EPA-450/3-77-030
October 1977
REVISION
OF EMISSION FACTORS
FOR PETROLEUM
REFINING
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
-------
EPA-450/3-77-030
REVISION OF EMISSION FACTORS
FOR PETROLEUM REFINING
by
CE. Burklin
Radian Corporation
8500 Shoal Creek Blvd.
Austin, Texas 78766
Contract No. 68-02-1889
Task No. 2
EPA Project Officer: CC. Masser
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
October 1977
-------
This report is issued by the Environmental Protection Agency to report
technical data of interest'to a limited number of readers. Copies are
available free of charge to Federal employees, current contractors and
grantees, and nonprofit organizations - in limited quantities - from the
Library Services Office (MD-35), Research Triangle Park, North Carolina
27711; or, for a fee, from the National Technical Information Service,
5285 Port Royal Road, Springfield, Virginia 22161.
This report was furnished to the Environmental Protection Agency by
Radian Corporation, 8500 Shoal Creek Blvd., Austin, Texas, in fulfillment
of Contract No. 68-02-1889, Task No. 2. The contents of this report
are reproduced herein as received from Radian Corporation. The opinions,
findings, and conclusions expressed are those of the author and not
necessarily those of the Environmental Projection Agency. Mention of
company or product names is not to be considered as an endorsement
by the Environmental Protection Agency.
Publication No. EPA-450/3-77-030
-------
TABLE OF CONTENTS
Page
1.0 SUMMARY 1
1.1 Process Description Review of AP-42 1
1.2 Emission Factor Review of AP-42 1
1.3 Test Plan Development . . . 2
2.0 INTRODUCTION 5
3.0 REVISION OF AP-42; SECTION 9.1, PETROLEUM
REFINERIES 6
3.1 Data Gathering Operations 6
3.2 Development of Improved Process
Descriptions 11
3.3 Emission Factor Review and Update 13
3.3.1 Boiler and Heater Emission Factors. 13
3.3.2 Catalytic Cracking Emission
Factors : 16
3.3.3 Fluid Coker Emission Factors 17
3.3.4 Compressor Engine Emission Factors. 18
3.3.5 Slowdown System Emission Factors .. 18
3.3.6 Vacuum Distillation Column
Condenser Emission Factors 19
3.3.7 Fugitive Emission Factors 20
3.3.8 Methane/Npn-methane Hydrocarbons .. 21
3.4 Data Supplied by EPA 22
4.0 TEST PLAN DEVELOPMENT 25
4.1 Emission Factor Background 25
4.1.1 Boilers and Heaters 26
4.1.2 Catalytic Crackers 28
4.1.3 Fluid Coking Units 30
4.1.4 Internal Combustion Engines 32
4.1.5 Slowdown Systems 33
111
-------
TABLE OF CONTENTS (Continued)
Page
4.1.6 Vacuum Distillation Column
Condensers 35
4.1.7 Glaus Plant Tail Gas 36
4'. 1.8 Storage and Loading Operations," .... 37
4.1.9 Fugitive Emission Sources 38
4.2 Sampling Program Costs . .. ._.. . . .... 40
"4.2.1 Sampling Strategy ................. 40
4.2.2 Point Source Emissions, . .....,,,.,,. 41
" "4.2.3 Fugitive Source Emissions;' . v-: ^:: . 42
4.2.4 Tankage, Emissions ....-,-,. ..,->.-;> ; -.- 43
4.3 Prioritization of Emission,,S(purces! . . . .;. . . 43
4.3.1 . High Priority . .-.vJ'-.V'.-;'../: . i 44
4.3.2 Intermediate Priority .........."... 45
;-. 4.3.3 Low Priority . . -. .'..... .-. . . .% 45
REFERENCES . .". '..'.'. '. . . .'V. 47
::APPENDIX A - REVISED AP-42 ....!...; A-l
iv
-------
1.0 SUMMARY
The program conducted by Radian in support of EPA's
Office of Air Quality Planning and Standards is summarized here
in Section 1. Radian provided assistance to the EPA in the
following two task areas:
1) Process descriptions and emission factors pre-
sented in the petroleum refinery section of
EPA document AP-42 were reviewed and expanded.
2) A testing strategy for refinery emission sources
with inadequate emission factors was developed.
1.1 Process Description Review of AP-42
The general descriptions of petroleum refining
objectives and the major processes used to
achieve these objectives were expanded.
Detailed process descriptions were prepared
for fourteen major refinery emission sources.
These descriptions were accompanied by a
discussion of emission characteristics and
applicable emission control technology.
1.2 Emission Factor Review of AP-42
Corrections, adjustments, and additions were
made to a major portion of the non-fugitive
emission factors.
-------
Fugitive emission factors were assembled into
a separate table. The fugitive emission factor
table lists emission factors for controlled
sources as well as emission factors for uncon-
trolled sources.
Where available, fugitive emission factors are
reported in more than one set of units to add
to their utility.
Where recent information indicated emission
factors were outdated, the emission factor accu-
racy ratings were lowered appropriately. Where
before all factors were rated A, most are now
rated B and C. Fugitive emission factors were
given a general rating of D.
1. 3 Test Plan Development
It was recommended that four refinery emission
sources be given high priority for emission
testing programs: fugitive, storage, blowdown,
and vacuum distillation column condensers. A
significant improvement can be gained in the
emission factor accuracy for fugitive, blowdown,
and vacuum distillation column condenser sources
Less improvement can be gained for storage
emission factors, but the magnitude of storage
emissions makes this improvement significant
also.
-2-
-------
Three refinery emission sources were given an
intermediate priority status: loading opera-
tions , internal combustion engines, and fluid
coking units. Emission factors for loading
operations are good, but the large volume of
loading losses makes their inaccuracies signifi-
cant. Internal combustion engine factors are of
less importance because internal combustion
engines are being phased out of use in the re-
fining industry. Fluid coking emissions are
small nationally, (%16% of the coking capacity)
but will become more important if the use of the
process increases.
Five emission sources were given low priority
status for testing programs because of their low
potential for significant accuracy improvements:
boilers, heaters, fluid catalytic crackers, mov-
ing bed catalytic crackers, and sulfur recovery
plants. Boilers and heaters have excellent
emission factors. Although emission factors are
not as good for fluid catalytic crackers, moving
bed catalytic crackers, and sulfur recovery
plants, these factors are not easily improved
because their inaccuracies are attributed to
variations in undefined process parameters.
The estimated cost for sampling point source
emissions is $100,000 per source category. A
fugitive emission source testing program is
estimated to cost a total of $1,400,000. It has
also been estimated that tankage emission sources
might be tested for $750,000. It should be noted
-3-
-------
that a broad study to characterize fugitive
emissions from petroleum refineries is currently
underway. The study is being conducted by Radian
Corporation for the Industrial Environmental
Research Laboratory of EPA, and includes the
testing of emissions from eleven categories of
fugitive emission sources in sixteen refineries.
-4-
-------
2.0 INTRODUCTION
Between 1955 and 1958 a task force consisting of
petroleum companies and air control agencies conducted the Los
Angeles Joint Project to determine the air emissions from petro-
leum refining operations. The refinery air emission factors
presently contained in Section 9.1 of AP-42, A Compilation of
Air Pollutant Emission Factors are based largely on this study.
However, there have been many technological improvements since
1958, and there have been many individual efforts to obtain
emission source testing data for the various processes and
equipment in refineries. Therefore, the objectives of this
study were 1) to revise the emission factors and process des-
criptions presented in AP-42 for refinery operations using
available information, and 2) to develop a comprehensive testing
strategy for emission sources where further source testing is
warranted.
The process descriptions were written to give a clear
understanding of overall refining activities with emphasis
placed on significant emission sources. The emission factors
were updated where possible. The revised factors reflect
improved technology and recent source testing data.
Having achieved an overview of the available source
testing data, a comprehensive testing strategy was developed
for sources whare further testing is needed. The strategy
development i.icl^ded an analysis of difficulty in testing,
relative emission contribution, available data, and accuracy of
the present factor.
-5-
-------
3.0 REVISION OF AP-42; SECTION 9.1, PETROLEUM REFINERIES
The major objective of this program was the develop-
ment of improved emission factors and process descriptions for
the petroleum refining industry. These improved factors and des-
criptions comprise the revisions to the petroleum refining
section of EPA Document AP-42, Compilation of Air Pollutant
Emission Factors. Section 3.1 discusses the data gathering
operations conducted by Radian for this program. Section 3.2
discusses the development of improved process descriptions and
Section 3.3 presents the methodologies used by Radian to update
the refinery emission factors. And Section 3.4 reviews the
testing information transmitted to Radian from the EPA. The
revised Section 9.1 of AP-42 on refinery emission sources is
presented in Appendix A.
3.1 Data Gathering Operations
To update the emission factors and process descriptions
for petroleum refineries contained in AP-42, it was necessary
to obtain all available information pertaining to petroleum
refining processes and emissions. Because this project concerns
both the processing and environmental aspects of refining,
information covering a broad range of topics was gathered.
Topical areas included general refinery technology information,
specific process data, and emission control technologies.
Sources of information were government agency reports, federal
and state publications and support documents, technical papers,
journals, industry news releases, process design data, previous
and current Radian studies, and contacts with government and
industrial personnel associated with petroleum refining and
its emissions.
-6-
-------
Information was gathered from the literature by using
computer assisted information services. The literature search
involved preliminary machine searches to several on-line
abstracting services, including Petroleum/Energy Business News
Index, Chemical Abstracts, NTIS/Governmental Reports Announce-
ments, and Engineering Index. Chemical Abstracts and Engineer-
ing Index provided the most useful information. Therefore,
searches were made of these two data bases using the broad
terms "Petroleum Refining", "Petroleum Products", "Refinery
Emissions", "Refinery Effluents", and "Refinery Wastes". Both
data bases were subsequently searched for the various refinery
processes. All searches were completed for the time period of
1972 to the present. Previous studies conducted by Radian
Corporation had extensively covered all data generated prior to
1972 on the subject of refinery processes and emissions. This
data was available to the project through the Radian library.
Printouts of titles obtained from Chemical Abstracts and abstracts
obtained from Engineering Index were scanned, and the pertinent
articles and/or abstracts were acquired.
Personnel in both industry and government were con-
tacted in an effort to obtain the latest information regarding
refinery processes, emission studies, and control technologies.
Table 3.1-1 lists the contacts made by Radian and reviews the
content of the discussions. As the table shows, very few re-
finery emission studies have been made recently. There are
several parallel studies currently underway or recently com-
pleted. These studies include investigations of emissions from
offshore platforms, tankers, tankage, and marketing. However,
no appreciable refinery emission source testing data is avail-
able at this time. Chuck Masser of EPA has furnished source
testing data for process heaters and boilers. PES had some
refinery source testing data that could not be released to us
in time for use in this study due to regulations concerning the
confidentiality of the data.
-7-
-------
TABLE 3.1-1. INDUSTRY CONTACTS
Name and Affiliation
Address and Phone
Subjects, Comments
KVB
Steve Cherry
17332 Irving
Tustin, Calif. 92680
(714) 832-9020
EPA - National Emissions
Inventory
Charles C. Masser
EPA, OAQPS, MDAD,
AMTB, SAS
Room 526, Mutual Bldg.
MD-14
Research Triangle Park,
North Carolina 27711
1) No refinery emis-
sion data available
other than what EPA
had already trans-
mitted to Radian
Corporation.
1) Supplied emissions
data for process
heaters and boilers.
2) Supplied NEDS
inventories.
EPA - Division of Sta-
tionary Source Enforcement
Jim Casey
Washington, B.C.
(202) 755-7927
1) Confidential refinery
source test results
which can't be re-
leased within the
time limits of our
study. Some ques-
tion also existed on
usefulness of the
data.
Pacific Environmental
Services
Artie Stein
1930 14th St.
Santa Monica, Calif.
90404
(213) 393-9449
1) Have done no testing
of refinery emission
sources.
2) Have received some
industry source test
results from industry
as a result of their
National Survey of
Refineries study; but
are confidential.
(Continued)
-8-
-------
TABLE 3.1-1. INDUSTRY CONTACTS (Continued)
Name and Affiliation
Address and Phone
Subjects, Comments
EPA - Chemical and
Petroleum Branch
Kent Hustvedt
Research Triangle Park, 1) No new refinery
Durham, N.C. 27711 emissions data
(919) 541-5371 available to his
knowledge.
API - Division of
Environmental Affairs
Edward Crockett
1801 K St. N.W.
Washington, D.C.
(202) 457-7084
1) They have canvassed
all members about
available test data
this year. No
appreciable refinery
emissions data was
turned up.
Western Oil and Gas
Association
Bob Harrison
Los Angeles, Calif.
(212) 486-7538
1) WOGA is not conducting
any studies on refin-
ery emissions sources.
2) WOGA is presently
studying emissions
from tanker opera-
tions, floating roof
tanks, and cone roof
tanks.
EPA - Chemical and
Petroleum Branch
Dick Burr
Research Triangle Park, 1) No new emissions data.
Durham, N.C. 27711
(919) 541-5371
Union Oil
Edwara Bloom
Dick Salsbury
Los Angeles, Calif.
(213) 486-7538
1) WOGA has plans to
conduct refinery
tests.
2) No current emissions
data.
