies LLC (Global) operates a petroleum terminal handling distillate and
residual oil products in Chelsea, Massachusetts. Global disco\ ered that its carbon bed (installed
as an odor control system) was not adequately sized lo control \ apors from its residual oil tanks.
As a result, Global conducted testing on the tank emissions lo in\ estigate the issue and
discovered that emissions from the residual oil tanks had been significantly underestimated. In
2007, Global submitted a self-disclosure to the Massachusetts Department of Environmental
Protection (MassDEP) and EPA Region 1 regarding these excess emissions
As a result of this incident, EPA Region I required two facilities (Global in South
Portland, Maine and Sprague Operating Resource I.I.C in Searsport, Maine) to conduct a series
of tests to quantify emissions from storage tanks to determine compliance with various Clean Air
Act requirements. These reports were re\ iewed in the context of determining whether changes
should be made to 1 lie existing emissions estimation methodologies for tanks in Chapter 7 of
AP-42.
5.1 Global C ompanies LLC-South Portland, Maine
Global operates a petroleum terminal handling distillate and residual oil products and
asphalt in South Portland, Maine LIW Region 1 issued a testing order to Global's South
Portland facility under Section I 14 of the Clean Air Act. Global tested headspace vapors from
Tank No. 9 in July and August 2012 (Eastmount, 2012b). Tank No. 9 stored liquid asphalt at the
time of the test. A temporary total enclosure (TTE), designed in accordance with EPA Method
204, was fitted over the single tank exhaust vent. The total hydrocarbons (THC) concentration,
measured with EPA Method 25A, and methane concentration, measured with EPA Method 18,
of the storage tank emissions were logged continuously for a period of 30 days. The testing
timeframe included a period of filling the storage tank so that both breathing and working losses
were represented. The methane emissions were subtracted from the THC readings to get a non-
methane total gaseous organic compound concentration, as a measure for VOC. Global
28
-------
DRAFT - Do not cite or quote
conducted similar testing on Tank No. 3 in April 2013 (Eastmount, 2013b). Tank No. 3 stored
No. 6 fuel oil at the time of the test. A TTE was fitted on the two exhaust vents. The THC and
methane concentrations of the emissions were logged continuously for a period of 15 days. The
testing timeframe included a period of filling the storage tank so that both breathing and working
losses were represented.
The measured VOC tank breathing losses when extrapolated over the course of a year
were 5.3 tons per year (tpy) for Tank No. 9 and 7.43 tpy for Tank No. 3. Hourly emissions
during the filling operation were observed to increase approximately 33% for Tank No. 9 and
100% for Tank No. 3 from the emissions observed during non-Ill ling operations.
While the data from the testing at the (ilohal facility are interesting, the equations for
estimating tank emissions in AP-42 are not based on a single emissions factor (EPA, 1995). AP-
42 Chapter 7 contains a series of complex equations thai rely heavily on silo-specific inputs,
including tank specific inputs (diameter, height, shell color and condition, lype of roof, etc.), fuel
specific inputs (vapor pressure, vapor molecular weight, ill rough put, etc.), and location specific
inputs (to determine average temperature and solar insolation information). In order for the
algorithms to provide reasonable estimates, the site specific information must be correct. In
order to determine u hether re\ isions to the AP-42 Chapter 7 equations are appropriate, targeted
testing would need to he performed to isolate each input parameter to determine how each
parameter affects the output of the equations Testing would need to include tanks of different
sizes, different places in the country, different times of year, different throughputs, different
colors, different storage materials, etc At this time, it does not appear that the necessary data to
perform this analysis is a\ailaMe
5.2 Sprague Operating Resource LLC- Searsport, Maine
Sprague Operating Resource LLC (Sprague) operates a tank terminal in Searsport, Maine
handling bulk liquid cargoes (including residual oil and asphalt), dry bulk products, and special
heavy lift projects. Sprague also received a testing order under Section 114 of the Clean Air Act
from EPA Region 1.
Sprague tested headspace vapors from Tank No. 3 in July and August 2012 (Eastmount,
2012a). Tank No. 3 stored No. 6 fuel oil at the time of the test. A TTE was fitted over the two
tank exhaust vents. The THC and methane concentrations were logged continuously for a period
29
-------
DRAFT - Do not cite or quote
of 30 days. The testing timeframe included a period of filling the storage tank so that both
breathing and working losses were represented. Sprague conducted similar testing on Tank No.
2 in May 2013 (Eastmount, 2013 a). Tank No. 2 stored liquid asphalt at the time of the test. A
TTE was fitted on the single exhaust vent. The THC and methane concentrations of the
emissions were logged continuously for a period of 15 days. The testing timeframe included a
period of filling the storage tank so that both breathing and working losses were represented.
The measured VOC tank breathing losses when extrapolated over the course of a year
were 10.6 tpy for Tank No. 3 and 4.2 tpy for Tank No 2 Hourly emissions during the filling
operation were observed to be approximately four times higher lor Tank No. 3 and decreased by
20 percent for Tank No. 2 from the emissions obser\ ed during non-Ill ling operations. These
emissions were much higher than expected, based on Sprague's emissions inventory estimates
for the years 2006-2009 (CITE). A review of the emissions in\ entory calculations revealed the
use of several default values.
Once again, while the data from the testing at the Sprague facility are interesting, in order
to conclude whether revisions to the AP-42 Chapter 7 equations are appropriate, targeted testing
would need to be performed to isolate each input parameter to determine how each parameter
affects the output of the equations. Testing would need to include tanks of different sizes,
different places in the country, different limes of year, different throughputs, different colors,
different storage materials, etc At this time, it does not appear that the necessary data to perform
this analysis is a\ ailahle Without conducting further testing, it does not appear possible to say
whether there are issues with the equations in AP-42 Chapter 7.
The testing performed at Sprague did highlight the need for site-specific data when using
the AP-42 Chapter 7 equations for estimating emissions from tanks. This is especially important
for a material like asphalt, for w liich no default data is available, or No. 6 fuel oil, which is often
mixed with more volatile cutter material. The majority of emissions in this type of material are
expected to come from the cutter stock, and as such, it is important to account for the vapor
pressure contributed by the cutter when determining the vapor pressure of the fuel oil (API,
2013).
30
-------
DRAFT - Do not cite or quote
Section 6
Proposed Conclusions
The following conclusions are based on a review of the reports cited in the Plaintiffs'
May 2013 complaint, along with other associated reports. a\ ail able documents, and recent tank
emissions testing. Based on our review, we propose Ihe following conclusions:
1. Emissions inventory estimates can underestimate actual emissions because they often
do not adequately account for site-specific conditions and do not generally account
for unusual emissions that occur as a result of process upsets, nial functions, startups
and shutdowns. Improving emissions in\ entoi y guidelines, as done by TCEQ (2013)
or as provided by the Refinery Protocol (RTI. 2d I I), appears to be the most effective
way to improve emissions in\ entoi ies
2. DIAL studies are useful lor identifying sources with high emissions. However, these
studies 1o date do not appear to be useful for de\ eloping emissions factors because:
a. The a\ ailable l)l.\l. study reports do not provide the necessary process
operating data by which to normalize the measured emissions in order to
de\elop an emissions factor
b It can be challenging to isolate a single emissions source within a site
boundary It is not possible to develop an emissions factor when the
emissions cannot be properly apportioned between the sources.
3. Raw data from passi\ e FTIR flare studies do include measurement data for CO, NOx,
and VOC It may be possible to use these data to revise emissions factors for CO and
NOx and propose an emissions factor for VOC from flares. Because passive FTIR
studies for flares have generally focused on determining flare CE and identifying
factors that affect flare CE, not all of the data are representative of normal operating
conditions. However, if the data are limited to those times when the flare was
meeting the recommended CE, it may be possible to use the data to develop
emissions factors. As such, the EPA has incorporated this data into the emissions
31
-------
DRAFT - Do not cite or quote
factor development process. We are proposing to revise section 13.5 of AP-42 to
incorporate the new emissions factors developed with these data.
4. For tanks, the equations in AP-42 can inaccurately estimate emissions when default
values are used inappropriately or when site-specific inputs are not entered into the
equations. It is important to develop site-specific vapor pressure information for
materials, like No. 6 fuel oil, which are routinely mixed with more volatile materials.
It is also important to note that the emissions estimation procedures only account for
emissions in normal operating scenarios. 1 -missions from tanks that are improperly
operated, defective (e.g. damaged floating root" rim seals and deck fittings), or in
disrepair cannot be accurately estimated using these methods
5. It does not appear possible to determine if the tank emissions estimation equations in
AP-42 Chapter 7 need to be revised In order to determine this, targeted testing
would need to be performed to isolate each input parameter to determine how each
parameter affects the output of the equations Testing would need to include tanks of
different sizes, locations, throughputs, colors, and content. Such testing would need
to occur at different times throughout the year At this time, it does not appear that the
necessary data to perform this analysis is available
32
-------
DRAFT - Do not cite or quote
Section 7
References
Allen, D. T. and V. M. Torres. 2011. TCEQ 2010 Flare Study Final Report. Prepared for the
Texas Commission on Environmental Quality. August 1. Available at:
www.tceq.texas. gov/assets/public/implementation/air/rules/Flare/2010flarestudv/2010-flare-
studv-final-report.pdf
API (American Petroleum Institute). 2013. Manual of I'd rote urn Measurement Standards,
Chapter 19.4. Evaporative Loss Reference Information ami S/>eciation Methodology. Third
Edition, Addendum 1, American Petroleum Institute No\ cm her 2013.
Chambers, A. and M. Strosher. 2006. Refinery / K-monst ration oj ()/>ncal Technologies for
Measurement of Fugitive Emissions andjor leak Detection. Prepared for Environment
Canada, Ontario Ministry of the Environment, and Alberta Environment March 31; revised
November 1. Available at:
www.environmentalintegritv.org'ixll'publications I-IP An I) Total Upset.pdf
Clean Air Engineering, Inc. 2010a. t'eijoiimince / esi oj a Steam-Assisted Flare with Passive
FTIR. Prepared for Marathon Pel ml cum Co . 1.1.('. Texas Refining Division, Texas City,
Texas. May. Available at'
www.tceq.texas uo\ assets nuMic implementation air/rules/Flare/2010flarestudv/mpc-txc.pdf
Clean Air Engineeri ng. Inc 2<)|( >h / 'eijormance / est of a Steam-Assisted Elevated Flare with
Passive FTIR-Detroit. Prepared lor Marathon Petroleum Co., LLC. Detroit Refinery,
Detraoit. Michigan. No\ ember A\ailahle at
wuu teen lexas uo\ assets nublie implementation,air/rules/Flare/2010flarestudv/mpc-
detroit pdf
Clean Air I jigineering, Inc 2< > I I /'///A' Test of a Steam-Assisted Elevated Flares - Port
Arthur. Prepared for Mint Mills Resources Port Arthur. LLC. Port Arthur Chemicals, Port
Arthur, Texas June I 7 A\ ailable at http://www.regulations.govA Docket Item No. EPA-
HQ-OAR-21) 11 )-t)6N2-<) 107
Coburn, J., and M. Icenhour. 2<)()K Preliminary Analysis of Short-term Variability in Storage
Vessel Emissions. Memorandum from Jeff Coburn and Melissa Icenhour, RTI International,
to Brenda Shine, EPA/OAQPS/SPPD. December 29, 2008.
Coburn, J. 2013a. Impact Estimates for Delayed Coking Units. Memorandum from Jeff Coburn,
RTI International, to Brenda Shine, EPA/OAQPS/SPPD. September 12, 2013. Available at
http://www.regulations.govA Docket Item No. EPA-HQ-OAR-2010-0682-0202.
33
-------
DRAFT - Do not cite or quote
Coburn, J. 2013b. Petroleum Refinery Sector Rule: Operating Limits for Flares. Memorandum
from Jeff Coburn, RTI International, to Andrew Bouchard and Brenda Shine, EPA/OAQPS.
December 12, 2013. Available at http://www.regulations.govA Docket Item No. EPA-HQ-
OAR-2010-0682-0206.
