United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
EPA-453/R-94-002b
November 1994
©EPA Gasoline Distribution Final
Industry (Stage I) - EIS
Background Information
for Promulgated Standards
-------
ENVIRONMENTAL PROTECTION AGENCY
Background Information
and Final
Environmental Impact Statement
for Gasoline Distribution Facilities
Prepared by:
Bruce C. Jordan (Date)
Director, Emission Standards Division
U.S. Environmental Protection Agency
Research Triangle Park, N.C. 27711
1. The final national emission standards for hazardous air
pollutants (HAP) will limit HAP emissions from existing and
new major source bulk gasoline terminals and pipeline breakout
stations, under section 112(d) of the 1990 Clean Air Act.
2. Copies of this document have been sent to the following
Federal Departments: Labor, Health and Human Services,
Defense, Transportation, Agriculture, Commerce, Interior, and
Energy; the National Science Foundation; the Council on
Environmental Quality; members of the State and Territorial
Air Pollution Program Administrators; the Association of Local
Air Pollution Control Officials; EPA Regional Administrators;
and other interested parties.
3. For additional information contact:
Mr. Stephen Shedd
Waste and Chemical Processes Group (MD-13)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Telephone: (919) 541-5397
4. Copies of this document may be obtained from:
U.S. Environmental Protection Agency's
Technology Transfer Network (TTN)
TTN modem access number: (919) 541-5742
TTN System Operator: (919) 541-5384
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: (703) 487-4650
-------
U.S. EPA Library (MD-35)
Research Triangle Park, NC 27711
Telephone: (919) 541-2777
-------
Gasoline Distribution Industry (Stage I)
Background Information for
Promulgated Standards
Emission Standards Division
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
November 1994
-------
This report has been reviewed by the Emission Standards Division of
the Office of Air Quality Planning and Standards, EPA, and approved
for publication. Mention of trade names or commercial products is
not intended to constitute endorsement or recommendation for use.
Copies of this report are available through the Library Services
Office (MD-35), U.S. Environmental Protection Agency, Research
Triangle Park NC 27711, (919) 541-2777, or from National Technical
Information Service, 5285 Port Royal Road, Springfield VA 22161,
(703) 487-4650.
ii
-------
TABLE OF CONTENTS
Section No. Page
1.0 SUMMARY 1-1
1.1 Summary of Changes Since Proposal 1-7
1.2 Summary of Reanalysis of Impacts of Final
Standards 1-7
1.3 References 1-11
2.0 GENERAL ISSUES RAISED IN PUBLIC COMMENTS 2-1
2.1 Spill Prevention 2-1
2.2 Emissions Averaging 2-3
2.3 Control Level 2-4
2 . 4 Other 2-6
2.5 References 2-10
3.0 APPLICABILITY ISSUES 3-1
3 .1 General 3-1
3.2 Screening Equations 3-2
3.3 Emissions Inventory 3-7
3.4 Change of Facility Status 3-9
3.5 Small Source Coverage 3-9
3.6 Potential to Emit 3-12
3.7 Refinery Bulk Terminals 3-16
3.8 Other 3-19
3.9 References 3-20
4.0 DEFINITIONS 4-1
4.1 Leak 4-1
4.2 Gasoline 4-2
4.3 Gasoline Tank Truck 4-3
4.4 New/Existing Facility 4-4
4.5 Contiguous Area 4-5
4.6 In VHAP Service 4-5
5.0 LOADING RACK STANDARDS 5-1
5.1 Emission Limit 5-1
5.2 Selection of Vapor Processor 5-9
5.3 Vacuum Assist Vapor Collection 5-13
5.4 Other 5-18
5.5 References 5-19
6.0 MONITORING OF OPERATIONS 6-1
6.1 Continuous Monitoring of Collection Systems .... 6-1
iii
-------
6.2 Continuous Monitoring of Processing Systems .... 6-1
6.3 Control Equipment 6-10
6.4 References 6-11
TABLE OF CONTENTS (Continued)
Section No. Page
7.0 CARGO TANK REQUIREMENTS 7-1
7.1 Emission Factors 7-1
7.2 Level of Control 7-3
7.3 Cargo Tank Testing 7-8
7.4 Other 7-12
7.5 References 7-15
8.0 STORAGE VESSEL STANDARDS 8-1
8.1 Determination of MACT Floor Control Level 8-1
8.2 Compliance Deadline 8-7
8 . 3 Other 8-8
9.0 EQUIPMENT LEAK STANDARDS 9-1
9.1 Emission Factors 9-1
9.2 Need for LDAR Standards 9-2
9.3 Other 9-5
9.4 References 9-6
10.0 BASELINE/ENVIRONMENTAL IMPACT CALCULATIONS 10-1
10.1 General 10-1
10.2 Gasoline RVP 10-2
10.3 References 10-3
11.0 TEST METHODS AND PROCEDURES 11-1
11.1 Alternative Test Methods 11-1
11.2 Testing of Cargo Tanks 11-1
12.0 REPORTING AND RECORDKEEPING 12-1
12.1 Storage of Records 12-1
12.2 Reporting Requirements 12-3
12.3 Other 12-5
13.0 COSTS/ECONOMIC IMPACTS 13-1
13.1 Storage Vessel Control Costs 13-1
13.2 Other 13-3
iv
-------
13.3 References 13-7
TABLE OF CONTENTS (Concluded)
Section No. Page
APPENDIX A CARGO TANK VAPOR LEAKAGE A-l
APPENDIX B ENVIRONMENTAL AND COST IMPACTS OF
STORAGE VESSEL CONTROLS B-l
APPENDIX C EMISSIONS FROM EQUIPMENT COMPONENTS AT BULK
TERMINALS AND PIPELINE BREAKOUT STATIONS . . . . C-l
APPENDIX D REVISIONS TO THE NATIONWIDE EMISSION REDUCTION
AND COST ESTIMATES D-l
v
-------
LIST OF TABLES
Table No. Page
Table 1-1 LIST OF COMMENTERS ON THE PROPOSED NESHAP FOR
THE GASOLINE DISTRIBUTION (STAGE I) INDUSTRY . 1-2
Table 1-2 SUMMARY OF MAJOR CHANGES MADE TO THE
STANDARDS SINCE PROPOSAL 1-8
Table 7-1 ALLOWABLE TANK TRUCK PRESSURE CHANGE 7-6
Table A-l 1988 TO 1992 BAAQMD TANK TRUCK LEAKAGE DATA . A-3
Table A-2 198 9 TO 1994 TANK TRUCK LEAKAGE DATA A-4
Table B-l STORAGE VESSEL DEGASSING AND REFILLING
EMISSIONS B-2
Table B-2 EMISSIONS BALANCING—INSTALL SECONDARY SEAL
(SUBPART Kb RIM SEALS) ON AN IFRT WITH A
VAPOR-MOUNTED PRIMARY SEAL B-4
Table B-3 EMISSIONS BALANCING—INSTALL SECONDARY SEAL
(SUBPART Kb RIM SEALS) ON AN EFRT WITH A
MECHANICAL-SHOE PRIMARY SEAL B-4
Table B-4 EMISSIONS BALANCING—INSTALL SECONDARY SEAL
(SUBPART Kb RIM SEALS) ON AN EFRT WITH A
LIQUID-MOUNTED PRIMARY SEAL B-4
Table B-5 EMISSIONS BALANCING—INSTALL SECONDARY SEAL
AND CONTROLLED FITTINGS (SUBPART Kb) ON AN
IFRT WITH A VAPOR-MOUNTED PRIMARY SEAL . . . B-6
Table B-6 EMISSIONS BALANCING—INSTALL SECONDARY SEAL
AND CONTROLLED FITTINGS (SUBPART Kb) ON AN
EFRT WITH A MECHANICAL-SHOE PRIMARY SEAL . .B-6
Table B-7 EMISSIONS BALANCING—INSTALL SECONDARY SEAL
AND CONTROLLED FITTINGS (SUBPART Kb) ON AN
EFRT WITH A LIQUID-MOUNTED PRIMARY SEAL . . .B-6
Table B-8 EMISSIONS BALANCING—INSTALL CONTROLLED
FITTINGS (SUBPART Kb) ON AN IFRT WITH A
VAPOR-MOUNTED PRIMARY SEAL AND SECONDARY
SEAL (SUBPART Kb RIM SEALS) B-7
Table B-9 EMISSIONS BALANCING—INSTALL CONTROLLED
FITTINGS (SUBPART Kb) ON AN EFRT WITH A
MECHANICAL-SHOE PRIMARY SEAL AND
vi
-------
SECONDARY SEAL (SUBPART Kb RIM SEALS) . . . . B-7
LIST OF TABLES (Continued)
Table No.
Table B-10
Table B-ll
Table B-12
Table B-13
Table B-14
Table B-15
Table B-16
Table B-17
Table B-18
Table B-19
Table B-20
Page
COST AND EMISSIONS IMPACT OF REQUIRING
CONTROLLED FITTINGS ON ALL GASOLINE STORAGE
TANKS B-ll
COSTS TO INSTALL FLOATING DECK, LIQUID-
MOUNTED PRIMARY SEAL, AND CONTROLLED
FITTINGS ON A 50' FRT B-12
COSTS TO INSTALL FLOATING DECK, LIQUID-
MOUNTED PRIMARY SEAL, AND CONTROLLED
FITTINGS ON A 100' FRT B-14
COSTS TO INSTALL SECONDARY SEAL AND
CONTROLLED FITTINGS ON A 78' EFRT
WITH A MECHANICAL-SHOE PRIMARY SEAL .... B-16
COSTS TO INSTALL SECONDARY SEAL AND
CONTROLLED FITTINGS ON A 100' EFRT
WITH A MECHANICAL-SHOE PRIMARY SEAL .... B-18
COSTS TO INSTALL SECONDARY SEAL AND
CONTROLLED FITTINGS ON A 50' IFRT
WITH A VAPOR-MOUNTED PRIMARY SEAL B-2 0
COSTS TO INSTALL SECONDARY SEAL AND
CONTROLLED FITTINGS ON A 100' IFRT
WITH A VAPOR-MOUNTED PRIMARY SEAL B-22
COSTS TO INSTALL CONTROLLED FITTINGS ON 50'
IFRT WITH A LIQUID OR MECHANICAL PRIMARY
SEAL—WITH AND WITHOUT DEGASSING COSTS . . . B-24
COSTS TO INSTALL CONTROLLED FITTINGS ON 100'
IFRT WITH A LIQUID OR MECHANICAL PRIMARY
SEAL—WITH AND WITHOUT DEGASSING COSTS . . . B-25
COSTS TO INSTALL CONTROLLED FITTINGS
ON 78' EFRT WITH A MECHANICAL-SHOE PRIMARY
SEAL AND SECONDARY SEAL—WITH AND WITHOUT
DEGASSING COSTS B-2 6
COSTS TO INSTALL CONTROLLED FITTINGS ON
100' EFRT WITH A MECHANICAL-SHOE PRIMARY
vii
-------
SEAL AND SECONDARY SEAL—WITH AND WITHOUT
DEGASSING COSTS B-28
Table B-21
Table No.
Table B-22
Table B-23
Table B-24
Table C-l
Table C-2
Table C-3
Table C-4
Table C-5
Table C-6
Table C-7
Table D-l
Table D-2
Table D-3
Table D-4
BULK TERMINAL MODEL TANK AND NATIONWIDE
COSTS TO MEET SUBPART Kb RIM SEAL
REQUIREMENTS (FLOOR) B-30
LIST OF TABLES (Concluded)
Page
BREAKOUT STATION MODEL TANK AND NATIONWIDE
COSTS TO MEET SUBPART Kb RIM SEAL
REQUIREMENTS (FLOOR) B-30
BULK TERMINAL MODEL TANK AND NATIONWIDE
COSTS TO MEET FULL SUBPART Kb REQUIREMENTS . B-31
BREAKOUT STATION MODEL TANK AND NATIONWIDE
COSTS TO MEET FULL SUBPART Kb REQUIREMENTS . B-31
REVISED EMISSION FACTORS FOR EQUIPMENT LEAKS . C-3
TERMINAL THROUGHPUTS AND COMPONENT
COUNTS C-3
OCCURRENCE OF COMPONENT TYPES AT BULK
TERMINALS C-5
EQUIPMENT COMPONENT POPULATIONS AT BULK
TERMINALS C-6
BREAKOUT STATION COMPONENT DATA C-7
REVISED EQUIPMENT COMPONENT POPULATIONS FOR
BREAKOUT STATIONS C-9
NATIONWIDE BASELINE EQUIPMENT EMISSIONS FOR
BULK TERMINALS AND BREAKOUT STATIONS C-9
SUMMARY OF NATIONWIDE EMISSION REDUCTIONS
AND COST IMPACTS OF THE FINAL RULE D-2
BASELINE EMISSIONS FROM PIPELINE BREAKOUT
STATIONS D-3
BASELINE EMISSIONS FROM BULK TERMINALS .... D-3
NUMBER OF AFFECTED SOURCE BULK TERMINALS
AND PIPELINE BREAKOUT STATIONS D-4
viii
-------
Table D-5 MODEL PLANT POTENTIAL TOTAL HAP EMISSIONS . . D-5
Table D-6 MODEL PLANT MAXIMUM INDIVIDUAL HAP
EMISSIONS D-5
Table D-7 EMISSION FACTORS FOR PIPELINE BREAKOUT STATION
STORAGE VESSELS D-7
Table D-8 EMISSION FACTORS FOR BULK TERMINAL STORAGE
VESSELS D-8
ix
-------
1 . 0 SUMMARY
On February 8, 1994, the U.S. Environmental Protection
Agency (EPA) proposed national emission standards for hazardous
air pollutant (HAP) emissions from bulk gasoline terminals and
pipeline breakout stations that are major sources of HAP's or are
collocated at plant sites that are major sources (59 FR 5868).
On August 19, 1994, EPA announced in the Federal Register the
availability of supplemental information regarding the level of
control and test procedures for gasoline cargo tanks and reopened
the public comment period (59 FR 42788) . These proposed
standards implemented section 112(d) of the Clean Air Act as
amended in 1990 (the Act). Public comments on the proposal and
the supplemental FR notice were requested at the time the
standards were proposed in the Federal Register. There were 48
comment letters (see Table 1-1) submitted by facility owners and
operators, trade associations, and State and local air pollution
control agencies. In addition, comments were received from a
control equipment supplier, a private environmental organization,
and one U.S. Government agency. Summaries of the comments that
were submitted, along with EPA's responses to these comments, are
presented in this document. This comment summary and the
Agency's responses served as the basis for the revisions made to
the standards between proposal and promulgation. This document
is volume II of "Gasoline Distribution Industry (Stage I) -
Background Information for Proposed Standards," EPA-453/R-94-
002a, January 1994 (later referred to as BID, Volume I) (docket
item III-B-1). This report also includes a discussion of the
1-1
-------
changes made to the regulatory analysis presented in BID, Volume
I.
1-2
-------
TABLE 1-1. LIST OF COMMENTERS ON THE PROPOSED
NESHAP FOR THE GASOLINE DISTRIBUTION
(STAGE I) INDUSTRY
W444444444444444444444444444444444444444444444444444444444444444444444444U
Item Number in
Docket A-92-38 Commenter and Affiliation
IV-D-1 R.L. Mikkelsen
Manager, Environmental Permits
Alyeska Pipeline
1835 South Bragaw Street
Anchorage, AK 99512
IV-D-2 Mr. Michael J. Hansel
Director, Environmental Strategic
Planning
KOCH Industries, Inc.
PO Box 64596
Saint Paul, MN 55184
IV-D-3 Mr. Dewey Mark
Executive Director
Independent Refiner/Marketers
Association
9901 IH 10 West, Suite 800
San Antonio, TX 78230
IV-D-4 Ms. Barbara J. Price
Vice President
Health, Environment and Safety
Phillips Petroleum Company
Bartlesville, OK 74004
IV-D-5 Ms. Melanie S. Kelley
V.P., Env., Safety & Public Affairs
Total Petroleum, Inc.
PO Box 500
Denver, CO 80201-0500
IV-D-6, IV-D-32 Mr. David S. Kircher
Manager, Engineering
Puget Sound Air Poll. Control Agency
110 Union Street, Suite 500
Seattle, WA 98101-2 038
IV-D-7 J.W. Casey
Manager, Environmental Support
Shell Oil Company
PO Box 4320
Houston, TX 77210-4320
IV-D-8 Sarosh J. H. Manekshaw
Director, Env. Safety & Health
Pennzoil Company
PO Box 2967
Houston, TX 77252-2967
IV-D-9 Mr. Harold L. Dinsmore
V.P., Vapor Control Systems
John Zink Company
PO Box 21220
Tulsa, OK 74121-1220
IV-D-10 Kelly A. Sakir
Demetriou, Del Guercio,
Springer & Moyer
801 South Grand Avenue
1-3
-------
Los Angeles, CA 90017-4613
))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))Q
1-4
-------
TABLE 1-1. (Continued)
))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))Q
Item Number in
Docket A-92-38 Commenter and Affiliation
IV-D-11
IV-D-12
IV-D-13
Environment and
IV-D-14
IV-D-15
IV-D-16
IV-D-17
IV-D-18
IV-D-19
IV-D-20
Operators
Ms. Sara Hutson, P.E.
WAID and Associates
14205 Burnet Road, Suite 500
Austin, TX 78728
0. Horton Cunningham
Vice President, Operations
TE Products Pipeline Company
PO Box 2521
Houston, TX 77252-2521
Ms. Deborah W. Gates
Vice President
alth
Ashland Petroleum Company
PO Box 391
Ashland, KY 41114
Mr. Walter Roy Quanstrom, V.P.
Env., Health & Safety Dept.
Amoco Corporation
PO Box 87703
Chicago, IL 60680-0703
B.R. Baker, V.P.
U.S. Supply & Logistics
Mobil Oil Corporation
3225 Gallows Road
Fairfax, VA 22037-0001
Mr. Paul C. Bailey, Jr.,
Director, Health &
Environmental Affairs
American Petroleum Institute
1220 L Street, Northwest
Washington, DC 20005
(summarizes comments in IV-D-22)
Norbert Dee
Director, Environ. Affairs
National Petroleum
Refiners Assoc.
1899 L Street NW, Suite 1000
Washington, DC 20036
Mr. George A. Walker
V.P., Health, Environment & Safety
Unocal Corporation
1201 West 5th St., PO Box 7600
Los Angeles, CA 90051
Mr. John Prokop
President and Counsel
Ind. Liquid Terminals Assoc.
1133 15th Street, N.W., Suite 650
Washington, DC 20005
Ms. Andrea Grant, Counsel
Independent Fuel Terminal
sociation
901 15th Street, N.W.,
1-5
-------
Suite 700
Washington, DC 20005-2301
))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))Q
TABLE 1-1. (Continued)
))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))Q
Item Number in
Docket A-92-38 Cormnenter and Affiliation
IV-D-21
IV-D-22
IV-D-23
IV-D-24
IV-D-25
Conse
IV-D-26
IV-D-27
IV-D-28
IV-D-29
IV-D-30
Robert H. Colby and
Donald F. Theiler
STAPPA/ALAPCO
444 North Capitol St. N.W.
Washington, DC 20001
Robert P. Strieter and
Ellen Siegler
American Petroleum Institute
1220 L Street, N.W.
Washington, DC 20005
Mr. Clifford J. Harvison
President
National Tank Truck Carriers
2200 Mill Road
Alexandria, VA 22314-4677
Mr. Charles L. Murphy
Senior Consultant
Conoco Inc.
PO Box 1267
Ponca City, OK 74602-1267
Mr. John W. Walton, Tech. Sec.
Tennessee APCB, Dept. of Env.
rvation
401 Church Street
Nashville, TN 37243-1531
Mr. Jerry Langley
Director of Health,
Environment & Safety
Williams Pipe Line Company
PO Box 34 4 8
Tulsa, OK 74101
Mr. David Driesen
Natural Resources Defense Council
1350 New York Avenue, N.W.
Washington, DC 20005
Mr. Kenneth M. Kunaniec, Eng. Mgr.
939 Ellis Street
San Francisco, CA 94109
Ms. Ann Farner
Director, Government Relations
Tosco Refining Company
2300 Clayton Road, Suite 1100
Concord, CA 94520-2100
Mr. Jeffrey L. Leiter &
Mr. Gregory M. Scott
Collier, Shannon, Rill & Scott
3050 K Street, N.W., Suite 400
Washington, DC 20007
1-6
-------
IV-D-31 Mr. Allen R. Ellet
Environmental Consultant
BP Oil Company
200 Public Square
Cleveland, OH 44114-2375
IV-D-32 (same as IV-D-6)
))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))Q
TABLE 1-1. (Continued)
))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))Q
Item Number in
Docket A-92-38 Commenter and Affiliation
IV-D-33 Mr. John J. Huber
Government Affairs Counsel
Petroleum Marketers Association of
America
1901 N. Ft Meyer Drive, Suite 1200
Arlington, VA 22209
IV-D-34 R.T. Richards
General Mgr., Env. and Product
Safety
Texaco Inc.
PO Box 509
Beacon, NY 12508
IV-D-35 Mr. William J. Doyle
Manager, Environmental Affairs
Marathon Oil Company
539 South Main Street
Findlay, OH 45840-3295
IV-D-36 Mr. David A. Buff, P.E.
KBN Engineering and Applied
Sciences, Inc.
1034 N.W. 57th Street
Gainesville, FL 32605
IV-D-37 Terri Thomas
Supervisor, Air Toxics Section
Ventura County Air Pollution
Control District
7 02 County Square Drive
Ventura, CA 93003
IV-D-38 Mr. Brian Bateman
Manager, Toxics Evaluation Section
Bay Area Air Quality Mgmt. District
939 Ellis Street
San Francisco, CA 94109
IV-D-39 Anthony R. O'Neill
V.P., Government Affairs
National Fire Protection Association
P.O. Box 9101
Quincy, MA 02269-9101
IV-D-40 Susan F. Tierney, Assistant
Secretary
Office of Policy, Planning &
Program Evaluation
Department of Energy
Washington, DC 20585
1-7
-------
))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))Q
1-8
-------
TABLE 1-1. (Concluded)
))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))Q
Item Number in
Docket A-92-38 Cormnenter and Affiliation
Additional comments on the notice
IV-D-47
IV-D-48
IV-D-4 9
IV-D-50
IV-D-51
IV-D-52
IV-D-53
IV-D-54
IV-D-55
of supplementary information:
Mr. John W. Walton, Tech. Sec.
Tennessee APCB, Dept. of Env.
Conservation
401 Church Street
Nashville, TN 37243-1531
Mr. Brent Tracy
Senior Environmental Engineer
Total Petroleum, Inc.
PO Box 500
Denver, CO 80201-0500
Mr. Paul Bailey
Director, Health &
Environmental Affairs
American Petroleum Institute
1220 L Street, Northwest
Washington, DC 20005
Mr. John Prokop
President and Counsel
Ind. Liquid Terminals Assoc.
1133 15th Street, N.W., Suite 650
Washington, DC 20005
Mr. B. R. Baker, V.P.
Mobil Oil Corporation
3225 Gallows Road
Fairfax, VA 22037-0001
Mr. William J. Doyle
Mgr., HES Policy & Analysis
Marathon Oil Company
539 South Main Street
Findlay, OH 45840-3295
Mr. Arthur Lee
Texaco Inc.
PO Box 50 9
Beacon, NY 12508
Ms. Deborah W. Gates
V.P., Environment & Health
Ashland Petroleum Company
PO Box 391
Ashland, KY 41114
Mr. Jerry Langley
Director Health, Env. & Safety
Williams Pipe Line Company
PO Box 34 4 8
Tulsa, OK 74101
W444444444444444444444444444444444444444444444444444444444444444444444444U
1-9
-------
1.1 SUMMARY OF CHANGES SINCE PROPOSAL
Several changes of varying importance have been made to the
standards since proposal. The majority of the changes were made
in response to the public comments, but some were made to improve
clarity or consistency. Table 1-2 provides a summary of the
major changes made in developing the final standards.
1.2 SUMMARY OF REANALYSIS OF IMPACTS OF FINAL STANDARDS
1.2.1 Alternatives
The regulatory alternatives are discussed in Section 5 of
BID, Volume I. These regulatory alternatives reflect the
different levels of emission control from which is selected the
approach that represents the best technology for continuous
emission reduction, considering cost, nonair quality health, and
environmental and economic impacts for gasoline distribution
facilities.
As discussed in Section 1.1, several major changes have been
made to the standards since proposal: 1) leak detection and
repair (LDAR) using a monitor has been deleted from the rule and
replaced with a sight, sound, and smell inspection program to
control equipment leaks; 2) vacuum assist vapor collection at new
bulk terminals has been deleted and replaced with an annual 1-
inch vapor tightness test to control cargo tank leakage, a year-
round cargo tank performance standard, and an internal vapor
valve test; and 3) the annual 3-inch vapor tightness test has
been replaced with the aforementioned 1-inch vapor tightness
test, year-round performance standard, and internal vapor valve
test to control cargo tank leakage at existing bulk terminals.