(Continued)
-9-
-------
TABLE 3.1-1. INDUSTRY CONTACTS (Continued)
Name and Affiliation
Address and Phone
Subjects, Comments
ARCO - Production
Engineering Department
John Hundley, Jr.
Bakersfield, Calif,
(805) 831-1600
1) ARCO has not con-
ducted any refinery
emission tests.
California Air Resources
Board
Jim Leach
Sacremento, Calif.
(916) 322-2745
2) ARCO has data on
emissions from pro-
duction equipment.
1) No recent refinery
emission studies.
Earlier data by KVB
already transmitted
to Radian.
Santa Barbara County Health
Department
John Laird
Santa Barbara, Calif,
(805) 967-2311
ARCO
Harvey Grimes
Harvey, Illinois
(312) 333-3000
1) No refinery emission
studies.
2) Test data on emis-
sions from offshore
platforms.
1) No refinery emission
studies.
2) Marketing loss
studies only.
-10-
-------
3.2 Development of Improved Process Descriptions
The process descriptions were written with two main
objectives. The first objective was to present a clear picture
of overall refining activities. The second was to emphasize
those refining processes which are the major sources of air
emissions. To accomplish these objectives it was necessary to
describe the processes in sufficient detail to clearly define
the process flows, major equipment items, and emission sources.
The descriptions included updated process information. It
was further necessary, because of space limitations, for the
process descriptions to be concise.
The description of the petroleum refining industry
was organized in the following manner:
1) statement of the refinery's overall objective,
2) definition and description of five general
process categories, and
3) definition and description of specific processes
that are significant air pollutant contributors.
The overall refinery description discusses products,
feedstocks, and objectives, and presents the diagram of an
example refinery flow scheme.
The individual refinery processes are categorized as
separation processes, conversion processes, treating processes,
product handling, and auxiliary facilities. The descriptions
of these five categories of processes discuss the products,
feedstocks, and objectives of the processes within the category
-11-
-------
Detailed descriptions are presented for the specific
refinery processes that are significant air pollutant contribu-
tors. These descriptions include products, feedstocks, pro-
cessing objectives, process flows, emission sources, and appli-
cable control technology.
Refining processes that are significant sources of
air pollutants are emphasized. The description of these pro-
cesses are more detailed than the descriptions presently con-
tained in AP-42. The expansion of the process descriptions
should enable a clearer understanding of the process and its
emission sources. Objectives, feedstocks and products are
discussed. A process flow scheme is presented to allow a
clear understanding of the process. Operating conditions
(e.g., high temperature and pressure) and equipment (e.g.,
process heaters and catalyst regenerators) important to consi-
derations of air pollutant sources are emphasized. Finally,
the types of air pollutants, emission sources, and applicable
control technology are discussed.
The sources of information for these descriptions
are recent texts, journals, periodicals, and Radian reports.
In general, specific process descriptions were based on the
most widely used process. Where two or more processes are in
wide use either two separate descriptions are included, or a
generalized description encompassing both processes is included.
Although refining processes are complex and unique to the
specific refinery needs, the generalized process descriptions
do allow an understanding of the process flows and the basic
processing equipment involved.
-12-
-------
3.3 Emission Factor Review and Update
As part of the revision of EPA Document AP-42, Radian
reviewed and updated the emission factors presented in AP-42
for refinery emission sources. Generally, this review and up-
date was based on information collected by Radian from an exten-
sive literature search and through contacts with the petroleum
industry and air control agencies. In this section the revisions
made by Radian to the refinery emission factors and the basis for
the revisions are presented. Information gathering methods were
discussed in Section 3.1. The revised refinery section of AP-42
is presented in Appendix A. Table 3.3-1 summarizes which emis-
sion factors were changed, and for what reasons. Section 3.4
reviews the testing information transmitted to Radian from the
EPA.
3.3.1 Boiler and Heater Emission Factors
Changes were made to both the carbon monoxide emission
factors and the hydrocarbon emission factors currently reported
in AP-42 for refinery boilers and heaters. The Los Angeles
Joint Project, upon which the existing refinery boiler and heater
emission factors were based, found the carbon monoxide emission
rates to be negligible (AT-040). However, later studies con-
ducted by EPA and Southwest Research Institute identified the
level of carbon monoxide from industrial boilers to be 210 Ibs/
1000 bbl oil burned or 17 lbs/105 ft3 gas burned (EN-071 Section
L.3 and 1.4}. Since refinery heaters operate under similar
conditions to refinery boilers, Radian assumed the carbon monoxide
rate for the two to be similar. The refinery boiler and heater
carbon monoxide emission rates were changed from negligible to
the above values.
-13-
-------
TABLE 3.3-1. SUMMARY OF EMISSION FACTOR REVISIONS
Emission
Source
Boilers &
Heaters
Cat. Cracker
w/CO boiler
Pollutant
CO
HC
HC
Aldehyde
Original Value
Negligible
140 lbs/1000 bbl
0.03 lbs/1000 ft3
220 lbs/1000 bbl
71 lbs/1000 bbl
Revised Value
210 lbs/1000 bbl
42 lbs/1000 bbl
0.003 lbs/1000 ft
Negligible
Negligible
Reason and
Source of Change
Studies by EPA &
SwRI (EN-071
Sec. 1.3 & 1.4)
3
Based on a reviei
of incinerator
Fluid Coker
Uncontrolled
Fluid Cokers
w/CO boilers
Reciprocating
Compressor
Engines
CO
HC
NO
Aldehydes
Ammonia
NO
Negligible
Negligible
Negligible
Negligible
Negligible
Not available
Not available
Not available
Not available
Not available
CO
HC
NO.
Negligible Not available
Negligible 0.43 lbs/1000 ft3
1.2 lb/1000 ft3 1-4 lbs/1000 ft3
0.9 lb/1000 ft3 3.4 lbs/1000 ft3
Gas Turbine
Compressor
Engines
S0x
CO
HC
NO^
None
None
None
None
Slowdown Sys-
tem uncontrolled
2s lbs/1000 ft3
0.12 lbs/1000 ft3
0.02 lbs/1000 ft3
0.3 lbs/1000 ft3
HC 300 lbs/1000 bbl 580 lbs/1000 bbl
impact on these
pollutants.
These pollutants
were not inves-
tigated in the
sources quoted.
These pollutants
were not inves-
tigated in the
sources.
Later studies by
SwRI developed
these revised
emission factors
(EN-071 Sec. 3.3)
based on exten-
sive testing.
These factors
did not previously
appear in this
table. Based on
SwRI studies
(EN-071 Sec. 3.3)
Lockheed Study
including more
extensive test-
ing (KL-081)
(Continued)
-14-
-------
TABLE 3.3-1. SUMMARY OF EMISSION FACTOR REVISIONS (Continued)
Emission
Source Pollutant
Slowdown sys-
tem controlled
w/flare
S0x
CO
HC
N()
Original Value
None
None
0.5 lb/1000 bbl
None
Revised Value
26.9 lb/1000 bbl
4.3 lb/1000 bbl
0.8 lb/1000 bbl
18.9 lb/1000 bbl
Reason and
Source of Change
Lockheed study
used to update
SOX, CO and He.
(KL-081) . Chass
Vac distillation HC
column condenser
130 lb/1000 bbl 50 lb/1000 bbl
and George used
to update NO
(CH-055).
Original factors
misquoted from
data sources
(AT-040, AM-055).
-15-
-------
The hydrocarbon emission factors currently used in
AP-42 were also developed in the Los Angeles Joint Project
between 1955 and 1958 (AT-040). This study determined the
hydrocarbon emission rates to be 0.026 lbs/1000 ft3 of gas and
142 lbs/1000 bbl of oil. Recent, more extensive studies have
been conducted by EPA on hydrocarbon emissions from commercial
and industrial boilers (EN-071 Section 1.3 and 1.4). These
studies determined the level of hydrocarbon emissions from
boilers to be approximately 0.003 lbs/1000 ft3 of gas and 42
lbs/1000 bbl of oil. Since refinery boilers and heaters
operate under similar conditions to industrial boilers, Radian
assumed the hydrocarbon emission rates to be similar. The
refinery boiler and heater hydrocarbon emission rates were
changed to the more recent EPA values presented above.
The EPA and SwRI studies, and EPA compliance test data
collected in 1974 by KVB (BA-291) verified the rest of the Los
Angeles Joint Project emission factors for boilers and heaters,
and these were left unchanged. Two heater tests conducted by
Battelle for the EPA reported particulate emissions from oil-
fired heaters to be approximately 15 to 20 lbs/1000 bbl. These
rates are significantly lower than all other rates reported in
available data. Because of the limited extent of these tests
it was decided not to change the heater emission rates until
more data becomes available.
3.3.2 Catalytic Cracking Emission Factors
The refinery emission factors currently reported in
AP-42 for both fluid bed (FCC) and moving bed (TCC) catalytic
crackers were primarily developed in the Los Angeles Joint
Project (AT-040). However, the particulate and carbon monoxide
-16-
-------
emission factors for FCC units were developed from later infor-
mation gathered by EPA in NSPS studies on FGC units (EN-072).
Based on this information, current AP-42 emission factors incorrectly
indicate that CO boilers have no impact on either hydrocarbon,
aldehyde, or ammonia emissions in FCC regeneration flue gas.
A review of EPA information on incinerators equipped with
afterburners indicates that hydrocarbons and aldehydes are
reduced to negligible quantities by secondary combustion equip-
ment (EN-071, Section 2). With this information as a bases,
Radian estimated that the CO boiler will combust the major
portion of the hydrocarbon, aldehyde, and ammonia emissions
in the regenerator off gas, thus lowering these emissions to
negligible levels.
3.3.3 Fluid Coker Emission Factors
The current AP-42 emission factor table for refin-
eries reports that carbon monoxide, hydrocarbon, nitrogen
oxide, aldehyde, and ammonia emissions are all negligible for
uncontrolled fluid cokers. However, a review of the literature
cited as the source for these emission values indicates that
only particulate emissions were measured on the uncontrolled
fluid coker and that the negligible emission values were
measured downstream from a carbon monoxide boiler (JO-086).
Based on this information, Radian changed the "negligible"
values to "NA" (not available) indicating that no information
is available en the emission of carbon monoxide, hydrocarbons,
aldehydes, nitrogen oxides, and ammonia from uncontrolled
fluid coking units.
-17-
-------
The negligible emission status was left unaltered
for carbon monoxide, hydrocarbon, aldehyde, and ammonia emis-
sions from fluid coking units equipped with CO boilers. It is
very likely that these emissions will be very low, if not
negligible.
3-3.4 Compressor Engine Emission Factors
The emission factors currently used in AP-42 for
reciprocating compressor engines in refinery service were
developed in the Los Angeles Joint Project conducted from 1955
to 1958 (AT-040). More extensive studies were conducted for the
American Gas Association by the Southwest Research Institute in
1974 and 1975 (UR-022, DI-142). These SwRI studies investigated
carbon monoxide, hydrocarbon, and nitrogen oxide emissions from
reciprocating compressor engines in the gas industry. The
test results obtained by SwRI indicated that the Joint Project
emission factors were low for all three pollutants. Radian
reviewed the SwRI testing program and found it to be a thorough
and accurate study. It was decided to substitute its results
for the reciprocating compressor engine emission factors
currently used by AP-42.
The SwRI studies also investigated SOX, CO, hydrocar-
bon, and NOX emissions from turbine compressor engines. These
emission factors are reported in Section 3.3.2 of AP-42 and are
added to the refinery emission factor table, Table 9.1-1.
3.3.5 Slowdown System Emission Factors
The emission factor currently presented in AP-42 for
refinery blowdown systems was developed in the Los Angeles Joint
Project between 1955 and 1958. The joint project found that
-18-
-------
uncontrolled blowdown system emissions averaged 300 Ibs of hydro-
carbons per thousand barrels of refinery feed. If the hydrocar-
bon vapors were flared or processed in vapor recovery systems,
these emissions were reduced to 5 Ibs per thousand barrels of
refinery feed (AT-040).
However, the Lockheed Missile and Space Company has
recently completed a study of blowdown systems for the EPA
(KL-081). This study was conducted with the assistance of
the American Petroleum Institute and eighteen refineries.
The study results indicate that in today's average refinery,
580 Ibs of hydrocarbons per thousand barrels of refinery feed
are vented to the blowdown system. In all cases studied, these
blowdown effluents are flared instead of vented to the atmos-
phere. The Lockheed study also measured the sulfur oxide,
carbon monoxide, and hydrocarbon emissions from the flare.
Because of the greater applicability of the Lockheed study to
today's refineries, Radian updated the refinery blowdown system
emission factors by using the Lockheed study results.
No testing information is available on the level of
NOX emissions from flares. Radian estimated nitrogen oxide
emissions from blowdown flares using the results of a nitro-
gen oxide emission study conducted by Chass and George in 1958
(CH-055). The Chass and George study was not specific to
flares but dealt with general combustion sources.
3.3.6 Vacuum Distillation Column Condenser Emission Factors
Radian changed the emission factors for vacuum dis-
tillation column condensers from a value of 130 Ibs of hydro-
carbons per thousand barrels of vacuum column feed to a value
of 50 Ibs of hydrocarbons per thousand barrels of vacuum
-19-
-------
column feed. AP-42 reports that the current emission factor of
130 lbs/1000 bbl of vacuum unit feed was developed by the Los
Angeles Joint Project Study. A review of the Joint Project
findings indicates that the range of emissions was from 0 to 130
Ibs of hydrocarbons/1000 bbl of vacuum unit feed (AT-040, AM-055)
However, most emissions fell in a. range of 15 to 50 lbs/1000 bbl
with the average being close to the higher figure.