Eastmount Environmental Services, LLC. 2012a. VOC and VOC HAP Emissions Testing from
Residual Oil Tank No. 3 Tank Headspace and Loading Operations at Sprague 's Sear sport,
Maine Terminal. Prepared for Sprague Operating Resources LLC. Searsport Terminal,
Searsport, Maine. October. Available at:
http://www.epa.gov/ttn/chief/consentdecree/Sprague Oil zip
Eastmount Environmental Services, LLC. 2012b. VO(' and !'()(' HAP Emissions Testing from
Asphalt Tank No. 9 Tank Headspace and Loading ()/KTaiions at Global's South Portland,
Maine Terminal. Prepared for Global Companies I.I.C South Portland Terminal, South
Portland, Maine. November. Available at'
http://www.epa.gov/ttn/chief/consentdeci ee (ilohal Asphalt.zip
Eastmount Environmental Services, LLC. 20 13a \'()C and I'OC HAP Emissions Testing from
Asphalt Storage Tank No. 2 Tank Headspace and Loading (Iterations at Sprague's
Searsport, Maine Terminal. Prepared for Sprague Operating Resources LLC. Searsport
Terminal, Searsport, Maine. July A\ailahlc at:
http://www.epa.gov/ttn/chief/consenldeciee Spramie Asnhall zip
Eastmount Environmental Sei\ices. LLC 2<) 13h V( H ' ami V( H ' HAP Emissions Testing from
Residual Oil Tank So. 3 tank /leails/Hice and / oading ()/K-rations at Global's South
Portland, Maine lerminal. Prepared for (ilohal Companies LLC. South Portland Terminal,
South Portland, Maine August A\ailaMe at
http://www.epa uo\ tin chief consenldecree (ilolxil Oil.zip
EPAd'S. Ln\ironmental Protection Agency) llM3 Hare Efficiency Study. Prepared for U.S.
En\ironmental Protection Agency. Industrial Environmental Research Laboratory, Research
Triangle I'ark, NC. July ll>K3 LPA-O00/2-83-052. Available at:
http://www.epa.gov/ttn chiel'ap42 ch 13/related/ref 01cl3s05 ianl995.pdf
EPA (U.S. Envi ronmental Protection Agency). 1985. Evaluation of the Efficiency of Industrial
Flares: Flare Head / K-sign and Gas Composition. Prepared for U.S. Environmental
Protection Agency. Air and Lnergy Research Laboratory, Research Triangle Park, NC.
September 1985. EPA-Mio 2-S5-106.
EPA (U.S. Environmental Protection Agency). 1995. Compilation of Air Pollutant Emission
Factors, Volume 1: Stationary Point and Area Sources, AP-42, Fifth Edition, U.S.
Environmental Protection Agency, Office of Air Quality Planning and Standards. January
1995. Available at: http://www.epa.gov/ttn/chief/ap42/index.html
EPA (U.S. Environmental Protection Agency). 2011. Petroleum Refinery Emissions Information
Collection, Part VI. Emissions Inventory Data. OMB Control No. 2060-0657, Approved
March 28, 2011. Available at http://www.regulations.govA Docket Item No. EPA-HQ-OAR-
2010-0682-0062.06.
34
-------
DRAFT - Do not cite or quote
EPA (U.S. Environmental Protection Agency). 2012. Parameters for Properly Designed and
Operated Flares. Prepared by U.S. Environmental Protection Agency, Office of Air Quality
Planning and Standards, Research Triangle Park, NC. April 2012. Available at:
http://www.epa.gov/airtoxics/flare/2012flaretechreport.pdf
EPA (U.S. Environmental Protection Agency). 2014. DRAFT Review of Emissions Test
Reports for Emissions Factors Development for Flares and Certain Refinery Operations.
Prepared by U.S. Environmental Protection Agency, Office of Air Quality Planning and
Standards, Research Triangle Park, NC. August 2014. Available at:
http://www.epa.gov/ttn/chief/consentdecree/draft report ef.pdf
INEOS ABS (USA) Corproation. 2010a. Passive Fourier Irans/orin Infrared Technology (FTIR)
Evaluation of POO 1 Process Control Device at INI A )S. 1HS (I SA) Corporation, Addyston,
Ohio Facility. July 2010. Available at at http://wi\w regulations uovA Docket Item No. EPA-
HQ-OAR-2010-0682-0159. .
INEOS ABS (USA) Corproation. 2010b. Passive lonrier Transform Infrared Technology
(FTIR) Evaluation ofPOOl Process Control / k-vice ai IX/A )S ABS (LS. 1 > ('orporation,
Addyston, Ohio Facility. Supplemental Report August 2<)|o Available at al
http://www.regulations.govA Docket Item No. EIW-IIQ-OA11-2010-0682-0159.
Lucas, B. 2007. Collection ofDetadei! Henzene Emissions / hila from 22 Petroleum Refineries.
Memorandum from Bob T.ucas. EPA SPPI). to I'PA Docket No. EPA-HQ-OAR-2003-0146.
August 20, 2<)i)7 A\ ailaMe at hup www regulations uo\ . Docket Item No. EPA-HQ-
OAR-2003-i) 14o-i »< > I 5
Milligan, P., F. Maitinsky. k (iood. and 1} I Nelson 2< >06. Evaluation Report: EPA Can
Improve Emission laciors / h-velopnieni ami \ lanagement. Office of Inspector General
Report \o 2<)()(i-P-()i)| 7 A\ailahle al hup www epa.gov/oig/reports/2006/20060322-
2006-P-"')') I 7 r>dI'
Randal I.I), and J. Cohu in 2<)|o ('iiiical Review of DIAL Emission Test Data for BP Petroleum
Refinery in lexas City, lewis. Prepared for U.S. Environmental Protection Agency, Office of
Air Quality Planning and Standards, Research Triangle Park, NC. EPA 453/R-10-002.
November. A\ ailable at: www.epa.gov/airtoxics/bp dial review report 12-3-10.pdf
Raun, L. and D. W I loyt 2d I I. Measurement and Analysis of Benzene and VOC Emissions in
the Houston Ship Channel. \rea and Select Major Stationary Sources Using DIAL
(Differential Absorption Light Detection and Ranging) Technology to Support Ambient HAP
Concentrations Reductions in the Community (DIAL Project). City of Houston, Bureau of
Pollution Control and Prevention. July 20. Available at:
www, greenhoustontx. gov/dial20110720.pdf
35
-------
DRAFT - Do not cite or quote
Robinson, R., T. Gardiner, and B. Lipscombe. 2008. Test Report: Measurements ofVOC
emissions from petrochemical industry sites in the Houston Area using Differential
Absorption Lidar (DIAL) during Summer 2007. Prepared for the Texas Commission on
Environmental Quality. Reference No. QBN1701-TCEQ-2007. February 6. Available at:
http://www.tceq.texas.gov/assets/public/implementation/air/am/contracts/reports/oth/200802
06-dial test report.pdf
RTI (RTI International). 2011. Emission Estimation Protocol for Petroleum Refineries.
Version 2.1.1, Final ICR Version Corrected. Prepared for U.S. Environmental Protection
Agency, Office of Air Quality Planning and Standards. Research Triangle Park, NC. May.
Docket Item No. EPA-HQ-OAR-2010-0682-0060 Most current version also available at
http://www.epa.gov/ttn/chief/efpac/protocol/.
Sertkaya, A., J. Renzaglia, and M. Thomas. 2013 Memorandum from Aylin Sertkaya, Jason
Renzaglia, and Mae Thomas, ERG, to Andrew IJouchard and IJrenda Shine, EPA. Flare
Performance Data: Summary of Peer Review ('omment and Additional Data Analysis for
Steam-Assisted Flares. July 15, 2013. A\ ailahlc at http://www.reuuhilions.govA Docket
Item No. EPA-HQ-OAR-2010-0682-0200.
Shell Global Solutions (US) Inc. 201 la Shell Deer I'ark Helming LP Deer Park Refinery East
Property Flare Test Report. Prepared l\\ Shell Global Solutions (US) Inc., Houston, TX
(April). Available at:
http://www.tceq.texas.gov/assets/|">uMic implementation air rules Flare/201 Oflarestudv/sdp-
epf-test.pdf
Shell Global Solutions (I S) Inc 2<>l II"). Shell / K-ei I'ark Site Deer Park Chemical Plant OP-3
Ground Flare Performance I est Report Prepared by Shell Global Solutions (US) Inc.,
Houston, TX May
Shine. 1} 2')')7 I'oieniial / <>\\ Hms of Reported I'OC Emissions from the Petroleum Refining
Industry Memorandum from Iirenda Shine, EPA/SPPD, to EPA Docket No. EPA-HQ-
0.\R-2<)()3-<)146 dated July 27. 2<)i)7 Available at http://www.regulations.govA Docket Item
No i :i\\-l IQ-OAR-20113-') 1>< > I < )
Smithers, B . McKay. J ; Van Ophem, G.; Van Parijs, K.; White, L. 1995. VOC Emissions from
External Floating Roof tanks: Comparison of Remote Measurements by Laser with
Calculation Methods. Prepared for CONCAWE Air Quality Management Group, Report No.
95/52, Brussels. January. Available at:
http://files.gamta.1t/aaa/Tipk/tipk/4 kiti%20GPGB/46.pdf
Sprague Operating Resource LLC. Searsport Summary 2005-2009 TRIReporting. Microsoft
Excel file. Created February 16, 2011. Available at:
http://www.epa.gov/ttn/chief/consentdecree/Searsport Summary 2005-2009 TRI
Reporting.xls
TCEQ (Texas Commission on Environmental Quality). 2013. 2012 Emissions Inventory
Guidelines. TCEQ Report No. RG-360/12. January. Available at:
http://www.tceq.texas.gov/publications/rg/rg-360 12/index.html
36
-------
DRAFT - Do not cite or quote
TNRCC (Texas Natural Resource Conservation Commission). 2000. Texas 2000 Air Quality
Study: Landmark Research in Texas on Ozone and Fine Particle Air Pollution. TNRCC
Report No. GI-260. (August - September). See also
http://www.utexas.edu/research/ ceer/texaq s/
37
-------
DRAFT - Do not cite or quote
Appendix A
COMPLAINT
MAY 2013
38
-------
Case l:13-cv-00621 Document 1 Filed 05/01/13 Page 1 of 20
UNITED STATES DISTRICT COURT
DISTRICT OF COLUMBIA
AIR ALLIANCE HOUSTON )
2409 Commerce Street, Suite A )
Houston, TX 77003; )
COMMUNITY IN-POWER AND }
DEVELOPMENT ASSOCIATION, INC \ r N
1301 Kansas Avenue; < SC -
Port Arthur, TX 77640; I
LOUISIANA BUCKET BRIGADE )
4226 Canal Street )
New Orleans, LA 70119; and )
)
TEXAS ENVIRONMENTAL JUSTICE )
ADVOCACY SERVICES )
6733 Harrisburg Boulevard ^
Houston, TX 77011; ]
Plaintiffs, \
)
v. )
)
Bob Perciasepe, in his official capacity as )
Acting Administrator, United States )
Environmental Protection Agency, )
1101A EPA Headquarters, Ariel Rios Building)
1200 Pennsylvania Avenue, NW J
Washington, D.C. 20460 )
Defendant. {
COMPLAINT FOR DECLARATORY AND INJUNCTIVE RELIEF
INTRODUCTION
1. Plaintiffs Air Alliance Houston, Community In-Power and Development
Association (CIDA), Louisiana Bucket Brigade, and Texas Environmental Justice Advocacy
Services (TEJAS), (collectively, "Plaintiffs") bring this action pursuant to section 304(a)(2) of
1
-------
Case l:13-cv-00621 Document 1 Filed 05/01/13 Page 2 of 20
the Clean Air Act ("CAA" or the "Act"), 42 U.S.C. § 7604(a)(2), to compel Defendant Bob
Perciasepe, Acting Administrator of the United States Environmental Protection Agency
("EPA"), to perform the nondiscretionary duties required by section 130 of the Act, 42 U.S.C. §
7430. Specifically, the Administrator has failed to review and, if necessary, revise the emission
factors for volatile organic compounds (VOCs), carbon monoxide, and nitrogen oxides at least
once every three years. 42 U.S.C. § 7430.
2. An emission factor is a "representative value" or "tool" used to estimate
emissions of a specific pollutant from an air pollution source. Emission factors are central to the
CAA, and are used to calculate roughly eighty percent of air emissions from virtually all
industrial sectors. Emissions data calculated from emission factors is used to: design regulations;
develop emission control strategies; establish emission limits and other applicable permit
requirements for major sources; guide enforcement priorities; and evaluate compliance with air
quality standards.
3. Refineries and petrochemical plants release VOCs and other toxic pollutants that
poses significant health risks to nearby communities and to the environment. VOCs are organic
chemicals that readily vaporize into the air, and can combine with nitrogen oxides to form ozone.