1.2.2 Environmental Impacts
The estimated environmental impacts of the proposed
standards were discussed in Section 6 of BID, Volume I. Changes
to these estimates have been made since proposal for several
reasons: 1) the equipment leakage emission factors have been
lowered significantly, 2) the controlled cargo tank leakage
emission factors have been lowered significantly, and 3) the
1-10
-------
estimated national average Reid vapor pressure of gasoline has
been lowered from 11.4 psia to 10.4 psia. Nationwide HAP and
1-11
-------
TABLE 1-2. SUMMARY OF MAJOR CHANGES MADE TO THE STANDARDS
SINCE PROPOSAL
Rule Section and Title
Rule Change (BID Section Reference)
§63.420
ADDlicabilitv
0 Screening equations revised (Sec. 3.2)
0 Potential to emit - operating limits
incorporated and reporting for large
nonmajor sources (Sec. 3.6)
0 Use of SIC codes in regulating refinery
terminals (Sec. 3.7)
0 Coverage of terminals at pipeline breakout
stations (Sec. 3.8)
§63.421
Definitions
° "In VHAP service" deleted (Sec. 4.6)
0 "Gasoline tank truck" changed to "Gasoline
cargo tank" (Sec. 4.3)
§63.422
Loadincr racks
0 Vacuum assist requirement deleted (Sec. 5.3)
0 Reduced emission factors for cargo tank
leakage (Sec. 7.1)
0 New floor determination for cargo tanks
(Sec. 7.2, App. A)
§63.423
Storacre vessels
0 New storage vessel floor determination (Sec.
8.1, App. B)
§63.424
EcruiDment leaks
0 Reduced emission factors for equipment leaks
(Sec. 9.1, App. C)
0 LDAR deleted, monthly visual inspection
added (Sec. 9.2)
0 Revised miscellaneous sources (housekeeping)
provision (Sec. 2.1)
$63,425 Test methods and
Drocedures
0 New procedures for testing cargo tanks (Sec.
7.3)
$63,426 Alternative means of
emission limitation
0 No change
§63.427
Continuous monitorincr
0 Eliminated 6-hour average values (Sec. 6.2)
0 Changed procedures for establishing
parameter values (Sec. 6.2)
$63,428 ReDortina and
recordkeeDina
0 Allowing 1 year for initial notifications
(Sec. 12.2)
0 Added records and reports for visual
inspection program (Sec. 9.2)
1-12
-------
§63.429 Delegation of authority
No change
1-13
-------
VOC emissions from affected bulk terminals and pipeline breakout
stations are estimated to decrease by approximately 2,100 Mg of
HAP and 36,000 Mg of VOC per year, respectively. This emissions
decrease will result in a reduction in the ambient air
concentrations of HAP and VOC in the vicinity of approximately
240 new and existing bulk terminals and 20 new and existing
pipeline breakout stations. The nationwide emission reductions
are discussed further in Appendix D.
1.2.3 Water Pollution Impacts
The overall impact of the final rule on water resources will
be negligible. None of the emission control technologies creates
a significant water discharge. As discussed in Section 6.2 of
BID, Volume I, only refrigeration condenser systems, which cool
and condense the vapors from the loading operation for liquid
recovery, when used for bulk terminal control, would create a
potential water pollution impact.
1.2.4 Solid Waste Impacts
As discussed at proposal, solid waste (spent carbon) may be
generated by bulk terminals which use a carbon adsorption system
to control loading rack emissions. As shown in Table 6-5 of BID,
Volume I, it is estimated that approximately 200 megagrams (Mg)
of solid waste per year would be generated from carbon disposal
at bulk gasoline terminals.
In addition, sludge (tank bottoms) will be generated from
storage vessels at breakout stations and bulk terminals that must
comply with the floor level of control (floating deck rim seal
requirements in 40 CFR part 60, subpart Kb). The EPA estimates
that at breakout stations, 95 external floating roof tanks
(EFRT's) with a mechanical primary seal only, 15 internal
floating roof tanks (IFRT's) with a vapor-mounted primary seal
only, and 28 fixed-roof tanks will have to be cleaned and
degassed earlier than their normal cleaning schedule to meet the
final rule. It is estimated that the amount of solid waste
generated from the cleaning of these tanks will be 3,985 Mg.
1-14
-------
Similarly, it is estimated that at bulk terminals 416 EFRT's
with a mechanical primary seal only, 154 IFRT's with a vapor-
mounted primary seal only, and 184 fixed-roof tanks will have to
be cleaned and degassed earlier than their normal cleaning
schedule to meet the final rule. The volume of solid waste
generated from these cleaning activities is estimated to total
approximately 9,750 Mg.
1.2.5 Energy Impacts
Energy impacts of the final rule have been estimated in the
form of gallons of gasoline saved. Energy savings were derived
by determining the liquid gasoline equivalent of the emission
reductions and assuming a gasoline density of 0.67 kg per liter.
Liquid gasoline is saved by controlling equipment leaks, storage
vessels, and cargo tanks since less product is allowed to
evaporate and escape. Gasoline is recovered at terminals when
carbon adsorption or refrigeration condenser systems are used to
control emissions from loading operations and from storage
vessels piped to vapor processors. For bulk terminals, it was
assumed that 25 percent of the emission reductions would be
processed using recovery devices (carbon adsorption,
refrigeration condenser). The remaining 75 percent of the
emission reductions would be achieved by vapor destruction
devices and would not provide an energy savings. The total
gasoline savings due to implementing the final rule is estimated
at 10.5 million gallons per year.
1.2.6 Other Environmental Impacts
As discussed in Section 6.5 of BID, Volume I, other
potential environmental impacts include noise impacts resulting
from the use of vapor processors at bulk terminals. As discussed
at proposal, the noise impacts of the promulgated action are
expected to be insignificant.
1.2.7 Cost and Economic Impacts
The estimated costs and economic impacts of the proposed
standards were discussed in Sections 7 and 8 of BID, Volume I.
1-15
-------
Changes in these impacts have occurred since proposal for several
reasons: 1) vacuum assist is no longer required at new facility
loading racks, and has been replaced with new cargo tank
certification and testing requirements; 2) the degassing/cleaning
cost for storage tanks has been reevaluated and estimated to be
significantly higher than proposed; 3) the costs of fitting
controls have been reevaluated and estimated to be significantly
higher than proposed; and 4) the LDAR program has been replaced
with a visual inspection program. The nationwide capital cost of
this regulatory action is estimated at approximately $117
million. The nationwide annual cost of this regulatory action is
estimated at approximately $15.5 million per year (including
reporting and recordkeeping costs). The nationwide environmental
and cost impacts are discussed further in Appendix D.
1.3 REFERENCES
III-B-1 Gasoline Distribution Industry (Stage I) -
Background Information for Proposed
Standards. U.S. Environmental Protection
Agency, Research Triangle Park, NC. EPA-
453/R-94-002a. January 1994.
1-16
-------
2.0 GENERAL ISSUES RAISED IN PUBLIC COMMENTS
2.1 SPILL PREVENTION
Comment: Several commenters stated that the proposed
requirement in §63.424(d) constituted a prohibition on spills
that would be impossible to comply with because, even with a
preventive maintenance plan, all spills cannot be eliminated.
The necessary steps required to clean up a spill would be
impossible to carry out without violating the current rule, thus
invalidating the current Spill Prevention Control and
Countermeasures (SPCC) plans already approved by EPA (e.g., open-
air oil/water separators). In addition, EPA has presented no
evidence that leaks or spills contribute significantly to HAP
emissions. As a result, the feasibility and cost of such a
requirement has not been demonstrated. Many terminals are
unmanned a majority of the time. It would be grossly unfair to
hold terminal owners or operators accountable for the action of
tank truck drivers that may cause gasoline to be spilled. The
spill prevention requirement should be deleted or at least
reworded to instruct facilities to take reasonable precautions to
prevent spills (IV-D-2, p. 10; IV-D-3, p. 2; IV-D-5, p. 2; IV-D-
13, p. 3; V-D-14, p. 10; IV-D-17, p. 8; IV-D-18, p. 8; IV-D-20,
p. 8; IV-D-22, p. 26; IV-D-29, p. 5; IV-D-34, p. 17; IV-D-35, p.
3; IV-D-36, p. 3).
Response: The EPA's proposed provision was worded:
"Owners or operators of bulk gasoline terminals and
pipeline breakout stations subject to the provisions
of this subpart shall not cause or allow gasoline to
be spilled, discarded in sewers, stored in open con-
tainers, or handled in any other manner that would
2-1
-------
result in vapor release to the atmosphere."
This provision is similar to EPA's "model rule" language that has
been offered to States for use in developing State implementation
plan rules for VOC sources. In State rules, the provision
typically prohibits discarding gasoline in sewers, storing it in
open containers, and handling it in any manner that would result
in evaporation (and loss). The proposed provision was not
intended to constitute a ban on spills, but rather a housekeeping
and good practices requirement that would address the smaller
potential HAP releases at a facility (which may be overlooked by
the SPCC or DOT plans). The EPA agrees that, in its proposed
form, the provision could be construed as a prohibition on any
form of accidental release or spill. While such a prohibition
may not be reasonable, the Agency maintains its intent to hold
owners and operators accountable for deliberate or careless
practices that could lead to emissions.
The commenters supplied recommended language for this
provision that they believe would create a workable requirement.
This suggested wording was consistent among the commenters, and
reads as follows:
"Owners or operators of bulk gasoline terminals
and pipeline breakout stations subject to the
provisions of this subpart shall not allow gasoline to
be intentionally handled in a manner that would result
in vapor releases to the atmosphere for extended
periods of time. This prohibition would not preclude
the following activities: (1) the use of passive
devices (i.e., dikes, troughs, pans, etc.) to contain
accidental gasoline spills or leaks; (2) the use of on-
site sewer systems to collect and transport gasoline to
reclamation/recycling devices such as oil/water
separators; (3) small volume open containers which are
used in routine maintenance and sampling activities;
and (4) maintenance/tank cleaning activities."
The EPA has considered these comments and, while agreeing in
principle that the proposed wording needed improvement, finds
that the term "intentionally" may lead to enforcement
2-2
-------
difficulties in determining intent. Also, the Agency does not
wish to provide the encompassing exclusions suggested by the
commenters, but instead to emphasize the positive measures that
should be taken to minimize miscellaneous vapor releases.
Therefore, the language in §63.424(g) of the final rule reads as
follows:
"Owners or operators shall not allow gasoline to
be handled in a manner that would result in vapor
releases to the atmosphere for extended periods of
time. Measures to be taken include, but are not
limited to, the following:
(1) Minimize gasoline spills;
(2) Clean up spills as expeditiously as
practicable;
(3) Cover all open gasoline containers with a
gasketed seal when not in use;
(4) Minimize gasoline sent to open waste
collection systems that collect and transport
gasoline to reclamation and recycling
devices, such as oil/water separators."
2.2 EMISSIONS AVERAGING
Comment: Three commenters believed emissions averaging to
be a reasonable approach to cost-effective control and also to be
acceptable under the Clean Air Act, and urged EPA to allow
sources to average emissions when determining compliance. The
Hazardous Organic NESHAP (HON) for the chemical production
industry is cited as allowing emissions averaging for facilities
regulated under the HON; also, the Administrator emphasized the
need for market-based regulatory approaches in her confirmation
hearings (IV-D-7, p. 10; IV-D-22, p. 74; IV-D-34, p. 21). One of
the commenters supported an averaging program that would allow
emissions trading within a facility (such as between storage
tanks and loading racks) and among different source categories
(i.e., terminals/breakout stations and refineries) (IV-D-22, p.
77). This commenter submitted a memorandum intended to
2-3
-------
demonstrate that EPA has the legal authority to implement MACT
standards as a "bubble" program (IV-D-22, App. J).
Response: In the proposal preamble, at 59 FR 5880, EPA
stated that it had considered including an emissions averaging
approach but was not able to identify any viable alternatives.
It was further stated that any alternative, to be acceptable,
would have to be protective of the environment and should lower
the cost of achieving a particular level of emission reduction.
The EPA requested data and comments that could be used to develop
an emissions averaging alternative in the final rule.
One of the commenters said that they were currently
conducting research on more effective ways to reduce emission
losses from storage tanks. The commenter further said that,
should this research lead to more effective control techniques at
reasonable cost, facilities would be able to provide additional
control for their storage tanks while retaining existing, less
efficient rack vapor recovery systems at a great cost savings.
While EPA is interested in learning more about effective
approaches that retain emission control and are enforceable,
these comments have not included any data or information ("viable
alternatives") that could serve as the object of EPA's analysis.
Any such information would have to include verifiable emissions
reduction data and comply with the legal limitations imposed upon
emissions averaging by the Act (as discussed, for example, in the
preamble to the final HON regulation at 59 FR 19425 ff., April
22, 1994). Since no specific information on the technological or
administrative aspects of a potential emissions averaging program
for this source category has been supplied, EPA has not included
provisions for such an approach in the final rule.
2.3 CONTROL LEVEL
Comment: Comments were received by the Agency on the
reopening for comment of the issue of calculation of the MACT
"floor" level of control for all major sources under a separate
notice (59 FR 11018, March 9, 1994). These comments are included
2-4
-------
here for completeness. Two commenters stated that, to be
consistent with the intent of the Clean Air Act amendments of
1990 (CAAA) , the floor emission limitation should be based on the
88th percentile or the average emission limitation achieved by
all members of the best performing 12 percent. Basing the floor
on the average of the top performing 12 percent, they felt, is a
misinterpretation of the CAAA (IV-D-22, p. 20, App. F; IV-D-34,
p. 21). In contrast to these comments, another commenter said
that EPA is mandated by the CAAA to establish the floor at the
average emission limitation achieved by the best performing 12
percent of the existing sources, and is required to establish the
floor at this level regardless of cost considerations. An
alternative method in which EPA would ignore the top 11 percent
of the existing sources and focus only on the source at the 88th
percentile would be an erroneous interpretation of the floor (IV-
D-27, p. 1) .
Response: This issue affects all HAP source categories
generally, and EPA has issued its decision and provided a
complete discussion of its arguments in a Federal Register notice
(59 FR 29196, June 6, 1994) .
Comment: One commenter noted that control measures proposed
for loading racks and gasoline tank trucks had cost effectiveness
values in excess of $20,000 per Mg, which they felt was a figure
warranted only in extreme nonattainment areas. The commenter
believes that the interpretation of section 112(d)(3) of the Act
regarding "best controlled" and "best performing" could allow for
cost effectiveness considerations if the HAP's being controlled
are of low toxicity. The commenter requested that EPA consider
whether the term "best" allows for cost effectiveness
considerations and reconsider whether the loading rack and tank
truck portions of the proposal represent "best performing"
controls (IV-D-40, p. 1).
Response: Section 112(d)(3) of the Act requires that at
least the "floor" level of control, the average emission
2-5
-------
limitation achieved by the best performing 12 percent of the
existing sources, must be applied to existing major sources. The
EPA's interpretation of the Act is that more stringent, or "above
the floor" controls, may be applied after "taking into
consideration the cost of achieving such emission reduction, and
any nonair quality health and environmental impacts and energy
requirements" [section 112(d)(2) of the Act]. The Agency has
intentionally avoided establishing a specific "bright line"
threshold for acceptable cost effectiveness, due to the many
other variables that must be considered in selecting among
regulatory alternatives, and the differences among the various
source categories. The impacts of the final rule provisions were
revised and are summarized in Section 1.2.
Comment: One commenter felt that many of the proposed
requirements are not warranted based on emission reductions and
that unreasonably stringent controls are being required in non-
problem areas. Extending current NSPS requirements to all
facilities is more appropriate (IV-D-26, p. 1).
Response: Section 112(d) (3) of the Act provides specific
direction, as discussed earlier in this section, to be used by
EPA in regulating major sources of HAP's. These include
determining the floor control level for existing facilities in
all areas of the country. In cases where NSPS standards were
applicable and constituted the floor, such requirements (e.g.,
storage vessels and some parts of vapor collection and testing at
loading racks) were adapted for use in this MACT regulation.
2.4 OTHER
Comment: Section 63.429(b) of the proposed rule states:
"The authority conferred in §63.426 and §63.427(a)(5) will not be
delegated to any State." (These two provisions concern approval
of alternative control and monitoring methods.) One commenter
said that this provision may unnecessarily restrict EPA's
delegation of authority and should be removed from the rule (IV-
D-3 8, p. 6).
2-6
-------
Response: The commenter (a local air quality control
agency) did not specify why this provision might pose a problem.
The EPA believes that the authority to approve alternative means
of emission limitation and monitoring approaches should be
retained by the Administrator to ensure that the alternatives
achieve the control level requirements of the Act and are
consistent in all areas of the country. Section 112(h)(3) of the
Act is also clear in granting this authority to the
Administrator.
Comment: One commenter stated that EPA should not use the
term "Stage I" for cargo tank loading at bulk distribution
facilities. This terminology has created confusion among both
agencies and industry. Stage I should be reserved for the
filling of stationary storage tanks at service stations (IV-D-28,
p . 2 ) .
Response: The EPA has traditionally used the term "Stage I"
to apply to controls on processes at terminals, bulk plants, and
service station storage tanks that do not include vehicle
refueling, or Stage II, controls. The term is used here to
distinguish the control measures considered in this rule from the
Stage II controls under development in response to other parts of
the Act [sections 182(b)(3) and 202(a)(6)]. Since no other
comments opposing this nomenclature were received, and in order
to maintain consistency with the proposal documentation, EPA has
retained the same terminology in publishing the final rule.
Comment: One commenter felt that EPA is inconsistent in
allowing the use of methyl tert-butyl ether (MTBE), a HAP, as a
gasoline oxygenate while attempting to control HAP's with MACT
standards. The EPA should recommend a phase-out period for the
use of any HAP currently used to satisfy the Federal oxygenated
fuel program (IV-D-28, p. 2).
Response: In 1985, MTBE was placed on the Toxic Substances
Control Act's (TOSCA) list because it was the most largely
produced chemical in the U.S. In 1990, MTBE was placed on the
2-7
-------
HAP list primarily for two reasons: 1) the health effects
testing being performed at that time provided inconclusive
results and 2) the majority of the population was exposed to MTBE
due to its presence in gasoline.
Section 211(m) of the Act specifies a minimum oxygen content
of 2.7 weight percent for oxygenated gasolines, while section
211(k) specifies a minimum oxygen content of 2.0 weight percent
for reformulated gasolines. The Federal oxygenated fuels program
is designed to reduce tailpipe carbon monoxide (CO) emissions
occurring from motor vehicles in the wintertime. The Federal
reformulated fuels program is designed to reduce motor vehicle
emissions of toxic and tropospheric ozone-forming compounds
occurring throughout the year. MTBE is the most prominent
gasoline additive used to satisfy the Act's oxygenate
requirements.
Preliminary air quality monitoring data from November 1,
1992 through January 31, 1993 in 20 areas implementing the
oxygenated fuels program for the first time show a 95 percent
reduction in the number of days exceeding the CO standard (IV-J-
12). In addition, adding MTBE to gasoline reduces the ratio of
other toxic pollutants in gasoline (e.g., benzene, a proven human
carcinogen) which reduces consumer exposure to those pollutants
(IV-J-11).
Although it is outside the scope of this MACT rule to
restrict the use of MTBE as an acceptable gasoline oxygenate, EPA
is continuing to research the overall environmental impacts of
its various programs to create the best results achievable within
the constraints of current technology and the marketplace. Also,
recent health effects testing has not determined whether MTBE
presents a significant health threat beyond that of other HAP's
found in gasoline (e.g., benzene). Consequently, in this
rulemaking EPA is not recommending a phase-out period for the use
of any HAP currently used to satisfy the Act's gasoline oxygenate
requirements.
2-8
-------
Comment: One commenter felt that EPA should delay
implementation of this rule to at least 270 days and preferably 1
year after promulgation. Independent gasoline marketers are
facing numerous other environmental compliance costs over the
next several years, including the RFG program and the financial
responsibility mandates for underground storage tanks (IV-D-30,
p. 2, 9). Another suggested that EPA change the wording in the
proposed rule [at §63.422(b) and §63.423] to indicate that
existing facilities must comply within 3 years after promulgation
of the final rule (as was stated in the proposal preamble) rather
than February 8, 1997 (IV-D-36, p. 2). A third commenter stated
that EPA has illegally required subject facilities to comply with
the proposed equipment leak detection and repair (LDAR) standards
prior to promulgation. This commenter and one other suggested
that all facilities be given at least 1 year from promulgation to
comply with the equipment leak standards (IV-D-5, p. 1; IV-D-18,
p . 6) .
Response: Section 112(i) (3) of the CAAA requires sources to
be in compliance as expeditiously as practicable but in no event
later than 3 years from the effective date (promulgation) of the
standards. An extension of up to 1 additional year may be
granted if necessary for the installation of controls. As
discussed in Section 12.2, the initial applicability notification
has been extended to be within 1 year of the effective date, and
equipment leak standards have been changed to a visual program
that for most terminals and stations is already in place. The
EPA feels it is reasonable to require equipment leak provisions
in advance of the general compliance schedule since it is a
simple program with insignificant costs. Therefore, the final
rule requires the leak standards to be implemented within 1 year
of promulgation.
Notwithstanding these considerations, several compliance
dates were incorrectly printed in the proposed rule. In
§63.422(b) and §63.423, the date for compliance to be achieved
2-9
-------
with the loading rack emission limit and storage vessel standards
was shown as February 8, 1997 (3 years after proposal). This
date should have been shown as 3 years after the date of
publication of the final rule in the Federal Register.
Similarly, in §63.424(b)(2) the date for compliance with the LDAR
requirements was given as August 8, 1994 (180 days after
proposal), instead of 180 days after publication of the final
rule. Editorial corrections to these dates were provided in a
correction notice published at 59 FR 10461 on March 4, 1994.
Comment: One commenter expressed concern that the
implementation of the requirements in the proposal could pose a
fire safety risk at the regulated facilities. For example, vapor
collection pipelines can serve as conduits for flame travel from
one point to another (e.g., between connecting tanks) (IV-D-39,
p. 1) .
Response: The control technologies proposed in this rule,
including vapor collection and processing systems, leak detection
and repair, and storage tank emission control are similar to
control requirements that have been in place at bulk distribution
facilities for many years. The safety record of these techniques
has, to the Agency's knowledge, been very good. The commenter
did not supply any additional information identifying specific
concerns. However, EPA is concerned about safety issues
expressed by commenters. Several other commenters expressed
concerns about the safety aspects of the proposed vacuum assist
system, and these comments are discussed in Section 5.3.
2.5 REFERENCES
IV-J-11 Oxygenated Gasoline in Alaska. Information
Bulletin. U.S. Environmental Protection
Agency and Alaska Department of Environmental
Conservation. December 29, 1992.
IV-J-12 U.S. EPA Press Release. John Kasper,
Director, Press Services Division. March 11,
1993 .
2-10
-------
3.0 APPLICABILITY ISSUES
3.1 GENERAL
Comment: One commenter objected to the use of equations to
assess the rule's applicability to a source and thought that the
proposed equations should be removed from the rule. The
commenter stated that section 112(a) of the Act and §63.2 of the
General Provisions are very clear in stating what does and does
not constitute a major source, and that sources should be allowed
to use any credible method to satisfy the requirement for an
applicability determination. The commenter also stated that the
equations were derived from a worst-case scenario and overstated
HAP emissions (IV-D-7, p. 1).
Response: The applicability equations are screening tools
and are intended to provide facilities with a quick and easy
method to determine their potential major source status. The
EPA's intention was that a small area source could determine its
nonapplicability status through the use of the appropriate
equation and would not have to perform an emissions inventory.
The equations were intentionally designed to overstate HAP
emissions so that all potential major sources would be determined
to be major. The screening equations cannot provide an accurate
estimation of HAP emissions for all facilities. The EPA will
retain these equations in the final rule as a screening tool to
identify area sources. However, the equations can only be used
by a particular facility if that facility (entire plant site)
emits HAP's only from the emission points identified in the
equation. As discussed in Section 3.2, EPA has modified and
3-1
-------
refined the equations per some of the commenters'
recommendations. These refinements will, in general, lower the
estimation of HAP's and allow more area source facilities to be
determined as non-major in the initial screening process.
Comment: One commenter requested that EPA clarify whether a
bulk terminal is a major source if it meets only one of the
applicability tests in §63.420, or whether it must meet both
tests (IV-D-20, p. 2).
Response: The EPA's intention is that all facilities use
the equations as an initial screening tool to determine their
potential major source applicability if HAP's at the plant site
are only from the emission points covered in the appropriate
equation. Any facility that uses the appropriate equation
correctly and finds itself not subject to the rule, does not have
to perform an emissions inventory as further documentation. Any
source which the screening equation determines to be a potential
major source may, if they choose, perform an emissions inventory
for a more precise estimation of HAP emissions. If the results
of the inventory show that in fact the facility (entire plant
site) is not a major source, the facility is not subject to the
rule. The results of the emissions inventory would therefore
supersede the results of the screening equations. If the results
of the inventory determine that the facility i_s a major source,
then the rule will be applicable to that facility.
3.2 SCREENING EQUATIONS
Comment: Several commenters felt that the development of
the applicability screening equations was not supported or
explained by EPA. As a result, the equations appear to be
entirely arbitrary (IV-D-14, p. 4; IV-D-15, p. 4; IV-D-36, p. 1).
Another commenter who has experience preparing emissions
inventories for bulk gasoline terminals in Texas said that, for
several terminals that do not exceed the 10/25 tons of HAP's per
year threshold, the applicability equation incorrectly indicates
that many of these terminals are major sources (IV-D-11, p. 1).
3-2
-------
Response: The development of the applicability screening
equations was discussed in the preamble to the proposed
standards, on page 5877, and was explained in more detail in a
memorandum that was included in the rulemaking docket
(item II-B-23). These equations were not arbitrary, but were
developed specifically to identify facilities that have the
potential to emit less than 10/25 tons per year of HAP and to
reduce the amount of effort needed to perform applicability
determinations. However, if a facility has other HAP emission
sources not considered in the equation, the equation will under-
predict emissions and cannot be used to determine if the facility
is a major source. In response to the comments, the equations
have been retained in the rule but have been revised to
accommodate the concerns of commenters and to make them more
accurate in their function as a screening tool. These
modifications and the new equations are discussed in detail in
the responses to the following comments and in a memorandum in
the docket (item IV-B-1).
Comment: Many commenters expressed support for the use of
screening equations as an aid in determining major source
applicability, but suggested the following modifications (shown
as numbered comments) to make the equations more representative:
1) Instead of using "worst-case" HAP-emitting gasolines to
derive the constants in the equations, use average parameters to
promote consistency between the equations and the rule (IV-D-7,
p . 2 ) .
2) Include an adjustment factor for facilities that do not
handle gasoline oxygenated with methyl tert-butyl ether (MTBE) or
provide an additional set of equations based on the HAP content
of conventional gasolines (IV-D-7, p. 2; IV-D-15, p. 8; IV-D-18,
p. 4; IV-D-22, p. 53; IV-D-34, p. 7; IV-D-35, p. 1; IV-F-1, p.
13) .
Response: At proposal, EPA developed the screening
equations based on a HAP to VOC ratio that was determined to
3-3
-------
represent the average MTBE content in reformulated and oxygenated
gasolines (see Table 3-2 of BID, Volume I). The HAP content used
to derive the equations does not represent the "worst-case" HAP
to VOC ratio. As seen in Test ID# Gil through G14 of Table C-3
of BID, Volume I, the MTBE content in gasoline ranged from 11.8
to 16.3 percent. Based on these data, EPA made an assumption
that the average MTBE content of reformulated and oxygenated
gasolines was 11.9 percent, which is slightly higher than the
lowest percentage found in the data. In addition, EPA assumed
that most facilities that handle higher MTBE content oxygenated
gasolines would also handle the lower MTBE content reformulated
gasolines. This approach is consistent with the Agency's intent
to avoid underestimating emissions in this screening process,
which could allow a major source to be deemed an area source and
thus improperly escape applicability of this rule. Facilities in
any case will have the opportunity to perform a full emissions
inventory in order to make a more accurate determination of their
status.