In addition to revising the vacuum distillation column
condenser emission factor, Radian also added a factor in the
units of 18 Ibs of hydrocarbons/1000 bbl of refinery feed.
Radian arrived at this factor using the typical vacuum unit feed
to refinery feed ratio experienced in today's refinery (CA-339).
Reporting the vacuum distillation column condenser emission fac-
tors in these units (in addition to the original units) will
increase the utility of the factors in cases where only the
refinery capacity is known.
3.3.7 Fugitive Emission Factors
Radian made no changes in the numerical values of the
emission factors reported in AP-42 for fugitive emission sources.
All values were checked against both the original findings of
the Joint Project (AT-040) and recent findings from petroleum
company testing (WO-099). The recent data on fugitive emission
rates indicate that good maintenance practices result in lower
emission factors than the current AP-42 factors. But there is
insufficient data for Radian to generate new factors.
However, Radian did change the format of the fugitive
emission factor table. Not only did Radian report the uncon-
trolled emission rates as before, but Radian also added the
controlled emission rates where available and listed the appli-
cable control technology. Radian also reported many of the
-20-
-------
fugitive emission factors in alternate units. As an example,
emissions from process drains and wastewater separators had
been reported in only units of "Ibs of hydrocarbons per thousand
gallons of wastewater". Radian also reported the emission fac-
tor in units of "Ibs of hydrocarbons per thousand barrels of
refinery capacity". Presenting the emission factors in several
units greatly increases their versatility.
3.3.8 Methane/Non-methane Hydrocarbons
Radian conducted a literature study to determine what
portion of the various refinery hydrocarbon emissions are methane,
It has been determined that methane does not contribute signifi-
cantly to the formation of photochemical oxidants or to smog.
However very little information was available on the composition
of individual refinery emissions. In a study for EPA, TRW has
reported that the overall composition of refinery hydrocarbon
emissions in the Los Angeles area is 0.3 wt 70 methane and 99.7
wt 7o non-methane (TR-107) . Rockwell International has also
investigated the hydrocarbon emissions from cat cracker regener-
ators and found them to contain less than 1% methane and greater
than 9970 non-methane hydrocarbons (GR-360) . No other hydrocarbon
composition information was available for specific refinery
emission sources, although some information was available for
similar emission sources in other industries. Another point
Radian considered is that much more extensive data is currently
being collected by Radian for the EPA on another program. This
data includes the methane concentration in hydrocarbon emissions
from numerous refinery emission sources. The data from this
program will be included in AP-42 at the completion of the pro-
gram. For these reasons, it was decided not to attempt to
develop methane/non-methane compositions for specific refinery
emissions. But Radian simply reported the overall methane/non-
methane composition of refinery emissions in this revision of the
refinery section of AP-42.
-21-
-------
3.4
Data Supplied by EPA
Five sets of emission data were received from the
National Air Data Branch of EPA for inclusion in this study
where applicable. Much of this information was found to be
incomplete or insufficient, for the development of new emission
factors. However, test data collected by KVB did prove to be
very helpful in verifying the accuracy of existing emission
factors. Table 3.4-1 summarizes the test data received from
the EPA.
TABLE 3.4-1. DATA SUPPLIED BY EPA
Source of Test Data Emission Source
Emissions
Lace Engineering
Texas Air Control
Board
Battelle Research
Labs
Battelle Research
Labs
KVB
Furnace
CO Boiler
Crude Heater
SO;
Particulates & S02
Particulates & S02
CO Boiler Particulates & S02
Furnaces & Boilers NO^
The first set of test data was for sulfur dioxide
emissions from a refinery furnace on a HDS depentanizer. These
tests were conducted by Lace Engineering at the Cosden Oil and
Chemical facility in Big Springs, Texas. The results of the
tests basically verified that the S02 emissions from a refinery
heater or furnace can be calculated directly and accurately
from the sulfur level in the fuel.
The second set of test data was sulfur dioxide and
particulate emission data from a CO waste heat boiler at the
Cosden Oil and Chemical facility in Big Spring, Texas.
-22-
-------
Difficulties were encountered with using these test results
because the samples were collected downstream of double stage
cyclones. No efficiencies were reported for the cyclones and
they were reported to be operating improperly. In addition,
only one set of particulate and one set of SOz tests were made.
Because of the many unknown and the limited quantity of data,
it was decided not to use these test results to modify the
existing EPA CO boiler emission factors. The EPA factors are
based on a much larger quantity of test results collected under
more controlled test conditions.
The third set of test data was collected by Battelle
Laboratories on a crude oil heater in Shell Oil Company's
Anacortes, Washington refinery. The sulfur dioxide emission
data again verified the accuracy of using a material balance
for calculating sulfur dioxide emission rates from the sulfur
content of the fuel. Two of the particulate test results were
not useable because of mixed fuel firing and difficulties in
assessing pollutant contributions by each fuel. The remaining
two particulate test results indicated particulate emissions
from fuel oil combustion are 0.37 and 0.45 lbs/1000 gallons
of fuel. These two values are significantly lower than the 10-20
lbs/1000 gallons reported in most other test results available.
Further investigation did not reveal the reason for these
largely varying values. It was decided that these two test
values were insufficient information to revise the existing EPA
emission factors.
The fourth set of test data were particulate and
sulfur dioxide emissions from a CO waste heat boiler at Texaco's
Anacortes, Washington refinery. These tests were also conducted
by Battelle Laboratories. Due to heavy firing of auxiliary
oil and gas fuels, it was not possible to accurately determine
-23-
-------
the contribution of each fuel to the net pollutants in the stack
gas.
The fifth set of data was a collection of nitrogen
oxide emission results collected by KVB for several West Coast
refineries. The data and test conditions for each of these
emission rates were in various states of completeness. The
three emission sources burning a single fuel and having complete
test information indicated NOX emissions from refinery fuel oil
boilers ranged from 50 to 60 Ibs per 1000 gallons of fuel oil.
This was found to be within the same range of the currently used
EPA emission factor of 2900 lb/1000 bbl (87 lb/1000 gal). The
high dependence of NOX emission rate on combustion conditions
makes it difficult to develop emission factors with greater
accuracy than exhibited by these test results.
-24-
-------
4.0 TEST PLAN DEVELOPMENT
Many of the emission factors presented in AP-A2 for
refinery emission sources are inaccurate. Depending on the
process, actual emissions may differ by as much as an order of
magnitude from the values calculated using AP-42. However, not
all of these inaccurate emission factors will warrant testing
programs initially. Some refinery emission sources with
inaccurate emission factors are not widely used or are currently
being phased out of the petroleum industry. For other emission
sources the accuracy of the factors cannot be improved because
of the erratic nature of the source's emission rates. And,
because of high sampling costs, it may not be cost effective to
sample some refinery emission sources.
This section of the final report discusses the merits
of conducting emissions testing programs for various refinery
emission sources. Section 4.1 reviews the origin and accuracy
of existing emission factors. Section 4.2 discusses the esti-
mated costs associated with testing these emission sources.
Finally Section 4.3 suggests a prioritization of refinery emis-
sion sources in order of sources most benefited by an emission
testing program.
4.1 Emission Factor Background
This section reviews the origin and accuracy of the
emission factors currently presented in EPA Document AP-42 for
refinery emission sources.
-25-
-------
4.1.1 Boilers and Heaters
Emission Impact
Steam boilers and process heaters are used in every
major refinery to supply process and utility steam, and to heat
process streams. The steam demand for a typical integrated
refinery is approximately 40,000 lb/1000 bbl of refinery feed.
To generate this amount of steam, a boiler capacity of 53 x 106
Btu/1000 bbl of refinery feed would be required. The heat
demand for a modern integrated refinery is approximately 270 x
106 Btu/1000 bbl of refinery feed. However, older, less effi-
cient refineries may have process heater demands approaching
600 x 106 Btu/1000 bbl of refinery feed. The total boiler and
heater fuel demand for the typical refinery is approximately
323 x 106 Btu/1000 bbl of refinery feed. This fuel demand is
generally supplied by refinery fuel gas but may also be supplied
by residual oil and other internally produced fuels (BU-185) .
Table 4.1-1 presents the estimated emission rates for
fuel gas fired boilers and heaters in a typical refinery having
a total heat demand of 323 x 10s Btu/1000 bbl of refinery feed.
As this table indicates, refinery boilers and heaters are poten-
tially a significant contributor to both the sulfur oxide and
the nitrogen oxide emissions from refineries. Sulfur contents
of fuel gas range from 0 to 1 percent. Sulfur oxide emissions
resulting from burning a 1% sulfur fuel gas are approximately
55 lbs/1000 bbl of refinery feed. Nitrogen oxide emissions are
approximately 71 lb/1000 bbl of refinery feed.
Emission Factor Accuracy
Emission factors for refinery boilers and heaters have
been given an emission factor rating of A, which indicates that
-26-
-------
TABLE 4.1-1. PRIORITIZATION OF REFINERY EMISSION SOURCES FOR EMISSION TESTING PURPOSES
Typical emission rates
(tb/IO* bbl refinery feed)
Part. SO CO He NO
EPA Estimated
emission Potential teat pro-
factor rating after grum expenses
rating emission testing ($)
CoMMnt
1. Rollers and Heaters
2. Fluid Cat. Cracking (controlled)
6 55 6 1 71
11 139 neg neg 20
1. Moving Bed Cat Cracking (controlled) neg 1 neg neg neg
4. Fluid Coking Units (controlled) neg neg neg neg neg
5. Reciprocating Engines
6. Slowdown Systems (w/flare)
J. Vac nlst Illation Column Condensers
neg 2 6 19 40
neg 27 4 1 19
18
8. Claim Plant Tail Gas (uncontrolled) neg 115 neg neg neg
9. Storage
10. Loading
II. Fugitive
470
37
290
A-B
B-C
100.0OO HlnlMl potential for
Improving factors
100,000 HI Rival potential for
Improving factors
100,000 Nlnlnal potential for
tuproving factors
100,000 Good potential for
improving factors,
source is locally signi-
ficant hut not nationally
significant
100,000 Good potential for
Improving factors, Major
NO^ source, but use is
declining
100,000 Good potential fnr
taiprovlng factors,
significant emission
source
100,000 Rood potential for
proving factors,
significant emission
source
100.000 Ninlmal potential for
improving factors. majn
emission source
750,000 Good potential for
iatproving factors,
emission source
i|or
100,000 Marginal potential for
Isiprovlng factors, sig-
nificant omission source
1,400,000 Kxcellent potential for
improving factors, major
emission source
Suggested
priority
low
low
low
inter-
medlnre
Infor-
med Inte
hlgli
High
low
high
Inter-
mediate
high
-------
they are considered very accurate. Each of the emission factors
except for the carbon monoxide and hydrocarbon emission factors
were developed in the Los Angeles Joint Project conducted from
1955 to 1958 (AT-040). Subsequent fuel combustion studies and
compliance tests conducted through 1974 have verified the boiler
and heater emission factors reported by the Los Angeles Joint
Project. These studies also indicate that carbon monoxide emis-
sions are approximately 210 lb/1000 bbl of oil burned or 0.02 lb/
1000 cu. ft. of gas burned; and that hydrocarbon emissions are
approximately 42 lb/1000 bbl of oil burned or 0.003 lb/1000 ft3
of gas burned.
Merits of Emission Testing
There will be very little value in conducting addi-
tional emissions testing on refinery boilers and process heaters
because the existing factors are accurate. Inaccuracies in
current boiler and heater emission factors are primarily due
to variations in operating conditions of the units and not to
insufficient source testing. Additional source testing is not
expected to change the emission factors significantly.
4.1.2 Catalytic Crackers
Emission Impact
Catalytic cracking is used in nearly every major
refinery in the United States. Of this catalytic cracking
capacity, over 93 percent is accomplished with fluid catalytic
crackers. On a national average 281 bbl of petroleum are pro-
cessed in fluid catalytic crackers per 1000 bbl of crude oil
entering the refinery. The national average moving-bed
catalytic cracker feed rate is 21 bbl per 1000 bbl of refinery
feed. Typical catalytic cracking emission rates are reported
in Table 4.1-1. These rates are based on the average catalytic
-28-
-------
cracking capacities and the controlled catalytic cracking
emission rates reported in EPA Document AP-42 (EN-071).
The catalytic cracking emission rates presented in
Table 4.1-1 indicate that controlled fluid catalytic cracking
units represent a major source of particulate, sulfur oxide,
and nitrogen oxide emissions within the typical refinery. These
emission rates will be almost doubled for those refineries with
catalytic cracking capacities of 500 bbl FCC feed per- 1000 bbl
refinery feed.
Moving bed catalytic cracking units do not normally
represent a very significant emission source within the typical
refinery because of their limited application and their low
emission rates on a per unit size basis.
Emission Factor Accuracy
The EPA emission factors for uncontrolled catalytic
cracking units were first developed in the Los Angeles Joint
Project which was conducted from 1955 to 1958. Later studies
by Ben G. Jones of Phillips Petroleum Company supported the
findings of the Joint Project. In 1973 the EPA reported the
results of NSPS studies on controlled and uncontrolled fluid
catalytic cracking emissions (EN-072). These studies centered
on carbon monoxide and particulate emissions. They verified
existing factors for uncontrolled sources and contributed to
the controlled emission factor values.