Ozone is a criteria pollutant that is responsible for respiratory ailments and increased hospital
admissions for coughing, chest pain, throat and nose irritation, lung inflammation and other
respiratory issues. Some VOCs are also toxic pollutants, such as benzene, 1,3-butadiene, and
hexane—all known carcinogens.
4. VOC emissions from petroleum refineries, petrochemical plants, and other
industrial sources are significantly underestimated because inaccurate emission factors are used
to quantify emissions from industrial flares ("flares"), liquid storage tanks ("tanks"), and
2
-------
Case l:13-cv-00621 Document 1 Filed 05/01/13 Page 3 of 20
wastewater collection, treatment, and storage systems ("wastewater treatment systems") at these
facilities. Numerous scientific studies have found that VOC emissions from these sources are
several orders of magnitude higher than emission factor estimates, in some cases measuring
VOC emissions 132 times above the estimated amount.
5. Under section 130 of the Act, 42 U.S.C. § 7430, the Administrator has a
mandatory duty to review and, if necessary, revise, emission factors used to estimate emissions
of VOCs from emission sources at least once every three years. The failure to comply with this
mandate compromises EPA's ability to implement the CAA in a manner that protects public
health and the environment. Yet, the Administrator has failed to perform the nondiscretionary
duty to review and, if necessary, revise the emission factors used to estimate VOC emissions
from flares, tanks, and wastewater treatment systems within the statutory timeframe.
6. With this action, Plaintiffs seek to compel the Administrator to expeditiously
complete a review of the VOC emission factors for flares, tanks, and wastewater treatment
systems, and, if necessary, revise these factors as required by section 130 of the Act, 42 U.S.C. §
7430.
JURISDICTION AND VENUE
7. This Court has jurisdiction over this action pursuant to 42 US.C. § 7604(a)(2)
(action arising under the CAA citizen suit provision), 28 U.S.C. § 1331 (federal question), and
28 U.S.C. § 1361 (mandamus). This Court may order the Administrator to perform the requisite
acts and duties, may issue a declaratory judgment and may grant further relief pursuant to 42
U.S.C. § 7604(a), (d) and 28 U.S.C. §§ 2201, 2202.
3
-------
Case l:13-cv-00621 Document 1 Filed 05/01/13 Page 4 of 20
8. Pursuant to section 304(a) of the CAA, 42 U.S.C. § 7604(a), "the district courts
shall have jurisdiction ... to order the Administrator to perform such act or duty [which is not
discretionary]."
9. Plaintiffs have a right to bring this action pursuant to section 304(a)(2) of the
CAA, 42 U.S.C. § 7604(a)(2), and the Administrative Procedure Act, 5 U.S.C. §§701 to 706.
10. By certified letter posted July 18, 2012, Plaintiffs sent Administrator written
Notice of Intent to Sue ("Notice") and have thereby complied with the notice requirements of
section 304(b)(2) of the CAA, 42 U.S.C. § 7604(b)(2), and 40 C.F.R. pt. 54. See Ex. A. More
than 60 days have passed since Plaintiffs provided Notice. The Administrator has not responded
to Plaintiffs Notice or remedied the alleged violations. Therefore, an actual controversy exists
between the parties.
11. Venue is vested in this Court under 28 U.S.C. § 1391(e) because the
Administrator resides in this district.
PARTIES
12. Plaintiff Air Alliance Houston is a nonprofit, non-membership organization,
incorporated and existing under the laws of the State of Texas, located in Houston, Texas. Air
Alliance Houston works to reduce air pollution exposure and related health effects on behalf of
communities in the Houston region. Air Alliance Houston is concerned that inaccurate
accounting of VOC emissions from the many petroleum refineries and petrochemical plants in
the Houston area undermine its efforts to reduce air pollution. Communities in the Houston
region are vulnerable to VOCs emitted from Houston facilities because of: the toxins they
contain and the smog they produce; the lack of accurate information about air pollutants,
concentrations, and resulting exposures has made, and continues to make, it difficult for
4
-------
Case l:13-cv-00621 Document 1 Filed 05/01/13 Page 5 of 20
Houstonians to determine how to best protect themselves; the inaccurate emissions data caused
by poor quality emissions factors has made, and continues to make, it difficult for community
members to effectively exercise their right to review and comment on CAA permits designed to
protect ambient air quality; and, the underreporting of emissions may expose community
members to pollutants at levels that are higher than the law allows and in concentrations
deleterious to human health.
13. Plaintiff Community In-Power and Development Association (CIDA) is a
nonprofit, membership corporation located in Port Arthur, Texas. CIDA advocates for its
members' environmental justice, social, and economic rights. CIDA is concerned about accurate
accounting of VOC emissions from petroleum refineries and petrochemical manufacturing plants
in the Port Arthur area. Members of CIDA are exposed to VOCs, and the related toxins and
smog, emitted from these facilities where they live and work; the lack of information about air
pollutants, concentrations, and resulting exposures has made, and continues to make, it hard for
them to determine how best to protect themselves; the inaccurate and poor quality of emissions
data has made, and continues to make, it difficult for them to effectively exercise their right to
review and comment on CAA permits designed to protect ambient air quality; and, the
underreporting of emissions may expose members to pollutants at levels that are higher than the
law allows and in concentrations deleterious to human health. CIDA brings this action on
behalf of itself and its members that live, work, and recreate near petroleum refineries and
petrochemical plants in the Port Arthur area in Southeast Texas.
14. Plaintiff Louisiana Bucket Brigade is a nonprofit, membership-based
environmental health and justice organization located in New Orleans, Louisiana. Its mission is
to work with communities to create Louisiana neighborhoods that are free of toxic air pollution
5
-------
Case l:13-cv-00621 Document 1 Filed 05/01/13 Page 6 of 20
from refineries and other industrial sources. Louisiana Bucket Brigade is concerned about
accurate accounting of VOC emissions from Louisiana's many petroleum refineries and
petrochemical manufacturing plants. Members of Louisiana Bucket Brigade are exposed to
VOCs, and the related toxins and smog, emitted from these facilities where they live and work;
the lack of information about air pollutants, concentrations, and resulting exposures has made,
and continues to make, it hard for them to determine how best to protect themselves; the
inaccurate and poor quality of emissions data has made, and continues to make, it difficult for
them to effectively exercise their right to review and comment on CAA permits designed to
protect ambient air quality; and, the underreporting of emissions may expose members to
pollutants at levels that are higher than the law allows and in concentrations deleterious to human
health. Louisiana Bucket Brigade brings this action on behalf of itself and its members that live,
work, and recreate near petroleum refineries and petrochemical plants in Louisiana.
15. Plaintiff Texas Environmental Justice Advocacy Services (TEJAS) is a nonprofit
corporation located in Houston, Texas. TEJAS's mission is to create sustainable, healthy
communities in the Houston Ship Channel region by educating individuals on health impacts
from environmental pollution and empowering individuals to promote enforcement of
environmental laws. TEJAS is particularly concerned about accurate accounting of VOC
emissions from petroleum refineries and petrochemical manufacturing plants in the Houston
Ship Channel. Members of TEJAS are exposed to VOCs and the related toxins and smog
emitted from these facilities where they live and work; the lack of information about air
pollutants, concentrations, and resulting exposures, has made, and continues to make, it hard for
them to determine how best to protect themselves; the inaccurate and poor quality of emissions
data has made, and continues to make, it difficult for them to effectively exercise their right to
6
-------
Case l:13-cv-00621 Document 1 Filed 05/01/13 Page 7 of 20
review and comment on CAA permits designed to protect ambient air quality; and, the
underreporting of emissions may expose members to pollutants at levels that are higher than the
law allows and in concentrations deleterious to human health. TEJAS brings this action on
behalf of itself and its members that live, work, and recreate near petroleum refineries and
petrochemical plants in the Houston Ship Channel.
16. The Plaintiffs and their members live, work, recreate, and breathe the air near
petroleum refineries and petrochemical plants that emit VOCs from flares, tanks, and wastewater
treatment systems. Plaintiffs' members have experienced, continue to experience, or are likely to
experience, harm to their health and to their environmental, recreational, aesthetic, and economic
interests due to the Administrator's ongoing failure to complete a review of emission factors as
required by section 130 of the Act, 42 U.S.C. § 7430.
17. Defendant Bob Perciasepe is the Acting Administrator of the EPA and in that role
is charged with the duty to review and, if necessary, revise, the emission factors in accordance
with Section 130 of the Act. 42 U.S.C. § 7430.
LEGAL BACKGROUND
18. The CAA was established "to protect and enhance the quality of the Nation's air
resources so as to promote the public health and welfare and the productive capacity of its
population" and "to initiate and accelerate a national research and development program to
achieve the prevention and control of air pollution." 42 U.S.C. § 7401(b).
19. A "primary goal" of the Act is "pollution prevention." Id. § 7401(c).
20. As part of the regulatory framework prescribed by the Act to accomplish these
objectives, EPA must establish "methods ('emission factors') used.. .to estimate the quantity of
7
-------
Case l:13-cv-00621 Document 1 Filed 05/01/13 Page 8 of 20
emissions of. . . volatile organic compounds . . . from sources of such air pollutants." 42 U.S.C.
§ 7430.
21. EPA must periodically review and revise these emission factors. Section 130
provides that "at least every 3 years [after Nov. 15, 1990], the Administrator shall review and, if
necessary, revise, the methods ('emission factors') used for purposes of [the CAA] to estimate
the quantity of emissions of. . . volatile organic compounds . . . from sources of such air
pollutants." 42 U.S.C. § 7430 (emphasis added). Section 130 requires that the Administrator
complete a review, and either make a formal determination that revision is not appropriate, or
revise the emission factors for VOCs within the statutory deadline. See id.
22. The timely review and, if necessary, revision of VOC emission factors is crucial
to EPA's ability to implement the CAA in a manner that is protective of public health. The EPA
recognizes that timely review and revision of emission factors is critical because new test data,
information, and technology can render existing emission factors obsolete or prove them to be
unreliable.
FACTUAL BACKGROUND
23. An emission factor is a "representative value" or "tool" used to estimate
emissions of a specific pollutant from an air pollution source. EPA regulations define an
"emission factor" as "the ratio relating emissions of a specific pollutant to an activity or material
throughput level." 40 C.F.R. § 51.50.
24. EPA has also defined "emission factor" as "a representative value that attempts to
relate the quantity of a pollutant released to the atmosphere with an activity associated with the
release of that pollutant." Office of Air Quality Planning & Standards, Office of Air and
Radiation, EPA, AP-42, Compilation of Air Pollutant Emission Factors Volume I: Stationary
8
-------
Case l:13-cv-00621 Document 1 Filed 05/01/13 Page 9 of 20
Point and Area Sources 1 (5th ed. 1995), available at www.epa.gov/ttn/chief/ap42/index.html
[hereinafter AP-42], EPA guidance documents define emission factors as "a tool that is used to
estimate air pollutant emissions to the atmosphere." Office of Air Quality Planning & Standards,
Office of Air and Radiation, EPA, EPA-454/R-95-015, Procedures for Preparing Emission
Factor Documents, 2-1 (1997), available at www.epa.gov/ttnchiel/efdocs/procedur.pdf.
25. The Compilation of Air Pollutant Emission Factors, or AP-42 as it is commonly
referred to, is the official compilation of emission factors and contains more than 1,700 emission
factors for over 200 air pollutants. AP-42, supra. The AP-42 emission factors are used by major
stationary sources to determine emissions from various pollution producing process units,
including flares, tanks, and wastewater treatment systems.
26. The EPA Locating and Estimating Air Toxics Emissions ("L&E") report series
compiles available information on source categories of toxic air emissions and identifies
potential release points and emission factors. Office of Air Quality Planning & Standards, Office
of Air and Radiation, EPA, Locating and Estimating Air Toxics Emissions from Sources of
(source category or substance) (2010), available at www.epa.gov/ttnchiel/le/. The emission
factors in the L&E report series cover toxic pollutants that are also VOCs, such as benzene and
toluene. Whereas AP-42 emission factors sometimes do not differentiate between different types
of VOCs, the emission factors in the L&E report series can be used to estimate emissions of
specific toxics that are also VOCs and create an inventory of toxic air emissions.
27. Industry uses emission factors to report air pollution to EPA and state regulatory
agencies. EPA and state agencies rely on this data to develop national, regional, state, and local
emissions inventories. These emission inventories are the primary tool that EPA and state
agencies use to develop emissions control strategies and make air quality management and
9
-------
Case l:13-cv-00621 Document 1 Filed 05/01/13 Page 10 of 20
permitting decisions. Basic Emissions Factors Information, Envtl. Prot. Agency,
www.epa.gov/ttn/chief/efpac/abefpac.html (last updated July 17, 2012).