The EPA agrees that the proposed emission factors
overestimate HAP emissions from facilities handling gasoline
without MTBE. As a result, EPA has included an adjustment factor
in the screening equations for facilities in this situation.
Facilities that handle, or anticipate handling, anv oxygenated or
reformulated gasoline containing MTBE as a component will not use
the adjustment factor in performing the calculations.
3) The EPA's assumption that annually certified and tested
cargo tanks with vapor control lose 10 percent of the displaced
vapors through leakage while loading is too high. This value
should be reevaluated when API completes its study (IV-D-7, p. 2;
IV-D-14, p. 6; IV-D-15, p. 8; IV-D-18, p. 4, IV-D-22, p. 55; IV-
D-34, p. 7, IV-D-35, p. 1; IV-F-1, p. 13).
Response: As discussed in Section 7 and Appendix A, the
assumption that cargo tanks in an annual test program leak 10
3-4
-------
percent has been reevaluated. The EPA has calculated a new
leakage rate that is much lower than the proposed figure, and the
revised calculations are shown in Appendix A and the value is
reflected in the revised equation in §63.420(a)(1).
4) Fixed-roof tanks connected to vapor recovery systems,
which may or may not contain internal pans, emit virtually no
HAP's and should either be excluded from the equations or a term
should be added to represent and quantify the low emissions
occurring from such tanks (IV-D-14, p. 6; IV-D-15, p. 8; IV-D-18,
p. 3; IV-D-28, p. 2; IV-D-34, p. 7, IV-D-35, p. 1; IV-D-35, p. 1;
IV-F-1, p. 13).
Response: The EPA agrees with the commenters and has added
a new expression, (1 - CE), to both screening equations. The
term "CE" represents the control efficiency of the control device
used to process vapors from the fixed-roof tank. The value of CE
must be documented by the facility as meeting the definition of
federally enforceable in subpart A of 40 CFR part 63 (General
Provisions). If the facility is not controlling emissions from
its fixed-roof tanks using a vapor control device, a value of
zero will be entered for the term "CE."
5) The factor Tes should be redefined to include only those
external floating roof tanks with seals that handle gasoline (IV-
D-7, p. 2 and IV-D-18, p. 3).
Response: The EPA agrees with this comment. The factor Tes
has been restated to include only those external floating roof
tanks with secondary seals that handle gasoline. This
qualification was inadvertently left out of the definition of Tes
at proposal.
6) The emission factors for pump seals and valves are too
high based on recent data collected at marketing facilities (IV-
3-5
-------
D-14, p. 6; IV-D-15, p. 8; IV-D-18, p. 4; IV-D-22, p. 53, IV-D-
35, p. 1) .
Response: The EPA has evaluated the new data and agrees
with this comment. The emission factors for pump seals and
valves have been revised as discussed in Section 9 and Appendix
C.
7) The equations should provide credits for facilities that
use LDAR programs, vacuum assist, or handle only conventional or
normal gasoline (IV-D-11, p. 2; IV-D-20, p. 3).
Response: As discussed in Section 9 and Appendix C, data
provided by industry show that the use of visual inspection
programs is just as effective as the use of instrument LDAR in
identifying equipment leaks at marketing terminals and breakout
stations. As a result, EPA will not grant credits to facilities
that currently use an LDAR program. The EPA has decided to not
require vacuum assist as explained in Section 5.3, due to Agency
concerns about the control effectiveness of vacuum assist
technology at bulk terminal loading racks. As a result, EPA also
will not provide emission credits for any facility using vacuum
assist technology.
8) Emission standards more stringent than the Federal NSPS
(40 CFR part 60, subpart XX) limit (35 mg/liter) should be
recognized (IV-D-18, p. 3; IV-D-28, p. 2, IV-D-34, p. 7; IV-D-38,
p . 2 ) .
Response: The term "EF" in the screening equation for bulk
terminals applies to any federally enforceable emission standard
in effect for the vapor processor. The concept of "federally
enforceable," defined in the General Provisions (§63.2), allows
emission standards or limitations more stringent than the NSPS
limit (see Section 6.2) .
3-6
-------
9) The screening equations should be modified beyond API's
recommendations to include tanks that store MTBE for infrequent
periods and durations (IV-D-12, p. 1).
Response: The EPA does not intend to regulate under this
rule storage vessels that store only MTBE or any other gasoline
component or additive. All the other non-gasoline liquids such
as MTBE will be studied under the forthcoming NESHAP source
category of "Non-Gasoline Liquid Distribution" under section 112
of the Act.
10) The equations do not provide a mechanism for
calculating emissions from "swing" tanks (tanks which store
gasoline only half the time). Guidance is requested on how to
estimate emissions from these tanks (IV-D-20, p. 3).
Response: In keeping with the intent of these equations as
an emission estimation screening tool, EPA has made the
simplifying assumption that vessels storing gasoline for any
period or periods during a year will be assumed to store gasoline
year round. As a result, the emissions from "swing" tanks will
be estimated in the same way as for tanks that store gasoline on
a continuous basis. Owners and operators should use the
emissions inventory approach, as specified in §63.420(a)(2) and
(b)(2), if these assumptions lead to a significant overestimation
of HAP emissions at their facility.
11) One commenter asked why the fugitive emission factors
and storage tank emission factors are different in the equations
for pipeline breakout stations and bulk terminals (IV-D-31, p.
4) .
Response: The emission factors for equipment components in
gasoline service at the two types of facilities are the same;
however, the amount of time the components are presumed to be
handling gasoline is different. As discussed in Chapter 5 of
BID, Volume I (III-B-1), products other than gasoline are sent
3-7
-------
through pipelines and stored at breakout stations. As a result,
many tanks, pumps, and valves are not in constant gasoline
service. The factors in the proposed screening equations were
adjusted on the basis of "equivalency factors" to reflect the
number of equipment components in gasoline service.
Additionally, the storage tank emission factors are different
because the size and turnover rates are not the same at the two
facility types. The development of these "equivalency factors"
is discussed in a memorandum prepared at proposal (II-B-23).
3.3 EMISSIONS INVENTORY
Comment: As a supplement to the emissions estimation
screening equations, §63.420(a)(2) and (b)(2) of the proposed
rule exempted those facilities "for which the owner or operator
has documented to the Administrator's satisfaction that the
facility is not a major source as defined in section 112(a)(1) of
the Clean Air Act." The proposal preamble on page 5877 indicated
that an "emissions audit" would have to be performed to satisfy
these provisions. One commenter felt that the rule provisions
should specifically state that the estimation of emissions for
the applicability determination is to be accomplished by means of
an emissions audit, as was stated in the preamble (IV-D-41, p.
3). Several other commenters found the term "emissions audit"
confusing, and questioned what EPA would consider acceptable for
demonstrating applicability. Some suggested that the familiar
term "emission inventory" be substituted because emission
inventories are common requirements and procedures are in place
under many State programs. Others requested that EPA define or
provide an approved methodology for conducting the emissions
audit. One commenter said that the public should have an
opportunity to comment on this guidance prior to EPA finalizing
the rule (IV-D-6, p. 1; IV-D-7, p. 3; IV-D-14, p. 6; IV-D-15, p.
4; IV-D-18, p. 4; IV-D-22, p. 62).
One commenter thought that EPA should eliminate the
requirement that a source determine its applicability status by
3-8
-------
means of an emissions audit. They felt such a requirement is
unnecessary and contrary to prohibitions in Executive Order 12866
since major sources, which are subject to Part 70 permitting, are
already required to determine their applicable regulatory
requirements and identify them in their permit applications (IV-
D-7, p. 3).
Response: In describing the formal means of documenting a
facility's major or area source status as an "emissions audit" on
pages 5870 and 5877 of the preamble, EPA was referring to the
calculation of a facility's potential to emit HAP's considering
federally enforceable controls. Such calculations are similar to
those already being prepared under many existing Federal and
State control programs. Therefore, the intent of the Agency was
in accord with the thoughts of the commenters. The discussion in
the preamble and the requirements in the final rule are intended
to clarify and simplify compliance with the rule and are not
known to be contrary to provisions of the part 70 permitting
requirements. The EPA feels that guidance on performing HAP
emissions inventories is not needed since the preparation of such
inventories is standard practice. The activities undertaken in
response to part 70 requirements are applicable and may relieve
the majority of the burden of fulfilling this inventory.
3.4 CHANGE OF FACILITY STATUS
Comment: Two commenters were concerned that the proposed
rule did not address the issue of a facility changing its
applicability status. A provision should be added allowing a
facility to be reclassified as an area source for reasons such as
reduced throughput and changes in handling oxygenated gasoline.
In that case, the facility should be relieved of the MACT
requirements (e.g., recordkeeping). Conversely, if a facility
that was once an area source is redesignated as a major source,
that facility should have 3 years to comply (IV-D-15, p. 11; IV-
D-22, p. 69) .
3-9
-------
Response: If the situation occurs that a source currently
complying with MACT becomes an area source due to conditions such
as reduced throughput, changes in type of gasoline handled, etc.,
the source must continue to follow all the MACT requirements to
ensure that the control equipment is maintained and working
properly and that emissions are being reduced. Facilities would
already have equipment installed and permits issued; therefore,
it is not logical that a possibly short-term situation should
allow the facility to disband control equipment and permit
conditions. However, the Agency is currently reviewing its broad
policy on all MACT standards related to changes in status from
major to area source as expressed by commenters. Agency
decisions on this issue will be provided in the near future.
3.5 SMALL SOURCE COVERAGE
Comment: Several commenters expressed support for EPA's
decision to not regulate area sources under the MACT proposal.
They pointed out that area sources do not pose a significant
threat to human health and therefore do not warrant regulation
(IV-D-2, p. 4; IV-D-3, p. 2; IV-D-8, p. 1; IV-D-14, p. 1; IV-D-
15, p. 11; IV-D-16, p. 3; IV-D-22, p. 47; IV-D-30, p. 3; IV-D-35,
p. 2; IV-F-1, p. 13). However, one commenter disagreed with
EPA's decision to not regulate area sources, stating that Stage I
vapor balance control at service stations is necessary to close
the loop on emissions from gasoline marketing sources. They
cited a report by the Northeast States for Coordinated Air Use
Management (NESCAUM) on exposure and health risk from service
station emissions as providing the necessary justification for
these controls (IV-D-21, p. 1).
Response: The EPA agrees with the first group of commenters
that area sources in the gasoline distribution source category do
not, based on the present data base, present a sufficient risk of
adverse effects to human health or the environment to warrant
regulation under these MACT standards. These sources are being
investigated, however, as part of the study of area sources in
3-10
-------
urban areas under section 112(k) of the 1990 amendments. The
NESCAUM report is being reviewed as part of that study.
Comment: One commenter said that some large bulk plants
exceed the 10/25 tpy major source threshold and should be
regulated (IV-D-38, p. 6).
Response: Since this commenter did not provide specific
data on the numbers, locations, or operational characteristics of
these facilities, EPA cannot respond directly. However, any
facilities of this nature would likely be regulated under the
case-by-case MACT requirements in section 112(j)(5) of the Act.
Comment: One commenter felt that EPA has not met its
statutory requirement to "promulgate regulations establishing
emission standards for each category or subcategory of maior
sources and area sources of hazardous air pollutants..." [CAAA
section 112(d)(1)]. The EPA's proposed rule regulates only 23
percent of all bulk terminals and pipeline breakout stations. In
order to comply with the CAAA of 1990, EPA should cover at least
50 percent of the emissions from the category. The EPA does not
discuss what portion of the total emissions are represented by
this 23 percent. If the rule does not cover at least 50 percent
of the facilities/emissions by bringing in smaller sources,
smaller bulk terminals will be at an advantage since they can
avoid the major capital investments that the larger terminals
will incur. Since smaller terminals can avoid regulation and the
capital investments for environmental compliance, an increasing
number of small uncontrolled terminals will be built and larger
terminals will go out of business, increasing emissions as a
consequence (IV-D-2, p. 4).
Response: In accordance with the requirements of the Clean
Air Act, these MACT standards regulate all existing and new major
source gasoline distribution facilities. The Administrator is
required under section 112(c)(3) to list "each category or
subcategory of area sources which the Administrator finds
presents a threat of adverse effects to human health or the
3-11
-------
environment (by such sources individually or in the aggregate)
warranting regulation ..." As indicated in a previous response,
area sources have not been determined to present a threat of
adverse effects under the requirements of section 112. Also, EPA
does not agree with the commenter that the Clean Air Act requires
50 percent of the total emissions to be addressed by these
standards. In any case, it would be difficult to construct an
equitable applicability approach (population of facilities to be
regulated) that would ensure that 50 percent of the emissions
were covered. Such an approach would necessarily have to include
some portion of the area sources and exclude the rest.
According to Agency calculations, even the smallest
terminals (by definition, gasoline throughput greater than 75,700
liters per day) would be major HAP sources if they did not
utilize at least some of the controls that are required by the
NESHAP. Many existing small terminals may, in fact, be exempt
from the requirements of the rule, but only because they already
have federally enforceable controls in place. However, these
terminals may not currently, and may not in the future, enjoy a
competitive advantage since there are economies of scale in
terminal operations. Furthermore, terminals, regardless of size,
are to some extent insulated from highly intense competition
because terminal distribution markets tend to be local. Finally,
new major source terminals are subject to the same requirements
regardless of size. These requirements, due to economies of
scale, are likely to have a greater impact on new small terminals
than large terminals.
Another point is that the overall cost of this NESHAP is not
a substantial percentage of the total cost to build and operate a
terminal. Therefore, the cost incentive, if one indeed exists,
is likely to be too small to create a widespread conversion of
large terminals to small terminals.
3.6 POTENTIAL TO EMIT
3-12
-------
Comment: One commenter felt that the rule was not clear in
explaining whether a facility's major source applicability is
determined from "potential to emit" (PTE) or actual emissions and
asked for clarification (IV-D-6, p. 1). Several commenters who
interpreted the rule to indicate that PTE should be used
expressed disagreement with EPA, and believed that basing major
source applicability on a source's PTE would draw into the
regulation many more sources than EPA has anticipated. They said
EPA should recognize that there are inherent limits in the
operational parameters (throughput, etc.) of gasoline
distribution facilities, and should base major source
determination on a source's actual emissions or at least a more
reasonable gasoline loading potential (IV-D-4, p. 2; IV-D-7, p.
3; IV-D-15, p. 4; IV-D-16, p. 2; IV-D-22, p. 63; IV-D-34, p. 21).
The American Petroleum Institute (API) recommended a scheme for
categorizing facilities based on actual emission rates that they
felt would alleviate the "potentially drastic consequences" of
applying the PTE definition. These categories are: I - actual
emissions exceed the major source threshold (10/25 tpy), so the
source is subject to all the provisions of the rule; II - actual
emissions are greater than 80 percent but less than 100 percent
of the major source amounts. The facility would have to certify
its area source status by obtaining a permit with enforceable
limits, submit annual certification of emission rates, and notify
of any change that could increase HAP emissions; III - actual
emissions are greater than 50 percent but less than or equal to
80 percent of the major source definition. The facility would
have to submit annual certification and provide notification of
any change; IV - actual emissions are 50 percent or less of the
major source cutoffs. This facility would only have to provide
notification of any changes affecting emissions (IV-D-16, p. 2;
IV-D-22, p. 66). Another commenter suggested that applicability
should be based on a combination of the potential to emit of the
3-13
-------
vapor recovery system and the actual emissions of the storage
tanks and fittings using EPA's TANKS software (IV-D-6, p. 1).
Response: At proposal, EPA did not use the term "potential
to emit" in the preamble discussion or in the proposed rule.
However, the proposed rule and discussion in the preamble did
reference the General Provisions (40 CFR part 63, subpart A),
which includes a definition for PTE. This definition of PTE is
as follows:
"Potential to emit means the maximum capacity of a
stationary source to emit a pollutant under its
physical and operational design. Any physical or
operational limitation on capacity of the stationary
source to emit a pollutant, including air pollution
control equipment and restrictions on hours of
operation or the type or amount of material combusted,
stored, or processed, shall be treated as part of its
design if the limitation or the effect it would have on
emissions is federally enforceable."
Terminals and breakout stations have many limitations that
affect emissions and some of these can vary according to gasoline
demand. Industry provided data showing many methods to calculate
maximum capacity, including total tank capacity, loading rack
pumping capacity, feeder pipeline pumping rate, etc. (IV-E-2).
Each of those methods of calculating capacity results in
different and conflicting PTE results. The EPA has decided to
provide an approach in the final rule that provides the facility
an opportunity to set some operational and physical limitations
that best fit its own operation only if all the HAP's emitted are
from affected gasoline operations. The EPA considered allowing
gasoline terminals and pipeline breakout stations with additional
HAP emissions from non-gasoline sources at the plant site to use
this approach. However, EPA believes that covering all
situations and other source categories under this rule would be
too complex and uncertain. Therefore, those sources would have
3-14
-------
to obtain enforceable conditions and limitations outside the
provisions of this rule.
Under this approach for plant sites emitting HAP's only from
affected gasoline operations, the bulk terminal or pipeline
breakout station facility can establish its potential to emit
through a combination of operational and physical limitations
that are otherwise federally enforceable outside the context of
this rule or that are made enforceable through compliance with
parameters included in the screening equation in this rule.
Examples of allowable federally enforceable limitations and
conditions are provided in the definitions section of the General
Provisions (§63.2). An example of limitations at bulk terminals
and breakout stations that are required to meet the definition of
federally enforceable outside the context of this rule are
emission limits on vapor processors that process emissions from
storage vessels and cargo tanks. Recordkeeping and reporting
requirements will be used to monitor compliance with all
limitations. Thus, the final rule allows the facility to limit
PTE by complying with the approved values of the physical or
operational parameters contained in the emission screening
equations, such as maximum throughput. This provides the
facility the most flexibility in operations without
overestimating PTE.
The proposed rule required facilities to either use a
specific emission estimation screening equation or prepare an
inventory of emissions for determination of their major or area
source status. The proposal further allowed area source
facilities to report their applicability findings and
calculations in their initial notifications to the Agency
[required under §63.9(b)]. After review and acceptance by the
Agency, the facility would have been considered an area source
and would not be subject to the control requirements of the rule.
Changes to the final rule establish certain facility parameters
used in the emission screening equation as new "physical or
3-15
-------
operational limitation(s) on the capacity of the stationary
source to emit a pollutant." Upon request, the owner or operator
of the bulk gasoline terminal or pipeline breakout station will
be responsible for demonstrating compliance with the facility's
applicability determination, including all assumptions,
limitations, and parameters used to calculate potential to emit
HAP's.
To monitor these limitations, certain facilities are
required in the final rule to annually certify that these
facility parameters are not being exceeded. It would be
burdensome and unnecessary for all facilities below the emissions
threshold for major sources to provide detailed reports and
records, and annually certify that changes have not occurred. As
suggested in the API comments, only facilities within 50 percent
of the emissions threshold for major sources will be required to
submit a detailed report of these calculations and assumptions
used in the calculations in an initial report, and then provide
annual certification that the established facility parameters are
not being exceeded. The remaining facilities will need to retain
a record at the facility of these calculations and notify the
Administrator of the use and results of the emission screening
equation. These records would remain at the facility for
inspection by the Administrator. If the PTE "limitations" are
exceeded or if the facility fails to keep records or report as
required, the facility will be in violation of this rule and may
in some cases be considered a major source and be subject to the
emission standards of the rule.
The final rule also requires the reports submitted
containing those limitations and certifications to be approved by
the Administrator and made available for public inspection. The
notifications and reports documenting those limitations must be
submitted to the Administrator within 1 year of the effective
date. The final rule allows facilities to change these
3-16
-------
parameters after submittal of the revised calculations and
approval by the Administrator.
If the facility becomes an area (nonmajor) source by
complying with the PTE enforceable limitations and conditions
established under the final rule, then the emission control
requirements of this rule would not apply. Furthermore, for
purposes of section 112 of the Act, it would not be a regulated
area source that would be required to have an operating permit
under 40 CFR part 70. In other words, being subject to the PTE
limitations in this rule does not in and of itself make the
facility subject to 40 CFR part 70. However, there may be other
reasons that the stationary source is required to comply with 40
CFR part 70.
The EPA believes the mechanisms provided in the final rule
for limiting PTE provide adequate safeguards for this source
category. However, EPA is still evaluating whether the general
approach taken in this rule will be appropriate for other source
categories.
3.7 REFINERY BULK TERMINALS
Comment: One commenter requested that, for bulk terminals
contiguous to refineries, EPA clearly define the separation
between terminal storage tanks and refinery storage tanks. These
terminals are usually fed from tanks located within the refinery
itself, often thousands of feet from the terminal. Refinery
tanks will be regulated by the NESHAP for petroleum refineries.
The commenter felt that tanks not located at the terminal itself
should be considered part of the refinery for the purposes of
regulation (IV-D-5, p. 4).
Several commenters were of the opinion that EPA should
distinguish the association and applicability of the gasoline
distribution MACT rule from the refinery MACT rule. Many
commenters believe that only cargo tank loading racks and cargo
tank leakage should be regulated at terminals that are
"contiguous to" refineries, whereas tankage and equipment leakage
3-17
-------
emissions should be regulated under the refinery MACT rule. One
suggested method to distinguish whether facilities are subject to
the refinery MACT rule or the gasoline distribution MACT rule is
to consult the applicable Standard Industrial Classification
(SIC) codes already assigned to these facilities (IV-D-2, p. 5;
IV-D-3, p. 2; IV-D-4, p. 2; IV-D-7, p. 3; IV-D-8, p. 2; IV-D-15,
p. 8; IV-D-17, p. 4; IV-D-22, p. 60).
Response: Terminals and pipeline facilities contiguous to
refineries are of two types. First, there are terminals and
pipeline facilities that are located within a contiguous area and
under common control, but are managed by the "marketing" or
"distribution" departments, though they are located on the same
property as a refinery. The other type are terminals and
pipeline facilities located among the refinery process units and
storage tanks and managed by the "refinery" management
departments. SIC codes are assigned and are currently being used
by these facilities to distinguish between equipment. Industry
commenters expressed a need to retain this separation because
they often have separate management for maintenance, capital
improvements, personnel, and operation of the assigned equipment.
This separation would keep the management of the air pollution
equipment under the same management structure as the surrounding
process equipment. The Agency agrees with the commenters that
maintaining this structure would be beneficial, because it will
increase the management of proper operation and maintenance of
the control equipment, decrease compliance costs, and improve the
reporting and recordkeeping and enforcement of this rule.
Since a final rule cannot refer to another standard that has
not been promulgated as a final rule, this change is not
incorporated into the final gasoline distribution rule. The
Agency, however, plans to carry out this change by modifying this
rule at the promulgation of the refinery MACT standards. The
refinery MACT standards (proposed at 59 FR 36130, July 15, 1994)
contain different requirements for equipment leaks and compliance
3-18
-------
schedules for storage tanks. The Agency will assess the
differences between these two rules after it considers public
comments on the refinery MACT proposal and develops the final
refinery MACT standards. Meanwhile, all provisions of the
gasoline distribution rule will be implemented as they are being
promulgated, since there are no requirements in the gasoline
distribution rule that must be implemented before the scheduled
promulgation of the refinery MACT standards. Independent of the
SIC code designation decision discussed above, EPA will make a
decision in the refinery MACT rule on the use of emissions
trading or averaging between the collocated gasoline distribution
and refinery sources.
Comment: Several commenters disagreed with the provision
that all bulk terminals contiguous to or under common control of
a major source petroleum refinery, regardless of the size of the
terminal, would be subject to the regulation. They felt that
such a provision is unfair and would put refinery terminals at a
competitive disadvantage with respect to bulk plants and non-
refinery terminals. They said that terminals should be regulated
according to their impact on the environment, not on their
location or ownership as proposed (IV-D-2, p. 4; IV-D-4, p. 2;
IV-D-5, p. 5; IV-D-17, p. 3; IV-D-29, p. 2; IV-D-34, p. 7). One
commenter requested that EPA not regulate small loading racks
that have a throughput of less than 20,000 gallons per day. Such
racks exist at refineries and are used infrequently, and
continuous emission monitoring at such racks would not be cost
effective (IV-D-8, p. 2). Another requested that the proposed
regulation apply only to storage operations not associated with
refineries, because refineries are not the major areas for
gasoline storage and distribution. Also, refinery operations are
already going to be covered by the refinery MACT rule (IV-D-10,
p. 1) .
Response: The Act requires the Agency to set standards for
major source categories of HAP emissions. Gasoline distribution
3-19
-------
facilities collocated with other HAP emission sources, such as
refineries, are considered part of a major source operation.
Refinery terminals with physical and operating characteristics
equal to non-refinery terminals may be non-uniformly impacted as
a result. However, the overall cost of the NESHAP is not a
substantial percentage of the total cost to build and operate a
terminal. Therefore, the cost advantage to unaffected non-
refinery terminals is likely to be too small to create widespread
demand shifts from refinery terminals to non-refinery terminals.
3.8 OTHER
Comment: Section 63.420(d) of the proposed rule stated the
following:
"The owner or operator of a bulk gasoline terminal
or pipeline breakout station subject to the provisions
of this subpart that is also subject to applicable
provisions of 40 CFR part 60, subparts K, Ka, Kb, W,
XX, and GGG of this chapter, or 40 CFR part 61,
subparts J and V of this chapter, shall comply only
with the provisions in each subpart that contain the
most stringent control requirements for that facility."
One commenter suggested that this section could cause some
confusion, and that the section should be modified to clarify
that facilities also subject to other Federal air standards shall
comply with the most stringent control requirements in each of
those subparts for facilities located within extreme, severe, or
serious ozone nonattainment areas (IV-D-36, p. 1).
Response: Proposed section 63.420(d) was included because
this MACT rule cross-references specific portions of these other
Federal standards. Facilities already required to comply with
all portions of those standards are not allowed to comply with
(in some cases) less stringent MACT standards. MACT standards
may be less stringent: compliance timing on tanks, the size of
affected facilities (terminals), etc. Since this standard does
3-20
-------
not cross-reference State regulations, it is inappropriate to
cite State or local regulations.