Because of their limited use, very little emission
data has been collected on moving bed catalytic cracking units.
-29
-------
Radian recommends assigning an emission factor rating
of B to catalytic cracking emission factors. Most test results
from a broad sampling of catalytic cracking units support the
EPA emission factors reported in Table 4.1-1. However, avail-
able test data also indicate that emissions from individual
catalytic cracking units can vary over a wide range. Therefore,
these factors are not very reliable when used to predict the
performance of individual units.
Merits of Emission Testing
The merits of conducting an emission test program on
catalytic cracking units are very minimal. As mentioned
previously, test results indicate that there is a wide range
in the actual level of emissions from catalytic cracking units.
This range of emission levels can be greater than + 5070.
Because the inaccuracies in catalytic cracking emission factors
are attributable to differences between units, additional
emission testing will not significantly improve the accuracy
of these emission factors.
4.1.3 Fluid Coking Units
Emission Impact
Fluid coking units are used in many refineries to
thermally crack heavy oils into naphthas and carbonaceous coke.
The contribution of coking emissions to the total emissions
from the typical refinery is negligible (Table 4.1-1). This
is due in part to a national fluid coking capacity of only
8 bbl/1000 bbl of refinery capacity. For refineries having
fluid coking capacities as high as 380 bbl/1000 bbl of refinery
capacity, the fluid coking unit becomes a major source of
-30-
-------
particulates and possibly of other criteria pollutants. Even
with emission controls, large fluid coking units are a major
source of emissions.
Emission Factor Accuracy
The emission factors for fluid coking units are con-
sidered to be of average quality and were given an emission
factor rating of C. The primary data source for the fluid
coking emission factors was a report by Mr. Ben G. Jones on
particulate control in Phillips Petroleum Company's Avon,
California refinery (JO-086). The study involved a limited
number of emission tests conducted on a single fluid coker in
1969 and 1970. Particulates and carbon monoxide were the only
pollutants tested. Because only a single fluid coker was
tested and there have been design changes in fluid cokers in the
past seven years, the potential for a large error exists in the
fluid coker emission factor now being used by EPA.
Merits of Emission Testing
There are definite merits in conducting an emission
testing program for fluid cokers. Currently, there are no
emission factors available for sulfur oxide, nitrogen oxide,
and hydrocarbon emissions from fluid cokers. The current parti-
culate and carbon monoxide emission factors can be improved
from a rating of C to a rating of B. It is not likely that a
single set of emission factors can be developed with an A rating
(excellent) because of the emission fluctuations that occur
between fluid cokers of various designs and operating conditions
-31-
-------
4.1.4 Reciprocating Engines
Emission Impact
Reciprocating engines fired with natural gas are used
in many refineries to run high pressure compressors. The
estimated reciprocating engine emissions for a typical refinery
are presented in Table 4.1-1. These emission estimates are based
on a national average reciprocating engine size of 13.3 MSCF
natural gas per thousand barrels of refinery feed (MS-001).
As Table 4.1-1 indicates, reciprocating engines
are a major source of both carbon monoxide and nitrogen oxide
emissions in the refinery. In many refineries where recipro-
cating engines play a major role in gas compression these
emission contributions may be doubled or tripled. The current
trend, however, is towards the decreased use of reciprocating
engines. Although natural gas has been a cheap, abundant
source of energy, lower reliabilities and increasing problems
with the cost and availability of natural gas have decreased
the use of reciprocating engines in recent years.
Emission Factor Accuracy
Reciprocating engines have been given an emission
factor rating of B (good). The emission factors used for
refinery engines prior to this study were developed in the Los
Angeles Joint Project conducted between 1955 and 1958 (AT-040).
However, reports on the reciprocating engine tests done for
the project indicate that they were inconclusive.
More intensive emission studies were conducted by
Southwest Research Institute in 1974 and 1975 on nitrogen oxide,
-32-
-------
hydrocarbon, and carbon monoxide emission rates (DI-142). These
studies found carbon monoxide and nitrogen oxide emissions to
be much higher than previously reported by the Joint Project.
Hydrocarbon emissions were approximately the same. The South-
west Research Institute emission factors appear to be very good
and were incorporated into the AP-42 revisions.
Merits of Emission Testing
Additional source testing of reciprocating engine
emissions has very few merits. The major pollutant species
in reciprocating engine emissions were included in the
recent SwRI studies. These factors are good, and the possi-
bility of increasing their accuracy through additional testing is
not definite. Emission rate variations among various engines
may be too great to allow improvements in emission factor
accuracy.
4.1.5 Slowdown Systems
Emission Impact
Slowdown systems are used in every refinery to safely
remove hydrocarbon liquids and vapors from downed or malfunc-
tioning equipment. A typical refinery vents approximately 580
pounds of hydrocarbons into the blowdown system per thousand
barrels of refinery feed. Many of the hydrocarbons vented into
the blowdown are recovered and recycled. The remainder are
commonly flared. Table 4.1-1 presents the emissions resulting
from flaring blowdown system hydrocarbons in the typical refin-
ery. Blowdown emissions, when flared, may be a major source
of sulfur oxide, carbon monoxide, and nitrogen oxide emissions.
If uncontrolled, the blowdown system is a major source of hydro-
carbon emissions in the refinery.
-33-
-------
Emission Factor Accuracy
The emission factors for blowdown systems were devel-
oped by Lockheed Missile and Space Company for the Environmental
Protection Agency with the Assistance of the American Petroleum
Institute (KL-081). Eighteen refineries were included in the
study. The quantities of hydrocarbons reported as being flared
ranged from 150 to 1800 pounds per thousand barrels of refinery
throughput and averaged approximately 580 pounds per thousand
barrels of refinery throughput. Lockheed calculated S02 emis-
sions produced by flaring the blowdown gases directly from blow-
down gas sulfur levels reported by the refineries. Carbon
monoxide and hydrocarbon emissions were calculated by Lockheed
from the data of Sussman, et. al., a limited flare study con-
ducted in 1958. The nitrogen oxide emissions were calculated
by Radian using information from a nitrogen oxide emission
study by Chass and George in 1958 on general combustion sources
(CH-055). Because of the wide range in hydrocarbon feed rates
to refinery flares, and because of the age of the combustion
data on flare emissions, the emission factors for controlled
refinery blowdown systems have been assigned an emission factor
rating of C (average accuracy).
Merits of Emissions Testing
The emission factors for controlled refinery blowdown
systems can be improved significantly by conducting an emission
testing program. However, the emission testing program will
not be able to improve significantly the accuracy of the esti-
mated rate of hydrocarbon venting into the blowdown system.
Actual blowdown rates for the refining industry vary over a very
wide range of values. The emission testing program will be most
effective in defining the emission rates of combustion products
-34-
-------
from known flaring blowdown rates. There is less variation
among combustion product emission rates for a given blowdown
rate than there is among blowdown rates. It should be noted
that conducting flare emission tests may prove to be a very
difficult task.
4.1.6 Vacuum Distillation Column Condensers
Emission Impact
Until recently, noncondensable vapors from the vacuum
units on vacuum distillation columns have been vented to the
atmosphere. Many refineries have now eliminated these emissions
completely by venting into either the fuel gas system or the blow-
down system. The emission rate of these vapors has been estimated
to average 50 Ibs per 1000 bbl of reduced crude feed to the vacuum
distillation column (AM-055). For the typical refinery this is
an equivalent emission rate of 18 Ibs per 1000 bbl of crude feed
to the refinery. Table 4.1-1 compares the contribution of various
emission sources to total refinery emissions, and indicates that
vacuum distillation emissions are of secondary importance.
Emission Factor Accuracy
The emission factor for hydrocarbon emission rates
from vacuum distillation column condensers has been assigned
an emission factor rating of C. This emission factor was
developed in the Los Angeles Joint Project conducted between
1955 and 1958 (AT-040, AM-055). Most of the vacuum distillation
column emission rates fell in a range from 15 to 50 Ibs per
1000 bbl of reduced crude feed to the vacuum distillation column.
Emission rates as high as 130 Ib per 1000 bbl were measured and
the average emission rate v/as close to the 50 Ib per 1000 bbl
of vacuum column feed. Since these emission rates have not been
-35-
-------
verified in recent years, they may not account for the higher
separation efficiency of today's distillation columns and the
new types of crude oils being refined today.
Merits of Emission Testing
An emission testing program for vacuum distillation
column condensers will likely improve the emission factor accu-
racy to an emission factor rating of B. However, greater
accuracy improvements will be limited by the wide range of
emission rates exhibited by vacuum distillation column conden-
sers .
4.1.7 Glaus Plant Tail Gas
Emission Impact
Although not discussed in the refinery section of
AP-42, the tail gases vented from a Glaus sulfur recovery plant
are a common refinery emission source. Assuming the imported
crude oil refined in the United States (4070) has a sulfur con-
tent of 1.5 wt7o, and that 80% of this sulfur is routed to a
three stage Glaus unit with a 9670 efficiency, then the average
sulfur dioxide emission rate from the typical U.S. refinery is
approximately 115 Ibs of SOa per 1000 bbl of refinery feed.
These values indicate that uncontrolled Glaus plant tail gas
emissions are a very significant source of sulfur dioxide
emissions in a refinery processing sour crude oils.
Emission Factor Accuracy
In Section 5.18 of AP-42, the emission factors for
Glaus plant tail gases have been given an emission factor rating
of D. The reasons contributing to this low rating are unknown.
-36-
-------
However, a review of emission test data on Glaus plants indi-
cates that the accuracy of these emission factors is good. A
large number of literature sources are available on sulfur oxide
emissions from Glaus plants. When applied to refineries, the
emission factor for Glaus plant tail gas probably has an emis-
sion factor rating of B.
Merits of Emission Testing
There is very little accuracy improvement to be
gained from conducting an emission source testing program for
Glaus plant emissions. The Glaus process is a well established
process for which there is extensive emission data available
from literature sources.
4.1.8 Storage and Loading Operations
Emission Impact
All petroleum liquids entering and leaving the petro-
leum refinery pass through the tank farm area of the refinery.
Here they are handled in storage and loading operations. As
indicated in Table 4.1-1, storage and loading operations in the
typical refinery are estimated to emit 500 pounds of hydrocar-
bons per thousand barrels of refinery throughput (BU-185).
These figures indicate that storage and loading operations are
the largest source of hydrocarbon emissions in the refinery.
Emission Factor Accuracy
The primary sources of hydrocarbon emission factors
for storage and loading operations are the factors and corre-
lations developed by the Evaporative Loss Committee of API
-37-
-------
between 1959 and 1962. These correlations are based on testing
results assembled by API. The reported accuracy of these corre-
lations at the time of their development was approximately ±257,
for storage operations and ±357. for loading operations. Emission
factors for gasoline and crude oil loading into tank trucks and
marine vessels have been updated recently as a result of current
studies by the petroleum industry and EPA. These factors are
expected to have even better accuracy.
Merits of Emission Testing
There have been many significant developments made in
recent years on storage tank design, especially in the area of
seals for floating-roof tanks. It is likely that the API
emission factor correlations do not adequately predict the
hydrocarbon emissions from newer storage tanks. Recent test
results also indicate that API storage and loading emission
correlations do not adequately deal with petroleum liquids
having low vapor pressures.
Because loading emission rates have just been updated,
there will be very little benefit derived from further testing
these emission sources. Storage emission factors will be the
most improved by an emission testing program. Testing of
storage emission sources will probably be able to improve sto-
rage emission factors from an emission factor rating of B to a
rating of A.
4.1.9 Fugitive Emission Sources
Emission Impact
The second largest source of hydrocarbon emissions
in the petroleum refinery is fugitive emission sources.
-38-
-------
Fugitive leaks from miscellaneous valves and fittings in the
typical refinery are estimated to be 290 Ibs per 1000 bbl of
refinery throughput (AT-040).
Emission Factor Accuracy
The accuracy of fugitive emission factors is considered
poor. Fugitive emission factors have been given an emission
factor rating of D. Most of the emission testing which formed
the basis for the fugitive emission factors was conducted in the
Los Angeles Joint Project between 1955 and 1958. Fugitive emis-
sion rates are very dependent on equipment type, equipment age,
housekeeping practices, and frequency of maintenance. Because
these parameters have all changed since the Joint Project was
conducted, the fugitive emission factors are suspected to be
very inaccurate. The results of a testing program conducted by
Meteorology Research Inc. for Atlantic Richfield Company also
indicate that fugitive emission factors for well maintained
facilities are substantially lower than current AP-42 factors
(WO-099). The MRI study tested a limited number of valves and
pump seals in a well maintained production facility. MRI found
that valve leaks average 0.00014 Ib/day-valve and pump seal leaks
average 0.00085 Ib/day-pump. Currently, used factors are 0.15
Ib/day-valve and 5 Ib/day-seal respectively. Although not con-
clusive, the MRI tests indicate that current fugitive factors
may be inaccurate for well maintained facilities.
Merits of Emission Testing
A well designed test program will be capable of improv-
ing the accuracy of current fugitive emission factors. However,
the improved fugitive emission factors may still have an error
of ±50% to ±757o. Fugitive emission sources are the most diffi-
cult for which to develop accurate factors or correlations.