28. "Emissions factors have long been the fundamental tool in developing national,
regional, state, and local emissions inventories for air quality management decisions and in
developing emissions control strategies. More recently, emissions factors have been applied in
determining site-specific applicability and emissions limitations in operating permits by federal,
state, local, and tribal agencies, consultants, and industry." Id. For example, emission factors are
used to calculate pollutant loadings used in the development of federally mandated air quality
plans designed to reduce smog and other pollutants. Emission factors may also be used to
determine compliance; EPA relies on the emission inventories—based on self-reported industry
emissions calculated using emission factors—to set an emission limit that industry then
demonstrates compliance with using emission factors. In addition, regulated industries often use
emission factors to determine if new or modified facilities will emit air pollution at levels that
require a CAA permit and best available pollution control technologies. Because poor quality
emission factors can significantly underestimate pollution emissions, the use of inaccurate
emission factors can result in the public being exposed to more air pollution than the law allows.
29. Petroleum refineries and petrochemical plants utilize flares, tanks, and wastewater
treatment systems—all of which emit significant quantities of harmful VOCs—in their
operations. Low income and minority communities suffer disproportionate health and
environmental impacts due to their proximity to these industrial sources, raising environmental
justice concerns.
30. Flares are used to control VOC releases from industrial operations, including
petroleum refineries and petrochemical plants by combusting excess gases—mostly
10
-------
Case l:13-cv-00621 Document 1 Filed 05/01/13 Page 11 of 20
hydrocarbons—to convert them into inert compounds. VOCs and other toxic pollutants are
released from flares as a result of incomplete combustion.
31. Liquid storage tanks are used in many industries that consume or produce organic
liquid, including petroleum refining and petrochemical manufacturing. Tanks emit significant
quantities of VOCs, some of which are toxic, such as benzene, toluene, and xylene.
32. Many industrial facilities, including petroleum refineries and petrochemical
plants, generate wastewater streams containing organic compounds. Emissions from wastewater
treatment systems are mostly fugitive VOCs and dissolved gases that evaporate from wastewater
surfaces left open to the air during some of the treatment processes.
33. VOC emissions endanger human health and the environment. VOCs are
substances that readily vaporize into the air, and include gaseous hydrocarbons and partially
oxidized hydrocarbons. VOCs and nitrogen oxides combine in a light-induced chemical reaction
to produce photochemical smog, an air pollution event that is characterized by high levels of
ground-level ozone. Ozone is a criteria pollutant known to endanger public health and the
environment.
At elevated levels, ozone has been shown in human laboratory
and/or community studies to be responsible for the reduction of
lung function, respiratory symptoms (e.g. cough, chest pain, throat
and nose irritation), increased hospital admissions for respiratory
causes, and increased lung inflammation. Animal studies have
shown increased susceptibility to respiratory infection and lung
structure changes. Ambient ozone has been linked to adverse
effects on agricultural crops and forests.
National Emission Standards for Hazardous Air Pollutants for Source Categories; National
Emission Standards for Hazardous Air Pollutants from Petroleum Refineries—Catalytic
Cracking (Fluid and Other) Units, Catalytic Reforming Units, and Sulfur Plant Units, 63 Fed.
11
-------
Case l:13-cv-00621 Document 1 Filed 05/01/13 Page 12 of 20
Reg. 48,890, 48,893 (proposed Sept. 11, 1998). Some VOCs are also toxic pollutants, such as
1,3-butadiene, toluene, and benzene—a known carcinogen.
34. The existing emission factors used to estimate VOC emissions from flares, tanks,
and wastewater treatment systems either significantly underestimate emissions or are rated poor
quality by EPA, potentially exposing communities to high levels of VOCs that are prohibited by
law and can have significant adverse health effects and pose grave risks to nearby communities.
35. The EPA has acknowledged, and scientific studies show, that the AP-42 emission
factors for flares, tanks, and wastewater treatment systems significantly underestimate VOC
emissions from these processes. See, e.g., Office of Inspector Gen., EPA, 2006-P-00017, EPA
Can Improve Emissions Factors Development and Management 11-12 (2006) (explaining that
for refineries "[t]he under-reporting was caused largely due to the use of poor quality emissions
factors"); Memorandum from Brenda Shine, EPA, to EPA (July 27, 2007) at 1, Docket ID No.
EPA-HQ-OAR-2003-0146-0010 ("This document provides the basis for our hypothesis that
there is a systematic low bias in reported emissions of VOC and air toxics from petroleum
refineries.").
36. Scientific studies conducted using remote sensing technology, including
Differential Absorption Lidar (DIAL) and Passive Fourier Transform Infrared (PFTIR), show
that VOC emissions from flares, tanks and wastewater treatment systems can be several orders of
magnitude higher than AP-42 emission factor estimates. See, e.g., David T. Allen and Vincent
M. Torres, Univ. of Tex. at Austin, Ctr. Energy & Envtl. Res., TCEQ 2010 Flare Study Final
Report (2011), available at
www.tceq.texas.gOv/assets/public/implementation/air/rules/Flare/2010flarestudy/2010-flare-
study-final-report.pdf; David Randall & Jeff Coburn, EPA, EPA 453/R-10-002, Critical Review
12
-------
Case l:13-cv-00621 Document 1 Filed 05/01/13 Page 13 of 20
of DIAL Emission Test Data for BP Petroleum Refinery in Texas City, Texas, at ES-2 tbl. 1
(2010), available at www.epa.gov/airtoxics/bp_dial_review_report_12-3-10.pdf (finding that
VOC emissions from several units exceeded emission estimates based on AP-42 emission
factors); Loren Raun & Dan W. Hoyt, Bur. Pollution Control & Prevention, City of Houston,
Measurement and Analysis of Benzene and VOC Emissions in the Houston Ship Channel Area
and Select Major Stationary Sources Using DIAL (Differential Absorption Light Detection and
Ranging) Technology to Support Ambient HAP Concentrations Reductions in the Community
(DIAL Project) (2011), available at www.greenhoustontx.gov/dial20110720.pdf (finding that
VOC and benzene emissions from tanks and wastewater treatment systems far exceeded
emission estimates based on unidentified emission factors that are presumed to be from AP-42);
Marathon Petroleum Co., LL.C., Performance Test of a Steam-Assisted Flare with Passive FTIR
(May 2010), available at
www.tceq.texas.gov/assets/public/implementation/air/rules/Flare/2010flarestudy/mpc-txc.pdf;
Marathon Petroleum Co., LL.C., Performance Test of a Steam-Assisted Elevated Flare with
Passive FTIR -Detroit (2010), available at
www.tceq.texas.gov/assets/public/implementation/air/rules/Flare/2010flarestudy/mpc-
detroit.pdf; Flint Hills resources Port Arthur, LL.C., PFTIR Test of Steam-Assisted Elevated
Flares-Port Arthur (2011), available at
www.epa.gov/compliance/resources/publications/civil/programs/caa/portarthur-report.pdf; Allan
Chambers & Mel Strosher, Alberta Research Council, Inc., Refinery Demonstration of Optical
Technologies for Measurement of Fugitive Emissions andfor Leak Detection (2006), available
at www.environmentalintegrity.org/pdf/publications/EIP_Att_D_Total_Upset.pdf.
13
-------
Case l:13-cv-00621 Document 1 Filed 05/01/13 Page 14 of 20
37. The L&E emission factors used to estimate air emissions of certain toxics,
including those that are also VOCs, from tanks and waste water treatment systems have been
rated poor or below average quality by EPA. EPA assigns each emission factor a rating of A
through E, with E being the poorest quality, based on certain data quality criteria. The emission
factors in the L&E report series for estimating emissions of 1,3 butadiene, benzene,
chlorobenzenes, toluene, methyl ethyl ketone, and xylene emissions from storage tanks and
wastewater treatment systems are either unrated or are rated D—below average, or E—poor.
38. In 2008, the City of Houston filed a Data Quality Act petition asking EPA to
correct the emission factors in AP-42 and the L&E report series that are used to estimate
emissions from petroleum refineries and petrochemical manufacturing plants.
39. In response to the petition, EPA committed to developing a comprehensive
protocol for the estimation of petroleum refinery emissions, providing a draft analysis of the
DIAL study conducted at the BP Amoco facility in Texas City, evaluating data from any future
remote sensing studies, and undertaking a review, and improvement, of existing emission factors
and methodologies for specific emission sources, including tanks and flares. While EPA has
completed several of these tasks, the Agency has not completed a Section 130 review of the
VOC emission factors for flares, tanks, and wastewater treatment systems, and either made a
formal determination that revision is not warranted or revised the emission factors.
40. The existing emission factors used to estimate VOC emissions from flares are
included in section 13.5 of AP-42. The Administrator has not completed a review, and either
made a formal determination that revision is not appropriate or revised these emission factors
since September 1991. The existing emission factors for flares in AP-42 significantly
underestimate actual VOC emissions from flares.
14
-------
Case l:13-cv-00621 Document 1 Filed 05/01/13 Page 15 of 20
41. The existing emission factors in AP-42 used to estimate VOC emissions from
flares are based primarily on a thirty-year old flare efficiency study conducted by EPA in 1983.
Despite the availability of more recent test data demonstrating that emission factors
underestimate VOC emissions from flares by overestimating flare efficiency by as much as 28%,
EPA has not completed the requisite review. In addition to EPA's own report on operating
parameters that affect flare combustion efficiency and performance, eleven studies on flare
efficiency have been conducted since 2006 that EPA can use to update the emission factors for
flares in AP-42. EPA's own report, published in 2012, found that certain operating parameters
can reduce flare efficiency, and established certain parameters that flares must operate within to
achieve and maintain combustion efficiency above 98%. At some facilities, EPA is already
requiring installation of remote sensing technologies that directly measure combustion
efficiency, and continuous monitoring of operating parameters that affect combustion efficiency.
See Consent Decree, United States v. BP Products N. A., Inc., Civil No. 2:12 CV 207, at app. D
(N.D. Ind. Sept. 28, 2012), available at
www.epa.gov/compliance/resources/decrees/civil/caa/whiting-cd.pdf; Consent Decree, United
States v. Marathon Petroleum Co., Civ. Action No. 2:12-cv-11544, at 41-51 (E.D. Mich. April 5,
2012), available at www.epa.gov/compliance/resources/decrees/civil/caa/marathonrefining-
cd.pdf. EPA's delay in completing the requisite review of the emission factors for flares is
inexcusable in light of the available information documenting the extent to which existing
emission factors undercount toxic VOC emissions from flares that nearby communities are
continually exposed to.
42. The emission factors used to estimate VOC emissions from tanks are included in
section 7.1 of AP-42 and the L&E report series. The Administrator has not completed a review,
15
-------
Case l:13-cv-00621 Document 1 Filed 05/01/13 Page 16 of 20
and either made a formal determination that revision is not appropriate or revised the emission
factors for tanks in AP-42 since at least 2006, and for the emission factors in the L&E report
series in over fourteen years. EPA last completed a review of the existing L&E emission factors
used to estimate benzene emissions from tanks in 1998; last reviewed the existing L&E emission
factors used to estimate emissions of chlorobenzenes, methyl eythyl ketone, toluene, and xylene
from tanks in 1994; last reviewed the existing L&E emission factors used to estimate emissions
of trichloroethylene in 1989; last reviewed the existing L&E emission factors used to estimate
emissions of ethylene oxide in 1986; and last reviewed the existing L&E emission factors used to
estimate emissions of acrylonitrile and chloroform in 1984. The existing emission factors for
tanks in AP-42 significantly underestimate actual VOC emissions, and the existing emission
factors in the L&E report series for estimating acrylonitrile, benzene, chlorobenzenes,
chloroform, ethylene oxide, methyl ethyl ketone, toluene, trichloroethylene, and xylene
emissions from tanks are either unrated or are rated D—below average, or E—poor.
43. At least three scientific studies conducted since 2006 shows that existing emission
factors significantly underestimate toxic VOC emissions from tanks. EPA's failure to complete
the requisite review of emission factors for tanks is inexcusable in light of the information
available to EPA, the significant amounts of toxic VOC emissions from tanks that are
unaccounted for by emission factors, and the danger these emissions pose to nearby
communities.