Comment: One commenter said that loading losses may occur
at pipeline breakout stations and should be included in the
NESHAP. Additionally, underground and pressure tanks have the
potential to emit HAP's and should not be neglected by this
regulation. Factors for such tanks should be included in the
applicability equation (IV-D-38, p. 6) .
Response: The EPA agrees with the commenter and has
clarified the final rule to include gasoline loading operations
that occur at facilities that include both a breakout station and
bulk terminal. Due to the complexity of such facilities, the
"simple" applicability screening equations could not be reworked
to include this situation and still retain the desired
simplicity. The final rule is modified such that any facility
that contains both a pipeline breakout station and gasoline
loading racks is not allowed to determine the applicability of
the rule using the equations in §63.420(a)(1) and (b)(1), but is
required to determine applicability based on the results of an
emissions inventory approved by the Administrator.
Subpart Kb does not exclude underground storage tanks (UST).
Under this rulemaking, UST's are viewed as fixed-roof tanks. As
required in subpart Kb, fixed-roof tanks have the option of being
controlled with rim seal controls/fitting controls or a closed
vent system and control device. For the purposes of this
rulemaking, major source terminals having UST's with a capacity
greater than or equal to 75 cubic meters must control such tanks
with a closed vent system and control device. As stated in
§60.110b(d)(2), pressure vessels are excluded from regulation if
the vessel is designed to operate in excess of 204.9 kPa and
without emissions to the atmosphere. Thus, any pressure vessel
storing gasoline or mixtures that contain gasoline is excluded
from the regulation provided the vessel meets the requirements
stated in §60.110b(d) (2) .
3-21
-------
The EPA believes that few underground storage tanks and
pressure vessels are used to store gasoline at bulk terminals.
In order to keep the applicability screening equations simple,
EPA has not included factors to account for emissions occurring
from UST's or pressure vessels at the affected facilities.
3.9 REFERENCES
II-B-23 Memorandum from Johnson, T., PES, Inc., to
Shedd, S., Chemicals and Petroleum Branch,
U.S. Environmental Protection Agency,
Research Triangle Park, NC. October 29,
1993. Development of applicability
equations.
III-B-1 Gasoline Distribution Industry (Stage I) -
Background Information for Proposed
Standards. EPA-453/R-94-002a. U.S.
Environmental Protection Agency, Research
Triangle Park, NC. January 1994.
IV-B-1 Memorandum from Johnson, T., PES, Inc., to
Shedd, S., Chemicals and Petroleum Branch,
U.S. Environmental Protection Agency,
Research Triangle Park, NC. August 31, 1994
Revised applicability equations.
IV-E-2 Memorandum from LaFlam, G., PES, Inc., to
Shedd, S., Chemicals and Petroleum Branch,
U.S. Environmental Protection Agency,
Research Triangle Park, NC. June 8, 1994.
EPA meeting with the American Petroleum
Institute.
3-22
-------
4.0 DEFINITIONS
4 . 1 LEAK
Comment: Several commenters stated that EPA has provided no
justification for the leak definition of 500 ppm above background
indicated in §63.425(a) of the proposed rule. They felt that
the concentration defining a leak should be the same as the NSPS
leak definition of 10,000 ppm (IV-D-5, p. 3; IV-D-17, p. 3; IV-D-
18, p. 9; IV-D-20, p. 9; IV-D-22, p. 81; IV-D-29, p. 4; IV-D-34,
p. 18).
Response: The 500 ppm instrument reading referred to by the
commenters applies during the performance test of the vapor
collection and processing systems. The selection of 500 ppm as
the instrument reading defining a leak was based on §61.242-11,
the standard for closed vent systems and control devices in 40
CFR 61, subpart V, the NESHAP for equipment leaks (fugitive
emission sources). This regulation defines "closed-vent system"
as "a system that is not open to atmosphere and that is composed
of piping, connections, and, if necessary, flow-inducing devices
that transport gas or vapor from a piece or pieces of equipment
to a control device." Both the proposed standards and §61.245(c)
prescribe Reference Method 21 as the accepted method for
performing the leak testing and determining compliance with the
500 ppm requirement. At the time 40 CFR 60, subpart XX, the NSPS
for bulk terminals, was promulgated, the equipment leak standard
was 10,000 ppm and in order to be consistent, the same value was
selected for §60.503(b) of subpart XX. However, the equipment
leak definition was subsequently lowered to 500 ppm. Therefore,
4-1
-------
EPA feels that the leak threshold applicable to equipment leaks
in subpart V should also be applied to closed vent systems under
this MACT standard.
The EPA believes that the number of components in the vapor
collection system is small, performance tests should only be
performed under leak-free conditions, and the difference in costs
to repair leaks would be minimal under the 500 ppm definition,
while the maximum reduction of HAP emissions would be
accomplished. Therefore, the 500 ppm leak definition has been
retained in the final rule for use during the emission testing of
the vapor collection and processing systems.
4.2 GASOLINE
Comment: Two commenters felt that aviation and reference
(research) fuels should be excluded from the rule, by defining
"gasoline" to specifically refer to road-use motor gasoline (IV-
D-4, p. 2; IV-D-22, p. 60). The EPA should also clarify that
gasoline blend components are not, by themselves, "gasoline."
Another commenter said that EPA should provide a definition for
gasoline that would clarify the rule's applicability to storage
vessels storing interface mix material at pipeline breakout
stations (IV-D-26, p. 2).
Response: The definition incorporated by reference at
proposal was the same as the definition promulgated in the NSPS
(40 CFR part 60, subpart XX) for bulk gasoline terminals on
August 8, 1983 (48 FR 37590). From §60.501 of this NSPS, the
definition of "gasoline" is "any petroleum distillate or
petroleum distillate/alcohol blend having a Reid vapor pressure
of 27.6 kilopascals or greater which is used as a fuel for
internal combustion engines" (27.6 kilopascals = 4 psia). This
NESHAP is intended to cover transfer and storage of gasoline and
of mixtures that contain gasoline. Some aviation fuels are
considered gasolines; however, jet fuels such as jet naphtha (JP-
4) or jet kerosene have RVP's well below the cutoff of 4 psia.
However, blend components that are stored separately from
4-2
-------
"gasoline" are not intended to be covered if they do not meet the
use criterion as a "fuel for internal combustion engines." For
tanks at pipeline breakout stations that handle intermix material
(the combined product mix at the interface between different
products conveyed in the pipeline), it would not be feasible to
specify in advance the percentage of gasoline that may be stored
in a particular tank. Therefore, any storage vessel of
sufficient size that is expected to contain gasoline (even as one
of possibly several components) is considered to be a gasoline
storage vessel and is subject to the control requirements of
§63.423 .
Due to these considerations, EPA has retained the same
definition for gasoline in the final rule.
4.3 GASOLINE TANK TRUCK
Comment: The proposed rule included the following
definition for "gasoline tank truck":
"Gasoline tank truck means a delivery tank truck
or railcar used at bulk gasoline terminals which is
loading gasoline or which has loaded gasoline on the
immediately previous load."
One commenter felt that EPA should delete the phrase, "or which
has loaded gasoline on the immediately previous load, " from this
definition. Terminal owners have no way of knowing what product
a tank truck carried on the previous load, especially trucks
owned by independent haulers who are filling at the terminal (IV-
D-20, p. 3) .
Response: The definition (with the added phrase "or
railcar") is the same as the definition contained in 40 CFR 60,
subpart XX. The purpose of the phrase referred to by the
commenter is to ensure that all displaced gasoline (HAP) vapors
are collected and routed to a vapor control system. If a driver
"switch loads" (loads a nonvolatile product, such as diesel fuel,
into a tank truck that carried gasoline on the previous load),
residual gasoline vapors would be expelled by the incoming
4-3
-------
product. In order to ensure that all gasoline vapors are
collected, many terminals outfit all of their loading positions
with vapor collection and require every truck loading to hook up
to the system. Another option is to restrict a certain rack or
racks to the loading of nonvolatiles, and to limit their use to
tank trucks dedicated to this type of product.
Due to these considerations, the same definition has been
retained in the final rule. However, the defined term has been
changed to "gasoline cargo tank," which is more general than
"tank truck" and is appropriate because of the inclusion of
railcars in the definition. The new term is also consistent with
the terminology utilized in the test procedures adapted from the
California cargo tank certification program (see Section 7.2).
4.4 NEW/EXISTING FACILITY
Comment: The EPA should define "new facility" and "existing
facility" in the rule itself rather than incorporating the
definitions by reference from the General Provisions.
Additionally, EPA should make clear that new facility
requirements do not apply to an existing facility unless the
modifications or improvements exceed 50 percent of the total cost
to build a terminal or pipeline breakout station (IV-D-19, p. 3;
IV-D-20, p. 6; IV-D-26, p. 2).
Response: In referring to subpart A of part 63 in proposed
§63.421 of the MACT regulation, EPA was incorporating the
definitions provided in §63.2 of the General Provisions
[promulgated in the Federal Register on March 16, 1994 (59 FR
12408)]. The EPA's approach of incorporating the substantial
requirements of the General Provisions (as well as other
requirements) into new regulations prevents duplication, provides
for consistency, and is considered practical and necessary. A
table has been added to the final rule to clarify the
applicability of the General Provisions under this regulation.
The definition provided for "new source" in §63.2 is "any
affected source the construction or reconstruction of which is
4-4
-------
commenced after the Administrator first proposes a relevant
emission standard under this part." "Existing source" means all
affected sources that are not new sources. The pertinent term
"reconstruction" consists essentially of the 50 percent
expenditure criterion cited by the commenter. In addition, it
must be technologically and economically feasible for the
reconstructed source to meet the relevant standards. The EPA
believes that these definitions make it clear that new facility
requirements apply only to newly constructed or reconstructed
facilities as defined in the General Provisions.
4.5 CONTIGUOUS AREA
Comment: Section 63.420(c) of the proposed rule provided
that the methods to determine applicability (equations and
emissions audit) do not apply to " ... bulk gasoline terminals or
pipeline breakout stations located within a contiguous area and
under common control of a petroleum refinery if the petroleum
refinery is a major source under section 112(a)(1) of the Act."
Two commenters suggested that EPA include a concise definition
for "contiguous" or "contiguous area" (IV-D-28, p. 3; IV-D-38, p.
7). One of the commenters suggested that "contiguous" be defined
as "being located on, or adjacent to, the property upon which the
refinery is located and is not separated by any road, street, or
highway maintained by the local government."
Response: The EPA believes that the definition offered by
the commenter, while certainly conveying the essence of the term
"contiguous," may not be applicable for all cases of interest.
Also, the Agency is unclear as to the critical distinction
presented by the presence of a road "maintained by the local
government." In essence, the EPA feels that developing a
definition could unnecessarily restrict the applicability
determinations made for this and other source categories.
Therefore, a specific definition for "contiguous" has not been
made part of this rule.
4-5
-------
The issue of applicability for bulk terminals that are
associated with major source refineries was discussed in detail
in Section 3.7.
4.6 IN VHAP SERVICE
Comment: The EPA should delete the definition of "in VHAP
service" because it does not apply to gasoline facilities. Also,
VHAP service is not mentioned anywhere in the rule (IV-D-29, p.
6) .
Response: The proposed rule provided a definition for "in
VHAP service" and "in VOC service" which equated these terms to
"in gasoline service." This was done to clarify that the
referenced equipment leak rules (40 CFR 60, subpart W and 40 CFR
61, subpart V) should be interpreted to apply only to equipment
handling gasoline vapors. Therefore, the definition was included
to make the referenced rules applicable to gasoline distribution
facilities. As discussed in Section 9.2, equipment leak
detection and repair requirements that rely on subparts V and W
are not being promulgated in the final MACT standards.
Therefore, the two definitions referred to above are not
necessary and have been deleted in developing the final rule.
4-6
-------
5.0 LOADING RACK STANDARDS
5.1 EMISSION LIMIT
Comment: Several commenters disagreed with the proposed
requirement that vapor control systems on loading racks at
existing facilities meet an emission limit of 10 mg/liter of
gasoline loaded. Many argued that the requirement was formulated
based on initial performance tests and not continuous performance
data and therefore does not represent the "floor." For this
reason, EPA has not demonstrated that existing systems
continuously meet 10 mg/liter. One commenter supplied terminal
test data on systems designed to meet an 80 mg/liter limit that
show the variability in vapor recovery unit operations. This
commenter pointed out that the floor level of control for
existing sources must be based on the "emission limitation"
achieved by the best performers, as directed by section 112(d)(3)
of the Act. The commenter felt that this language indicates it
is the regulatory limit, or at least the average performance of
controls over the life of operation, that should dictate the
proper floor level of control. Others suggested that it is too
expensive to replace existing units to meet 10 mg/liter
considering the small emissions savings. Several recommended
that existing processors be regulated at the current NSPS
standard of 35 mg/liter, because the NSPS limit is more
representative of the average emission performance of loading
rack control systems (IV-D-3, p. 2; IV-D-4, p. 3; IV-D-5, p. 2;
IV-D-7, p. 5; IV-D-8, p. 2; IV-D-12, p. 1; IV-D-13, p. 3; IV-D-
14, p. 6; IV-D-17, p. 7; IV-D-18, p. 3; IV-D-20, p. 5; IV-D-22,
5-1
-------
p. 73, App. I; IV-D-34, p. 15; IV-D-35, p. 4). Several
commenters expressed support for EPA's decision to regulate
existing loading racks to meet 10 mg/liter (IV-D-21, p. 2; IV-D-
28, p. 3; IV-D-38, p. 3).
Response: Section 112(d)(3) of the Act requires HAP
emission standards for existing sources that are no less
stringent, and may be more stringent than the average emission
limitation that is achieved in practice by the best performing 12
percent of the existing sources. This minimum control level
required for existing sources is termed the "floor." In order to
support setting this floor level, the Agency collected test
information on facilities in several areas throughout the
country, subject to various regulatory emission limits (80, 35,
and 10 mg per liter of gasoline loaded). These data were
summarized in Table 4-1 of BID, Volume I. The data showed that
all of the vapor processors at terminals regulated by the 10
mg/liter standard met the 10 mg/liter limit; however, EPA
estimates that less than 3 percent of terminals are subject to a
10 mg/liter emission limitation (BID Volume I, Table 3-11). In
addition to these terminals, the majority (about 70 percent) of
control systems at terminals regulated by the 35 mg/liter NSPS
standard achieved less than 10 mg/liter. Additional data
collected in Table 3-11 of the proposal BID show that
approximately 40 percent of the terminals are subject to the 35
mg/liter standard. Based on this information, EPA concluded that
nearly 30 percent of the vapor processors are capable of
achieving the 10 mg/liter limit. Therefore, the average emission
limitation achieved by the best performing 12 percent of the
existing sources is a 10 mg/liter standard, so 10 mg/liter was
selected as the floor control level for existing bulk gasoline
terminals.
The EPA believes the data show that the 10 mg/liter emission
level is achievable and can be maintained over the life of a
vapor control system. Commenters did not provide any data to
5-2
-------
demonstrate that the systems could not be maintained to meet this
limit over extended periods of time. Also, the available data do
not indicate any consistent degradation in performance with time.
The test data provided by one commenter (for systems installed to
meet 80 mg/liter) do indeed reflect a variability in performance
over a 13-year period (test results were between 7 and 54
mg/liter at two bulk terminals). However, the test results
improve with time for one of the terminals (average for the first
7 years = 43 mg/liter, average for the last 6 years = 26
mg/liter).
Based on the above discussion, the regulatory limits are not
a reliable predictor of the actual performance of these systems.
Systems are currently, and have been for some time, manufactured,
installed, warranted, and operated to achieve 10 mg/liter. In
addition, the majority of the systems designed for the NSPS level
(35 mg/liter) are achieving 10 mg/liter or lower. Therefore, the
NSPS limit of 35 mg/liter cannot be presumed to be the floor
control level. The EPA is obligated by the Clean Air Act to set
standards no less stringent than the floor level of control for
existing sources.
Comment: One commenter said that existing control devices
already do a good job of controlling HAP's, because HAP's found
in gasoline are larger, heavier molecules and are preferentially
captured by both carbon adsorption and refrigeration systems. As
a result, HAP's are removed to very low levels regardless of the
VOC limit of the vapor recovery unit. The commenter said that
sampling results at a number of their terminals support this
assertion, and said that they would provide the results from this
sampling to the docket as soon as possible (IV-D-31, p. 2).
Response: The results of a sampling program conducted by
the commenter at one bulk terminal were received by EPA following
the close of the public comment period on the proposal (IV-D-46).
Two sets of 1-hour sample data were collected at the inlet and
outlet of the vapor processor using passivated Summa canisters.
5-3
-------
Analysis of the vapor samples for HAP's was by gas
chromatograph/mass spectrometry utilizing direct cryogenic
concentration. A liquid gasoline sample was also collected from
one of the gasoline storage tanks at the facility, for gas
chromatographic analysis of the liquid and headspace vapor HAP
content.
The data in this report demonstrate very good control
efficiencies (>99 percent) for MTBE, benzene, toluene,
ethylbenzene, xylenes, hexane, isooctane, cumene, and
naphthalene. However, the data were not collected for a number
of terminals (which would allow comparisons among different types
of control systems with different emission levels), but apply to
the control system at a single facility. Thus, it is impossible
to verify the commenter's assertion that HAP's are removed to
very low levels "regardless of the VOC emission limit of the
recovery unit." Also, there is no indication in the report of
the type of vapor processor in use at the facility or its overall
VOC control efficiency. The Agency upon examining these data
finds that they are insufficient to draw the type of conclusion
indicated in this comment.
The EPA has determined the new source MACT and existing
source MACT floor control levels on the basis of data reflecting
the best performing control systems. The Agency presumes that
these best performing systems will achieve maximum control of all
VOC, including those which are designated as HAP's. The data
show that new control systems and many older systems can achieve
an emission level of 10 mg of total organic compounds per liter
of gasoline loaded and, therefore, the standard for new and
existing sources has been set at this level.
Comment: Two commenters were concerned with secondary
emissions resulting from combustion units (IV-D-28, p. 3; IV-D-
38, p. 4). One of the commenters suggested the following
revisions to the 10 mg/liter standard: 1) for non-destructive
control technologies, the total emissions of VOC (or POC) shall
5-4
-------
not exceed 10 mg/liter loaded; and 2) for incineration devices,
the combined emissions of VOC (or POC), NOx, and CO shall not
exceed 10 mg/liter. Such a provision would not make incinerators
the system of choice and would not increase the generation of
secondary pollution (NOx and CO) .
Response: The EPA agrees with the concept of reducing all
pollutant emissions. However, under section 112 regulatory
control is to be placed on HAP's and not on criteria pollutants.
Even if such an approach could be considered, the Agency does not
have the data that would be needed to set an appropriate emission
limit or limits. Finally, it is not clear how these suggested
limits would discourage the use of incineration (unless the limit
were not achievable by these systems).
Comment: One commenter said that §63.422(b) of the proposed
rule is not clear as to whether the 10 mg/liter limit for loading
racks includes fugitive emissions. The EPA should indicate in
the rule that the emission limit applies only to the process
stream outlet of the control device (IV-D-7, p. 6).
Response: The proposed rule in §63.422(b) applied the 10
mg/liter emission limit to "emissions to the atmosphere from the
loading racks and the vapor collection and processing system."
This provision was intended to have the same applicability as
current State and Federal new source regulations (processor
outlet emissions as measured in the performance test). However,
in reconsidering this provision in light of this comment, the
Agency agrees that the provision as worded may be subject to
misinterpretation. Therefore, the provision has been reworded in
the final rule to apply to "emissions to the atmosphere from the
vapor collection and processing systems." The EPA believes that
this clarifies the intent of the provision. It should be
recognized that, prior to each performance test, the vapor
collection and processing systems must be monitored for fugitive
leakage and the leaks repaired, but any leakage found in this
5-5
-------
survey need not be included in the total "loading rack emissions"
that are to be compared to the regulatory emission limit.
Regarding fugitive or leakage emissions in general, Section
9.2 discusses the new monthly visual inspection program that will
be used to detect equipment leaks at gasoline marketing
facilities.
Comment: A few commenters stated that methane and ethane,
which are not ozone precursors, should be excluded from the total
organic emissions measured to determine compliance. Carbon
adsorbers and refrigeration units (which do not control methane)
would have difficulty meeting the proposed emission limit stated
in terms of total organic compounds (TOC). One of the commenters
pointed out that in the past the standard has been in terms of
volatile organic compounds (VOC). Another stated that a limit
based on TOC would require that monitoring/testing and laboratory
equipment, as well as established test methods, be changed. This
commenter suggested that a limit on VOHAP and/or VOC might be
appropriate, while another commenter felt that TOC should be
replaced with VOC or POC (IV-D-9, p. 11; IV-D-28, p. 3; IV-D-38,
p. 4; IV-F-1, p. 26).
Response: The selection of TOC as a surrogate parameter for
the VOC content in the emission stream was explained in the BID
for the promulgated NSPS for bulk gasoline terminals (IV-A-1).
Briefly, it was stated in that document that the control
technologies in use at bulk terminals do not selectively control
VOC (photochemically reactive compounds), but rather all of the
organic compounds making up gasoline vapors. Furthermore, the
NSPS and NESHAP emission limits are based on test data and test
procedures that measure TOC, and the test methods specified in
both types of Federal standards measure TOC. Therefore, to
reflect accurately the performance of the control technologies
selected as MACT and to be consistent with the data base and test
methods on which the emission limit is based, the emission limit
is expressed in terms of total organic compounds, rather than
5-6
-------
VOC. Only two non-VOC (and non-HAP) constituents, methane and
ethane, occur in any appreciable quantity in gasoline vapors
(approximately 2 percent). Section 60.503(c)(6) of subpart XX
allows an owner or operator to subtract these two compounds in
calculating the emission rate for comparison to the emission
limit.
Comment: One commenter felt that the loading rack emission
standard should be stated in terms of mass emissions (mg) per
unit volume (liter) of all the products loaded rather than only
liters of gasoline loaded. Non-gasoline products are also loaded
at bulk terminals and their displaced organic vapors are piped to
the same control device. Also, switch loading (i.e., loading a
nonvolatile product into a tank truck that has just carried
gasoline) may occur and will contribute to the mass measured at
the outlet (IV-D-9, p. 12; IV-F-1, p. 26).
Response: Units for a bulk terminal standard were first
specified by EPA in the 1977 bulk terminal CTG (II-A-4). The
recommended RACT emission limit was 80 mg hydrocarbon per liter
of gasoline loaded. Similar units were used for the bulk
terminals NSPS, promulgated on August 8, 1983 (48 FR 37578). The
NSPS emission limit is 35 mg TOC per liter of gasoline loaded.
The rationale for considering only the volume of gasoline loaded
was explained in the preamble to the proposed bulk terminal NSPS,
at 45 FR 83140 (December 17, 1980).
To summarize that discussion, the selection of gasoline
volume loaded is tied to the practice of "switch loading" at many
bulk terminals, and its possible effect on emission test results.
There are two principal types of switch loading of concern with
regard to testing during tank truck loading. First, gasoline may
be loaded into a tank that has carried a nonvolatile product,
such as diesel fuel, on the previous load. The tank would
contain essentially no organic vapors, so the emissions during
loading of that tank would be negligible. Second, a product such
as diesel fuel may be loaded into a tank which has carried
5-7
-------
gasoline on the previous load. The VOC vapors from the previous
load of gasoline would be displaced by the incoming product.
If it is assumed that the frequency of each of these two
switch loading cases would be about equal, the quantity of
emissions could be accounted for by considering only the volume
of gasoline dispensed during a given time period. This approach
to determining emissions at a terminal would simplify the test
procedure. If the liquid volume of all products dispensed into
gasoline tank trucks during the performance test were considered,
then the liquid volume not displacing gasoline vapors would have
to be subtracted from the total volume loaded in order to
correlate the TOC mass emitted with the corresponding liquid
volume. This procedure would require that each driver loading
during the test be asked which product was carried on the
previous load. Based on the information obtained, only the
loadings displacing gasoline vapors would be added together to
obtain the total volume to be used in the calculations. However,
since the accuracy of this information could be subject to
question, and it may require extra test personnel to question the
drivers, this procedure is not considered to be the most
practical method of conducting the performance test.
The procedure in which only the volume of gasoline loaded
during the test is considered relies on a known quantity which
can be obtained directly from dispensing meters, instead of
relying on uncertain data. The two cases of switch loading
essentially cancel each other in terms of their effects on the
test results. Therefore, the standards have been developed to
require that emissions be calculated in terms of the total volume
of gasoline dispensed during the performance test, and they also
specify an emission limit in terms of total gasoline loaded.
Since excessive instances of switch loading have the potential to
affect the test results by increasing the apparent emission
level, especially if there are extra unbalanced cases of
nonvolatile product loadings into tanks containing gasoline
5-8
-------
vapors, switch loading should be minimized during the performance
test.
Comment: One commenter recommended that, due to the
stringency of the 10 mg/liter standard, existing facilities be
"grandfathered" in accordance with the federally enforceable
emission standard applicable at the time the vapor control system
was installed (similar to 40 CFR 60, subparts K, Ka, and Kb tank
construction provisions). The commenter also recommended that
any existing facility that undergoes reconstruction or
modification be given 5 years to comply (IV-D-36, p. 2).
Response: As noted earlier, under the Clean Air Act EPA is
obligated to set a floor level for existing sources and so may
not "grandfather" such facilities that are subject to MACT
standards. Also, as discussed in Section 2.4, section 112 (i) (3)
of the Act requires compliance within 3 years after promulgation
(with an extension of an additional 1 year possible).
5.2 SELECTION OF VAPOR PROCESSOR
Comment: One commenter, while agreeing that carbon
adsorption and enclosed flame combustion systems are capable of
achieving the proposed 10 mg/liter emission limit for tank truck
loading, recommended that EPA not consider open-flame flares to
be an acceptable control technology to meet the 10 mg/liter
emission limit. First, such flares cannot be tested for
compliance; and second, these flares may not be efficient enough
to meet the standard under various loading rates and VOC inlet
concentrations. In contrast, thermally controlled vapor
combustion units would be appropriate for meeting the 10 mg/liter
limit. The commenter stated as an example that an open-flame
flare with 98 percent destruction efficiency at an inlet
hydrocarbon concentration of 60 percent will yield a 33 mg/liter
emission rate. In addition, open-flame flares which were
enclosed in a stack so that they could be tested averaged about
19 mg/liter (IV-D-9, p. 2; IV-F-1, p. 20, 31). Another commenter
5-9
-------
suggested that open-top flares be regulated by equipment design
standards rather than a mass loading standard (IV-D-36, p. 1).