-39-
-------
There are numerous fugitive sources of various equipment types
in the typical refinery. As a result, the fugitive emission
rates will vary widely across the refinery, but the average
fugitive emission rates for each refinery should vary less
between refineries.
4. 2 Sampling Program Costs
The costs associated with sampling emission sources
within refineries vary greatly, depending on such factors as
accuracy of results, size of sources, difficulty of sampling
task, and number of pollutants associated with the source. In
this section the problems and estimated costs associated with
sampling some of the major refinery emission sources are dis-
cussed.
4.2.1 Sampling Strategy
An effective sampling program begins with the estab-
lishment of a well designed sampling strategy. The initial step
in establishing a sampling strategy is to identify the "choice
parameters" associated with the source to be sampled. "Choice
parameters" are those variables which directly impact the rate
of emissions from the source. Typical choice parameters might
include fuel type, unit type, unit size and age, and operating
temperatures and pressures. These choice parameters form the
source categories which must be included when selecting the
sources to be tested. The proper selection of choice para-
meters is very important to the collection of statistically
accurate results.
After the identification of choice parameters,
"correlating parameters" must be identified for the emission
source to be tested. "Correlating parameters" are additional
-40-
-------
variables which have a secondary impact on the emission rate.
They will be measured and included in the correlations if neces-
sary. Typical correlating parameters might include refinery
location, equipment location or configuration, and possibly some
of the choice parameters listed above.
Once the choice and correlating parameters are iden-
tified for the emission source, they are used in a statistical
experimental design procedure for selecting the specific emis-
sion tests to be conducted. The statistical experimental design
procedure applies a systematic orderly procedure for selecting
the optimum combination of parameters to be tested. This pro-
cedure minimizes the number of tests required to obtain statis-
tically accurate results.
4.2.2 Point Source Emissions
Point source emissions are emitted from a limited
number of easily identifiable stacks or vents. They include
such sources as boiler flue gas stacks and Glaus plant tail gas
vents. Point sources are the easiest category of refinery
emission sources to test because of their limited number and
easily identifiable point of emissions. A wide range of pollu-
tants are emitted from refinery point sources; including parti-
culates, sulfur oxides, mercaptans, hydrogen sulfide, nitrogen
oxides, carbon monoxide, hydrocarbons (methane and non-methane),
aldehydes, and ammonia.
In a program recently conducted for the EPA, Radian
Corporation developed the costs for sampling five point sources
in a study of sixteen refineries. A statistical analysis of the
sources indicated that including sixteen refineries would opti-
mize the cost-effectiveness of collecting accurate emission data,
-41-
-------
The cost for the five point source test program was $500,000
or approximately $100,000 per point source (for all 16 refin-
eries). This cost includes setting up the program, sampling,
analyzing for seven pollutants, and presenting the results in
a final report. This is a typical point source sampling program
cost, and many sources will be more expensive or less expensive
to test. The cost per point source also drops as the size of
the test program increases due to the consolidation of some costs
4.2.3 Fugitive Source Emissions
Fugitive refinery emission sources are small miscel-
laneous hydrocarbon leaks scattered throughout the refinery.
These emissions are very difficult to identify because they are
not normally emitted from a well defined stack or vent. Some
examples of fugitive refinery emission sources include waste
water drains, cooling towers, valves, and pipe fittings.
In a study for EPA Radian developed the cost for a
sampling program to identify the emissions from eleven fugitive
emission sources (EPA Contract 68-02-2147). These eleven
sources include:
in-line valves
open-end valves
flanges
pressure-relief devices
pump seals
unit drains
compressor seals
cooling towers
wastewater facilities
ditches
sumps
-42-
-------
Approximately six hundred samples would be collected in sixteen
refineries in order to quantify the emissions from these fugi-
tive emission sources. The cost of the fugitive sampling pro-
gram was estimated to be $1,400,000. Species characterization
of th^ fugitive hydrocarbon emissions was not included in this
cost estimate. The estimated time period required for conduct-
ing the fugitive sampling program would be 1.25 years.
4.2.4 Tankage Emissions
Storage tank testing costs were not included in either
the point source cost estimates or the fugitive source cost esti-
mates. Storage tank testing is very difficult and poses some
very unique problems. Refinery storage tanks are generally too
large to be tested by enclosing and measuring the emissions
directly. They must be tested either by testing scaled down
models and scaling up the test results, or by conducting sensi-
tive product sampling and calculating weathering rates based
on changes in product composition.
Based on the above tankage testing procedures the cost
of a tankage emission testing program is estimated to be approx-
imately $750,000. This cost estimate includes test plan develop-
ment, sampling, analysis, and reporting of final results.
4.3 Prioritization of Emission Sources
The prioritization of emission sources for the pur-
pose of developing a sampling plan is a very difficult task.
Things to be considered in prioritizing emission sources include
national impact, regional impact, budget, availability of samp-
ling teams, potential for collecting meaningful results, and
growth trends. This section discusses the general prioritization
-43-
-------
of refinery emission sources. The two things considered are the
relative impact of the source's emissions and the potential for
improving current criteria pollutant emission factors. The
refinery emission source prioritization presented here does not
consider any pollutants other than the five EPA criteria pollu-
tants; particulates, sulfur oxides, carbon monoxide, hydrocarbons,
and nitrogen oxides. A distinctly different prioritization
order would be generated if toxic or carcinogenic pollutants
were to be considered.
4.3.1 High Priority
The highest priority for refinery emission sources
is given to emission sources for which an emission testing
program will definitely improve the accuracy of currently used
emission factors. Sources included in the highest priority
category are blowdown systems, vacuum distillation column con-
densers, storage tanks, and fugitive emissions. Fugitive
emission sources are given high priority because they represent
the second largest source of refinery emissions and the source
with the greatest potential for emission factor improvement.
Although storage tanks have a good emission factor rating, they
have also been included in the highest priority category.
Storage tanks represent the largest source of hydrocarbon
emissions in the refinery and small improvements in their
emission factor accuracy have significant impact on the accuracy
of the total refinery emissions. Although not as large an emis-
sion source, blowdown emissions and vacuum distillation column
condensers are also considered high priority for testing pro-
grams. These two sources have a large potential for accuracy
improvement at a relatively lower cost than most of the other
sources.
-44-
-------
4.3.2 Intermediate Priority
The intermediate priority sources are emission sources
for which testing programs will be less cost effective than the
testing programs for high priority sources. This lower cost
effectiveness may be attributable to either lower potential for
improvements in accuracy or to higher testing costs. The three
sources which are rated intermediate priority are fluid coking
units, internal combustion engines, and loading operations.
Fluid coking units are considered intermediate priority because
of their low impact on a national level. Although they are high
particulate and carbon monoxide emitters, fluid cokers are not
used very extensively today. But, they are gaining in popularity
and may become significant emission sources in the future.
Internal combustion engines are also considered intermediate
priority sources. Although internal combustion engines are
significant emission sources with a large potential for accuracy
improvement, the trend is towards their declining usage. They
will be contributing less to refinery emissions in the future.
Loading emission sources have been given intermediate priority
because of the questionable potential for improving their emis-
sion factors. Many loading emission sources have been tested
recently and additional testing programs may have only a small
effect on current emission factor accuracy.
4.3.3 Low Priority
Low priority ratings have been given to boilers,
heaters, fluid bed catalytic crackers, moving bed catalytic
crackers, and Glaus plant tail gases. Regardless of the emis-
sion rates from each of these sources, they were given the low
priority rating because of the minimal potential which exists
for improving their emission factor by emission testing programs.
For some factors, only small improvements can be achieved because
-45-
-------
the emission rates are already well characterized. For other
inaccurate factors, small potential exists for improvement
because they can't be better characterized using simple emission
factors. However, emission testing should not be ruled out for
these sources. Although accurate factors cannot be developed
for some sources, testing often leads to the development of
accurate complex correlations which incorporate the important
parameters.
-46-
-------
REFERENCES
AM-055 American Petroleum Inst., Committee on Refinery
Environmental Control, Hydrocarbon Emissions from
Refineries, API Publication No. 928, Washington, D.C.,
1973.
AM-084 American Petroleum Inst., Evaporation Loss Committee,
Evaporation Loss in the Petroleum Industry, Causes
and Control. API Bull. 2513. Washington. D.C.. 1959.
AT-040 Atmospheric Emissions from Petroleum Refineries.
A Guide for Measurement and Control, PHS No.763,
Washington, D.C., Public Health Service, 1960.
BA-291 Bartz, D. R., et al., Control of Oxides of Nitrogen
from Stationary Sources in the South Coast Air Basin
(of California), final report, PG 237 688. Tustin. CA,
KVB Engineering, Inc., 1974.
BR-199 Brown, R. A., H. B. Mason, and R. J. Schreiber,
Systems Analysis Requirements for Nitrogen Oxide
Control of Stationary Sources, Final Report, EPA
650/2-74-091, PB 237 367, Contract No. 68-02-1318,
Task 3, Mountain View, CA, Aerotherm/Acurex Corp., 1974,
BU-185 Burklin, C. E., et al., Control of Hydrocarbon Emis-
sions fromPetroleum Liquids, Contract No.68-02-1319,
Task 12, EPA 600/2-75-042, PB 246 650, Austin, Texas,
Radian Corporation, Sept. 1975.
CA-339 Cantrell, Ailleen, "Annual Refining Survey", Oil Gas
J. 1976 (March 29), 124.
CH-055 Chass, Robert L. and Ralph E. George, "Contaminant
Emissions from the Combustion of Fuels", J. APCA 10,
34-43 (1960).
CR-040 Crumlish, Wm. S., "Review of Thermal Pollution Prob-
lems, Standards and Controls at the State Government
Level", Presented at the Cooling Tower Inst. Symp.,
New Orleans, Jan. 30, 1966.
DA-069 Danielson, John A., comp. and ed., Air Pollution
Engineering Manual, 2nd ed., AP-40, Research Triangle
Pk., N.C., EPA, Office of Air & Water Programs, 1973.
-47-
-------
REFERENCES (Continued)
DI-142 Dietzmann, Harry E. and Karl J. Springer, Exhaust
Emissions from Piston and Gas Turbine Engines Used
in Natural Gas Transmission, San Antonio, Southwest
Research Institute, Jan.1974.
EN-071 Environmental Protection Agency, Compilation of Air
Pollutant Emission Factors, 2nd ed. with supplements,
AP-42, Research Triangle Park, N.C., 1973.
EN-072 Environmental Protection Agency, Office of Air and
Water Programs, Office of Air Quality Planning and
Standards, Background Information for Proposed New
Source Standards: Asphalt Concrete Plants. Petroleum
Refineries, Storage Vessels, Secondary Lead Smelters
and Refineries, JSrass or Bronze Ingot Production
Plants. IrpVand Steel Plants, Sewage Treatment "Plants,
Vols. 1 and 2, Research Triangle Park, N.C. , 1973.
EN-090 Environmental Protection Agency, Office of Adminis-
tration, Guide for Compiling a. Comprehensive Emission
Inventory, APTD-1135, Research Triangle Park, N.C.,
1972.
GR-360 Griscom, R. W. , Regional Air Pollution Study (RAPS)
100% Completion Report Point and Area Source Organic
Emission_Inventory, EPA Contract No. 68-02-20937 Task
No. 108-1, Creve Coeur, MO, Rockwell International,
Atomics International Division, Air Monitoring Center,
Oct. 1977.
JO-086 Jones, Ben G., "Refinery Improves Particulate Control",
Oil Gas J. 69 (26), 60-62 (1971).
KL-081 Klett, M. G. and J. B. Baleski, Flare Systems Study,
EPA 600/2-76-079, EPA Contract No. 68-02-1331, Task 3,
Huntsville, Ala., Lockheed Missiles and Space Co.,
Inc., March 1976.
MA-216 Marchant, William H., "Heater Air & Noise Pollution
Controlled in Refineries", Oil Gas J. 1973 (Jan. 1),2.
MS-001 MSA Research Corp., Hydrocarbon Pollutant Systems
Study Vol. 1, Stationary Sources, Effects and Control,
PB-219-073, APTD 1499, Evans City, PA, 1972.
-48-
-------
REFERENCES (Continued)
NA-006 National Air Pollution Control Techniques Advisory
Committee, Control Techniques for Sulfur Oxide Air
Pollutants, NAPCA Pub. No. AP-52, January 1969.
NA-308 "National Petroleum Refiners Assn. (NPRA), '74 Panel
Views Processes", Hydrocarbon Proc. 54(3), 127-38 (1975),
OL-044 Olson, H. N. and K. E. Hutchison, "How Feasible are
Giant, One-Train Refineries?", Oil Gas J. 70(1), 39
(1972).
PE-140 Petrolite Corp., Impurities in Petroleum, Long Beach,
CA, Petreco.
TR-107 Trijonis, John C. and Kenneth W. Arledge, Utility of
Reactivity Criteria In Organic Emission Control
Strategies, Application to the Los Angeles Atmosphere.
EPA 600/3-76-091, EPA Contract No. 68-02-1735,
Redondo Beach, CA, TRW Environmental Services, Aug.
1976.
UR-022 Urban, Charles M. and Karl J. Springer, Study of Exhaust
Emissions from Natural Gas Pipeline Compressor Engines,
Project PR-15-61, San Antonio, TX, Southwest Research
Institute, February 1975.