44. The existing emission factors used to estimate VOC emissions from wastewater
treatment systems are included in section 4.3 of AP-42 and the L&E report series. The
Administrator has not completed a review, and either made a formal determination that revision
is not appropriate or revised the emission factors for wastewater treatment systems in AP-42
16
-------
Case l:13-cv-00621 Document 1 Filed 05/01/13 Page 17 of 20
since at least 2006, and for the emission factors in the L&E report series in over fourteen years.
EPA last completed a review of the existing L&E emission factors used to estimate benzene
emissions from wastewater treatment systems in 1998; last reviewed the existing L&E emission
factors used to estimate 1,3 butadiene emissions from wastewater treatment systems in 1996; last
reviewed the existing L&E emission factors used to estimate emissions of chlorobenzenes and
xylene emissions from wastewater treatment systems in 1994; and last reviewed the existing
L&E emission factors used to estimate emissions of chloroform, epichlorohydrin and ethylene
dichloride from wastewater treatment systems in 1984. The existing emission factors for
wastewater treatment systems in AP-42 significantly underestimate actual VOC emissions, and
the existing emission factors in the L&E report series for estimating benzene, 1,3 butadiene,
chlorobenzenes, chloroform, epichlorohydrin, ethylene dichloride, and xylene emissions from
wastewater treatment systems are either unrated or are rated D—below average, or E—poor.
45. Notwithstanding the poor quality of these emission factors and EPA's own
acknowledgments and scientific data that makes clear that these emission factors can
significantly undercount the emissions nearby communities are exposed to, the Administrator has
failed to complete a review and make necessary revisions of these emission factors within the
statutory timeframe in accordance with section 130 of the CAA. In light of EPA's continued
failure to act, Plaintiffs issued a notice of intent to sue EPA for failure to comply with its
statutory duties under section 130 of the CAA on July 18, 2012. See Ex. A.
CAUSES OF ACTION
46. Plaintiffs re-allege and incorporate the allegations of all foregoing paragraphs.
47. The Administrator has failed to review and, if necessary, revise the existing
emission factors for flares in AP-42 since 1991. The Administrator's ongoing failure to
17
-------
Case l:13-cv-00621 Document 1 Filed 05/01/13 Page 18 of 20
complete a Section 130 review, and either make a final determination that revision is not
appropriate or revise the VOC emission factors for flares in AP-42 within the statutory
timeframe constitutes a "failure of the Administrator to perform any act or duty under this
chapter which is not discretionary with the Administrator" within the meaning of section
304(a)(2) of the CAA, 42 U.S.C. § 7604(a)(2).
48. The Administrator has failed to review and, if necessary, revise the existing
emission factors for tanks in AP-42 and L&E since at least 2006. The Administrator's ongoing
failure to complete a Section 130 review, and either make a final determination that revision is
not appropriate or revise the VOC and other emission factors for tanks in AP-42 and L&E within
the statutory timeframe constitutes a "failure of the Administrator to perform any act or duty
under this chapter which is not discretionary with the Administrator" within the meaning of
section 304(a)(2) of the CAA, 42 U.S.C. § 7604(a)(2).
49. The Administrator has failed to review and, if necessary, revise the existing
emission factors for wastewater treatment systems in AP-42 and L&E since at least 2006. The
Administrator's ongoing failure to complete a Section 130 review, and either make a final
determination that revision is not appropriate or revise the VOC and other emission factors for
tanks in AP-42 and L&E within the statutory timeframe constitutes a "failure of the
Administrator to perform any act or duty under this chapter which is not discretionary with the
Administrator" within the meaning of section 304(a)(2) of the CAA, 42 U.S.C. § 7604(a)(2).
PRAYER FOR RELIEF
WHEREFORE, Plaintiffs respectfully request that this Court:
A. Declare that the Administrator's failure to complete a review of the VOC
emission factors for flares in AP-42, and either make a final determination that revision is not
18
-------
Case l:13-cv-00621 Document 1 Filed 05/01/13 Page 19 of 20
appropriate or revise the emission factors within the statutory timeframe, constitutes a "failure of
the Administrator to perform any act or duty under this chapter which is not discretionary with
the Administrator" within the meaning of section 304(a)(2) of the CAA, 42 U.S.C. § 7604(a)(2);
B. Declare that the Administrator's failure to complete a review of the VOC
emission factors for tanks in AP-42 and L&E, and either make a final determination that revision
is not appropriate or revise the emission factors within the statutory timeframe, constitutes a
"failure of the Administrator to perform any act or duty under this chapter which is not
discretionary with the Administrator" within the meaning of section 304(a)(2) of the CAA, 42
U.S.C. § 7604(a)(2);
C. Declare that the Administrator's failure to complete a review of the VOC
emission factors for wastewater treatment systems in AP-42 and L&E, and either make a final
determination that revision is not appropriate or revise the emission factors within the statutory
timeframe, constitutes a "failure of the Administrator to perform any act or duty under this
chapter which is not discretionary with the Administrator" within the meaning of section
304(a)(2) of the CAA, 42 U.S.C. § 7604(a)(2);
D. Order the Administrator to complete the required Section 130 reviews and to
either revise the VOC emissions factors for flares, tanks, and wastewater treatment systems in
AP-42 and L&E, or make a final determination that such revision is not appropriate, pursuant to
section 130 of the CAA, 42 U.S.C. § 7430, in accordance with expeditious deadlines specified by
this Court;
E. Retain jurisdiction of this action to ensure compliance with this Court's decree;
F. Award Plaintiffs the costs of this action, including attorney's fees; and
G. Grant such other relief as the Court deems just and proper.
19
-------
Case l:13-cv-00621 Document 1 Filed 05/01/13 Page 20 of 20
DATED: May 1, 2013
Respectfully submitted,
/s/ Jennifer Peterson
Jennifer Peterson (D.C. Bar No. 978352)
Environmental Integrity Project
One Thomas Circle, Suite 900
Washington, DC 20005
(202) 263-4449
j peterson@environmentalintegrity. org
/s/ Whitney Ferrell
Whitney Ferrell* (D.C. Bar No. 1013459)
Environmental Integrity Project
One Thomas Circle, Suite 900
Washington, DC 20005
(202) 263-4456
wferrell@environmentalintegrity.org
*Motion to appear pro hac vice pending
Counsel for Air Alliance Houston,
Community In-Power and Development
Association, Louisiana Bucket Brigade,
and Texas Environmental Justice Advocacy
Services
20
-------
Case l:13-cv-00621 Document 1-1 Filed 05/01/13 Page 1 of 2
CIVIL COVER SHEET
JS-44 (Rev. 3/13 DC)
I. (a) PLAINTIFFS
Air Alliance Houston
Community In-Powerand Development Association, Inc.
Louisiana Bucket Brigade
Texas Environmental Justice Advocacv Services ~
(b) COUNTY OF RESIDENCE OF FIRST LISTED PLAINTIFF 88888
DEFENDANTS
Bob Perciasepe, in his official capacity as Acting
Administrator, United States Environmental Protection Agency
COUNTY OF RESIDENCE OF FIRST LISTED DEFENDANT 11001
(EXCEPT IN U.S. PLAINTIFF CASES)
(IN U.S. PLAINTIFF CASES ONLY)
NOTE- W LAND CONDEMNATION CASES. USE THE I.< KTATION Ol I HE TRACT Ol- L. JsTD INVOLVED
(c) ATTORNEYS (FIRM NAME, ADDRESS, AND TELEPHONE NUMBER)
Environmental Integrity Project
One Thomas Circle, Suite 900
Washington, D.C. 20005
(202) 263-4456
ATTORNEYS (IF KNOWN)
Jennifer Peterson, D.C. Bar No. 978352
Whitney Ferrell, D.C. Bar No. 1013459 (pending motion to
appear pro hac vice)
II. BASIS OF JURISDICTION
(PLACE AN x IN ONE BOX ONLY)
O 1 U.S. Government Q 3 Federal Question
Plaintiff (U.S. Government Not a Party)
© 2 U.S. Government O 4 Diversity
Defendant (Indicate Citizenship of
Parties in item III)
III. CITIZENSHIP OF PRINCIPAL PARTIES (PLACE AN x in one box for
PLAINTIFF AND ONE BOX FOR DEFENDANT) FOR DIVERSITY CASES ONLY!
PTF DFT PTF DFT
Citizen of this State ' O ' Incorporated or Principal Place O 4 O 4
of Business in This State
Citizen of Another State C>2 C>2 incorporated and Principal O 5 Qs
Place of Business in Another State
Citizen or Subject of a Q 3 Q 3 _
Foreign Country Foreign Nation O 6 O 6
IV. CASE ASSIGNMENT AND NATURE OF SUIT
(Place an X in one category, A-N, that best represents your Cause of Action and one in a corresponding Nature of Suit)
0 A. Antitrust
1 1410 Antitrust
0 B. Personal Injury/
Malpractice
1 1310 Airplane
I 1315 Airplane Product Liability
I 1320 Assault, Libel & Slander
I 1330 Federal Employers Liability
I 1340 Marine
I 1345 Marine Product Liability
1 1350 Motor Vehicle
I 1355 Motor Vehicle Product Liability
I 1360 Other Personal Injury
I 1362 Medical Malpractice
I 1365 Product Liability
I 1367 Health Care/Pharmaceutical
Personal Injury Product Liability
I 1368 Asbestos Product Liability
© c. Administrative Agency
Review
I I 151 Medicare Act
Social Security
I 1861 HIA (1395ff)
I I 862 Black Lung (923)
I I 863 DIWC/DIWW (405(g))
I I 864 SSID Title XVI
I I 865 RSI (405(g))
Other Statutes
I I 891 Agricultural Acts
[ X I 893 Environmental Matters
I I 890 Other Statutory Actions (If
Administrative Agency is
Involved)
O D. Temporary Restraining
Order/Preliminary
Injunction
Any nature of suit from any category
may be selected for this category of case
assignment.
* (If Antitrust, then A governs)'
O E. General Civil (Other)
OR
O F. Pro Se General Civil
Real Property
I 1210 Land Condemnation
I 1220 Foreclosure
I 1230 Rent, Lease & Ejectment
I 1240 Torts to Land
I 1245 Tort Product Liability
I 1290 All Other Real Property
Personal Property
I 1370 Other Fraud
I 1371 Truth in Lending
I 1380 Other Personal Property
Damage
I 1385 Property Damage
Product Liability
Bankruptcy
I |422 Appeal 27 USC 158
I 1423 Withdrawal 28 USC 157
Prisoner Petitions
I I 535 Death Penalty
~ 540 Mandamus & Other
~ 550 Civil Rights
I 1555 Prison Conditions
I 1560 Civil Detainee - Conditions
of Confinement
Property Rights
I 1820 Copyrights
I 1830 Patent
I 1840 Trademark
Federal Tax Suits
I 1870 Taxes (US plaintiff or
defendant)
I 1871 IRS-Third Party 26 USC 7609
Forfeiture/Penalty
I 1625 Drug Related Seizure of
Property 21 USC 881
I 1690 Other
Other Statutes
I 1375 False Claims Act
I 1400 State Reapportionment
I 1430 Banks & Banking
I 1450 Commerce/ICC
Rates/etc.
I 1460 Deportation
I 1462 Naturalization
Application
I 1465 Other Immigration
Actions
I 1470 Racketeer Influenced
& Corrupt Organization
I 1480 Consumer Credit
I 1490 Cable/Satellite TV
I 1850 Securities/Commodities/
Exchange
I 1896 Arbitration
I 1899 Administrative Procedure
Act/Review or Appeal of
Agency Decision
I 1950 Constitutionality of State
Statutes
I 1890 Other Statutory Actions
(if not administrative agency
review or Privacy Act)
-------
Case l:13-cv-00621 Document 1-1 Filed 05/01/13 Page 2 of 2
O g.