Response: The EPA's selection of the floor level of control
for existing sources was based on test data collected for carbon
adsorption, thermal incineration, and refrigeration condenser
type vapor processors. The thermal incineration devices in the
data base consisted of systems with enclosed combustion areas,
which allowed samples of the exhaust gases to be collected and
analyzed for hydrocarbon content. The open-flare type of
combustion device referred to by the commenter does not have an
exhaust stack to allow for the collection and analysis of the
emitted gases. As a result, the test methods applicable to bulk
terminal control systems (Methods 25A and 25B) cannot be
utilized, and a direct compliance determination relative to the
applicable emission limit cannot be made.
The use of open-flame flares at bulk terminals, and the need
to specify design/performance requirements for these devices, was
recognized by the Agency when these MACT standards were proposed.
In the preamble, page 5872, it was stated:
"Due to the inherent inability to measure mass
emissions from elevated flares (elevated flare's flame
is open to atmosphere and therefore the emissions
cannot be routed through stacks), these test methods
are not applicable. Therefore, the Agency has
established performance requirements for flares. These
performance requirements, including a limitation on
visible emissions, are provided in §63.11 of the
proposed General Provisions, which specifies Method 22
for determining visible emissions from this hard to
test type of flare."
Since this preamble was prepared, the final General Provisions
have been promulgated (59 FR 12408, March 16, 1994). Section
63.11(b) provides requirements for flares, which include:
(1) monitoring to assure that they are operated and maintained in
conformance with their designs; (2) flares shall be steam-
assisted, air-assisted, or non-assisted; (3) flares shall be
operated at all times when emissions may be vented to them;
5-10
-------
(4) no visible emissions, except for periods not to exceed a
total of 5 minutes during any 2 consecutive hours; (5) presence
of a flame at all times and shall be monitored using a
thermocouple or equivalent device to detect the presence of a
pilot flame; (6) net heating value of the gas being combusted at
300 Btu/scf or greater if the flare is steam-assisted or air-
assisted, or with a net heating value of 200 Btu/scf or greater
if the flare is non-assisted; and (7) a maximum gas exit velocity
that depends on the net heating value.
The EPA has analyzed these criteria in light of the emission
limit in these standards and the conditions occurring at bulk
terminals, and has attempted to answer two questions. First, do
the loading conditions at bulk terminals allow the §63.11(b)
criteria to be satisfied? Second, are open-flame flares that
comply with the criteria in §63.11(b) capable of achieving the 10
mg/liter emission limit?
The net heating value of the entering gas stream is a
variable that depends on the degree of saturation of hydrocarbons
in air that occurs in individual tank trucks. This concentration
can typically vary from 10 to 60 volume percent. Assuming that
butane is characteristic of the entering air-vapor mixture, the
heating value would vary from approximately 450 to over 2,000
Btu/scf. These values are well in excess of the minimum values
prescribed in §63.11(b). Finally, for a typical vapor
displacement rate of 600 gal/min (80 cfm), the exit velocity
would be under 60 ft/sec (the maximum allowable value) for any
flare tip cross-sectional area greater than 3.2 square inches:
[(80 ft3/min)/(60 ft/sec)/(60 sec/min) x 144 in2/ft2 = 3.2 in2].
Therefore, it appears that open-flame flares can be designed to
satisfy or exceed the criteria in the General Provisions under
the conditions existing at bulk terminals.
With regard to the second point, the first commenter stated
that, at EPA's assumed 98 percent VOC destruction efficiency,
open-flame flares will not meet the 10 mg/liter limit. The
5-11
-------
critical point is the commenter's assumption that EPA assumes a
98 percent efficiency for these flares. In fact, the average
efficiency of flares that meet the provisions of §63.11(b) is
well over 99 percent. The 98 percent figure is considered to be
a minimum efficiency for a well-designed flare. Additional test
data supplied by this commenter showed the results of enclosing a
flare burner in a refractory lined cylinder so that test
measurements could be made. The VOC emissions were determined to
range from 10.2 to 31.5 mg/liter. The EPA believes that these
test results may not necessarily be representative of the
performance of all enclosed-flame flares because locating the
burner within a stack mav create a situation in which the
combustion process is deprived of the oxygen needed for complete
combustion, thereby reducing the control efficiency and
increasing the emission rate.
In summary, data are not available to quantify the control
performance of open-flame flares used at bulk terminals.
However, EPA believes that such devices can be designed and
operated to achieve the emission limit in this rule by complying
with the control device requirements of §63.11(b). It should be
noted that EPA is not, in stating this conclusion, necessarily
advancing open-flame flares as the preferred choice for this
application.
Comment: One commenter said that a loading rack vapor
recovery system exists which may comply with the proposed
emission standard but cannot demonstrate compliance using the
test methods and procedures in proposed §63.425. In this system,
vapors are collected during loading and routed to a vapor holding
tank equipped with a diaphragm (bladder). The vapors are routed
off site to the fuel gas system of a nearby oil and gas
production facility and piped to numerous combustion devices
associated with the production facility to produce useful energy.
The commenter said that such a system cannot be tested using the
methods in §63.245 and suggested that an alternative compliance
5-12
-------
demonstration method be included for such a system (IV-D-37, p.
1) •
Response: The proposal discussed and analyzed several of
the most prominent types of vapor control systems in use at bulk
terminals. The EPA acknowledges that variations of these basic
systems are possible that may achieve good control of gasoline
vapor emissions. Such "alternative" systems that satisfy the
requirements of the standards would be approvable. For example,
the systems must be leak-free and must create a backpressure less
than 18 inches of water column in the cargo tank vapor space
during loading. Further, combustion devices must be testable
using approved methods or must meet the minimum requirements in
§63.11(b) of the General Provisions (flare requirements). Each
such device receiving vapors from the loading racks would have to
meet these criteria, or one device could be dedicated for this
purpose and that device would have to be tested or must meet the
requirements of the General Provisions.
The promulgated standards cannot include detailed compliance
information for all alternative systems, but EPA will evaluate
and, if warranted, approve such systems on a case-by-case basis.
Comment: A few commenters said that certain features of the
proposed rule contradict EPA's policy of favoring the use of
product recovery techniques and pollution prevention. Setting a
monitored operating parameter value that effectively lowers the
emission limit and requiring continuous monitoring of exhaust
vent TOC are serious disincentives for sources to install carbon
units. As a result, combustion type processors will become the
system of choice, and this will lead to an increase in N0X and CO
emissions (IV-D-14, p. 14). One commenter said that all six Bay
Area (California) terminals that use incineration have also
installed a recovery unit upstream from the incinerator, and
recommended that EPA mandate product recovery devices at all
loading racks, including those using incineration. The commenter
believes that, at the very least, an abatement system which
5-13
-------
incorporates product recovery should be MACT for new sources (IV-
D-3 8, p. 5).
Response: The floor for existing sources was established
using test data from carbon adsorption, combustion, and
refrigeration condenser type systems. There is no provision in
the Act that allows the exclusion of a particular control
technology in setting the floor.
The EPA expects that, as in the past, facilities will select
specific vapor control systems on the basis of a number of
considerations, in addition to simply the initial installed cost
of the system. Continuous monitoring (discussed in Section 6) is
required for any type of processor selected, and may consist of
an outlet VOC concentration monitor or an operating parameter
monitor. Also, due to the value of product recovered, recovery
type devices may often show a lower net annualized cost than the
combustion type. Therefore, EPA believes that facilities will
continue to select both recovery and non-recovery systems on the
basis of several factors including, but not limited to, those
cited by the commenter.
5.3 VACUUM ASSIST VAPOR COLLECTION
Comment: Many commenters expressed opposition to the
proposal to require use of "vacuum assist" technology at new bulk
terminal loading racks. Most of the commenters believe that
annual vapor tightness testing and repair is adequate to control
cargo tank leakage emissions. The following concerns were
expressed:
1) Some commenters expressed safety concerns; e.g.,
concerning the potential for fires and cargo tank implosion. One
of them said that internal tank vacuums can (and already do)
damage the internal compartment heads of cargo tanks by reversing
those heads and weakening the tank's outer shell, which
compromises product retention capability (IV-D-5, p. 3; IV-D-7,
p. 6; IV-D-20, p. 6; IV-D-22, p. 44; IV-D-23, p. 4; IV-D-26, p.
5-14
-------
2; IV-D-28, p. 4; IV-D-30, p. 7; IV-D-35, p. 6; IV-D-49, p. 3;
IV-D-54, p. 2).
2) Several do not believe that vacuum assist technology has
been demonstrated as "achievable in practice." The technology
has been used in only one State (Texas) and has not been tested
under various climatic conditions, such as combined low
temperatures and high humidity levels (IV-D-7, p. 6; IV-D-22, p.
44; IV-D-26, p. 2; IV-D-34, p. 16; IV-D-49, p. 3; IV-D-53, p. 2).
3) The complexity of the loading system will increase.
Also, due to rapid fluctuations in gasoline flow rates and the
requirement to maintain a vacuum at all times during loading,
nuisance shutdowns of the loading operation could be a problem
(IV-D-26, p. 2; IV-D-36, p. 2; IV-F-1, p. 23; IV-D-55, p. 1).
4) One commenter said that such a system may adversely
affect the efficiency of the vapor control device because air can
leak into the vapor collection system and dilute the inlet VOC
concentration (IV-D-9, p. 7). Another felt that volatilization
of fuel in the cargo tank would be increased due to the vacuum,
sending more vapors to the control device. This would require a
larger device which may have greater emissions, and more solid
waste impact for the case of a carbon system (IV-D-48, p. 2).
5) Vacuum assist systems will increase electrical power
consumption 15 to 400 percent depending on the type of emission
control device used (IV-D-9, p. 5; IV-F-1, p. 23).
6) Vacuum assist is unnecessary, because cargo tanks do not
leak enough during loading to justify vacuum assist as a means of
reducing the losses. Recent API data show that cargo tank
leakage has been significantly reduced since the EPA study
performed in 1978 (IV-D-2, p. 9; IV-D-3, p. 2; IV-D-15, p. 4; IV-
D-17, p. 9; IV-D-26, p. 2; IV-D-28, p. 4; IV-D-35, p. 6; IV-D-49,
p. 2; IV-D-54, p. 2).
7) The system will control losses from the cargo tank only
during the loading operation, and will not address losses from
the cargo tank in transit and while operating at bulk plants and
5-15
-------
service stations. Therefore, a rigorous testing and inspection
program will have overall increased emission reduction benefits
(IV-D-28, p. 5; IV-D-38, p. 2; IV-D-49, p. 3).
8) Vacuum assist is very expensive and not cost effective
(IV-D-7, p. 6; IV-D-22, p. 44; IV-D-35, p. 6; IV-D-49, p. 2).
9) If EPA decides to mandate a cargo tank vapor tightness
program similar to the one in California, vapor leakage from
cargo tanks will decrease and there will be even less reason to
require vacuum assist systems at new bulk terminals (Williams).
The EPA should certainly not combine the California standards and
the vacuum assist requirement, because this would be an option
above the floor (IV-D-48, p. 1).
Response: The vacuum assisted vapor collection system was
proposed for new source bulk terminal loading racks to control
HAP emissions due to vapor leaks from cargo tanks (tank trucks
and railcars) during gasoline loading operations. This system
creates a negative pressure in the vapor collection system during
loading to ensure that vapors will not be forced out into the air
through any leakage points. Following proposal and the receipt
of public comments, EPA published supplemental information and
reopened the public comment period on this new information in a
Federal Register (FR) notice dated August 19, 1994 (59 FR 42788).
In that notice, EPA referred to the proposed requirement for
vacuum assist and stated that, in the Agency's consideration of
the vacuum assist system for new sources, California's test
requirements for vapor tightness would also be considered. The
California program, along with specific comments and EPA
responses on the details of that program, is discussed in
Sections 7.2, 7.3, and 7.4 of this document.
The proposal rationale was based on the following
information. Vacuum assist systems are in use at a few bulk
gasoline terminals (in addition to the annual vapor tightness
test for cargo tanks) in Texas, so it meets the Act requirement
to consider the best controlled similar source in establishing
5-16
-------
the floor level of control for new terminals. Since less than 1
percent of terminals use this vacuum assist system, it is not
considered the floor for tank truck leakage at existing
terminals. Annual vapor tightness testing of the cargo tanks was
considered at proposal to be the floor for existing terminals
(this floor determination has been modified on the basis of
information received in the public comments, as discussed in the
supplemental FR notice and in Section 7.2). Based on field tests
in the late 1970's, an annual vapor tightness testing program was
estimated to reduce the leakage rate from baseline levels at 30
percent leakage to about 10 percent leakage. The vacuum assist
system was estimated to reduce that 10 percent leakage rate under
the annual vapor tightness test program by nearly 100 percent.
Industry sources had expressed concerns before proposal
regarding the operational reliability of a vacuum assist system,
especially under extreme cold weather conditions. Those
commenters also believed that the system could present a safety
hazard if excess negative pressures were developed within a tank
truck fuel compartment. To the Agency's knowledge, the systems
in operation have not experienced any significant problems, and
one of the systems has been operating for over 3 years. These
systems contain safety pressure relief devices in combination
with the pressure-vacuum vents already installed on each tank
truck compartment. However, safety concerns are important to the
Agency. The EPA specifically requested comment at proposal,
including technical documentation and data where available, on
the reliability, effectiveness, safety aspects, and any other
issue concerning vacuum producing equipment for bulk terminal
vapor collection systems. No technical documentation or data on
installed systems was provided during the comment period.
As discussed in Appendix A, the leakage emission factor for
controlled cargo tanks under an annual vapor tightness program
was adjusted to reflect current data on the frequency with which
tank trucks pass the test on the first attempt. Emissions lost
5-17
-------
from cargo tanks under test programs with a pressure decay limit
of 3 inches H20 are now estimated to be 1.3 percent of total
vapor displaced during loading operations (just under 99 percent
collection efficiency). The leakage emission rate from these
cargo tanks is calculated to average 0.7 and 2 milligrams of HAP
per liter of gasoline loaded for normal and oxygenated gasolines,
respectively. In California, where an annual pressure decay
limit of 1 inch H20 is in effect, the emission losses during
loading are estimated at 0.8 percent (slightly over 99 percent
collection). The corresponding HAP emission factors are 0.4 and
1.3 mg/liter of HAP for normal and oxygenated gasolines,
respectively. At proposal, the leakage emission rate was
estimated to be a 10 percent loss (90 percent collection
efficiency). Thus, while vacuum assist systems were previously
thought to have the potential to capture an additional 10 percent
of the loading emissions, they now appear to have the potential
to capture about 1 percent.
The EPA shares commenters' concerns that the emission
control achieved with the vacuum assist system is uncertain. The
Agency's uncertainty centers on the system's effectiveness in
accurately maintaining a slight vacuum to collect a small leak (1
percent of the volume displaced to the collection system) while
handling the variability of flows and pressures and limiting the
ingestion of air into the system to a degree where it does not
affect the control effectiveness of the processor. The vapor
volume collected by the system and internal pressures within the
vapor collection system vary widely throughout the day. Each
cargo tank loading and displacing vapors influences the pressures
and flows in the system. Terminals operate on demand, just like
gasoline service stations. The number of tanks loading at any
given time varies from none, to a few, to 10 or more tanks.
Additionally, vapor processor control efficiency may be adversely
influenced by increased amounts of air sent to the control
system. A vacuum assist system draws additional air into the
5-18
-------
system. Even small malfunctions in the system would be likely to
increase emissions above the 1 percent control target. Finally,
the Agency agrees that it lacks sufficient information to
determine whether conditions prevailing outside of Texas may
affect the control performance of vacuum assist methods.
The proposal of vacuum assist was based on the minimum
baseline (floor) at which standards may be set. Under section
112(d)(3) of the Act, the floor for new sources "shall not be
less stringent than the emission control that is achieved in
practice by the best controlled similar source, as determined by
the Administrator." The Administrator has determined that
emission control is not being achieved in practice given the
technical uncertainties about achieving emission reduction from
this source as discussed in the previous paragraph.
Consequently, the proposed vacuum assist requirement for new bulk
terminals has been deleted from the final rule.
5.4 OTHER
Comment: One commenter questioned whether a performance
test of a controlled loading rack conducted prior to promulgation
of the standards could be used to satisfy the loading rack
testing requirements. It was suggested that any existing
performance tests that were conducted in accordance with 40 CFR
part 60, subpart XX and that demonstrated compliance with the 10
mg/liter standard should be deemed to satisfy the new loading
rack requirements. Such a provision could be limited to
performance tests conducted within the past 5 years, or to vapor
recovery systems equipped with a monitoring device that meets the
requirements of the rule and demonstrates that performance has
not degraded since the test (IV-D-6, p. 1). Another commenter
felt that EPA should waive the performance test requirements for
all California terminals because California State law requires a
certification test on each bulk terminal. One District also
conducts two 24-hour tests annually at each facility (IV-D-28, p.
9) .
5-19
-------
Response: The final rule incorporates the performance
testing requirements of the General Provisions. In summary,
§63.7 requires that a performance test be conducted within 180
days after the date the standard first becomes effective for a
new source and 180 days after the compliance date for existing
sources. This testing allows the facility to establish an
operating parameter value for the vapor processing system in
response to §63.427, as well as to demonstrate compliance with
the loading rack emission limit in §63.422. Testing will follow
the development of a site-specific test plan and a performance
evaluation test plan for the continuous monitoring system (CMS)
which must be approved by EPA before the performance testing is
carried out. Due to these specific requirements associated with
the initial performance test, the results of previous tests would
not be adequate to fulfill the MACT requirements. For the same
reasons, the routine tests performed in California would not
satisfy the initial test requirements under this MACT rule.
However, subsequent tests performed according to EPA approved
methods and procedures may be able to be used to fulfill multiple
regulatory requirements.
5.5 REFERENCES
II-A-4
IV-A-1
IV-D-46
Control of Hydrocarbons from Tank Truck
Gasoline Loading Terminals. U.S.
Environmental Protection Agency, Research
Triangle Park, NC. EPA-450/2-77-026.
October 1977 .
Bulk Gasoline Terminals - Background
Information for Promulgated Standards. U.S.
Environmental Protection Agency, Research
Triangle Park, NC. EPA-450/3-80-038b.
August 1983.
Letter from Ellett, A.R., BP Oil Company, to
Air Docket Section, U.S. Environmental
Protection Agency, Washington, DC. August 5,
1994. Transmits final report detailing
5-20
-------
results of sampling program at gasoline
distribution terminal.
5-21
-------
6.0 MONITORING OF OPERATIONS
6.1 CONTINUOUS MONITORING OF COLLECTION SYSTEMS
Comment: One commenter stated that the burden of compliance
for tank trucks at new sources maintaining a vacuum during
loading was placed on the terminal in proposed §63.428(e). The
commenter suggested the following language for this provision:
"Owners and operators of new proprietary bulk gasoline terminals
(terminals that maintain and operate their own fleet of trucks)
subject to provisions of the subpart shall submit to the
administrator, a quarterly report of all instances in which a
vacuum is not maintained in a gasoline tank truck during loading"
(IV-D-36, p. 4).
Response: As discussed previously in Section 5.3, the
proposed requirement for new sources to install vacuum assisted
vapor collection has been deleted from the regulation.
Therefore, the provision offered by the commenter would not be
applicable in the final rule.
6.2 CONTINUOUS MONITORING OF PROCESSING SYSTEMS
Comment: One commenter stressed that, while continuously
monitoring a key operating parameter of a vapor processing device
may serve as a guide to warn of potential problems and to gauge
efficient operation, such monitoring would not be sufficient to
assure compliance with the pertinent emission standard. This
commenter and others were concerned that a value of the monitored
process variable could be selected that is more stringent than
necessary to indicate compliance with the proposed 10 mg/liter
emission standard. They felt that requiring a facility to
6-1
-------
continuously maintain a value determined during an initial
performance test and then consider the facility out of compliance
if it exceeds that value would be unfair. It is highly probable
that during an initial performance test the vapor control device
while operating at a particular value will perform much better
than the emission limit (IV-D-9, p. 7; IV-D-11, p. 3; IV-D-14, p.
11; IV-F-1, p. 24). One of the commenters said that, as an
example, thermally controlled combustion systems do not require
elevated temperatures all of the time to achieve 10 mg/liter.
The commenter recommended that, for these units, a single high
temperature value not be set because assist fuel gas consumption
would be very high and the unit would be made to operate at
control efficiencies substantially higher than the standard.
One commenter suggested that facilities be allowed to use an
extrapolative method to predict the operating parameter value at
the regulated emission standard based upon the operating
parameter value associated with the lower emission level recorded
during the performance test. Such an allowance is needed because
it is usually not possible to operate a vapor processing system
at maximum design conditions (IV-D-34, p. 18). Another commenter
recommended that the operating parameter value be set by the
least stringent parameter value obtained during the test while
the unit is in compliance with the standard (IV-D-5, p. 3).
Response: Section 114(a) (3) of the Act requires enhanced
monitoring and compliance certification of all major stationary
sources. The annual compliance certifications certify whether
compliance has been continuous or intermittent. Enhanced
monitoring shall be capable of detecting deviations from each
applicable emission limit or standard with sufficient
representativeness, accuracy, precision, reliability, frequency,
and timeliness to determine if compliance is continuous during a
reporting period. The monitoring in this regulation satisfies
the requirements of enhanced monitoring.
6-2
-------
The required performance test is a minimum of 6 hours in
duration, with outlet organic concentration and flow rate data
recorded every 5 minutes. While it seems reasonable to base the
selection of the parameter range or limit on a 6-hour period to
be consistent with the length of the test (as the Agency did at
proposal), the Agency has decided this is too long a period to
calculate a meaningful average on a continuous basis. One
commenter requested that EPA consider using an extrapolative
method (not specified by the commenter) using a single high
temperature, or setting the parameter based on data just meeting
the 10 mg/liter standard. As noted at proposal, EPA proposed
that site-specific monitoring parameter values be used to account
for the different types and designs of control equipment
available and the site-specific facility operating conditions.
The proposal required a performance test recording 5-minute
readings of outlet concentrations and flow rates while
continuously recording the specified parameter values. An
engineering assessment of those data, along with the
manufacturer's recommendations, could be used to find the
appropriate parameter value, monitoring frequency, and averaging
time that is equivalent to the emission standard. This approach,
which is incorporated into the final rule, is the most
straightforward way of accounting for both the emission standard
and the variability of the control equipment design and facility
operations. Under this approach, the Agency is allowing some
latitude for the method by which the parameter range of the "not
to exceed" limit is developed under the final standards. The
engineering assessment and manufacturer's recommendations must be
documented (recorded in facility files) and reported to the
Administrator for approval.
Comment: One commenter felt that requiring terminals to
continuously calculate rolling 6-hour averages of operating
parameters is burdensome and should be removed from the
regulation. Such a provision would require terminals to have a
6-3
-------
microprocessor-driven system or a dedicated computer, which is an
onerous requirement for a terminal (IV-D-5, p. 3). Two
commenters asked that a provision be inserted that would allow a
facility to manually calculate the 6-hour averages from readings
taken by the monitor. However, the monitoring requirement will
be difficult to comply with since many terminals are frequently
or totally unmanned (IV-D-18, p. 6, 9; IV-D-22, p. 82). One of
these commenters requested that the 6-hour rolling averages be
based on data obtained at a minimum time span of 15-minute
intervals (IV-D-22, p. 82). Another commenter said that
calculating a 6-hour average value to demonstrate compliance is
not applicable for flares because the flame does not burn
continuously (IV-D-36, p. 4).
Response: As discussed in the previous response, EPA has
decided to delete the requirement to continuously calculate 6-
hour averages of the operating parameter value. The EPA does not
believe that a microprocessor would be necessary to record and
analyze the monitoring data required under the final rule,
because no specific time averages are required. However, the
averaging time for determining the operating parameter value
should not exceed the time period of the performance test as
specified in §60.503(c)(1). With regard to the last comment,
only data recorded during the loading of gasoline cargo tanks
would be taken into account, and any type of processor would have
to be in operation during these periods.
Comment: One commenter asked that EPA clarify whether the
operating parameter values established in the initial performance
test would form the sole basis for future compliance
determinations at the facility, or if repeat testing would
establish new values or could even be used by the facility to
establish new parameter values reflecting new operating
conditions (IV-D-22, p. 80).
Response: The EPA does not foresee that the initial
performance test would necessarily establish parameter values
6-4
-------
that would be in effect for the life of the loading rack control
system. The proposed and final rules allow for multiple tests
under §63.425(b), and all procedures required under the initial
test would have to be satisfied for the repeat test. However,
this process must take place prior to any enforcement action or
notification issued for non-compliance with the current operating
parameter value. In addition, §63.425(c) specifies that the
owner or operator must document the reasons for any change in the
operating parameter value since the previous performance test.
Comment: Two commenters felt that the rule should provide
more flexibility in monitoring for compliance. Alternative
monitoring methods should be allowed, if such methods demonstrate
continuous compliance. For some units there is no known
continuous emissions monitoring (CEM) method for measuring all
the necessary parameters to assure continuous compliance with the
proposed standard (IV-D-29, p. 3; IV-D-36, p. 4).
A few commenters also recommended that EPA allow carbon
adsorption systems to have continuous parametric emission
monitors rather than continuously monitoring the exhaust VOC
concentration. One pointed out that monitoring the
concentration of total organics in the outlet gas is an extremely
poor surrogate for the 10 mg/liter limit (IV-D-14, p. 11).