WO-099 Woffinden, George J., Total Hydrocarbon Emission
Measurements of Valves and Compressors at ARCO's
Ellwood Facility, Report No. 911 115 1661, Altadena,
California, January 1976.
-49-
-------
APPENDIX A
REVISED AP-42
9.1 Petroleum Refining
9.1.1 General Description
9.1.2 Process Emission Sources and Control Technology
9.1.3 Fugitive Emission Sources and Control Technology
A-l
-------
9.1 PETEOLEUM REFINING1
9.1.1 GENERAL DESCRIPTION
The petroleum refining industry is primarily involved
in the conversion of crude oil into more than 2500 refined pro-
ducts including liquefied petroleum gas, gasoline, kerosene,
aviation fuel, diesel fuel, fuel oils, lubricating oils, and
feedstocks for the petrochemical industry. Petroleum refinery
activities start with crude storage at the refinery, include
all petroleum handling and refining operations, and terminate
with storage of the refined products at the refinery.
The petroleum refining industry employs a wide variety
of processes for the conversion of crude oil to finished petro-
leum products. A refinery's processing flow scheme selection is
largely determined by the composition of the crude oil feedstock
and the chosen slate of petroleum products. The example refinery
flow scheme presented in Figure 9.1-1 shows the general processing
arrangement used by U.S. refineries for major refinery processes.
The arrangement of these processes will vary among refineries and
few, if any, refineries employ all of these processes. Petroleum
refining processes having direct emission sources are presented
in bold line boxes.
In general, refinery processes and operations can be
divided into five categories:
1) Separation processes
a. atmospheric distillation
b. vacuum distillation
c. light ends recovery (gas processing)
A-2
-------
NOTE: gg DENOTES
PROCESSING UNITS
E*d.O"INO PROCESS
HEATERS.
REFINING PROCESSES
HAVING DIRECT
EMISSION SOURCES
ARE IN BOLD
LINE BOXES.
r~MYOHJOEN~l
I PRODUCTION
PIANT
HVODOO* N
TO ItAVY HYDROCARBON
STOHAUt t BLENDING
TO MIDDLE DISTILLATE
STORAGE t BLENDING
TO OA3OLME I PETROCHEMICAL
STORAGE t BLENDING
FIGURE 1-1 SCHEMATIC OF AN EXAMPLE INTEGRATED PETROLEUM REFINERY
-------
2) Petroleum conversion processes
a. cracking (thermal and catalytic)
b. reforming
c. alkylation
d. polymerization
e. isomerization
f. coking
g. visbreaking
3) Petroleum treating processes
a. hydrodesulfurization
b. hydrotreating
c. chemical sweetening
d. acid gas removal
e. deasphalting
4) Feedstock and product handling
a. storage
b. blending
c. loading
d. unloading
5) Auxiliary facilities
a. boilers
b. wastewater treatment
c. hydrogen production
d. sulfur recovery plant
e. cooling towers
f. blowdown system
g. compressor engines.
These refinery processes are defined in the following section
along with their emission characteristics and applicable emission
control technology.
A-4
-------
Petroleum Separation Processes
The first phase in petroleum refining operations is
the separation of crude oil into its major constituents using
three petroleum separation processes: atmospheric distillation,
vacuum distillation, and light ends recovery. Crude oil consists
of a mixture of hydrocarbon compounds including paraffinic,
naphthenic, and aromatic hydrocarbons plus small amounts of
impurities including sulfur, nitrogen, oxygen, and metals.
Refinery separation processes use such techniques as distillation,
stripping, and absorption to separate these crude oil constituents
into common boiling point fractions.
Petroleum Conversion Processes
Product demand and economic considerations require that
less valuable components of crude oil be converted to more valuable
products by using the petroleum conversion processes. To meet
the demands for high octane gasoline, jet fuel, and diesel fuel,
lower value residual oils, fuel oils, and light ends are often
converted to gasolines and other light fractions. The cracking,
coking, and visbreaking processes are used to break large petro-
leum molecules into smaller petroleum molecules. On the other hand,
polymerization and alkylation processes are used to combine
small petroleum molecules into larger ones. Isomerization and
reforming processes primarily rearrange the structure of petroleum
molecules to produce higher value molecules of a similar molecu-
lar size.
Petroleum Treating Processes
Petroleum treating processes stabilize and upgrade
petroleum products by separating them from less desirable
A-5
-------
petroleum products and by removing objectionable elements from
petroleum products and feedstocks. Treating processes used to
stabilize products and remove undesirable elements such as sulfur,
nitrogen and oxygen include hydrodesulfurization, hydrotreating,
chemical sweetening and acid gas removal. Treating processes
employed primarily for the separation of petroleum products
include such processes as deasphalting. Desalting is used to
remove salt, minerals, grit, and water from crude oil feed
stocks prior to refining. And asphalt blowing is used to poly-
merize and stabilize asphalt, improving its weathering character-
istics.
Feedstock and Product Handling
The refinery feedstock and product handling operations
consist of storage, blending, loading, and unloading activities.
All feedstocks entering the refinery and all products leaving the
refinery are subject to the refinery handling operations.
Auxiliary Facilities
Auxiliary facilities include a wide assortment of
processes and equipment which are not directly involved in the
refining of crude oil, but which perform functions vital to the
operation of the refinery. These include boilers, wastewater
treatment, hydrogen plants, cooling towers, sulfur recovery units,
etc. Products from auxiliary facilities (clean water, steam,
process heat, etc.) are required by the majority of refinery
process units and are not limited to any one part of the refinery.
A-6
-------
9.1.2 PROCESS EMISSION SOURCES AND CONTROL TECHNOLOGY
This section presents descriptions of those refining
processes which are significant air pollutant contributors.
Process flow schemes, emission characteristics, and emission
control technology are discussed for each process. Table 9.1-1
lists the emission factors for direct process emissions in the
petroleum refinery. The following process emission sources
are discussed in this section on petroleum refining emissions :
vacuum distillation,
catalytic cracking,
thermal cracking processes,
utility boilers,
heaters,
compressor engines,
blowdown systems, and
sulfur recovery.
Vacuum Distillation
Topped crude withdrawn from the bottom of the atmos-
pheric distillation column is composed of high boiling point
hydrocarbons which decompose and polymerize to foul equipment
when distilled at atmospheric pressures. In order to further
separate topped crude into components, it must be distilled in
A-7
-------
TABLE 9.1-1. EMISSION FACTORS FOR PETROLEUM REFINERIES
i
00
Type of Process
_
Boiler* and Process He.itcrs
lb/10'bbl oil burned
kg/101 liters oil burned
lb/10'ft' gas burned
kg/io'm1 gas burned
Fluid Catalytic Cracking
U.iits8
Uncontrolled
lb/10'bbl fresh feed
kg/10' liters fresh feud
Electrostatic Preclpitator
and CO boiler
lb/10'bbl fresh feed
kg/101 liters fresh feed
Movln^-bed C.ltalytic
Cracking Unitsb
lb/10'bbl fresh feed
kg/101 liters fresh feed
Fluid Coklnfi Units'1
Uncontrolled
lb/10'bbl fresh feed
Electrostatic Prectpicator
and CO boiler
lb/101 bbl fresh feed
k^/10J liters fresh ieed
Delayed Coking Units
Particulates
840
2.4
0.02
0.32
242
(93 to 340)h
0.695
(0.267 to 0.976)
451
(7 to 150)
0.128
(0.020 to 0.428)
17
0.049
523
1 en
1 . J\J
6.85
0.0196
HA
Sulfur Total* Nitrogen
Oxides Carbon Hydro- Oxides
(SO 2) Monoxide Carbons (NOz) Aldehyde* Ammonia
6.720SC 210d 42d 2.900 25 Neg*
19. 2S 0.6d 0.12d 8.3 0.071 Neg
2.sf 0.02d 0.003d 0.23 0.003 Nee
32s 0.32d 0.048d 3.7 0.048 Neg
493 13,700 220 71.0 IS 54
(100 to 525) (37.1 to 145.0)
1.413 39.2 0.630 0.204 0.054 0.155
(0.286 to 1.505) (0.107 to 0.416)
493 Keg Neg 71. 01 Neg Neg
(100 to 525) (37.1 to 145.0)
1.413 Neg Neg 0.204J Neg Neg
(0.286 to 1.505) (0.107 to 0.416)
60 3,800 87 5 12 6
0.171 10.8 0.250 0.014 0.034 0.017
NA NA HA NA NA NA
MA NA MA hi A MA UA
HA nn HA tin. ptJ\ N,A
NA Neg Neg NA Neg Neg
NA Neg Neg NA Neg Neg
NA NA NA NA KA NA
Emission
Factor
Rating
A
A
A
A
8
B
B
B
B
B
C
C
C
C
NA
(Continued)
-------
TABLE 9.1-1. EMISSION FACTORS FOR PETROLEUM REFINERIES (Continued)
Type of Process 1
Compressor Engines
Reciprocating Engines
lb/101 ft1 gas burned
kg/lO'm1 gas burned
Cas Turbines
lb/101 ft1 gas burned
i^./10'm1 gas burned
Slowdown Systems
Uncontrolled
lb/101 bbl refinery
feed
kp./lO1 liters refinery
fcsd
and flaring
Ib/lO^bl refinery
feed
kg/101 liters refinery
feed
Vacuum Distillation0
Column Condensers
Uncontrolled
lb/101 bbl refinery
feed
kg/101 liters refinery
feed
lb/10'bbl vacuum feed
kg/ 101 liters vacuum feed
Controlled
Part iculates
Neg
Neg
Neg
Neg
Neg
Neg
Neg
Neg
Neg
Neg
Neg
Neg
Neg
Sulfur Total* Nitrogen
Oxides Carbon Hydro- Oxides
(802) Monoxide Carbons (NO;) Aldehydes Ammonia
2s 0.43 1.4 3.4 0.1 0.2
32s 7.02 21.8 55.4 1.61 3.2
2s 0.12 0.02 0.3 NA NA
32s 1.94 0.28 4.7 NA NA
Neg Neg 580 Neg Neg Neg
Neg Neg 1.662 Keg Neg Neg
26.9 4.3 0.8 18.9 N*g Neg
0.077 0.012 0.002 0.054 Neg Neg
Neg Neg 18 Neg Meg Neg
Neg Neg 0.052 Neg Neg Neg
Neg Neg 50 (0-130) Neg Neg Neg
Neg Neg 0.144 Neg Neg Neg
Neg Neg Neg Neg Neg Neg
Emission
Factor
Rating
B
B
B
B
C
C
C
C
C
C
C
C
C
(Continued)
-------
TABLE 9.1-1. EMISSION FACTORS FOR PETROLEUM REFINERIES (Continued)
Sulfur Toml* Nitrogen KmlxKlun
Oxide* Carbon Hydro- Oxides Factor
Type of Process Part Iculales (SO;) Monoxide Carbons (HOZ) Aldehydes Ammonia Rat IIIK
Claim I'lani and_Tall (ins See Section 5.18
Treatment
Overall, leas than 1 percent by weight of the total hydrocarbon emissions are methane.
'' Reference 2
S * Fuel oil sulfur content (weight percent): factors based on 100 percent combustion of sulfur
to SO, and assumed density of 336 Ib/bbl (0.96 kg/liter).
Carbon monoxide and hydrocarbon factors taken from Tables 1.3-1 and 1.4-1 of this manual.
'' Nc(.l Iglhle emission
s - Refinery gas sulfur content (Hi/1000 ft1): factors based on 100 percent combustion of sulfur to SO;.
" KcliTtMiccx 2 through 8
Numbers In parenthesis Indicate range of values observed.
Under the New Source Performance Standards, controlled FCC regenerators will have paniculate
emissions lower than 19 lb/101 bbl fresh feed.
May be higher due to the combustion of ammonia.
Reference 5
1 NA, Not Available
'" Reference 9 and 10
" Reference 2, 11
" Reference 2, 12, 13
-------
a vacuum distillation column at a very low pressure and in a
steam atmosphere.
In the vacuum distillation unit topped crude is heated
with a process heater to temperatures ranging from 700°F to 800°F,
The heated topped crude is flashed into a multi-tray vacuum dis-
tillation column operating at vacuums ranging from 0.5 psia to
2 psia. In the vacuum column the topped crude is separated into
common boiling point fractions by vaporization and condensation.
Stripping steam is normally injected into the bottom of the
vacuum distillation column to assist in the separation by lower-
ing the effective partial pressures of the components. Standard
petroleum fractions withdrawn from the vacuum distillation
column include lube distillates, vacuum oil, asphalt stocks,
and residual oils. The vacuum in the vacuum distillation column
is normally maintained by the use of steam ejectors but may be
maintained by the use of vacuum pumps.
The major sources of atmospheric emissions from the
vacuum distillation column are associated with the steam ejectors
or vacuum pumps. A major portion of the vapors withdrawn from
the column by the ejectors or pumps are recovered in condensers.
Historically, the noncondensable portion of the vapors has been
vented to the atmosphere from the condensers. There are approxi-
mately 50 pounds of noncondensable hydrocarbons per thousand
barrels of topped crude processed in the vacuum distillation
column.2'12'13 A second source of atmospheric emissions from
vacuum distillation columns is combustion products from the
process heater. Process heater requirements for the vacuum
distillation column are approximately 37,000 Btu per barrel of
topped crude processed in the vacuum column. Process heater
emissions and their control are discussed later in this section.