Habeas Corpus/
2255
I 1530 Habeas Corpus - General
1 1510 MotionA'acate Sentence
I 1463 Habeas Corpus - Alien
Detainee
O H. Employment
Discrimination
| | 442 Civil Rights — Employment
(criteria: race, gender/sex,
national origin,
discrimination, disability, age,
religion, retaliation)
*(lf pro se, select this deck)*
0 I. FOIA/Privacy Act
1 1895 Freedom of Information Act
I 1890 Other Statutory Actions
(if Privacy Act)
*(If pro se, select this deck)*
0 J. Student Loan
1 1152 Recovery of Defaulted
Student Loan
(excluding veterans)
0 K. Labor/ERISA
(non-employment)
1 1710 Fair Labor Standards Act
I 1720 Labor/Mgmt. Relations
I 1740 Labor Railway Act
I 1751 Family and Medical
Leave Act
I 1790 Other Labor Litigation
I 1791 Empl. Ret. Inc. Security Act
0 L. Other Civil Rights
(non-employment)
1 1441 Voting (if not Voting Rights
Act)
I |443 Housing/Accommodations
I 1440 Other Civil Rights
I 1445 Americans w/Disabilitics -
Employment
I 1446 Americans w/Disabilitics -
Other
I 1448 Education
O M. Contract
I 1110
I 1120
I 1130
I 1140
I 1150
I 1153
I 1160
~ 190
~ 195
~ 196
Insurance
Marine
Miller Act
Negotiable Instrument
Recovery of Overpayment
& Enforcement of
Judgment
Recovery of Overpayment
of Veteran's Benefits
Stockholder's Suits
Other Contracts
Contract Product Liability
Franchise
0 N. Three-Judge
Court
1 I 441 Civil Rights - Voting
(if Voting Rights Act)
V. ORIGIN
© 1 Original O 2 Remand O 3 Remanded from O 4 Reinstated or Q 5 Transferred from Q 6 Multi-district O' Appeal to
Proceeding from State Appellate Court Reopened another district Litigation District Judge
Court (specify) from Mag. Judge
VI. CAUSE OF ACTION (CITE THE U.S. CIVIL STATUTE UNDER WHICH YOU ARE FILING AND WRITE A BRIEF STATEMENT OF CAUSE.)
Action brought under 42 U.S.C. 7604(a)(2) against defendant for failure to review VOC emission factors for flares, tanks, g
VII. REQUESTED IN
COMPLAINT
CHECK IF THIS IS A CLASS
ACTION UNDER F.R.C.P. 23
DEMAND $
JURY DEMAND:
Check YES only if demanded in complaint
YES I I NO I X" I
VIII. RELATED CASE(S)
IF ANY
(See instruction)
YES I I NO| K I
If yes, please complete related case form
DATE: May 1,2013
SIGNATURE OF ATTORNEY OF RECORD (A / r[ ^-A^/C 7 /7
/
r/ -
INSTRUCTIONS FOR COMPLETING CIVIL COVER SHEETV&44
Authority for Civil Cover Sheet
The JS-44 civil cover sheet and the information contained herein neither replaces nor supplements the filings and services of pleadings or other papers as required
by law, except as provided by local rules of court. This form, approved by the Judicial Conference of the United States in September 1974, is required for the use of the
Clerk of Court for the purpose of initiating the civil docket sheet. Consequently, a civil cover sheet is submitted to the Clerk of Court for each civil complaint filed.
Listed below are tips for completing the civil cover sheet. These tips coincide with the Roman Numerals on the cover sheet.
I. COUNTY OF RESIDENCE OF FIRST LISTED PLAINTIFF/DEFENDANT (b) County of residence: Use 11001 to indicate plaintiff if resident
of Washington, DC, 88888 if plaintiff is resident of United States but not Washington, DC, and 99999 if plaintiff is outside the United States.
III. CITIZENSHIP OF PRINCIPAL PARTIES: This section is completed only if diversity of citizenship was selected as the Basis of Jurisdiction
under Section II.
IV. CASE ASSIGNMENT AND NATURE OF SUIT: The assignment of a judge to your case will depend on the category you select that best
represents the primary cause of action found in your complaint. You may select only one category. You must also select one corresponding
nature of suit found under the category of the case.
VI. CAUSE OF ACTION: Cite the U.S. Civil Statute under which you are filing and write a brief statement of the primary cause.
VIII. RELATED CASE(S), IF ANY: If you indicated that there is a related case, you must complete a related case form, which may be obtained from
the Clerk's Office.
Because of the need for accurate and complete information, you should ensure the accuracy of the information provided prior to signing the form.
-------
Case l:13-cv-00621 Document 1-2 Filed 05/01/13 Page 1 of 1
AO 440 (Rev. 12/09; DC 03,10) Summons in a Civil Action
United States District Court
for the
District of Columbia
Air Alliance Houston, et. al.
Plaintiff
V.
Bob Perciasepe, in his official capacity as Acting
Administrator, U.S. Environmental Protection Agency
Defendant
Civil Action No.
SUMMONS IN A CIVIL ACTION
To: (Defendant's name and address) Bob Perciasepe, in his official capacity as
Acting Administrator, United States
Environmental Protection Agency,
1101A EPA Headquarters, Ariel Rios Building
1200 Pennsylvania Avenue, NW
Washington, D.C. 20460
A lawsuit has been filed against you.
Within 21 days after service of this summons on you (not counting the day you received it) — or 60 days if you
are the United States or a United States agency, or an officer or employee of the United States described in Fed. R. Civ.
P. 12 (a)(2) or (3) — you must serve on the plaintiff an answer to the attached complaint or a motion under Rule 12 of
the Federal Rules of Civil Procedure. The answer or motion must be served on the plaintiff or plaintiffs attorney,
whose name and address are: Jennifer Peterson
Environmental Integrity Project
One Thomas Circle, Suite 900
Washington, DC 20005
If you fail to respond, judgment by default will be entered against you for the relief demanded in the complaint.
You also must file your answer or motion with the court.
ANGELA D. CAESAR, CLERK OF COURT
Date:
Signature of Clerk or Deputy Clerk
-------
Case l:13-cv-00621 Document 1-3 Filed 05/01/13 Page 1 of 14
Exhibit A
-------
E
Case l:13-cv-00621 Document 1-3 Filed 05/01/13
ENVIRONMENTAL
INTEGRITY PROJECT
Page 2 of 14
1 Thomas Circle, Suite 900
Washington, DC 20005
main: 202-296-8800
fax: 202-296-8822
www.environmentalintegrity.org
July 18,2012
Via Certified Mail
Ms. Lisa P. Jackson
Administrator
U.S. Environmental Protection Agency
Ariel Rios Building
1200 Pennsylvania Avenue, NW
Washington, DC 20406
jackson.lisa@epa.gov
Re: Notice of Intent to Sue for Violations of Nondiscretionary Duties to Review and Revise
Emission Factors under Section 130 of the Clean Air Act Every Three Years
Dear Administrator Jackson,
We are writing on behalf of Air Alliance Houston, Texas Environmental Advocacy
Services ("TEJAS"), Community In-power and Development Association, Inc. ("CIDA"), and
Louisiana Bucket Brigade ("Plaintiffs") to provide you with notice of our intent to file suit
against Administrator Jackson, in her official capacity as Administrator of the U.S.
Environmental Protection Agency (EPA), for failure to perform nondiscretionary duties under
section 130 of the Clean Air Act (CAA), 42 U.S.C. § 7430, which requires the Administrator to
review and, if necessary, revise emission factors for industrial flares (AP-42, section 13.5), liquid
storage tanks (AP-42, section 7.1; L&E documents), and wastewater collection, treatment, and
storage systems (AP-42, section 4.3; L&E documents) at least once every three years.
Under section 130 of the CAA, 42 U.S.C. § 7430, the Administrator has a mandatory
duty to review and, if necessary, revise, the emission factors used to estimate emissions of
carbon monoxide (CO), volatile organic compounds (VOCs), and oxides of nitrogen (NOx) from
emission sources at least once every three years. The Administrator has failed to perform the
nondiscretionary duty to review and, if necessary, revise, at least once every three years,
emission factors used to estimate emissions of CO, VOCs, and NOx from flares, tanks, and
wastewater treatment systems. EPA has not reviewed emission factors for flares since 1991,
emission factors for wastewater treatment systems have not been reviewed since 1998, and
emission equations for tanks have not been reviewed since 2006.1
Accurate accounting of air pollutant emissions is the linchpin of the CAA. Air emissions
data is used to: design regulations, develop emission reduction control strategies, determine
emission limits and applicable permit requirements for major sources, guide enforcement
1 See EPA, Compilation of Air Pollutant Emission Factors: Stationary Point and Area Sources (1995), a\>ailable at
www.epa.gov/ttn/cliief/ap42/index.html [hereinafter AP-42],
1
-------
Case l:13-cv-00621 Document 1-3 Filed 05/01/13 Page 3 of 14
decisions, achieve air quality goals, and protect communities from toxic exposure.2 Outdated
emission factors can grossly underestimate emissions of air pollutants from petroleum refineries,
chemical plants, and other industrial sources.3 Emissions from petroleum refineries pose grave
risks to nearby communities, many of which are low income and minority communities, because
refinery emissions contain hazardous air pollutants such as benzene, which is a known
carcinogen.4 While the City of Houston filed a Data Quality Act petition in 2008 asking EPA to
correct data quality errors in emission factors for petroleum refineries and chemical
manufacturing plants, the EPA has yet to do so.5 The lack of data means that cost-effective
opportunities to reduce pollution are hidden in plain sight, which may force reliance on more
expensive alternatives. The duty to timely review and revise section 130 emission factors is
critical to EPA's ability to implement the CAA in a manner that is protective of public health.
Section 304 of the CAA, 42 U.S.C. § 7604(a)(2), authorizes citizen actions "against the
Administrator where there is an alleged failure of the Administrator to perform any act or duty
under this chapter which is not discretionary with the Administrator." Citizens must provide
notice to the Administrator at least sixty days before filing a civil suit under section 130. 42
U.S.C. § 7604(b)(1)(A); 40 C.F.R. § 54.2. In accordance with Section 304 of the CAA, this
letter serves to notify you that Plaintiffs intend to file suit in federal district court any time
beginning sixty days from the postmarked date of this letter. 42 U.S.C. § 7604; 40 C.F.R. §
54.2(d).
I. Background: Section 130 Emission Factors
Emission factors are used to estimate pollutant emissions when source-specific test data
is not available.6 Emission factors and emission inventories are fundamental tools of air quality
management used by EPA to plan and implement air pollution control programs under the
CAA.7 Emission factors are used to calculate around eighty percent of national emissions from
virtually all sources of air pollution.8 Thus, reliable emission factors that accurately estimate
2 Office of Inspector General, EPA, 2006-P-00017, EPA Can Improve Emissions Factors Development and
Management 1 (2006), available at www.epa.gov/oig/reports/2006/20060322-2006-P-00017.pdf [hereinafter EPA,
2006 Inspector General Report] ("Quantifying air emissions is a vital aspect of air pollution programs. Regulatory
authorities and others use emissions values in: (1) developing emissions inventories, (2) identifying and evaluating
control strategies, (3) determining applicability of permit and regulatory requirements, and (4) assessing risks.").
3 Id. at 8 ("EPA officials have identified the inappropriate use of emissions factors for key environmental decisions,
such as permit limits and the level of air pollution control equipment installed at specific facilities, resulting in the
release of significant amounts of unidentified and uncontrolled emissions.").
4 See EPA, Toxics Release Inventory Explorer, available at www.epa.gov/triexplorer (last visited July 17, 2012).
5 Letter from Bill White, Mayor, City of Houston, Tex., to Information Quality Guidelines Staff, EPA, Request for
Correction of Information under the Data Quality Act and EPA's Information Quality Guidelines (July 9, 2008),
available at www.greenhoustontx.gov/reports/epaletter20080709.pdf.
6 See AP-42, supra note 1, at 1.
7 EPA, 2006 Inspector General Report, supra note 2, at 4 ("Emission factors are used to develop the emissions data
that are the cornerstone of a host of important environmental decisions made by EPA .... including] . . . facility
permitting, development of control strategies, and compliance and enforcement decisions.").
8 Id. (citing U.S. Gov't Accountability Office (GAO), GAO-Ol-46, EPA Should Improve Oversight of Emissions
Reporting by Large Facilities 3 (2001), available at www.gao.gov/new.items/d0146.pdf ("In preparing emissions
reports,. .. large facilities rely primarily on estimates and extrapolation instead of directly measuring their pollutant
emissions. To estimate their annual emissions of each pollutant, most facilities use industry- and pollutant specific
2
-------
Case l:13-cv-00621 Document 1-3 Filed 05/01/13 Page 4 of 14
emissions are imperative to EPA's ability to make air quality management decisions that are
protective of public health.