Another stated that monitoring of carbon regeneration frequency,
as adopted in New Jersey, should be allowed as an alternative to
CEM (IV-D-29, p. 3). It was also suggested by a number of
commenters that monitoring the vacuum level during carbon bed
regeneration would be an effective and far less expensive means
of gauging system performance. The peak vacuum level is an
indication that the working capacity of the carbon bed has been
reestablished, the emission control device does not have any
vacuum leaks within the system, and the settings of the
components related to the regeneration system are adjusted
properly. This monitoring would provide a proactive position to
correct the problem rather than reacting once the emission limit
6-5
-------
has been exceeded. The vacuum monitor would continuously measure
the vacuum on the carbon bed during the regeneration cycle and
could signal an alarm to the facility when 25 inches of mercury
vacuum was not achieved during the last 5 minutes of regeneration
for two consecutive cycles. Additionally, the system could be
set to shut down the loading operation if three consecutive
regeneration cycles of the carbon bed do not reach 25 inches of
mercury during the last 5 minutes of the cycle (IV-D-7, p. 10,
IV-D-9, p. 7; IV-D-15, p. 6; IV-D-22, p. 78; IV-D-34, p. 19; IV-
F-1, p. 24).
Two commenters said that continuous emission monitors are
unnecessary for carbon adsorption systems and recommended that
annual performance tests be allowed to demonstrate compliance
(IV-D-13, p. 2; IV-D-17, p. 6). Another commenter said that
averaging emissions data annually would be a more reliable method
of determining compliance. In lieu of the continuous monitoring
requirements, the commenter recommended requiring terminals to
conduct performance tests every 3 years; to monitor pumps,
valves, and operating gauges, and retain the records for EPA
inspection; and allow carbon adsorbers the provision that if the
regeneration cycle for the carbon is 15 or 30 minutes, the unit
is considered to be in compliance (IV-D-20, p. 12).
Response: The EPA believes that continuous monitoring of a
parameter of the control system is necessary for ensuring that
the system is properly maintained and operated and that the
emission limit of the standards is being achieved continuously
(see first response in this section). The use of periodic
performance testing to accomplish this would be inadequate, since
an out-of-compliance system could operate for a long period of
time before the problem was recognized and corrected. For
similar reasons, annual averaging of emissions data would be
unacceptable.
The Agency recognizes that vapor control systems may have a
variety of operating parameters that could indicate their proper
6-6
-------
operation and whether the emission limit being achieved. At
proposal [§63.427(a)], EPA's best judgment concerning parameters
suitable for this purpose was exhaust organic concentration for
carbon adsorption and downstream temperature for refrigeration
condenser and thermal oxidation. Proposed §63.427(a)(5),
however, allowed the monitoring of an alternative operating
parameter upon demonstrating to the Administrator's satisfaction
that the alternative parameter indicates continuous compliance.
One industry commenter provided a technical report (IV-D-45)
presenting continuous operating data for a carbon bed vapor
control system installed at one of their bulk terminals. Two
types of continuous monitors have been installed on this control
system. One of these monitors provides a continuous record of
the exhaust hydrocarbon concentration, measured as equivalent
butane percentage or parts per million (ppm) of propane. A
second monitor continuously records the vacuum pressure achieved
in each carbon bed during the carbon regeneration cycles. The
recording system on the vacuum monitor keeps track of the
sequencing of the carbon beds and the vacuum level achieved
during regeneration.
A permit limit of 20,500 ppm propane (1-hour average) is in
effect for this control installation. During a 30-day test on
the system, the highest 1-hour average concentration was 13,806
ppm (and the highest 15-minute concentration was 19,360 ppm).
These values occurred on the day with the highest gasoline
throughput. During the test period, a consistent bed vacuum
pattern was observed; the regeneration vacuum quickly rose to the
specified 25 inches of mercury and peaked between 27 and 28
inches.
The report concluded that the CEM output (exhaust
concentration) can be correlated with the vacuum being pulled on
the activated carbon. This conclusion was based on the good
performance of the system during the test; i.e., the
concentration never exceeded the permit limit and the bed vacuum
6-7
-------
consistently peaked at the proper levels. The report further
concludes that the carbon bed regeneration recorder has benefits
not only when used as a means of compliance, but also as a
maintenance tool. Due to its simplicity and low purchase and
maintenance costs, the report characterizes this system as an
"effective, reliable, flexible, and cost efficient replacement
for CEM's."
The EPA has reviewed this technical report and the included
data. While the data indicate that both exhaust concentration
and peak vacuum pressure were within acceptable limits during the
test, the report does not demonstrate that any correlation exists
between these parameters. In other words, it is unclear from
these data whether a decreased vacuum level (for example, 20
inches of mercury) would correlate with a significant increase in
hydrocarbons (and HAP emissions) at the outlet. Further, there
is no indication that vacuum level relates to the emission level
in milligrams per liter (the units of the regulatory limit),
which is required for this parameter to be used as an indicator
of compliance. These additional test data would be necessary in
order for an operating parameter value to be established under
the final rule.
The EPA believes that the parameters discussed at proposal
are appropriate for typical vapor processors. As a particular
example, the Puget Sound Air Pollution Control Agency (PSAPCA) in
Washington State prescribes continuous exhaust VOC concentration
monitors for carbon adsorption systems at bulk terminals within
its jurisdiction, and has found this approach useful in
monitoring the operation of these systems (IV-E-5).
The provision allowing for an alternative operational or
emission parameter is retained in §63.427(a)(5) of the final
rule. This provision allows an owner or operator to present data
and supporting information for parameters (including carbon bed
vacuum) other than those listed in the rule. However, the data
must demonstrate that the parameter is an indicator of compliance
6-8
-------
with the emission limit and must establish and support a specific
operating parameter value.
Comment: One commenter felt that continuous monitoring
should not be required at terminals and pipeline breakout
stations that are not in continuous use. Monitoring should only
be required when the systems are in operation. Continuous
monitoring of vapor control systems which are not operating
continuously wastes time and money, and is an administrative
burden (IV-D-19, p. 5).
Response: The EPA recognizes that loading operations are a
noncontinuous batch-type operation, and that most vapor
processors operate on demand to process vapors as they are
released from the loading racks. The monitoring systems required
under these standards must have the ability to record the
operating parameter value during times that cargo tanks are
actually loading. Monitoring and recording this parameter at
other times would not provide information concerning emissions
control at the facility, and so would not be required. The
language of §63.428(f) has been worded to clarify that monitoring
and recording are required only for times when a gasoline cargo
tank is actually loading at the facility.
Comment: One commenter stated that continuous emissions
monitoring is necessary to assure compliance but should revolve
around more stringent local standards, not the federally
enforceable limit. The EPA has not recommended any monitoring
parameters for the three control technologies utilized to achieve
the 10 mg/liter standard (IV-D-28, p. 9).
Response: The term "federally enforceable" is defined in
the final General Provisions (§63.2) as follows:
"Federally enforceable means all limitations and
conditions that are enforceable by the Administrator
and citizens under the Act or that are enforceable
under other statutes administered by the Administrator.
Examples of federally enforceable limitations and
conditions include, but are not limited to:
6-9
-------
(1) Emission standards, alternative emission
standards, alternative emission limitations, and
equivalent emission limitations established pursuant to
section 112 of the Act as amended in 1990;
(2) New source performance standards established
pursuant to section 111 of the Act, and emission
standards established pursuant to section 112 of the
Act before it was amended in 1990;
(3) All terms and conditions in a title V permit,
including any provisions that limit a source's
potential to emit, unless expressly designated as not
federally enforceable;
(4) Limitations and conditions that are part of an
approved State Implementation Plan (SIP) or a Federal
Implementation Plan (FIP);
(5) Limitations and conditions that are part of a
Federal construction permit issued under 40 CFR 52.21
or any construction permit issued under regulations
approved by the EPA in accordance with 40 CFR part 51;
(6) Limitations and conditions that are part of an
operating permit issued pursuant to a program approved
by the EPA into a SIP as meeting the EPA's minimum
criteria for Federal enforceability, including adequate
notice and opportunity for EPA and public comment prior
to issuance of the final permit and practicable
enforceability;
(7) Limitations and conditions in a State rule or
program that has been approved by the EPA under subpart
E of this part for the purposes of implementing and
enforcing section 112; and
(8) Individual consent agreements that the EPA has
legal authority to create."
The EPA has several reasons for including a requirement that
the emission limits in MACT standards must be federally
enforceable. To summarize, the purposes of these requirements
(as discussed more fully in the preamble to the final General
Provisions at 59 FR 12414) are: (1) to make certain that limits
on a source's capacity are, in fact, part of its physical and
operational design, and that any claimed limitations will be
observed; (2) to ensure that an entity with strong enforcement
capability (i.e., the Federal government) has legal and practical
means to make sure that such commitments are actually carried
out; and (3) to support the goal of the Act that EPA should be
6-10
-------
able to enforce all relevant features of the air toxics program
as developed pursuant to section 112. Federal enforceability is
both necessary and appropriate to ensure that State and local
limitations and reductions are actually incorporated into a
source's design and followed in practice.
As discussed previously, §63.427(a) of the proposed rule
specified particular parameters that are to be monitored
continuously for carbon adsorption, refrigeration condenser, and
thermal oxidation systems. In addition, §63.427(a)(5) allowed
the monitoring of an alternative operating parameter to those
mentioned in the rule upon a demonstration that the alternative
parameter is an indicator of continuous compliance. The final
rule retains these provisions. The EPA believes that terminal
operators and equipment manufacturers are in the best position to
identify and recommend such alternatives; however, EPA must
review and approve any such recommendations.
6.3 CONTROL EQUIPMENT
Comment: One commenter stated that, because no equipment
operates correctly at all times, EPA should provide some
allowance for excess emissions or exceedances of the standards
due to unforeseen or unpreventable occurrences (IV-D-36, p. 2).
Response: The EPA understands that even well-designed and
maintained equipment is subject to malfunctions. The General
Provisions in §63.6(e) contain detailed provisions dealing, in
particular, with periods of startup, shutdown, and malfunction of
process and pollution control equipment. Each affected facility
is expected to develop and implement a written startup, shutdown,
and malfunction plan that describes, in detail, procedures for
operating and maintaining the source during periods of startup,
shutdown, and malfunction and a program of corrective action for
malfunctioning process and air pollution control equipment used
to comply with the emission limit of the standard. In addition,
§63.8(c) specifies operation and maintenance requirements for
continuous monitoring systems. These plans and requirements are
6-11
-------
intended to reduce emissions, especially during startup,
shutdown, and malfunction events at the facility, and do provide
the requested allowance needed for unpreventable occurrences.
Comment: One commenter pointed out that short-term air
quality standards allow one exceedance per year at any one
receptor, which translates into an exceedance rate of 6.6 percent
over a 24-hour average time standard. The EPA should allow a
similar exceedance rate at loading racks (IV-D-36, p. 2).
Response: The commenter did not explain why there should be
a correlation between exceedances of the ambient air quality
standards and excess emissions events at the loading racks.
However, as discussed in the previous response, exceedances due
to unpreventable upsets at a facility are accounted for in the
operation and maintenance provisions of the General Provisions
(and, thus, these standards). The Agency believes that these
provisions adequately address the issue raised by the commenter.
6.4 REFERENCES
IV-E-5 Telecon. LaFlam, G., PES, Inc., with Pade,
G., Puget Sound Air Pollution Control Agency.
July 15, 1994. PSAPCA's continuous
monitoring requirements for bulk terminals.
6-12
-------
7.0 CARGO TANK REQUIREMENTS
7.1 EMISSION FACTORS
Comments: Several commenters felt that EPA's assumption at
proposal that tank trucks that have passed the EPA Method 27
annual vapor tightness test leak 10 percent of their emissions
during controlled loading is outdated and inaccurate.
Consequently, the baseline emissions calculated for tank trucks
are grossly overstated. New data suggest that very few tank
trucks leak due to today's better construction standards and the
test requirements in effect under current Federal and State rules
(IV-D-2, p. 9; IV-D-7, p. 2; IV-D-14, p. 12; IV-D-15, p. 3; IV-D-
22, p. 42; IV-D-30, p. 6; IV-D-34, p. 14; IV-F-1, p. 13). One
commenter provided calculations indicating that, under the
proposed pressure decay standard (which is the same as the
subpart XX NSPS requirement), a typical controlled tank truck
would have a leakage emission factor for loading of 5.6 mg/liter
(at the maximum 18 inches of water backpressure) (IV-D-28, p. 5).
Another commenter estimated, on the basis of test failure rate
data from the BAAQMD and several oil companies, that the overall
average leak rate is 0.88 percent of the total volume of vapors
displaced during the loading of tank trucks connected to a vapor
recovery device (IV-D-22).
Response: The EPA's estimate of 10 percent vapor leakage
from emission sources in tank trucks while loading at controlled
loading racks was based on data collected in 1978 on 27 tank
trucks in California (II-A-12, II-A-25). These tank trucks were
under a State requirement to be certified annually in a vapor
7-1
-------
tightness test, and time periods ranging from 4 days to a full
year had elapsed since the last certification test for these
trucks. The volume losses among the trucks varied from 0.1 to
35.8 percent, with the average leakage being about 10 percent.
The data from these tests were further described, and the 10
percent figure derived, in the BID for the proposed NSPS for bulk
gasoline terminals (II-A-14).
The commenter who supplied the 0.88 percent overall leakage
estimate relied upon vapor volume loss data for individual tank
trucks reported in the 1978 study, and combined these data with
test failure rate data from BAAQMD (pressure test data) and from
several oil companies (combustible gas detector results gathered
during loading rack performance tests). Based on an assumption
that a leak definition of 10,000 ppm is equivalent to a 1 percent
loss of vapors through leakage, the commenter determined that the
average leak rate for tanks with leakage rates over 1 percent
("failing" tanks) was 12.1 percent, while the average leak rate
for the remaining, "passing" tanks was 0.5 percent. On the basis
of the failure rate data, the overall failure rate during 1989 to
1994 was found to be 3.3 percent. Combining the average leak
rate figures with these failure prevalence data, the commenter
arrived at the overall leak rate for all tank trucks of 0.88
percent.
The EPA recognizes and agrees with the commenter that the
available data indicate that overall vapor leakage rates from
tank trucks subject to a regular test and repair program using
the pressure decay procedure have been reduced over the past 16
years. However, the use of concentration data to estimate a
volume leakage rate, as the commenter has done, is uncertain. In
addition, neither EPA nor industry have access to current data
for several areas throughout the country that would allow a
national average calculation of this volume leakage to be made.
Therefore, any numerical result derived from the existing data
would be at best a broad estimate, which would not account for
7-2
-------
the full range of truck ages, ownership scenarios, and local
control programs.
In spite of these limitations, EPA has made an estimate
which it feels more closely reflects actual overall emissions
under a vapor-tight cargo tank program than the emission factor
used for the proposal. The Agency's new emission factor, 0.8
percent of the total vapors displaced or 8 mg of VOC/liter, is
based on the use of a volume loss equation found in EPA's tank
truck CTG (II-A-9) combined with the test failure rate data
submitted by one commenter and measured leakage from trucks that
failed the test. This new emission factor represents the
emissions after control to the level of the final standards as
discussed in the following section. Appendix A presents more
details on the calculation of this new emission factor.
7.2 LEVEL OF CONTROL
Comment: Two commenters recommended that EPA implement the
cargo tank vapor tightness program in effect within the State of
California since 1977. The California standard requires annual
certification that cargo tanks meet 5-minute pressure and vacuum
decay standards of 1 inch of water. Based on a Bay Area survey
of 200 tank truck owners which quantified actual pressure change
values, California is proposing to lower this standard to 0.5
inch of water.
In addition, the same commenters recommended that cargo
tanks be required to meet a year-round inspection standard of 2.5
inches of water in 5 minutes. Using a simple, accurate field
test method, the Bay Area has implemented a comprehensive
outreach program for cargo tanks since 1986. Operators are
instructed in the source test method, and participate in an
ongoing inspection and maintenance program. Participation is
voluntary, and the incentives are reduced penalties for
violations if documentation shows a history of regular tests and
maintenance on the tanks. Both commenters recommended that EPA
adopt California's proposed annual Method 27 limit of 0.5 inch
7-3
-------
and a year-round standard of 2.5 inches of water (IV-D-28, p. 6;
IV-D-38, p. 4).
As discussed in Section 5.3, following proposal EPA
published supplemental information pertaining to the cargo tank
vapor tightness program in California and requested comment on
EPA's determination that these standards are the new and existing
source floor for this MACT rule. The EPA received the following
comments on the floor determination and on the level of control
that is appropriate for controlling cargo tank leakage.
Five commenters felt that the existing California standards
should be specified for cargo tanks at new sources, but would be
inappropriate for existing sources (IV-D-49, p. 1,4; IV-D-51, p.
1; IV-D-52, p. 1; IV-D-53, p. 3; IV-D-54, p. 2). These
commenters based their opinion on the conclusion that EPA had
inappropriately based its floor determination on the gasoline
throughput, or number of tank trucks operating in California.
They felt that, since the legal responsibility for compliance
would be on the terminal owner or operator, the basis should be
the number of terminals in California. One commenter said that
this figure is 71, out of a total of 1,125 terminals nationwide
(6.3 percent). Since this value is less than the required 12
percent, applying this control level to existing sources would be
an "above the floor" option. Thus, a cost effectiveness analysis
should be provided to justify the California standards as the
existing source floor. Another commenter stated that the
California Highway Patrol, which administers California's cargo
tank testing program, does not include vapor tightness testing in
its 44-point program for inspecting out-of-State cargo tanks.
The commenter felt that this issue could impact the foundation
upon which EPA had reopened the proposal action (IV-D-56, p. 1).
Two commenters had no objections to incorporating the California
standards for both new and existing sources (IV-D-50, p. 2; IV-D-
55, p. 1).
7-4
-------
Several commenters responded to EPA's request for comments
on whether the level of control for cargo tanks at new and
existing facilities should be based on the existing or the
proposed California standards. Commenters were unanimous in
asserting that only the existing, and not the proposed,
California standards should be considered (IV-D-48, p. 1; IV-D-
49, p. 4; IV-D-52, p. 1; IV-D-53, p. 3; IV-D-54, p. 2; IV-D-55,
p. 1). Two of the commenters felt that BAAQMD's survey of 200
tank truck owners was not sufficiently representative to indicate
that the more stringent proposed standards should be applied.
One of them questioned whether data exist to show that a large
percentage (91 percent passing the 0.5-inch level) were achieving
the proposed California control level by August 1992, as required
by section 112(d)(3)(A) of the Act (18 months before proposal of
a standard) (IV-D-48, p. 1). Another commenter said the proposed
requirements should not be adopted because: 1) the testing in
the survey has not been properly peer reviewed, 2) the proposal
has yet to be adopted by the California Air Resources Board
(ARB), and 3) there is no conclusive demonstration of any
significant difference between the current and proposed standards
(IV-D-53, p. 3). Two other commenters echoed that there is no
basis for considering the more stringent standards because the
effect on tank truck emissions is unknown (IV-D-49, p. 4; IV-D-
54, p. 2). Finally, one commenter requested that EPA consider
the proposed California standards for new and existing
facilities, feeling that this would standardize regulations
nationwide and result in lower costs for equipment and remove
some burden from the California ARB (IV-D-47).
Response: The California ARB and the California air
pollution control districts have been implementing cargo tank
leakage standards since the late 1970's. Currently, all cargo
tanks transporting gasoline in California, including tank trucks
from neighboring States that operate in California, must meet the
California standards and are checked by the California air
7-5
-------
pollution control districts. In summary, they include three
major standards: an annual certification test, a year-round
pressure standard for the tank and its vapor piping and hoses,
and a year-round pressure standard for the cargo tank's internal
vapor valve (IV-E-3). The annual certification standards include
initially pressurizing, and later evacuating the tank and
associated vapor piping and hoses, to 18 inches of water and to 6
inches of water, respectively. In 5 minutes the maximum pressure
change can be no more than the values shown in Table 7-1.
Further details on the performance requirements and test
TABLE 7-1. ALLOWABLE CARGO TANK TEST PRESSURE OR
VACUUM CHANGE
Cargo Tank or
Compartment
Capacity, liters
(gal)
Annual Certification-
Allowable Pressure or
Vacuum Change in 5
Minutes, mm H20
(in. H20)
Allowable
Pressure
Change in 5
Minutes at Any
Time, mm H20
(in. H20)
9,4 64 or more
(2,500 or more)
25 (1.0)
64 (2.5)
9,463 to 5,678
(2,499 to 1,500)
38 (1.5)
76 (3.0)
5,679 to 3,785
(1,499 to 1,000)
51 (2.0)
89 (3.5)
3,782 or less
(999 or less)
64 (2.5)
102 (4.0)
procedures used in the California program were discussed in EPA's
supplemental FR notice (59 FR 42788). The EPA's Control
Techniques Guideline (CTG) document for tank trucks (II-A-9) and
the bulk terminal NSPS (40 CFR part 60, subpart XX) contain
annual pressure and vacuum test levels of initial pressures and
test duration which are the same as California's. However, a
less stringent pressure change of 3 inches of water column is
allowed for all tank trucks under the NSPS, the CTG, and the
proposal.
7-6
-------
In the August 19, 1994 supplemental FR notice, EPA stated
that the gasoline throughput in California accounts for nearly 12
percent of the national gasoline consumption (13.46 billion
gal/yr out of 117.9 billion gal/yr) (III-B-1, Table D-2).
Essentially all of this gasoline would be transported by tank
trucks, which include both California and out-of-State cargo
tanks, all of which are subject to the State's vapor tightness
standards (IV-E-8). For this reason, it was assumed that about
12 percent of the national tank truck population is under a
requirement for annual certification and periodic testing in
accordance with the California vapor tightness standards.
On the basis of public comments, however, EPA has examined
the effect of considering the number of terminals in California
on the floor determination. As pointed out by one of the
commenters, California terminals account for 6.3 percent of the
national total. In determining the floor for existing sources,
EPA looks at emission limitations achieved by each of the best
performing 12 percent of existing sources, and averages those
limitations (59 FR 29196). In this case, the "best performing"
cargo tanks are presumed to be those subject to the most
stringent vapor tightness standards. The Agency interprets
"average" to mean a measure of central tendency such as the
arithmetic mean, mode, or median. It can be seen here that on
the basis of the number of terminal facilities, the California
standards meet this test by constituting certainly the 94th
percentile or median, and mode. Therefore, even when the number
of terminals is used in the floor determination, the existing
California standards constitute the floor level of control for
cargo tanks at existing bulk terminals affected by the final MACT
standards. As proposed and discussed in Section 5.4, it has also
been determined that the same tests can be applied to railcars
since they are similar sources. Therefore, the final rule
incorporates the existing California standards for cargo tanks
7-7
-------
(tank trucks and railcars) loading at existing and new
facilities.
Commenters had several concerns on the level of control for
cargo tanks. In the supplemental notice, EPA had discussed
promulgating cargo tank leakage control levels based either on
the existing or the proposed California certification annual leak
rate, 1 in. H20 or 0.5 in. H20 pressure change, respectively.
Some commenters questioned the data collected on the number of
tank trucks meeting the lower proposed California standard as not
representative, not peer reviewed, and not providing a conclusive
demonstration of increased emission reduction. Also, some
commenters were concerned that the proposed standards based on
those data have not at this time been adopted by the California
ARB. The EPA shares the commenters' concerns and is reluctant to
move forward and recommend a final standard based on data the
California ARB has not acted on by adopting and implementing the
standards that have been proposed within the State. Thus, the
Agency is setting the level of cargo tank leak standards for new
and existing facilities on the basis of the existing California
standards.
Comment: One commenter suggested several strategies to
reduce tank truck vapor leakage. These suggestions included the
following: 1) a phase-out of top loading of gasoline; 2) a
minimum inside diameter of 4 inches for vapor return hoses and
cargo tank vapor plumbing; 3) at new sources, a maximum pressure
drop across the rack-to-control device vapor collection system of
5 inches of water during peak loading; 4) at new sources, a
maximum allowable backpressure in the cargo tank of 10.0 inches
of water; and 5) the use of a differential pressure gauge across
in-line flame arrestors, with a maximum allowable pressure drop
specified (IV-D-38, p. 2).
Response: The commenters only suggested these requirements
and indicated that California's Bay Area district plans to
introduce these changes in the near future. Top loading of
7-8
-------
gasoline has not been allowed in the Bay Area since 1980.
Current EPA rules do not specifically disallow top loading.
However, rules requiring vapor recovery and leak-free systems
indirectly rule out the use of top loading since such systems
cannot meet these criteria. The EPA has no additional
information on the number of facilities implementing such
measures or on their environmental impacts. Therefore, these
changes were not incorporated into the final rule.
7.3 CARGO TANK TESTING
Comment: Several commenters expressed opinions concerning
the proposed requirement for annual vapor tightness testing of
tank trucks loading at major source terminals. One commenter
expressed agreement with the requirement for annual testing of
both tank trucks and railcars (IV-D-21, p. 2). Others felt that
the requirement was unnecessary. One stated that terminal owners
and operators should not have to test for-hire tank trucks owned
by others (IV-D-7, p. 11). Other commenters stated that, since
leakage rates have declined over the years, the vapor tightness
testing is unnecessary and these requirements are duplicative of
current Federal and State regulations (IV-D-30, p. 6; IV-D-34, p.
13). Two commenters believe that current DOT testing programs,
with modifications if necessary, sufficiently address the leakage
problem (IV-D-18, p. 5; IV-D-30, p. 7). One of these felt that
inspections by State highway enforcement agencies should be
allowed (IV-D-18, p. 5).
Response: The EPA's data indicate that the vapor loss from
tank trucks connected to a vapor collection system may average 30
percent of the potential vapor transferred to the system in areas
where no regular vapor tightness testing program is in effect.
As discussed in Section 7.1, tank trucks subject to an annual
test requirement lose an estimated 0.8-1.3 percent of the
displaced vapors. Based on these figures, the annual test is
very effective and necessary for reducing this source of HAP
emissions.
7-9
-------
Further, the test does not duplicate DOT programs or Federal
and State requirements. As pointed out in BID, Volume I, Section
4.1.4.2, the current DOT leakage test does not verify the
integrity of some portions of the vapor containing equipment; on
the other hand, DOT allows EPA's Method 27 to be used as an
alternative. Many State rules apply only in certain areas,
leaving some major source HAP facilities potentially unregulated.
As discussed earlier in Section 7.2, annual tank truck testing is
considered the floor for existing sources and it is therefore the
minimum that can be required. These MACT standards establish a
uniform set of requirements, including the use of EPA Reference
Method 27, for tank trucks loading at these facilities in all
areas of the country.
Comment: One commenter felt it would be a burden to
"police" the trucks to verify they have vapor tightness
documentation, and thought that the terminal may need to be
staffed 24 hours per day to comply (IV-D-18, p. 5).