Fugitive hydrocarbon emissions from leaking seals and fittings
A-ll
-------
are also associated with the vacuum distillation unit, but these
are minimized by the low operating pressures and low vapor pres-
sures in the unit. Fugitive emission sources are also discussed
later in this section.
Control technology applicable to the non-condensable
emissions vented from the vacuum ejectors or pumps include venting
into blowdown systems or fuel gas systems, and incineration in
furnaces or waste heat boilers.2'lz'l3 These control technolo-
gies are generally greater than 99 percent efficient in the control
of hydrocarbon emissions, but they also contribute to the
emission of combustion products.
Catalytic Cracking
Catalytic cracking, using heat, pressure, and cata-
lysts, converts heavy oils into lighter products with product
distributions favoring the more valuable gasoline and distil-
late blending components. Feedstocks are usually gas oils from
atmospheric distillation, vacuum distillation, coking, and de-
asphalting processes. These feedstocks typically have a boiling
range of 650-1000°F. All of the catalytic cracking processes
in use today can be classified as either fluidized-bed or moving
bed units.
Fluidized-bed Catalytic Cracking (FCC) - The FCC
process uses a catalyst in the form of very fine particles which
act as a fluid when aerated with a vapor. Fresh feed is pre-
heated in a process heater and introducted into the bottom of a
vertical transfer line or riser with hot regenerated catalyst.
The hot catalyst vaporizes the feed bringing both to the desired
reaction temperature (880-980°F). The high activity of modern
catalysts causes most of the cracking reactions to take place in
the riser as the catalyst and oil mixture flows upward into the
A-12
-------
reactor. The hydrocarbon vapors are separated from the catalyst
particles by cyclones in the reactor. The reaction products
are sent to a fractionator for separation.
The spent catalyst falls to the bottom of the reactor
and is steam stripped as it exits the reactor bottom to remove
adsorbed hydrocarbons. The spent catalyst is then conveyed to
a regenerator. In the regenerator, coke deposited on the catalyst
as a result of the cracking reactions is burned off in a control-
led combustion process with preheated air. Regenerator tempera-
ture is usually 1100 - 1250°F. The catalyst is then recycled
to be mixed with fresh hydrocarbon feed.
Moving-bed Catalytic Cracking (TCC) - In the TCC
process catalyst beads (^0.5 cm) flow by gravity into the top
of the reactor where they contact a mixed phase hydrocarbon feed.
Cracking reactions take place as the catalyst and hydrocarbons
move concurrently downward through the reactor to a zone where
the catalyst is separated from the vapors. The gaseous reaction
products flow out of the reactor to the fractionation section
of the unit. The catalyst is steam stripped to remove any
adsorbed hydrocarbons. It then falls into the regenerator where
coke is burned from the catalyst with air. The regenerated
catalyst is separated from the flue gases and recycled to be
mixed with fresh hydrocarbon feed. The operating temperatures
of the reactor and regenerator in the TCC process are comparable
to those in the FCC process.
Air emissions from catalytic cracking processes are
1) combustion products from process heaters, and 2) flue gas from
catalyst regeneration. Emissions from process heaters are dis-
A-13
-------
cussed later in this section. Emissions from the catalyst
regenerator include hydrocarbons, oxides of sulfur, ammonia,
aldehydes, oxides of nitrogen, cyanides, carbon monoxide, and
particulates (Table 9.1-1). The particulate emissions from FCC
units are much greater than those from TCC units because of the
higher catalyst circulation rates used.2'3'5
FCC particulate emissions are controlled by cyclones
and/or electrostatic precipitators. Particulate control effi-
ciencies are as high as 80 to 85 percent.3'5 Carbon monoxide
wasteheat boilers reduce the carbon monoxide and hydrocarbon
emissions from FCC units to negligible levels.3 TCC catalyst
regeneration produces similar pollutants to FCC units but in
much smaller quantities (Table 9.1-1). The particulate emissions
from a TCC unit are normally controlled by high efficiency
cyclones. Carbon monoxide and hydrocarbon emissions from a TCC
unit are incinerated to negligible levels by passing the flue
gases through a process heater fire-box or smoke plume burner.
In some installations sulfur oxides are removed by passing the
regenerator flue gases through a water or caustic scrubber.2'3'5
Thermal Cracking
Thermal cracking processes include visbreaking and
coking which break heavy oil molecules by exposing them to high
temperatures.
Visbreaking - Topped crude or vacuum residuals are
heated and thermally cracked (850-900°F, 50-250 psig) in the
visbreaker furnace to reduce the viscosity or pour point of the
charge. The cracked products are quenched with gas oil and
flashed into a fractionator. The vapor overhead from the
fractionator is separated into light distillate products.
A-14
-------
A heavy distillate recovered from the fractionator liquid can
be used as a fuel oil blending component or used as catalytic
cracking feed.
Coking - Coking is a thermal cracking process which
is used to convert low value residual fuel oil to higher value
gas oil and petroleum coke. Vacuum residuals and thermal tars
are cracked in the coking process at high temperature and low
pressure. Products are petroleum coke, gas oils and lighter
petroleum stocks. Delayed coking is the most widely used process
today, but fluid coking is expected to become an important pro-
cess in the future.
In the delayed coking process heated charge stock is
fed into the bottom section of a fractionator where light ends
are stripped from the feed. The stripped feed is then combined
with recycle products from the coke drum and is rapidly heated
in the coking heater to a temperature of 900-1100°F. Steam
injection is used to control the residence time in the heater.
The vapor-liquid feed leaves the heater, passing to a coke drum
where, with controlled residence time, pressure (25-30 psig),
and temperature (750°F), it is cracked to form coke and vapors.
Vapors from the drum return to the fractionator where the thermal
cracking products are recovered.
In the fluid coking process, typified by Flexicoking,
residual oil feeds are injected into the reactor where they are
thermally cracked, yielding coke and a wide range of vapor pro-
ducts. Vapors leave the reactor and are quenched in a scrubber
where entrained coke fines are removed. The vapors are then
fractionated. Coke from the reactor enters a heater and is
devolatilized. The volatiles from the heater are treated for
fines removal and sulfur removal to yield a particulate free,
A-15
-------
low sulfur fuel gas. The devolatilized coke is circulated from
the heater to a gasifier where 9570 of the reactor coke is
gasified at high temperature with steam and air or oxygen. The
gaseous products and coke from the gasifier are returned to the
heater to supply heat for the devolatilization. These gases
exit the heater with the heater volatiles through the same fines
removal and sulfur removal processes.
From available literature it is unclear what emissions
are released and where they are released. Air emissions from
thermal cracking processes include coke dust from decoking
operations, combustion gases from the visbreaking and coking
process heaters, and fugitive emissions. Emissions from the
process heaters are discussed later in this section. Fugitive
emissions from miscellaneous leaks are significant because of
the high temperatures involved, and are dependent upon equipment
type and configuration, operating conditions, and general main-
tenance practices. Fugitive emissions are also discussed later
in this section. Particulate emissions from delayed coking
operations are potentially very significant. These emissions
are associated with removing the coke from the coke drum and
subsequent handling and storage operations. Hydrocarbon emis-
sions are also associated with cooling and venting the coke drum,
prior to coke removal. However, comprehensive data for delayed
coking emissions have not been included in available literature.1*
Particulate emission control is accomplished in the
decoking operation by wetting down the coke.5 Generally, there
is no control of hydrocarbon emissions from delayed coking.
However, some facilities are now collecting coke drum emissions
in an enclosed system and routing them to a refinery flare.1*'5
A-16
-------
Utilities Plant
The utilities plant supplies the steam necessary for
the refinery. Although the steam can be used to produce elec-
tricity by throttling through a turbine, it is primarily used
for heating and separating hydrocarbon streams. When used for
heating, the steam usually heats the petroleum indirectly in
heat exchangers and returns to the boiler. When used in direct
contact operations the steam can serve as a stripping medium,
or a process fluid. Steam may also be used in vacuum ejectors
to produce a vacuum.
The emissions from boilers, and applicable emission
control technology are discussed in much greater detail in
Chapter 1.0.
Sulfur Recovery Plant
Sulfur recovery plants are used in petroleum refin-
eries to convert hydrogen sulfide (H2S) separated from refinery
gas streams into the more disposable by-product, elemental sulfur,
The emissions from sulfur recovery plants and their
control are discussed in Section 5.18.
Slowdown System
The blowdown system provides for the safe disposal of
hydrocarbons (vapor and liquid) discharged from pressure relief
devices.
A-17
-------
Most refining processing units and equipment subject to
planned or unplanned hydrocarbon discharges are manifolded into
a collection unit, called the blowdown system. By using a series
of flash drums and condensers arranged in decreasing pressure,
the blowdown is separated into vapor and liquid cuts. The
separated liquid is recycled into the refinery. The gaseous
cuts can either be smokelessly flared or recycled.
Uncontrolled blowdown emissions primarily consist of
hydrocarbons, but can also include any of the other criteria
pollutants processed by the refinery. The emission rate in a
blowdown system is a function of the amount of equipment
manifolded into the system, the frequency of equipment discharges,
and the blowdown systems"s controls.
Emissions from the blowdown system can be effec-
tively controlled by combustion of the non-condensables in a
flare. To obtain complete combustion or smokeless burning,
(as required by most states) steam is injected in the combustion
zone of the flare to provide turbulence and to inspirate air.
Steam injection also reduces NO emissions by lowering the flame
X
temperature. Controlled emissions are listed in Table 9.1-1.2'11
Process Heaters
Process heaters (furnaces) are used extensively in
refineries to supply the heat necessary to raise the temperature
of feed materials to reaction or distillation temperature. They
are used in many processes throughout the refinery.
Process heaters are designed to raise petroleum fluid
temperatures to a maximum of about 950°F. The fuel burned may
A-18
-------
be refinery gas, natural gas, residual fuel oils, or combina-
tions, depending on the economics, operating conditions, and
pollution requirements. The process heaters may also use carbon
monoxide-rich regenerator flue gas as fuel.
All the criteria pollutants are emitted from process
heaters. The quantity of these emissions is a function of the
type of fuel burned, the nature of the contaminants in the fuel,
and the heat duty of the furnace. Sulfur oxide emissions can be
controlled by fuel desulfurization or flue gas treatment. Carbon
monoxide and hydrocarbons can be limited by better combustion
efficiency. Current technology is investigating four general
techniques or modifications for the control of NOX emissions.
These include combustion modification, fuel modification, alter-
nate furnace design, and flue gas treatment. Several of these
NOX control techniques are presently being applied to large
utility boilers, but their applicability to process heaters is
undefined.2'1"
Compressor Engines
Many older refineries use reciprocating and gas tur-
bine engines fired with natural gas to run high pressure com-
pressors. Natural gas has traditionally been a cheap abundant
source of energy. Examples of refining units operating at high
pressure include hydrodesulfurization, isomerization, reforming,
and hydrocracking units. Internal combustion engines are less
reliable and harder to maintain than steam engines or electric
motors. For this reason and because of increasing natural gas
costs very few such units have been installed in the last few
years.
A-19
-------
The major source of emissions from compressor engines
are combustion products in the exhaust gas. These emissions
include carbon monoxide, hydrocarbons, nitrogen oxides, aldehydes,
and ammonia. Sulfur oxides may also be present depending on the
sulfur content of the natural gas. All of these emissions are
significantly higher in exhaust from reciprocating engines than
from turbine engines.
The major emission control technique applied to com-
pressor engines is carburetion adjustment similar to that applied
on automobiles. Catalyst systems similar to those applied to
automobiles may also be effective in reducing emissions but are
currently undefined for this application.
9.1.3 FUGITIVE EMISSION SOURCES AND CONTROL EQUIPMENT
This section presents descriptions of refinery pro-
cesses and operations which are significant sources of fugitive
emissions. Process flow schemes, emission characteristics and
emission control technology are discussed for each process.