AP-42 is the EPA's official compilation of air pollutant emission factors and contains
more than 1,700 rated emission factors for over 200 air pollutants.9 AP-42 emission factors
represent long-term average emissions and testing is generally done under normal operating
conditions that do not account for conditions that may cause short-term fluctuations in
emissions.10 In addition, the EPA Locating and Estimating (L&E) documents compile available
information on source categories of toxic air emissions and identify potential release points and
emission factors.11
Although the CAA does not define "emission factor," the EPA has defined it as a
"representative value" or "tool" used to estimate emissions of a specific pollutant from an air
pollution source.12 EPA regulations define "emission factor" as "the ratio relating emissions of a
specific pollutant to an activity or material throughput level." 40 C.F.R. § 51.50; see also Air
Emissions Reporting Requirements, 73 Fed. Reg. 76,539-01, 76,554 (Dec. 17, 2008). The AP-
42 defines "emission factor" as "a representative value that attempts to relate the quantity of a
pollutant released to the atmosphere with an activity associated with the release of that
pollutant."13 EPA guidance documents broadly define "emission factor" as "a tool that is used to
estimate air pollutant emissions to the atmosphere,"14 Thus, "emission factor" includes emission
estimation tools or equations that are used to estimate emissions from liquid storage tanks and
wastewater treatment systems.
II. Section 130 Emission Factors for Flares, Tanks, and Wastewater Treatment
Systems
The Administrator has not timely reviewed or revised emission factors used for
estimating emissions from flares, tanks, and wastewater treatment systems. The EPA recognizes
that the timely review and revision of emission factors is critical because there is a "moving
target" aspect to emission factor development in that new information and processes can render
existing emission factors obsolete or prove them to be unreliable.15 The current AP-42 emission
emissions factors .... EPA's data show that, nationally, emissions factors are used for about 80 percent of
emissions determinations.").
9 AP-42, supra note 1, at 1.
10 See id. at 4 ("emission factors essentially represent an average of a range of emission rates"); Letter from
Elizabeth Craig, Acting Assistant Administrator, U.S. EPA, to Bill White, Mayor, Houston, Tex. 3 (Apr. 7, 2009)
("These factors are designed to be representative values relating the quantity of a pollutant released to the
atmosphere under normal operating conditions with an activity associated with the release of that pollutant.")
(emphasis added).
11 EPA, Locating & Estimating (L&E) Documents, available at www.epa.gov/ttnchiel/le/ (last visited July 17,
2012) [hereinafter L&E Documents].
12 Id. at 1; EPA, EPA-454/R-95-015„Procedures for Preparing Emission Factor Documents, 2-1 (1997) [hereinafter
EPA, Procedures for Preparing Emission Factor Documents], available at
www.epa.gov/ttnchiel/efdocs/procedur.pdf (emphasis added).
13 AP-42, supra note 1, at 1.
14 EPA, Procedures for Preparing Emission Factor Documents, supra note 12, at 2-1.
15 Office of Inspector General, EPA, No. 6100306, Emission Factor Development 12-13 () (1996), available at
www.epa.gov/oig/reports/1996/emisrept.pdf [hereinafter EPA, 1996 Inspector General Report]
3
-------
Case l:13-cv-00621 Document 1-3 Filed 05/01/13 Page 5 of 14
factors used to estimate emissions from flares, tanks, and wastewater treatment systems can
significantly undercount emissions of CO, VOCs, and NOx from refineries and petrochemical
plants.16 The EPA has acknowledged, and scientific studies conducted using remote sensing
technology, including Differential Absorption Lidar (DIAL) and Passive Fourier Transform
Infrared (PFTIR), have consistently shown that actual emissions from these sources can be
several orders of magnitude higher than emission factor estimates.17 See Attachment A.
Three separate DIAL studies at refineries in North America have shown that emission
factors for flares, tanks, wastewater treatment systems, and several other processes significantly
underestimate emissions. An EPA review of a 2008 DIAL test at the BP Texas City petroleum
refinery found that actual emissions from several units exceeded emission factor estimates for
emissions from tanks and flares.18 For flares, EPA found that actual average emissions were six
times higher than the average hourly emissions in the emissions inventory report.19 For storage
tanks, EPA found that actual emissions were, in some cases, at least three to seven times higher
than emission factor estimates.20 A 2006 DIAL test at a refinery in Alberta Canada found that
actual emissions from storage tanks exceeded emission factor estimates for benzene and VOCs.21
A 2011 DIAL test at Shell Deer Park measured emission concentrations of benzene and VOCs
that far exceeded emission factor estimates.22 The DIAL test results found: actual VOC
emissions from tanks were underestimated by a factor of 132; actual benzene emissions from
tanks were underestimated by a factor of 93; actual VOC emissions from wastewater treatment
systems were underestimated by a factor of 108; and actual benzene emissions from wastewater
treatment systems were underestimated by a factor of 67.23
16 Alan Chambers & Mel Strosher, Refinery Demonstration of Optical Technologies for Measurement of Fugitive
Emissions and for Leak Detection (2006) [hereinafter Alberta DIAL Study]; Rod Robinson, Tom Gardiner & Bob
Lipscombe, National Physical Laboratory, Measurements Of VOC Emissions From Petrochemical Industry Sites In
The Houston Area Using Differential Absorption Lidar (DIAL) During Summer 2007: DRAFT FOR COMMENT,
27-28 (2008) [hereinafter BP Texas City DIAL Study]; David Randall & Jeff Coburn, EPA, EPA 453/R-10-002,
Critical Review of DIAL Emission Test Data for BP Petroleum Refinery in Texas City, Texas, ES-1 (2010)
[hereinafter EPA, Review of BP Texas City DIAL Emissions Test Data]; Dan Hoyt et. al., City of Houston Bureau
of Pollution Control and Prevention, Measurement and Analysis of Benzene and VOC Emissions in the Houston
Ship Channel Area and Select Major Stationary Sources Using DIAL (Differential Absorption Light Detection and
Ranging) Technology to Support Ambient HAP Concentrations Reductions in the Community (DIAL Project) 92
(2011) [hereinafter Shell Deer Park DIAL Study].
17 See e.g., Shell Deer Park DIAL Study, at 92 ("Based on the current data and associated statistics, the true
emissions may be underestimated by a factor of as much as 132 for VOCs"); National Emission Standards for
Hazardous Air Pollutants from Petroleum Refineries, 72 Fed. Reg. 50,716, 50,725-26 (Sept. 4, 2007) (stating that
EPA review of the data indicates "inherent uncertainty in the development and use of emission factors");
Memorandum from Brenda Shine, EPA, to EPA Docket No. EPA-HQ-OAR-2003-0146, Potential Low Bias of
Reported VOC Emissions from the Petroleum Refining Industry, (July 27, 2007) [hereinafter EPA, Memorandum
from Brenda Shine], available at http://www.regulations.gov/#!documentDetail;D=EPA-HQ-OAR-2003-0146-
0010; EPA, 2006 Inspector General Report, supra note 2, at 11-12.
18 EPA, Review of BP Texas City DIAL Emissions Test Data, supra note 16, at ES-2 & tbl. 1.
19 Id. atES-5.
20 Id. at ES-1, ES-4 ("On average, the DIAL results for external floating roof tanks storing crude oil were at least 3
to 7 times higher than estimates that used conditions at the time of the DIAL testing.").
21 Alberta DIAL Study, supra note 16, at 27.
22 Shell Deer Park DIAL Study, supra note 16, at 1, 92.
23Id.
4
-------
Case l:13-cv-00621 Document 1-3 Filed 05/01/13 Page 6 of 14
A. Flares
The emission factors used to estimate emissions from flares in section 13.5 of AP-42
were last revised in September 1991.24 According to the EPA website, the most recent flare
efficiency study was conducted in July 1983.25 The emission factor for VOCs from flares is 0.14
lb/MMBtu; the emission factor for CO from flares is 0.37 lb/MMBtu; and the emission factor for
NOx from flares is 0.068 lb/MMBtu.26 Pollutants of concern from flaring include carbon
particles (soot), unburned hydrocarbons, CO, other partially burned and altered hydrocarbons,
NOx, and SO2.27 Because it has been over twenty years since EPA last revised emission factors
for flares, the Administrator has failed perform the nondiscretionary duty required by section 130
of the CAA, 42 U.S.C. § 7430.
B. Tanks
The emission equations in section 7.1 of AP-42 are the only emissions estimating tool
EPA has for tanks.28 The major pollutant of concern from Liquid Storage Tanks are VOCs.29
The emission estimation equations in section 7.1 of AP-42 were developed by the American
Petroleum Institute (API).30 These emission equations are the basis of the software program
TANKS that is used to generate site-specific emission factors and estimate emissions from liquid
storage tanks.31 EPA recognizes that industry has an interest in developing conservative
emission equations because "there is a financial benefit to industries to use emission factors that
produce low emission estimates."32 Yet, despite the increased risk of generating biased or
unrepresentative emission estimates,33 the EPA has only reviewed emission equations for tanks
once since they were first developed by API fifteen years ago.34 The emission equations used to
24 AP-42, supra note 1, at ch. 13: Miscellaneous Sources, available at www.epa.gov/ttn/chief/ap42/chl3/index.html
(last visited July 17, 2012).
25 Id.
26Id. at 13.5-4.
21 Id. at 13.5-3.
28 AP-42, supra note 1, at ch. 7, available a/www.epa.gov/ttn/chief/ap42/ch07/final/c07s01.pdf.
29 Id.
30 Id. at 7.1-9 ("These procedures are valid for all petroleum liquids, pure volatile organic liquids, and chemical
mixtures with similar true vapor pressures").
31 Id.-, see also TANKS Emission Estimation Software, available at www.epa.gov/ttn/chief/software/tanks/ (last
visited July 17, 2012).
32 See, e.g. EPA, Memorandum from Brenda Shine, supra note 17; EPA, 1996 Inspector General Report, supra note
2, at 21 ("Industries may be more inclined to participate in a partnership that would result in lower emission factors
because of the uses of these factors. Emission factors are sometimes used to determine whether a source needs to
obtain a construction or Title V operating permit, and estimate annual emissions for the purpose of determining
annual permit fees under the Title V permit program. These uses of emission factors provide industry with a
financial incentive to use emission factors that produce low emission estimates. Emission factors that produce low
emission estimates may allow a source to avoid obtaining a permit and implementing required emission controls. In
addition, sources that obtain Title V permits would pay lower annual fees when these fees are based on emission
factors that result in lower emission estimates.").
33 Shell Deer Park DIAL Study, supra note 16, at 1 ("Emission factors used to estimate emissions from the
Southwest Tanks VOCs produced the most potential underestimated emissions compared to the DIAL measured
emissions, off by a factor of 132."); see also id. at 92, 99.
34 See EPA, Emission Factor Documentation for AP-42 Section 7.1: Organic Liquid Storage Tanks Final Report
(2006), available at www.epa.gov/ttn/chief/ap42/ch07/bgdocs/b07sO 1 .pdf.
5
-------
Case l:13-cv-00621 Document 1-3 Filed 05/01/13 Page 7 of 14
estimate emissions from tanks in section 7.1 AP-42 were last reviewed in November 2006.35
While the review falls outside of the statutory three year timetable, it is also unclear whether the
2006 review satisfies the substantive requirements of section 130.36 In addition, it has been over
fourteen years since EPA last reviewed L&E emission factors for VOC emissions from storage
tanks.37 EPA last reviewed L&E emission factors for benzene emissions from storage tanks in
1998, and L&E emission factors for methyl ethyl ketone, toluene, and xylene emissions from
storage tanks in 1994.38 Because it has been over five years since EPA last reviewed AP-42
emission equations for tanks, and over fourteen years since EPA last reviewed L&E emission
factors for tanks, the Administrator has failed to perform the nondiscretionary duty required by
section 130 of the CAA, 42 U.S.C. § 7430.
C. Wastewater Treatment Systems
Methodologies for estimating emissions from wastewater treatment systems were last
reviewed in February 1998.39 Section 4.3 of AP-42 provides emission calculation methodologies
for estimating air emissions from wastewater treatment systems. The Surface Impoundment
Modeling System (SIMS) is a computer program that can be used to estimate emissions of
organic compounds from wastewater treatment systems.40 SIMS uses mass transfer correlations
to predict VOC emissions from industrial waste water.41 EPA recognizes that "in some cases, []
orders-of-magnitude differences may result between actual and estimated emissions, depending
on differences in source configurations, control equipment, and operating practices."42
36 See id. at 1-1, 6-1 (While EPA states that the 2006 document is a background report, clarifying that the "purpose
of this report is to provide background information to support revisions to AP-42 Section 7.1, Organic Liquid
Storage Tanks," the document also "summarize[s] the major changes made since the previous version of Section
7.1~Organic Liquid Storage Tanks (September 1997) of AP-42.").