Response: The cargo tank documentation requirements in
these standards were proposed and are now promulgated to apply in
the same way, and in fact are the same requirements, as those in
40 CFR 60, subpart XX, the NSPS for bulk gasoline terminals. The
cargo tank owner (who may be either the terminal owner or an
outside trucking or rail company), in order to load a particular
gasoline cargo tank at an affected terminal, would have to outfit
the tank for vapor collection and include the annual vapor
tightness test in the tank's maintenance schedule. Terminal
owners and operators would not have to "police" the loading
activities by manning the loading racks. The method for assuring
that untested cargo tanks do not load gasoline involves
examination of loading records and notification of cargo tank
owners if unauthorized loadings are made. At bulk terminals
(even automated computer-billed terminals), some hard copy
manifest is given to the driver of a for-hire tank truck or
railcar to verify the date and the type and amount of product
7-10
-------
loaded. The driver keeps this copy for his company's records and
a copy is often returned to the terminal to cross-check the
computer billing. This record containing the tank ID number
allows the terminal operator to identify each tank that loaded if
he desires to do so in the future. Under these standards, the
operator is required to periodically cross-check the tank
identification with the vapor tightness documentation kept on
file at the terminal. This cross-checking is required within 2
weeks of the loading.
If the terminal discovers that an unauthorized cargo tank
has received gasoline, the terminal operator is required to
notify the tank owner, indicating that only vapor-tight (tested)
tanks may load gasoline at that terminal. Steps would then have
to be taken to assure that the unauthorized cargo tank does not
reload at the terminal until documentation of a successful vapor
tightness test has been provided. Terminals have flexibility in
the manner in which these steps are taken as necessary in order
to minimize potential disruptions to the terminal operations.
Comment: In response to EPA's FR notice announcing the
availability of supplemental information and the reopening of the
comment period on the new information (59 FR 42788, August 19,
1994), commenters offered opinions concerning the two field tests
used in California to verify compliance with the year-round
performance standard for cargo tanks: the nitrogen pressure
decay test and the combustible gas detector (CGD) test. Two
commenters felt that the nitrogen test should be voluntary and
not a mandatory requirement for the terminal operator (IV-D-49,
p. 3; IV-D-52, p. 2). One commenter on the CGD test thought that
this test should be used only as a screening tool to search for
leaks, but it should not be used to issue violations, due to its
questionable correlation with the actual leakage rate as
determined in pressure decay testing (IV-D-49, p. 3). Another
commenter discussed data showing the failure of the CGD test in
correctly identifying the compliance status of tank trucks, and
7-11
-------
agreed that the test should be considered a screening tool. This
commenter also felt that the use of this test during loading is
inherently unsafe, because a hatch cover could blow open while
the tester is on top of the tank truck. The preferred method is
for personnel to do the testing after the actual filling process
has ceased (IV-E-7).
Response: In the supplemental FR notice, EPA concluded that
the existing California standards are the floor for cargo tanks
loading at existing bulk terminals. The nitrogen test is part of
the California program. The test is utilized in two ways.
First, the local control districts perform the test to verify the
compliance of cargo tanks with the year-round performance
standard. When performed as a compliance evaluation, this test
of course is not "voluntary." The second way the test is applied
is by terminal facilities themselves, which install the test
apparatus at the facility and perform the testing under their own
schedule as part of their internal tank truck maintenance
program. This is a voluntary activity which is intended to catch
leaks and to maintain an acceptable level of compliance. Despite
the terminal's own level of test (maintenance) activity, any
gasoline cargo tank is subject to testing by the control agency
at any time. The EPA is including the same nitrogen test in the
final rule. As in the California program, EPA encourages
affected source terminals (and cargo tank operators who will be
loading at these facilities) to institute programs designed to
ensure that continuous year-round compliance is maintained.
The EPA agrees that the leak detection test using a
combustible gas detector is best characterized as a screening
tool, and the reported low correlation of this test with the
quantitative pressure loss tests is a concern to the Agency. For
this reason, both the California standards and this final rule do
not consider a cargo tank out of compliance and subject to
penalty on the basis of failure of the CGD test alone. However,
a tank failing this test will not be allowed to load until it
7-12
-------
demonstrates compliance by passing a pressure decay test (before
any leak repair is performed on the tank). Any tank failing the
decay test will be penalized and then required to certify to the
annual certification standard (with a new 12-month clock
starting) before it is authorized to load gasoline at a regulated
facility.
7.4 OTHER
Comment: Two commenters suggested that EPA delete the
proposed requirements for railcars, re-do the analysis, and focus
on regulating railcar loading as a separate subcategory. They
said EPA has presented little information regarding controls for
railcar loading. The commenters believe that the top 12 percent
of existing terminals where railcar loading exists probably do
not control railcar loading; therefore, "uncontrolled" should be
the floor. Railcar loading at new terminals should not be
regulated beyond the current NSPS subpart XX standard of 35
mg/liter of gasoline loaded, if any standard is required at all
(IV-D-2, p. 8; IV-D-17, p. 4).
Response: The EPA agrees that the data base for facilities
loading gasoline into railcars, including data on vapor control
operations at these facilities, is limited. However, vapor
control with a 10 mg/liter emission limit is being achieved in
practice at facilities in parts of California, Texas, and
Washington. Although there are obvious construction differences
between tank trucks and railcars, both are equipped with similar
cargo tanks, have the same displacement losses of HAP, and
employ similar loading and vapor control equipment. Since this
standard is currently being practiced at some facilities, the
emission limit of 10 mg/liter is the appropriate MACT control
level applicable to new railcar loading sources. Also, since the
floor for existing tank truck loading operations is an emission
limit of 10 mg/liter, the similar operation of railcar loading
with the same emission characteristics can also be presumed to be
capable of achieving the same 10 mg/liter level of emission
7-13
-------
control. Therefore, EPA is promulgating the 10 mg/liter standard
for all cargo tanks, which includes both tank trucks and
railcars.
In addition to this comment, a few miscellaneous comments
were received in response to EPA's proposal to incorporate
California's cargo tank vapor tightness program into these MACT
standards.
Comment: One commenter supported the adoption of the 1-inch
pressure drop annual standard and the 2.5-inch pressure drop
year-round standard for all cargo tanks regardless of tank
capacity as a compliance option to be used at the discretion of
the terminal owner or operator to streamline applicability and
recordkeeping (IV-D-53, p. 3).
Response: The California 5-minute annual certification
pressure change ranges from 1 to 2.5 inches H20, and the year-
round standard ranges from 2.5 to 4.0 inches H20, depending on
the capacity of the cargo tank or compartment being tested. The
lower levels of allowable pressure change correspond to the
largest cargo tanks (2,500 gallons and up). For smaller tanks,
the higher pressure losses that are specified correspond to an
equivalent vapor leakage rate. Therefore, application of the
lower pressure change limits to the smaller tanks would be a more
stringent option, and would be above the floor. For this reason,
EPA is promulgating the California limits in this final rule.
However, as a compliance option intended to simplify these
procedures, individual companies or facilities may choose to
apply a single (the most stringent) level to all sizes of cargo
tanks.
Comment: The same commenter felt that applying this cargo
tank proposal to existing sources would create a problem of
inequitable treatment for bulk terminals nationwide. Tank truck
owners and operators would prefer to load at unaffected terminals
due to less burdensome testing and paperwork requirements. The
commenter suggested that such a problem would not be created if
7-14
-------
the existing DOT and NSPS subpart XX requirements were extended
to major HAP sources that are not already covered under those
standards (IV-D-53, p. 4).
Response: As discussed in Section 7.2, EPA has determined
the California level of control to be the floor for all gasoline
cargo tanks (tank trucks and railcars) loading at new and
existing bulk terminals. The differences between the testing and
recordkeeping requirements under the NSPS versus this MACT
program are relatively minor; however, additional emission
reductions are expected under the MACT standards due to increased
maintenance and repair activity on cargo tanks. The EPA does not
believe that a significant number of cargo tank operators will
forfeit their access to major bulk terminals in exchange for a
slightly less stringent maintanance program for their fleet.
Comment: Two commenters contended that mandating the
California testing standards for existing major source terminals
will create a de facto new nationwide cargo tank testing
standard. Maintaining appropriate records as to which cargo
tanks may load at individual terminals would be excessively
burdensome and would unacceptably reduce critical transportation
and terminal operational flexibility. Accordingly, terminals and
truck companies would be compelled to test, unnecessarily, all
cargo tanks to the California standards (IV-D-49, p. 4; IV-D-51,
p . 2 ) .
Response: The response to this comment is similar to the
previous response. Under the NSPS subpart XX, bulk terminals
must keep records indicating their own or for-hire tank trucks
that are certified as vapor-tight. The EPA is not aware of this
practice being an excessive burden for the NSPS terminals. Many
cargo tank operators would undoubtedly certify their equipment to
the more stringent levels in order to increase their flexibility
to load at any terminal facility. However, many others would
probably not find this necessary or desirable. To the extent
that operators choose to apply enhanced maintenance standards,
7-15
-------
additional HAP emission reductions should be realized.
7.5 REFERENCES
II-A-9 Control of Volatile Organic Compound Leaks
from Gasoline Tank Trucks and Vapor
Collection Systems. U.S. Environmental
Protection Agency, Research Triangle Park,
NC. E PA-450/2-78-051. December 1978.
II-A-12 Norton, R.L., Pacific Environmental Services,
Inc. Evaluation of Vapor Leaks and
Development of Monitoring Procedures for
Gasoline Tank Trucks and Vapor Piping.
Prepared for U.S. Environmental Protection
Agency, Research Triangle Park, NC. EPA-
450/3-79-018. April 1979.
II-A-14 Bulk Gasoline Terminals - Background
Information for Proposed Standards. U.S.
Environmental Protection Agency, Research
Triangle Park, NC. EPA-450/3-80-038a.
December 1980.
II-A-25 Leak Testing of Gasoline Tank Trucks. Scott
Environmental Technology, Inc. Prepared for
U.S. Environmental Protection Agency,
Research Triangle Park, NC. August 1978.
III-B-1 Gasoline Distribution Industry (Stage I) -
Background Information for Proposed
Standards. U.S. Environmental Protection
Agency, Research Triangle Park, NC. EPA-
453/R-94-002a. January 1994.
IV-D-42 Letter from Strieter, R.P., American
Petroleum Institute, Washington, DC, to
Shedd, S.A., U.S. Environmental Protection
Agency, Research Triangle Park, NC. May 12,
1994. Transmits information on storage tank
emissions and revised tank truck fugitive
emission rates.
IV-E-3 Telecon. LaFlam, G., PES, Inc., and Shedd,
S., U.S. Environmental Protection Agency,
7-16
-------
Chemicals and Petroleum Branch, with Loop,
J., California Air Resources Board. June 14,
1994. State of California's gasoline cargo
tank certification and testing requirements.
IV-J-14 Ward, C.C. Gaseous Fuels. In: Marks'
Standard Handbook for Mechanical Engineers.
New York, McGraw-Hill. 1978. pp. 7-21 to
7-24 .
7-17
-------
8.0 STORAGE VESSEL STANDARDS
8.1 DETERMINATION OF MACT FLOOR CONTROL LEVEL
Comment: Several commenters claimed that the discussion in
the proposal concerning the "floor" level of control for storage
vessels was inadequate and unclear. The EPA's conclusion was
that the NSPS requirements of 40 CFR part 60, subpart Kb
constituted the floor for storage vessels at existing sources.
One commenter stated that EPA had not performed an adequate
evaluation to establish the subpart Kb deck rim seal requirements
as the floor (IV-D-15, p. 9). Several other commenters believed
that EPA had demonstrated only that subpart Kb's floating roof
rim seal requirements are the floor for existing sources, but not
the additional subpart Kb requirement to install controls on the
roof deck fittings (IV-D-3, p. 1; IV-D-7, p. 7; IV-D-8, p. 4; IV-
D-12, p. 1; IV-D-15, p. 9; IV-D-16, p. 2; IV-D-17, p. 5; IV-D-22,
p. 29; IV-D-31, p. 3; IV-D-34, p. 16; IV-D-35, p. 4; IV-F-1, p.
9, 12). Other commenters expressed concerns regarding the cost
effectiveness of controlling the deck fittings (IV-D-3, p. 1; IV-
D-8, p. 4; IV-D-15, p. 9; IV-D-17, p.7; IV-D-22, p. 38; IV-D-34,
p. 17; IV-D-35, p. 50).
Response: At proposal, EPA required gasoline storage tanks
at existing facilities to comply with the storage vessel
standards in NSPS 40 CFR part 60, subpart Kb (NSPS subpart Kb).
NSPS subpart Kb specifies closure devices between the wall of the
storage vessel and the edge of the floating roof ("rim seals")
and the installation of gaskets on specified lids and other
openings in the floating deck ("controlled fittings"). The EPA
8-1
-------
also proposed the same NSPS subpart Kb requirements as the floor
for tanks at new facilities. NSPS subpart Kb is the most recent
(1984) new source performance standard for new gasoline and
certain other liquid storage tanks.
Regarding the comments concerning the floor determination
for rim seal requirements for existing sources, EPA continues to
maintain its previous conclusion that the NSPS subpart Kb rim
seal requirements are the floor for gasoline distribution
facilities as proposed and discussed in the proposal notice (59
FR 5868). The EPA believes it did perform a proper evaluation,
and the commenter did not provide any data or information to
support a change in the finding that NSPS subpart Kb rim seals
are the floor level of control. The rim seal requirements in the
CTG's and the requirements of the NSPS (40 CFR part 60, subparts
K, Ka, and Kb) are identical for both internal and external
floating roofs having mechanical and liquid-mounted primary
seals. The rim seal requirements for the CTG's and the NSPS
subparts Ka and Kb differ only for internal or external floating
roofs equipped with a third type of seal, a vapor-mounted primary
seal. Vapor-mounted rim seals must be equipped with secondary
seals to meet the requirements of NSPS subpart Ka and the CTG for
external floating roof tanks and to meet requirements of NSPS
subpart Kb for internal floating roof tanks. (NSPS subpart Kb
does not allow any vapor-mounted seals on external floating
roofs.) Approximately 35 percent and 40 percent, respectively,
of external floating roof tanks at breakout stations and bulk
terminals have primary and secondary seals meeting some level of
NSPS and CTG control. Approximately 40 percent and 70 percent of
internal floating roof tanks at breakout stations and bulk
terminals, respectively, are equipped with floating roofs with
seals meeting some level of NSPS and CTG control.
Vapor-mounted seals are the only source of uncertainty in
concluding that over 35 percent of all gasoline storage tanks
meet NSPS subpart Kb rim seal requirements, because no data are
8-2
-------
available to estimate the number of tanks with vapor-mounted
seals. However, since this seal type is one of three types
available, it is reasonable to assume that one-third of the tanks
may have vapor-mounted seals. Thus, the Agency concludes that
over 20 percent of the tanks currently meet NSPS subpart Kb
control levels. Additionally, analysis of the tank survey for
the proposed refinery MACT standard (59 FR 36130, July 15, 1994)
found that the rim seal requirements of the NSPS subpart Kb was
the floor for existing storage tanks at refineries. Those
refinery tanks are same type of storage tanks as those used at
terminals and breakout stations; however, on average they
normally contain liquids of lower volatility. Therefore, it is
reasonable to assume that a higher percentage of gasoline storage
tanks would be better controlled due to the value of the gasoline
that would be subject to loss through evaporation. Based on
these arguments, EPA concludes that the rim seal requirements in
the most recent NSPS standard (40 CFR part 60, subpart Kb)
represent the average limitation achieved by the best performing
12 percent of sources.
The EPA, however, does agree with the commenters' statements
that the discussion in the proposal preamble did not support the
NSPS subpart Kb fitting control requirements set in 1984 for new
tanks as part of the floor for storage vessels at existing
facilities. The EPA did not have access to any data regarding
the number of gasoline storage vessels that are equipped with
controlled fittings. The commenters also did not provide any
data or information on the number of storage vessels with or
without fitting controls for these subcategories. Information
obtained in the tank survey conducted for the refinery MACT
standards was inconclusive regarding the use of controlled
fittings on storage vessels. As a result, EPA has no data to
support the conclusion that controls on tank fittings are part of
the floor for existing sources. Therefore, EPA has determined
the existing source floor for fittings as "uncontrolled."
8-3
-------
The Agency has considered controlled fitting requirements as
an option providing the maximum degree of reduction in HAP
emissions ("above the floor") as required by the Act. The
Administrator is required under section 112(d) to set emission
standards for new and existing sources of HAP's that require the
maximum degree of reduction in emissions of HAP's that is
achievable, taking into consideration the cost of achieving the
emission reduction, any nonair quality health and environmental
impacts, and energy requirements. New tanks at new or existing
facilities since 1984 are meeting the deck fitting control
requirements in 40 CFR part 60, subpart Kb and, therefore, these
requirements are achievable. Controlling fittings to that level
is also considered the maximum degree of emission reduction.
Emission reductions and costs for controlled fittings
were analyzed on both a per model storage vessel and a nationwide
basis using two typical size and throughput vessels, and
different potential HAP contents in gasoline. Additionally,
installation of controlled fittings on many tanks requires
degassing and cleaning of the tanks. Industry reports that
storage vessels are degassed and cleaned at least every 10 years
for safety inspections and requested that EPA require all
retrofits (fittings and rim seals) on storage tanks to occur
simultaneously. Therefore, EPA's new analysis included two
options, with and without degassing and cleaning costs. If
fitting controls were required in the first 3 years, the cost
impact for this standard would include the cost for tank
degassing and cleaning along with the cost of installing
controlled fittings if a tank's routine safety inspection would
not have occurred during that 3-year time period. The option of
waiting until the next routine tank degassing and cleaning would
avoid the additional costs of cleaning and degassing as an impact
of this standard since the activity would have occurred anyway.
A discussion and presentation of the model tank analysis of
fitting controls is included in Appendix B.
8-4
-------
Installing controlled fittings on floating roof tanks,
without degassing and cleaning costs, would result in a cost
savings due to the value of gasoline vapor prevented from
evaporating through openings in the floating roof deck. The
capital costs of installing deck fitting controls on the model
tanks, without the cost of degassing and cleaning of the tanks,
ranged in the analysis from $1,200 to $2,800, annualized costs
ranged from a savings to a cost of $340 per year, and the cost
effectiveness ranged from a savings to a cost of $7,500 per
megagram of HAP reduced. When controlled deck fitting
installation costs included degassing and cleaning costs, the
capital costs ranged from $21,000 to $67,000, annualized costs
ranged from $4,000 to $14,000 per year, and the cost
effectiveness ranged from $25,000 to $300,000 per megagram of HAP
reduced. Calculation of product price increases under either
option showed them to be insignificant (less than 0.05 cent per
gallon). In conclusion, installing controlled deck fittings is
significantly less costly if it can be done at the next scheduled
tank degassing and cleaning.
The Agency has decided to require installation of controlled
deck fittings on each existing external floating roof storage
tank that is required to be degassed and taken out of service for
the purpose of replacing or upgrading rim seals to meet 40 CFR
60, subpart Kb requirements. Since these tanks must be degassed
and cleaned and have plant maintenance personnel on site, it is
reasonable to require installation of the fitting controls at the
same time. A national impact analysis was performed on this
requirement. Table D-l in Appendix D presents the results of the
national analysis on storage tanks and other emission sources at
bulk terminals and pipeline breakout stations. Installing
fitting controls on external floating roof tanks is estimated to
reduce 66 megagrams per year of HAP at an annualized cost savings
of $93,000 per year.
8-5
-------
The cost analyses show that installing controlled fittings
when installing or replacing rim seals on existing external
floating roof tanks involves a small capital cost (approximately
$2,000 per tank), with an annualized cost savings, and
insignificant change in gasoline prices. Given these low costs
and the simplicity of these control measures when tanks are
otherwise out of service, EPA has concluded that fitting controls
are practical and affordable for existing external floating roof
storage tanks. These controls also prevent pollution and
conserve energy by preventing liquid gasoline from evaporating.
Having given full consideration to the directives in the Act, the
Administrator is requiring existing facilities to control the
deck fittings when replacing or installing rim seals on external
floating roof storage tanks to comply with the requirements in
the final rule. Given the small national HAP emission reduction,
the Agency has decided not to require fitting controls on
existing internal floating roof storage tanks. While EPA is not
at this time requiring these controls nationally on internal
floating roofs, EPA encourages industry to consider the
installation of these controls on a case-by-case basis. All new
storage tanks at both new and existing facilities are already
required under NSPS requirements of 40 CFR part 60, subpart Kb to
install these same fitting controls. Those NSPS requirements are
cross-referenced and are therefore part of the final rule. This
level of control for roof deck fittings for new sources and for
existing external floating roof tanks upgrading to rim seal
requirements under this rule, is the same level as proposed on
February 8, 1994. The storage vessel compliance period is
discussed and analyzed in the next section.
While the final rule does not require fitting controls for
existing internal floating roof storage tanks or the existing
external floating roof storage tanks currently meeting the rim
seal requirements in this rule, the Agency believes it is
appropriate and recommends the inspection, repair, and upgrading
8-6
-------
of gasketing materials on fittings in the tank roof when any
storage tank is taken out of service. It is a major part of the
normal safety and maintenance procedure to inspect, repair, and
upgrade the physical and mechanical condition of all the tank
components. Additionally, requiring fitting controls to be
installed on all tanks will reduce additional air toxics and
volatile organic compounds, and will upgrade all tanks to the
same level of control. An effective mechanism for getting
controlled fittings in place on all tanks is the combination of
this rule, the air toxics programs under section 112(1) of the
Act, and the national ambient air quality programs for control of
ambient ozone under the Act.
8.2 COMPLIANCE DEADLINE
Comment: Several commenters said that the proposed 3-year
compliance period for storage vessels is unreasonable and is more
stringent than the compliance schedule in other Federal
regulations. To install the required controls, tanks would have
to be taken out of service, cleaned, and degassed. Requiring all
storage tanks to comply in a 3-year period could potentially
disrupt the nation's gasoline supply, causing a gasoline
shortage, especially in light of the new reformulated/oxygenated
fuel requirements. One commenter stated that limited contractor
resources could make the schedule logistically unworkable.
Additionally, the cleaning and degassing of a storage tank
creates an air emissions event that in many cases will exceed the
emission reductions resulting from the new controls (e.g., the
retrofit of internal floating roof already meeting subpart Ka
seal requirements). All of the commenters suggested that EPA
relax the compliance schedule and allow storage tank owners and
operators to comply at the next scheduled tank inspection or in
10 years, whichever comes first. One of the commenters felt that
a 10-year period has been established as an integral part of the
floor for existing sources, and therefore EPA must perform a cost
8-7
-------
effectiveness analysis to support a 3-year compliance period.
This commenter recommended that, should EPA not allow up to 10
years for compliance for all tanks currently equipped with
floating roofs, at a minimum internal floating roof tanks
currently meeting NSPS subpart Ka requirements should be provided
a compliance period up to 10 years, or the next regular
inspection cycle, whichever occurs first (IV-D-2, p. 8; IV-D-3,
p. 1; IV-D-5, p. 4; IV-D-6, p. 2; IV-D-7, p. 7; IV-D-8, p. 4; IV-
D-12, p. 1; IV-D-14, p. 9; IV-D-15, p. 9; IV-D-16, p. 2; IV-D-17,
p. 3; IV-D-18, p. 6; IV-D-22, p. 33; IV-D-29, p. 4; IV-D-30, p.
7; IV-D-31, p. 3; IV-D-32, p. 2; IV-D-34, p. 17; IV-D-35, p. 4;
IV-D-36, p. 3; IV-F-1, p. 10).
Response: Section 112(i)(3) of the Act requires the
Administrator to establish a compliance date which shall provide
for compliance as expeditiously as practicable, but in no event
later than 3 years after the effective date (promulgation) of the
standards. In addition, the Administrator (or a State with a
program approved under title V) may issue a permit which grants
up to a 1-year extension to comply with the standards if an
additional period is necessary for installation of controls.
However, some commenters suggest that taking a tank out of
service before its normal cleaning and inspection schedule to
comply with the regulation may generate more emissions than the
added controls would reduce or control in the 3-year period.
To determine whether any tanks should be allowed an
extension of the compliance time to achieve the maximum degree of
reduction in emissions of HAP's, EPA compared the emission
reductions achieved by the controls (i.e., rim seals and fitting
controls) to the emissions generated from degassing and cleaning
of fixed-roof and internal and external floating roof tanks for
various tank diameters and gasoline turnover rates. The results
of this analysis showed that degassing and cleaning emissions do
not exceed the emission reductions from tanks complying with the
final rule within the required 3-year compliance period. The
8-8
-------
analysis did show net emissions increases for some very large
tanks either installing secondary seals without installing
fitting controls, or installing fitting controls alone. However,
the final standards require a facility to install fitting
controls when installing secondary rim seals, and no tanks are
required to install fitting controls alone. A discussion of
emissions generated from tank cleaning and degassing, as well as
an analysis of the environmental impacts of storage vessel
controls, is presented in Appendix B.
8.3 OTHER
Comment: One commenter felt that the proposed rule overlooked
the distinction between "breakout tanks" and bulk terminal
"storage tanks." A "breakout tank" is a tank used to:
(a) relieve surges in a hazardous liquid pipeline system, or
(b) receive and store hazardous liquid transported by a pipeline
for reinjection and continued transportation by pipeline. The
commenter felt this distinction has been historically recognized
by EPA and the Department of Transportation (DOT) in developing
regulations, and clarifies the distinction between facilities
regulated by these two agencies. In its preamble to proposed
changes to the SPCC regulation plan, EPA noted that EPA regulates
facilities with bulk storage tanks, while DOT regulates breakout
tanks (56 FR 54612). The commenter pointed out that the
regulations developed by DOT impose design and maintenance
specifications for breakout tanks (§49 CFR sections 195.192 and
192.264) (IV-D-1, p. 1).
Response: The EPA proposed to regulate storage tanks at
both gasoline bulk terminals and breakout stations to the level
of the 40 CFR part 60, subpart Kb requirements. The commenter
cited a proposed rule entitled "Oil Pollution Prevention; Non-
transportation related Onshore and Offshore facilities" (56 FR
54612), which was promulgated under section 311(j)(1)(C) of the
Clean Water Act, as amended by the Oil Pollution Act of 1990.