Emission factors for both uncontrolled and controlled fugitive
emission sources are listed in Table 9.1-2. The following
fugitive emission sources are discussed in this section on
petroleum refining emissions:
wastewater systems,
cooling towers,
pipeline fittings,
relief valves,
pump and compressor seals,
asphalt blowing,
blind changing,
sweetening,
A-20
-------
I
tsD
TABLE 9.1-2. FUGITIVE HYDROCARBON EMISSION FACTORS FOR PETROLEUM REFINERIES
EMISSION FACTOR RATING: D
a,b
Emission Source
Process drains and
waste water separators
Cooling towers
Pipeline valves
and flanges
Vessel relief valves
Pump seals
Uncontrolled
Emission Factor Units Emissions
lb/103gal wastewater
kg/103 liters wastewater
lb/103bbl refinery feed°
kg/103 liters refinery feed
lb/106gal cooling water
kg/106liters cooling water
lb/103bbl refinery feed
kg/103 liters refinery feed
Ib/day-valve
kg/day-valve
lb/103 bbl refinery feed
kg/103 liters refinery feed
Ib /day-valve
kg/day-valve
lb/103 bbl refinery feed
kg/103 liters refinery feed
Ib/day-seal
kg/day-seal
lb/103 bbl refinery feed
kg/103 liters refinery feed
5
0.6
200
0.6
6
0.7
10
0.03
0.15
0.07
28
0.08
2.4
1.1
11
0.03
5
2.3
17
0.05
Controlled
Emissions
0.2
0.024
10
0.03
d
NA
NA
NA
NA
NA
NA
NA
NA
Neg
Neg
Neg
Neg
3
1.4
10
0.03
Applicable
Control Technology
Vapor recovery systems
and/or separator covers
Minimization of oil
leaks into cooling
water system through
good housekeeping and
maintenance
Good housekeeping and
maintenance
Rupture discs up stream
of relief valves and/or
vent to blowdown system
Mechanical seals, dual
seals, purged seals
(Continued)
-------
I
N>
TABLE 9.1-2. FUGITIVE HYDROCARBON EMISSION FACTORS FOR PETROLEUM REFINERIES
EMISSION FACTOR RATING: D (Continued)
Emission Source
Compressor Seals
Asphalt blowing
Blind changing
Miscellaneous: sampling,
non-asphalt blowing,
Uncontrolled
Emission Factor Units Emissions
Ib/day-seal
kg/day-seal
lb/103bbl refinery feed
kg/103 liters refinery feed
Ib/ton of asphalt
kg/metric ton of asphalt
lb/103bbl refinery feed
kg/103 liters refinery feed
lb/103bbl refinery feed
kg/103 liter refinery feed
9
4
5
0.014
60
30
0.3
0.001
10
0.03
Controlled
Emissions
NA
NA
NA
NA
Neg
Neg
Neg
Neg
NA
NA
Applicable
Control Technology
Mechanical Seals, dual
seals, purged seals
Scrubber, incinerator
Line flushing, use of
"line" blinds, blind
insulation with gate
valves
Good housekeeping and
maintenance
(sweetening), purging
etc.
Storage
Loading
See Section 4.3
See Section 4.4
a References 2, 4, 12, 13
Overall, less than 1 percent by weight of total hydrocarbon emissions are methane.
c Refinery feed is defined as the crude oil feed rate to the atmospheric distillation column.
NA - These factors are not available.
-------
storage, and
transfer operations.
Sweetening
Sweetening of distillates is accomplished by the con-
version of mercaptans to alky1-disulfides in the presence of a
catalyst. The conversion process may be followed by an extrac-
tion step for the removal of the alkyl-disulfides.
In the conversion process sulfur is added to the sour
distillate with a small amount of caustic and air. This mix-
ture is then passed upward through a fixed-bed catalyst counter-
current to a flow of caustic entering at the top of the vessel.
In the conversion and extraction process the sour
distillate is prewashed with caustic and then is contacted with
a solution of catalyst and caustic in the extractor. The
extracted distillate is then contacted with air to convert
mercaptans to disulfides. After oxidation the distillate is
settled, inhibitors are added, and the distillate is sent to
storage. Regeneration is accomplished by mixing caustic from
the bottom of the extractor with air and separating the disul-
fides and excess air.
The major source of air emissions are fugitive hydro-
carbon emissions generated when the distillate product is con-
tacted with air in the "air blowing" step. These emissions are
dependent upon equipment type and configuration as well as on
operating conditions and maintenance practices.'*
A-23
-------
Asphalt Blowing
The asphalt blowing process polymerizes asphaltic resi-
dual oils by oxidation, increasing their melting temperature and
hardness to achieve an increased resistance to weathering.
The oils containing a large quantity of polycyclic
aromatic compounds (asphaltic oils) are oxidized by blowing
heated air through a preheated batch mixture or, in the con-
tinuous process, by passing hot air countercurrent to the oil
flow. The reaction is exothermic, and quench steam is sometimes
needed for temperature control. In some cases ferric chloride or
phosphorus pentoxide is used as a catalyst to increase the
reaction rate and impart special characteristics to the asphalt.
Blowing is stopped when the asphalt reaches the desired product -
specifications.
Air emissions from asphalt blowing are primarily
fugitive hydrocarbon vapors vented with the blowing air. The
quantities of emissions are small because of the prior removal
of volatile hydrocarbons in the distillation units, but the
emissions may contain hazardous polynuclear organics . 2 ' **'13 '1 5
Emissions from asphalt blowing can be controlled to
negligible levels by vapor scrubbing, by incineration, or by a
combination of the two . **'* 3
Storage
All refineries have a feedstock and product storage
area, termed a "tank farm", which provides surge storage capa-
city to insure smooth, uninterrupted refinery operations.
Individual storage tank capacities range from less than 1000
A-24
-------
barrels to more than 500,000 barrels, and total tank farm storage
capacities commonly range from several days to several weeks.
Storage tank designs, emissions, and emission control technolo-
gies are discussed in detail in Section 4.3.
Transfer Operations
Although most refinery feedstocks and products are
transported by pipeline, many feedstocks and products are trans-
ported by trucks, rail cars, and marine vessels. The refinery
feedstocks and products are transferred to and from these trans-
port vehicles in the refinery tank farm area using specialized
pumps and piping systems. The emissions from transfer operations
and applicable emission control technology are discussed in much
greater detail in Section 4.4.
Wastewater Treatment Plant
All refineries employ some form of wastewater treat-
ment to upgrade the quality of water effluents such that they
can be safely returned to the environment or reused within the
refinery.
The design of wastewater treatment plants is compli-
cated by the diversity of refinery pollutants including oil,
phenols, sulfides, dissolved solids, suspended solids, and toxic
chemicals. Although the wastewater treatment processes employed
by refineries vary greatly, they generally include neutralizers,
oil-water separators, settling chambers, clarifiers, dissolved
air flotation systems, coagulators, aerated lagoons, and activated
sludge ponds. Refinery water effluents are collected from various
processing units and conveyed through sewers and ditches to the
wastewater treatment plant. Most of the wastewater treatment
processing occurs in open ponds and tanks.
A-25
-------
The main components of atmospheric emissions from waste-
water treatment plants are fugitive hydrocarbons and dissolved
gases which evaporate from the surfaces of wastewaters residing
in open process drains, wastewater separators, and wastewater
ponds (Table 9.1-2). Treatment processes which involve the exten-
sive contacting of wastewater with air such as aeration ponds
and dissolved air flotation create an even greater potential for
atmospheric emissions.
The control of wastewater treatment plant emissions
involves covering wastewater systems where emission generation
is greatest (such as covering API separators and settling basins)
and removing dissolved gases from wastewater streams with sour
water strippers and phenol recovery units prior to their con-
tact with the atmosphere. These control techniques can poten-
tially achieve greater than 90 percent reduction of wastewater
system emissions.13
Cooling Towers
Cooling towers are used extensively in refinery cooling
water systems to transfer waste heat from the cooling water to
the atmosphere. The only refineries not employing cooling towers
are those with once through cooling. The increasing scarcity of
large water supplies required by once through cooling is contri-
buting to the disappearance of that form of refinery cooling.
In the cooling tower warm cooling water returning from refinery
processes is contacted with air by cascading through packing.
Heat in the cooling water is transferred to the air. Cooling
water circulation rates for refineries commonly range from 0.3
to 3.0 gpm/bbl per day of refinery capacity.2'16
A-26
-------
Atmospheric emissions from the cooling tower consist
of fugitive hydrocarbons and gases stripped from the cooling
water as the air and water come into contact. These contaminants
enter the cooling water system from leaking heat exchangers and
condensers. Although the predominant contaminant in cooling
water is hydrocarbons, dissolved gases such as H2S and NHs may
also be found in cooling water (Table 9.1-2).2'1*
Control of cooling tower emissions is accomplished
by reducing contamination of cooling water through the proper
maintenance of heat exchangers and condensers. The effectiveness
of cooling tower controls is highly variable, depending on
refinery configuration and existing maintenance practices.1*
Miscellaneous Fugitive
Miscellaneous fugitive emission sources are generally
defined as hydrocarbon emission sources which are not associated
with a particular refining process, but which are scattered
throughout the refinery. Fugitive emission sources include
valves, flanges, pipe fittings, pump and compressor seals, blind
changing, and sample line purging. Hydrocarbon emissions from
fugitive emission sources are attributable to the evaporation
of leaked or spilled petroleum liquids and gases. Normally the
control of fugitive emissions involves the minimization of leaks
and spills through equipment changes, procedural changes, and
improved housekeeping and maintenance practices. Fugitive
emissions which are localized can often be controlled by incinera-
tion or vapor recovery systems.
A-27
-------
REFERENCES
1. Burklin, C.E., R.L. Sugarek, and F.C. Knopf, Revision of
Emission Factors for Petroleum Refining, Radian Corporation,
Austin, Texas. Prepared for the U.S. Environmental Protec-
tion Agency, Research Triangle Park, N.C., Report No.
EPA-450/3-77-030, October, 1977.
2. Atmospheric Emissions from Petroleum Refineries. A Guide
for Measurement and Control. PHS No. 763, Washington, D.C.,
Public Health Service, 1960.
3. Environmental Protection Agency, Office of Air and Water
Programs, Office of Air Quality Planning and Standards,
Background Information for Proposed New Source Standards:
asphalt concrete plants, petroleum refineries, storage
vessels, secondary lead smelters and refineries, brass or
bronze ingot production plants, iron and steel plants,
sewage treatment plants, Vol. 1, EPA Report No. APTD-1352a,
Research Triangle Park, N.C., 1973.
4. Danielson, John A., comp. and ed. Air Pollution Engineering
Manual, 2nd ed., AP-40, Research Triangle Park, N.C., EPA,
Office of Air and Water Programs (1973).
5. Jones, Ben G. "Refinery Improves Particulate Control",
The Oil and Gas Journal, 69 (26): 60-62, June 28, 1971.
6. "Impurities in Petroleum, in Petreco Manual, Long Beach,
Petrolite Corp. 1958, p. 1.
A-28
-------
REFERENCES (Continued)
7. Control Techniques for Sulfur Oxide in Air Pollutants,
Environmental Protection Agency, Office of Air Programs,
Research Triangle Park, N.C., Publication Number AP-52,
January 1969.
8. Olson, H.N. and K.E. Hutchinson, "How Feasible are Giant,
One-Train Refineries?" The Oil and Gas Journal. 70(1):
39-43, January 3, 1972.
9. Urban, C.M. and K.J. Springer, Study of Exhaust Emissions
From Natural Gas Pipeline Compressor Engines, Southwest
Research Institute, San Antonio, Texas, prepared for
American Gas Association, Arlington, VA., February 1975.
10. Dietzmann, H.E. and K.J. Springer, Exhaust Emissions From
Piston and Gas Turbine Engines Used in Natural Gas Trans-
mission, Southwest Research Institute, San Antonio, Texas,
prepared for American Gas Association, Arlington, VA.,
January 1974.
11. Klett, M.G., and J.B. Galeski, Flare Systems Study, Lockheed
Missiles & Space Company, Huntsville, Alabama. Prepared for
the U.S. Environmental Protection Agency, Report No.
EPA 600/2-76-079, March, 1976.
12. American Petroleum Inst., Evaporation Loss Committee,
Evaporation Loss in the Petroleum Industry, Causes and
Control, API Bull, 2513, Washington, D.C., 1959.
13. American Petroleum Inst., Committee on Refinery Environmental
Control, Hydrocarbon Emissions from Refineries, API Publica-
tion No. 928, Washington, D.C., 1973.
A-29
-------
REFERENCES (Continued)
14. Brown, R.A., H.B. Mason, and R.J. Schreiber, Systems Analysis
Requirements for Nitrogen Oxide Control of Stationary Sources,
Aerothenn/Acurex Corporation, Mountain View, CA. Prepared
for the U.S. Environmental Protection Agency, Report No.
EPA 650/2-74-091, 1974.
15. Hangebrauck, R.P., et al., Sources of Polynuclear Hydro-
carbons in the Atmosphere. 999-AP-33, Public Health Service,
1967.
16. Crumlish, Wm. S., "Review of Thermal Pollution Probelms,
Standards and Controls at the State Government Level",
Presented at the Cooling Tower Institute Sytnp. , New Orleans,
Jan. 30, 1966.
A-30
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-450/3-77-030
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Revision of Emission Factors for
Petroleum Refining
5. REPORT DATE
October, 1Q77
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
C. E. Burklin
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
10. PROGRAM ELEMENT NO.
Radian Corporation
8500 Shoal Creek Blvd.
Austin, Texas 78766
11. CONTRACT/GRANT NO.
68-02-1889
Task Order #2
12. SPONSORING AGENCY NAME AND ADDRESS
13. TYPE OF REPORT AND PERIOD COVERED
U.S. Environmental Protection Agency
Office of Air and Waste Management
Office of Air Quality Planning and Standards
PaH(' Nointh Carolina 27711
Final Report
14. SPONSORING AGENCY CODE
15. S
16. ABSTRACT
The refinery air pollutant emission factors presently contained in
Section 9.1 of AP-42, A Compilation of Air Pollutant Emission Factors are primarily
based on the results of the Los Angeles Joint Project conducted from 1955 to 1958.
Since that time there have been many process and equipment developments in the
petroleum refining industry. There have also been several individual and control
agency emission testing programs for refinery emission sources. This report presents
the results of an in-depth study to revise and update the emission factors and
process descriptions presented in AP-42 for the petroleum refining industry. The
revisions were to be made only using available information. A testing strategy
was also developed for testing refinery emission sources where further source
testing is warranted.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
Emission Factors
Petroleum Refining
Emissions
Air Pollutants
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report!
Unclassified
21. NO. OF PAGES
20. SECURITY CLASS {This pagej
Unclassified
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE
------- |