37 The following pollutants are VOCs: benzene, methyl ethyl ketone, toluene, and xylene.
38 EPA, EPA-454/R-98-011, Locating and Estimating Air Emissions from Sources of Benzene, 6-41 (1998)
[hereinafter L&E Benzene Documents], available at www.epa.gov/ttn/chief/le/benzene_ptl.pdf.
www.epa.gov/ttn/chief/le/benzene_pt2.pdf; EPA, EPA-454/R-93-046, Locating and Estimating Air Emissions from
Sources of Methyl Ethyl Ketone, (1994), available a/http://www.epa.gov/ttn/chief/le/mek_l&e.pdf; EPA, EPA-
454/R-93-047, Locating and Estimating Air Emissions from Sources of Toluene, 6-20 (1994), available at
http://www.epa.gov/ttn/chief/le/toluene.pdf; EPA, EPA-454/R-93-048, Locating and Estimating Air Emissions from
Sources of Xylene (1994) [hereinafter L&E Xylene Documents], available at
http://www.epa.gov/ttn/chief/le/xylene.pdf.
39 AP-42, supra note 1, at ch. 4: Evaporation Loss Sources, available at
www.epa.gov/ttn/chief/ap42/ch04/index.html.
40 Id. at 4.3-17 (SIMS program and user manual can be downloaded from EPA's CHIEF electronic bulletin board).
41 Id. (SIMS requires, at a minimum, waste water flow rate and component surface area. Default values are provided
for all other inputs, however where site-specific information is available it should be entered in place of default
values for a more accurate emissions estimate).
42 L&E Benzene Documents, supra note 38, at 1-3; see also id., at 6-25 - 6-26 (providing that wastewater emissions
from petroleum refinery process units can be estimated by multiplying the average flow factor, the volatile HAP
concentrations, and the fraction emitted for the specific process unit capacity); see also id. at 6-27 (providing non-
AP42 wastewater emission factors for oil/water separators, air flotation systems, and sludge dewatering units at
petroleum refineries).
6
-------
Case l:13-cv-00621 Document 1-3 Filed 05/01/13 Page 8 of 14
In addition, it has been over fourteen years since EPA last reviewed L&E emission
factors for VOC emissions from wastewater treatment systems.43 EPA last reviewed L&E
emission factors for benzene emissions from wastewater treatment systems in 1998; L&E
emission factors for xylene emissions from wastewater treatment systems in 1994; and L&E
emission factors for carbon tetrachloride, epichlorohydrin, and ethylene dichloride emissions
from wastewater treatment systems in 1984.44 Because EPA has not reviewed AP-42 SIMS or
L&E emission factors for wastewater treatment systems in over fourteen years, the Administrator
has failed perform the nondiscretionary duty required by section 130 of the CAA, 42 U.S.C. §
7430.
III. Failure of the Administrator to Perform a Nondiscretionary Duty
Section 130 of the CAA, 42 U.S.C. § 7430, requires the Administrator to "at least every 3
years . . . review and, if necessary, revise, the methods ("emission factors") used for purposes of
[the CAA] to estimate the quantity of emissions of carbon monoxide, volatile organic
compounds, and oxides of nitrogen from sources of such air pollutants." This nondiscretionary
duty to "review and, if necessary, revise" emission factors under section 130 includes a duty to
make a determination - yes or no - as to whether revision of the emission factor is appropriate.45
The Administrator has not completed a review of emission factors for flares, tanks, or
wastewater treatment systems within three years of the prior review, as required by section 130
of the CAA, 42 U.S.C. § 7430 46 Specifically, the emission factors for tanks have not been
reviewed since 2006, the emission factors for flares have not been reviewed since 1991, and
emission factors for wastewater treatment systems have not been reviewed since 1998.
Therefore, EPA is in violation of the Act for its failure to comply with the mandatory duties
imposed by section 130 of the Act.
43 The following pollutants are VOCs: benzene, carbon tetrachloride, epichlorohydrin, ethylene dichloride, and
xylene.
44 L&E Benzene Documents, supra note 38, at 6-27; L&E Xylene Documents, supra note 38; EPA, EPA-450/4-84-
007b, Locating and Estimating Air Emissions from Sources of Carbon Tetrachloride, (1984), available at
http://www.epa.gov/ttn/chief/le/carbtet.pdf; EPA, EPA-450/4-84-007j, Locating and Estimating Air Emissions from
Sources of Epichlorohydrin, (1984), available at http://www.epa.gov/ttn/chief/le/epichlor.pdf; EPA, EPA-450/4-84-
007d, Locating and Estimating Air Emissions from Sources of Ethylene Dichloride 82 (1984), available at
http://www.epa.gov/ttn/chief/le/ethyldi.pdf.
45 See Envtl Def. Fund v. Thomas, 870 F.2d 892, 894-95 (2d Cir. 1989) (holding that "the Administrator has a non-
discretionary duty to make some formal decision whether to revise || NAAQS" under section 109 of the CAA. 42
U.S.C. § 7409(d), which requires the Administrator to "complete a thorough review of the criteria published under
Section 108 ... and promulgate such new standards as may be appropriate" every 5 years); Our Children's Earth
Found, v. U.S. Envtl. Prot. Agency, 527 F.3d 842, 849 (9th Cir. 2008) (holding that the duty to determine whether
revision is appropriate is implicit in EPA's non-discretionary duty to review and "if appropriate, revise" effluent
limitation guidelines once every five years under section 304 of the CWA, 33 U.S.C. 1311(d)).
46 EPA issued an emission estimation protocol for petroleum refineries in 2011 that outlines methodologies for
estimating emissions from flares, tanks, and wastewater treatment systems. EPA, Emission Estimation Protocol for
Petroleum Refineries (2011), available at
www.epa.gov/ttn/chief/efpac/protocol/Emission_Estimation_Protocol_for_Petroleum_Refinerie_052011.pdf.
However, the protocol does not discharge EPA of its obligations under section 130; EPA has neither conducted a
review to determine whether revisions to the emission factors are necessary, nor revised any emission factors. See
id.at ch.l, ch.3, ch.6, ch.7.
7
-------
Case l:13-cv-00621 Document 1-3 Filed 05/01/13 Page 9 of 14
Conclusion
The Administrator is required to review and, if necessary, revise section 130 emission
factors at least once every three years. However, EPA has failed to review emission factors for
tanks since 2006, for flares since 1991, and for wastewater collection, treatment and storage
since 1998. Plaintiffs intend to sue the Administrator for failure to perform the nondiscretionarv
duty under section 130 of the CAA and seek to compel EPA to review and revise the AP-42
emission factors used to estimate emissions of VOCs, NOx, and CO from flares, tanks, and
wastewater treatment systems to accurately reflect emissions from refineries, petrochemical
plants, and other industries. If you have any questions regarding the allegations in this notice or
would like to discuss this matter further, please contact us at the number or email below.
Respectfully submitted,
W
Sparsh Khandeshi
Environmental Integrity Project
One Thomas Circle, Suite 900
Washington, DC 20005
202.263.4456 or 202.263.4446
wferrell@environmentalintegrity.org
skhandeshi@environmentalintegrity.org
Counsel for:
Matthew Tejada Juan Parras
Executive Director Executive Director
Air Alliance Houston TEJAS
2409 Commerce Street 6733 Harrisburg Boulevard
CC {Via Certified Mail):
Eric H. Holder, Jr.
Attorney General of the United States
U.S. Department of Justice
950 Pennsylvania Avenue, NW
Washington, DC 20530-0001
Houston, TX 77003
Houston, TX 77011
Hilton Kelley
Executive Director
CIDA
1301 Kansas Avenue
Port Arthur, TX 77642
Anne Rolfes
Founding Director
Louisiana Bucket Brigade
4226 Canal Street
New Orleans, LA 70119
8
-------
Case l:13-cv-00621 Document 1-3 Filed 05/01/13 Page 10 of 14
Gina McCarthy
Assistant Administrator for the Office of Air and Radiation
U.S. Environmental Protection Agency
Ariel Rios Building
1200 Pennsylvania Avenue, NW
Washington, DC 20406
McCarthy.gina@epa.gov
Peter Tsirigotis
Sector Policies and Programs Division Director
U.S. Environmental Protection Agency
Ariel Rios Building
1200 Pennsylvania Avenue, NW
Washington, DC 20406
9
-------
SElMDtP,: COMPl.f-:n: 71-US SECIOH
ument
¦ Complete items 1,2, and 3. Also complete
item 4 if Restricted Delivery is desired.
¦ Print your name and address on the reverse
so that we can return the card to you,
¦ Attach this card to the back of the mailpiece,
or on the front if space permits.
1. Article Addressed to:
D. Is delivery address different from item 1?
If YES, enter delivery address below:
~ No
B. Received by ( Printed Name)
ciscx
U.S.
£V'. ev (kxVWvoy
\"LOO 4rct_,
v VDo "lOMOlo
3. Service Type
^pCCertified Mail
~ Registered
~ Insured Mail
~ Express Mail
~ Return Receipt for Merchandise
~ C.O.D.
4. Restricted Delivery? (Extra Fee)
~ Yes
2. Article Number
(Transfer from service labt
7D12 D47D DUDE 1S7D Bb74
PS Form 3811, February 2004
Domestic Return Receipt
102595-02-M-154C
-------
SENDEE: COMPi.rrE Ti-'iS SEC' lO.V
COMfl.nriz Tl-iiS SECTION 0 V CM/VERY
Bi Hi ¦¦ a a -attsf Ma-rB iFirelgTt Hlr 1 hi "hM
¦ Complete items 1, 2, and 3. Also complete
item 4 if Restricted Delivery is desired.
¦ Print your name and address on the reverse
so that we can return the card to you.
¦ Attach this card to the back of the mailpiece,
or on the front if space permits.
A. Signature
v D A9ent
~ Addressee
B. Received by ( Printed Name) 'J'- C. Date of Delivery
1. Article Addressed to:
oqT
D. Is delivery address different from item 1? ~ Yes
If YES, enter delivery address below: ~ No
Mo,
3. Service Type
~^Certified Mail ~ Express Mail
11 Registered ~ Return Receipt for Merchandise
~ Insured Mail ~ C.O.D.
4. Restricted Delivery? (Extra Fee) ~ Yes
2. Article Number - ,
(Transfer from service label) 7011 USD ODDS 3354 t3L] CI b
PS Form 3811, February 2004
Domestic Return Receipt
102595-02-M-1540
-------
EWDER: COMPLETE THIS SECTiON
ument 1
Complete items 1,2, and 3. Also complete
item 4 if Restricted Delivery is desired.
Print your name and address on the reverse
so that we can return the card to you.
Attach this card to the back of the mailpiece,
or on the front if space permits.
X
1. Article Addressed to:
GXvOCX
Kt * P-cAv
U-S. GPA
{Loo ?eAoSo^v>cvWex Ai^i
, Oo lOMOb
of 14
jfignal
c
B. Received by (Printed Name)
D. Is delivery address different from item 1 ?
If YES, enter delivery address below:
3. Service Type
Q Certified Mail
~ Express Mail
~ Registered
~ Return Receipt for Merchandise
~ Insured Mail
~ C.O.D.
4. Restricted Delivery? (Extra Fee)
~ Yes
2. Article Number-
so
(Transfer from service label)
7015 D47D DDD5 1B70 flbSO
PS Form 3811, February 2004
Domestic Return Receipt
102595-02-M-1540
-------
Track & ConfirmCase l:13-cv-00621 Document 1-3 Filed 05/01/13 Page 14 of 14
GET EMAIL UPDATES
YOUR LABEL NUMBER
SERVICE
First-Class Mail®
STATUS OF YOUR ITEM DATE & TIME
Delivered July 23, 2012, 8:49 am
Notice Left (No
Authorized Recipient
Available)
Arrival at Unit
Facility
Acceptance
July 22, 2012, 12:16 pm
July 22, 2012,10:22 am
July 19,2012
Depart USPS Sort
Facility
FYocessed at USPS Juiy 18, 2012, 9:07 pm
Origin Sort Facility
Dispatched to Sort
July 18, 2012, 5:41 pm
July 18, 2012, 3:39 pm
LOCATION
WASHINGTON, DC 20460
WASHINGTON, DC 20460
WASHINGTON, DC 20018
CAPITOL
HEIGHTS, MD 20790
CAPITOL
HEIGHTS, MD 20790
WASHINGTON, DC 20005
WASHINGTON, DC 20005
Expected Delivery By:
July 19, 2012
Certified Mail™
Return Receipt
------- |