This proposed rulemaking would establish requirements for Spill
8-9
-------
Prevention and Control Countermeasures (SPCC) Plans to prevent
oil spills by non-transportation related onshore and offshore
facilities into the waters of the United States or adjoining
shorelines. The following definitions were proposed:
Breakout tank means a container that is
part of a pipeline facility regulated by the
Department of Transportation and is used
solely for the purpose of compensating for
pressure surges or to control and maintain
the flow of oil through pipelines. Such
tanks are frequently in-line.
Bulk storage tank means any container
used to store oil. These tanks are used for
purposes including, but not limited to, the
storage of oil prior to use, while being
used, or prior to further distribution in
commerce.
In that rulemaking, EPA excluded facilities under the
jurisdiction of DOT (transportation-related onshore and offshore
facilities including breakout tanks) from regulation based on a
previous EPA and DOT Memorandum of Understanding (MOU) dated
November 24, 1971, which established the responsibilities of EPA
and DOT for the purposes of administering their respective spill
prevention programs. The commenter also cited 49 CFR §195.192
and §192.264 as regulations developed by DOT for the
transportation of hazardous liquids and pipeline facilities.
Although the purpose of the MOU between EPA and DOT was to
delegate responsibility for administering their respective spill
prevention programs, for the purposes of this rulemaking, EPA
does not consider the MOU between EPA and DOT to be applicable to
regulations developed to control air emissions. As an example,
40 CFR part 60, subpart Kb, which is the NSPS for volatile
organic liquid storage vessels built or reconstructed after 1984,
does not exempt "breakout tanks" from regulation. The EPA
considers breakout tanks to be subject to the subpart Kb
requirements. For this final rulemaking, EPA will continue to
8-10
-------
regulate storage tanks at pipeline breakout stations (or
"breakout tanks") to control HAP emissions. At the same time,
EPA has modified the definition of "pipeline breakout station" in
§63.421 to more fully describe and distinguish breakout stations
from bulk terminal facilities.
Comment: One commenter felt that the exemption for certain
pressure vessels in §60.110b(d)(2) should be retained in the
reference to NSPS subpart XX (IV-D-4, p. 2).
Response: Section 60.110b(d)(2) of NSPS subpart Kb exempts
pressure vessels designed to not operate in excess of 204.9 kPa
and without emissions to the atmosphere. The EPA agrees that
storage vessels meeting these specifications should not be
covered by this rulemaking, so this exemption is retained in the
final rule.
Comment: One commenter felt that EPA should not specify
vapor control technology (internal sealed floating roofs) for
fixed-roof storage tanks, but should instead develop allowable
vapor emission standards. This would recognize the use of vapor
balancing through a common vapor collection pipe with the loading
racks, a practice which may have some advantages over internal
floating roofs. The commenter stated that the overall vapor
control efficiencies of these two approaches are comparable (IV-
D-9, p. 10; IV-F-1, p. 25).
Response: NSPS subpart Kb allows the use of a closed vent
system and control device if the system meets the following
specifications stated in section 60.112b(a)(3):
"The closed vent system shall be designed to
collect all VOC vapors and gases discharged from the
storage vessel and operated with no detectable
emissions indicated by an instrument reading of less
than 500 ppm above background and visual inspections,
as determined in part 60, subpart W, section
60.485(b).
The control device shall be designed and operated
to reduce inlet VOC emissions by 95 percent or greater.
If a flare is used as the control device, it shall meet
the specifications described in the general control
8-11
-------
device requirements (section 60.18) of the General
Provisions."
For the purposes of this rulemaking, EPA proposed and is
promulgating that a closed vent system and control device meeting
the requirements in §112b(a)(3) are equivalent and an acceptable
alternative control technology to reduce HAP emissions from
storage vessels.
Comment: One commenter stated that OSHA's confined-space
regulations make conducting annual primary and/or secondary seal
gap measurements on internal floating roof tanks unnecessarily
burdensome and unwarranted. The commenter suggested that visual
inspections of seal condition be performed annually (or
quarterly), and actual seal gap measurements be conducted every 5
years. More frequent inspections should be reserved only for
tank seals greater than 10 years old (IV-D-18, p. 6).
Response: The EPA is requiring the same storage tank
inspection requirements stated in NSPS subpart Kb (§60.113b).
Subpart Kb does not require annual seal gap measurements for
internal floating tanks as the commenter suggests. Instead,
subpart Kb requires facilities to perform annual visual
inspections of internal floating roof tanks equipped with a
liquid-mounted or mechanical shoe-mounted primary seal and
secondary seal (if one is in service) through manholes and roof
hatches on the fixed roof. Owners or operators of external
floating roof tanks must measure the gaps between the tank wall
and primary seal at least once every 5 years. Similarly, tank
owners or operators must measure the gaps between the tank wall
and secondary seal at least once per year.
Comment: One commenter was concerned with the requirement
in proposed §63.425(e) [referencing §60.113b(a)(2)] to repair
storage tanks within 45 days following an inspection. This
commenter believed that this time period is insufficient, and
that repairs should be made on an annual basis, or at least "as
soon as practicable" (IV-D-20, p. 11).
8-12
-------
Response: As stated earlier, subpart Kb rim seal
requirements are considered to be the floor for storage vessels.
As proposed, EPA will continue to require owners and operators of
storage vessels to comply with the testing and procedures
requirements of §60.113b to ensure continuous compliance with the
rule. It should be noted that §60.113b contains the provision
that if a failure detected during an inspection cannot be
repaired within 45 days and if the vessel cannot be emptied
within 45 days, the owner or operator may request a 30-day
extension from the Administrator in the inspection report
required in §60.115b(a)(3). The request for an extension must
document that alternate storage capacity is unavailable and
specify a schedule of action which will ensure that the control
equipment will be repaired or the vessel will be emptied as soon
as possible.
8-13
-------
9.0 EQUIPMENT LEAK STANDARDS
9.1 EMISSION FACTORS
Comment: Several commenters strongly objected to EPA's use
of 1980 refinery data to estimate emissions from equipment
(pumps, valves, etc.) at bulk terminals and pipeline breakout
stations. These commenters were in support of using the new
American Petroleum Institute (API) data gathered at several bulk
terminals. These data indicate that leakage from bulk terminal
equipment is very small and that the refinery emission factors
overestimate these emissions greatly. The commenters pointed out
that EPA's emission factors were extremely high and that this
would lead to incorrect calculations of applicability status and
baseline emissions (IV-D-4, p. 3; IV-D-7, p. 8; IV-D-11, p. 2;
IV-D-12, p. 1; IV-D-13, p. 2; IV-D-14, p. 1; IV-D-16, p. 2; IV-D-
18, p. 7; IV-D-19, p. 2; IV-D-22, p. 7; IV-D-26, p. 2; IV-D-30,
p. 5; IV-D-31, p. 3; IV-D-34, p. 10; IV-D-35, p. 4; IV-F-1, p. 6,
12, 17) .
Response: At proposal, EPA used the refinery equipment
emission factors in publication AP-42 (II-A-17) to estimate
emissions from equipment components at marketing terminals and
pipeline breakout stations. The API supplied new data contained
in Appendices A and C of their comments (IV-D-22) which indicated
that corresponding emission factors for marketing terminals and
breakout stations are over 99 percent lower. The EPA has
reviewed the data submitted by API. In May 1994, EPA released a
draft report presenting new correlation equations for marketing
terminals using the API data. The Agency is still reviewing and
9-1
-------
analyzing the API data to determine new EPA emission factors.
For the purposes of this analysis and completion of the final
rule, API's suggested emission factors are being used because in
the Agency's judgment these new factors better reflect emissions
from this source category than the 1980 refinery data. The EPA
intends to issue new EPA emission factors in the near future.
9.2 NEED FOR LDAR STANDARDS
Comment: Several commenters expressed disagreement with the
proposal to require a leak detection and repair (LDAR) program at
major source bulk terminals and pipeline breakout stations. Many
believe that the baseline emissions resulting from fugitive leaks
are much smaller than EPA has estimated. Consequently, the
commenters considered EPA's estimated emission reductions due to
an LDAR program to be greatly overstated. As a result, the LDAR
program would provide an extremely small benefit (if any) and is
"infinitely cost-ineffective." In lieu of an LDAR program, many
commenters believe that a mandatory visual inspection program
(similar to existing programs at many terminals) would be more
appropriate. The API performed a leak rate survey at bulk
terminals, including both terminals where an LDAR program was in
effect and terminals that were not carrying out a formal LDAR
program. The API's conclusion was that there was no
statistically significant difference in the leak rates found at
the two groups of terminals. The commenters concluded that LDAR
programs are more appropriate for refineries, where equipment
handles fluids at higher temperatures and pressures (IV-D-2, p.
9; IV-D-3, p. 1; IV-D-4, p. 2; IV-D-5, p. 2; IV-D-7, p. 9; IV-D-
12, p. 1; IV-D-13, p. 2; IV-D-14, p. 1; IV-D-15, p. 1; IV-D-16,
p. 1, 2; IV-D-17, p. 2; IV-D-18, p. 7; IV-D-19, p. 2; IV-D-20, p.
7; IV-D-22, p. 12; IV-D-26, p. 2; IV-D-30, p. 4; IV-D-31, p. 3;
IV-D-34, p. 11; IV-D-35, p. 3; IV-F-1, p. 5, 16). One commenter
expressed support for the imposition of an LDAR program at
gasoline distribution facilities (IV-D-21, p. 2).
9-2
-------
Response: Before proposal of this MACT regulation, EPA
learned that few existing terminals and pipeline breakout
stations (less than 1 percent) routinely use a portable organic
vapor analyzer (OVA) to carry out leak detection and repair
programs (LDAR) on their gasoline handling equipment. As a
result, the "floor" for control of equipment leaks at existing
terminals was found to be periodic visual inspections (no formal,
federally enforceable inspection program). A monthly LDAR
program using an OVA was determined to be in practice at a few
terminals associated with refineries and therefore was determined
to be the floor for equipment at new terminals and breakout
stations. As stated earlier, EPA in the proposal analysis used
the refinery emission factors in AP-42 to calculate baseline
emissions from equipment leaks at existing facilities and
analyzed instrument LDAR as an "above the floor" option. The EPA
found LDAR to be cost effective; however, the Agency noted that
there were industry concerns with the refinery factors and so did
not select the higher emission reduction alternative (monthly
instead of quarterly LDAR). As discussed in the previous section
of this chapter, after reviewing equipment leak data submitted by
API, EPA agrees that the equipment leak factors at marketing
terminals are much lower than the refinery factors, resulting in
much lower potential emission reductions due to an LDAR program.
As a result of this determination, the cost effectiveness of a
formal instrument LDAR program has been found to be much less
favorable for gasoline marketing facilities.
The new gasoline distribution equipment leak data submitted
by API showed only a slight difference (0.2 percent) between
emission factors at facilities performing periodic LDAR (with an
instrument) and facilities with a periodic visual program. Based
on its review of these data, EPA agrees with API's assessment
that this difference is statistically insignificant. Therefore,
EPA is in agreement with the majority of commenters that periodic
9-3
-------
visual inspection and LDAR programs achieve essentially equal
emission reductions for these facilities.
Industry submitted survey information that 81 percent of
terminal facilities are implementing some type of periodic visual
inspection program. The survey data did not show the frequency
of visual inspections, but API has stated that current industry
periodic visual programs range in frequency from daily to
quarterly. The API suggested a quarterly program and provided
language to make it enforceable and verifiable through
recordkeeping. The suggested API program included: 1) a
quarterly determination of leaks by visual, audible, and
olfactory inspection of pumps and valves; 2) a log book listing
all of the equipment in gasoline service; 3) note all non-
inspected equipment; 4) if a leak is detected: (a) repair as soon
as practical (considering safety) , (b) if the leak cannot be
repaired immediately, then the leak must be repaired or the
equipment replaced within 15 calendar days, unless not practical
for reasons stated in the log book, or (c) when possible, use of
the leaking equipment is to be suspended; 4) annual checks of log
book by facility supervisor; and 5) quarterly logs and records of
annual checks retained for 5 years and accessible within 3
business days.
The NSPS for bulk terminals [40 CFR part 60, subpart XX,
§60.502(j)] requires monthly inspection of loading racks as
follows:
"(j) Each calendar month, the vapor collection
system, the vapor processing system, and each loading
rack handling gasoline shall be inspected during
loading of gasoline tank trucks for total organic
compounds liquid or vapor leaks. For the purposes of
this paragraph, detection methods incorporating sight,
sound, or smell are acceptable. Each detection of a
leak shall be recorded and the source of the leak
repaired within 15 calendar days after it is detected."
The visual inspection program in the final rule is similar to
these NSPS provisions; however, the provisions have been expanded
9-4
-------
based on suggestions of the commenters and certain requirements
in existing Federal LDAR regulations. As in the NSPS, a monthly
inspection using sight, sound, and smell is required. Each
detection of a leak is to be recorded in a log book. Leaks must
be repaired as soon as practicable, but with the first attempt at
repair made no later than 5 calendar days after detection, and
repair completed within 15 days after detection. Delay of repair
is allowed upon demonstration to EPA that timely repair is not
feasible. Full records of each inspection are required,
including for each leak a record of the date of detection, nature
of the leak and detection method, dates of repair attempts and
methods used, and details of any delays of repairs.
The final rule contains a requirement for both new and
existing facilities to perform a visual inspection of equipment
on a monthly basis because it is achieved in practice on the same
and similar equipment as required under the above 40 CFR part 60,
subpart XX requirements [§60.502(j) ] and at some facilities that
are covered under monthly LDAR programs in response to 40 CFR
part 60, subparts W and GGG, and 40 CFR part 61, subparts J and
V. As noted earlier, the emission reductions resulting from
these visual inspection programs have not been established, so
the emission benefits cannot be quantified other than to say that
periodic inspections achieve low emission levels. The nationwide
annual cost for monthly visual inspections under the final rule
is estimated to be $43,000 per year.
9.3 OTHER
Comment: One commenter felt that, since the leak incidence
rate at terminals in Southern California is less than 1 percent,
only leaking components should be tagged and recorded.
Complicated systems that require records on all components are
not justified (IV-D-19, p. 5).
One commenter recommended that existing facilities be
required to conduct LDAR monitoring at least as frequently as, or
more frequently than, new facilities. Since leaks would be more
9-5
-------
likely from the older pumps and valves, monthly monitoring at
existing facilities would be more beneficial (IV-D-41, p. 2).
One commenter asked whether a 1-month old pump at an
existing facility would have to be monitored monthly or quarterly
under the LDAR program (i.e., would the new or existing facility
requirements apply?). This commenter felt that facilities should
be allowed to offer alternative monitoring schedules in order to
increase efficiency and save costs (IV-D-19, p. 4).
One commenter disagreed with the implementation date for
LDAR of August 8, 1994 as indicated in the preamble to the
proposed standards. The commenter felt that the appropriate date
would be 180 days after promulgation of the final rule (IV-D-8,
p. 4). Two other commenters felt that at least 1 year should be
provided to set up an LDAR program (IV-D-18, p. 6; IV-D-36, p.
3) .
One commenter felt that the frequency of monitoring at
existing facilities should be reduced to once every 6 months for
equipment that does not leak for 2 consecutive quarters (IV-D-29,
p . 5) .
Proposed §63.424 references §60.482-1 to 60.482-10 (NSPS
subpart W) as the basic requirements for an equipment leak
control program. Two commenters felt that references to "heavy
liquid service" in these NSPS provisions should be removed since
this MACT standard deals only with gasoline. Also, since control
devices are addressed in §63.422 of the proposed rule, reference
to controls should be removed in §60.482-10 (IV-D-5, p. 2; IV-D-
17, p. 2) .
One commenter felt that it should be stated that the
requirements of §60.482-1 to §60.482-10 of subpart W are not
applicable to equipment in vacuum service, as long as the
equipment is identified. This commenter also pointed out that
the provisions of §60.482-3 relate to compressors, so the rule
should not reference this section (IV-D-36, p. 3).
9-6
-------
Response: All of the comments in this section pertain to
the proposed LDAR program. As discussed in Section 9.2, the LDAR
requirement has been deleted from the rule and replaced with a
monthly visual inspection program; therefore, response is not
necessary.
9.4 REFERENCES
II-A-17 Compilation of Air Pollutant Emission
Factors, Fourth Edition (AP-42). Section
9.1.3, Fugitive Emissions and Controls at
Petroleum Refineries. U.S. Environmental
Protection Agency, Research Triangle Park,
NC. September 1985.
9-7
-------
10.0 BASELINE/ENVIRONMENTAL IMPACT CALCULATIONS
10.1 GENERAL
Comment: One commenter identified several factors that may
have led EPA to overestimate baseline emissions from loading
racks and storage vessels. First, the commenter challenged EPA's
data base indicating that 13 percent of gasoline storage tanks
are uncontrolled fixed-roof tanks (with a consequently high
emission factor). The commenter believes that there are fewer of
this type of tank because of the product quality, safety, and
financial issues that discourage the use of these tanks. Second,
EPA did not use the most recent tank emission estimation methods
(1992 AP-42, Chapter 12), which may have led to an overstatement
of emissions. Third, the commenter expressed that EPA calculated
loading rack baseline emissions using the regulatory limits in
effect for control systems, whereas actual controlled emissions
are lower than these limitations indicate. The commenter urged
EPA to revise its baseline emissions estimates on the basis of
corrected emission factors (IV-D-14, p. 8).
Response: The EPA has not located any new data or
information indicating the percentage of uncontrolled fixed-roof
tanks located at gasoline distribution facilities and therefore
has no basis to change the assumption used in previous EPA
studies that 13 percent of gasoline storage tanks are
uncontrolled fixed-roof tanks. Also, the commenter did not
provide any data that would allow EPA to make a better estimate.
The EPA has recalculated its baseline storage tank emissions
using the TANKS software referenced in Chapter 12 of EPA's
10-1
-------
Compilation of Air Pollutant Emission Factors, AP-42 (IV-A-2).
The revised nationwide storage tank baseline emissions are 6,370
Mg of HAP per year at pipeline breakout stations and 4,930 Mg of
HAP per year at bulk terminals. The assumptions and methodology
used to calculate the nationwide storage tank emissions are
discussed further in Appendix D.
Regarding the calculation of loading rack baseline emissions
at proposal, EPA did calculate baseline emissions based on actual
emissions, rather than regulatory limits. The EPA first
estimated the percentage of existing gasoline throughput
occurring in each of the regulated control level areas (see Table
D-3 of BID, Volume I). As seen in Table 4-1 of BID, Volume I,
the test data that EPA evaluated showed that many processors
regulated in the 80, 35 and 10 mg/liter control level area
perform much better than these standards. For example, some of
the processors regulated to 80 mg/1 performed much better than 80
mg/1, 35 mg/1, 10 mg/1, and in some cases better than 5 mg/1.
Similarly, the data suggested that vapor processors regulated in
35 and 10 mg/1 control level areas also performed better than
these standards. As a result, to calculate the baseline
emissions, a portion of the throughput in each regulatory control
level area was redistributed into more stringent control level
areas. Also, a portion of the throughput was allocated into a
hypothetical 5 mg/1 control level area to account for the systems
operating below 10 mg/1 (see Table D-20 of BID, Volume I).
10.2 GASOLINE RVP
Comment: One commenter felt that EPA has overestimated the
Reid vapor pressure (RVP) of 1998 (base year) gasoline, leading
to overestimation of baseline emissions from loading racks and
storage vessels. The commenter stated that current gasoline
pools average about 10.0 psia RVP, and are expected to have lower
average values in 1998. Due to EPA's estimate of 11.4 psia as
the national annual average RVP, the commenter concluded that EPA
10-2
-------
overestimated these emissions by approximately 20 percent (IV-D-
14, p. 8) .
Response: At proposal, EPA's Emission Standards Division
(ESD) estimated that 68 percent of the gasoline throughput
occurred in the summer and had an average RVP of 10.2 psia and
that 32 percent of the gasoline throughput occurred in the winter
and had an average RVP of 14.0 psia, resulting in a nationwide
average RVP of 11.4 psia. Based on analyses generated by EPA's
Office of Mobile Sources (OMS), EPA has lowered its estimate of
the nationwide RVP of gasoline from 11.4 psia to 10.4 psia. A
Federal Register notice dated February 16, 1994 indicated a 1990
gasoline summer RVP of 8.7 and a winter RVP of 11.5 (IV-I-4).
According to OMS, 39.6 percent of the gasoline throughput occurs
in the summer while 60.4 percent occurs in the winter (IV-E-4).
Although the assumed percentages of summer and winter throughputs
were different for the OMS and ESD data, for the purpose of the
gasoline distribution MACT rulemaking, the proposed summer and
winter RVP's of 10.2 and 14.0 psia, respectively, were lowered to
9.3 and 12.8 psia, resulting in a nationwide average RVP of 10.4
psia.
10.3 REFERENCES
IV-A-2
IV-E-4
IV-I-4
Compilation of Air Pollutant Emission
Factors, Fourth Edition (AP-42). U.S.
Environmental Protection Agency, Research
Triangle Park, N.C. Chapter 12, Storage of
Organic Liquids. October 1992.
Telecon. Johnson, T., Pacific Environmental
Services, Inc. with Korotney, D., EPA:OMS.
May 17, 1994. Reid Vapor Pressure of
Gasoline.
40 CFR 80. Regulation of Fuel and Fuel
Additives: Standards for Reformulated and
Conventional Gasoline. 59 FR 7716-7878.
February 16, 1994.
10-3
-------
11.0 TEST METHODS AND PROCEDURES
11.1 ALTERNATIVE TEST METHODS
Comment: Two commenters felt that test methods adopted by
local districts that have stricter emission standards should be
included in the rule as equivalent methods. These commenters
stated that EPA Methods 25A and 25B are not accurate enough to
verify compliance with these strict limits (IV-D-28, p. 9; IV-D-
38, p. 6).
Response: Both EPA Test Methods 25A and 25B have sufficient
accuracy at the level of the emission standard in this
regulation. Other methods may be acceptable as alternatives
provided they meet the criteria in Method 301 of appendix A to 40
CFR part 63.
Comment: One commenter suggested that the wording of
§63.425(d) be changed to read: " ... shall comply with the test
methods and procedures in §60.485(b) through (g) of this chapter,
or as approved by the local administrator" (the phrase following
the comma was added by the commenter). These test methods
pertain to the performance of an LDAR program for equipment leaks
(IV-D-36, p. 4).
Response: As discussed previously in Section 9.2, the
proposed requirement for sources to implement an instrument based
LDAR program has been deleted from the regulation. Therefore,
the provision offered by the commenter would not be applicable in
the final rule.
11.2 TESTING OF CARGO TANKS
11-1
-------
Comment: One commenter recommended that EPA adhere to DOT
methods for the detection of defects in the cargo tanks of tank
railcars. The commenter felt that it would be unreasonable to
subject tank cars transporting gasoline to the same tank
tightness testing criteria as cargo tank trucks, because of the
construction and operational differences between the two types of
vehicles. DOT's proposed rules, once promulgated, will provide
adequate integrity testing of all tank cars, so the additional
testing criteria under the gasoline MACT regulation are
unnecessary (IV-D-15, p. 7).
Response: The EPA proposed that railcars annually pass the
EPA Method 27 pressure and vacuum test before loading gasoline at
existing and new major source facilities. Additionally, EPA
proposed for new facilities the use of a loading rack vacuum
assist system, in addition to the proposed annual pressure and
vacuum test, to further control leakage from cargo tanks (tank
trucks and railcars).
The commenter referred to a rule proposed by the Department
of Transportation (DOT) which focuses on the use of non-
destructive testing methods to identify potential failures of
tank car shells, welds, and fittings. The purpose of this
proposed rule, "Detection and Repair of Cracks, Pits, Corrosion,
Lining Flaws, Thermal Detection Flaws and Other Defects of Tank
Car Tanks" (58 FR 48485), is to increase the confidence that
critical tank car defects (e.g., structural fatigue, crack
propagation, and corrosion) will be detected and to enhance the
safe transportation of hazardous material in tank cars. The rule
proposes that tank cars used in hazardous materials service
(other than chlorine) be inspected at 10-year intervals.
Included in the railcar testing is a leakage pressure test [see
proposed §180.509(j)(1)] which requires the following:
"At minimum, each tank car facility shall perform
a leakage pressure test on the tank fittings and
appurtenances. The leakage pressure test must include
11-2
-------
product piping with all valves and accessories in place
and operative, except that during the pressure test the
tank car facility shall remove or render inoperative
any vent devices set to discharge at less than the test
pressure. Test pressure must be maintained for at
least 5 minutes. Leakage test pressure must not be
less than 50% of the tank test pressure."
According to the Federal Railroad Administration, the
leakage pressure test may be either a hydrostatic or a
pressurized air test performed at pressures as high as 300 psi.
Although the leakage pressure test requires a high test pressure,
it is performed only once every 10 years. The EPA considers a
10-year test frequency too long to ensure continuous vapor
tightness of the railcar. In addition, the test allows certain
pressure relief vents to be capped off or closed during testing.
To ensure that railcars are vapor-tight during loading, EPA
believes that all product piping, vapor piping, and pressure
relief devices should remain operative during testing to ensure
vapor tightness of all the equipment. As proposed, EPA is
requiring that railcars annually pass the EPA Method 27 pressure
and vacuum tests before loading gasoline at existing and new
major source facilities, as well as be subject to additional
testing as noted in Section 7.
11-3
-------
12.0 REPORTING AND RECORDKEEPING
12.1 STORAGE OF RECORDS
Comment: A few commenters stated that the proposed
recordkeeping requirements would be onerous and excessive. One
stated that 5-year records of control system monitoring data
would be voluminous and more than could be analyzed (IV-D-20, p.
14), while another found no justification for the requirement to
retain tank tightness testing records for 5 years (IV-D-23, p.
4). One commenter felt that monitoring records should be
retained for no more than 2 years (IV-D-36, p. 4). Another
commenter believed that it should suffice to preserve actual
emissions data for the previous 12 months and to keep exception
and repair reports for 5 years (IV-D-18, p. 9).
Response: In the final rule, every effort has been made to
reduce the recordkeeping and reporting burden associated with the
rule. The EPA has examined all of the proposed requirements and
has streamlined the rule to include only those necessary to
ensure compliance.
As discussed in the preamble to the final General Provisions
to part 63 (59 FR 12408, March 16, 1994), EPA believes that the
5-year records retention requirement is reasonable and needed for
consistency with the part 70 operating permit program and the 5-
year statute of limitations, on which the permit program based
its requirement. This retention period will also allow EPA to
establish a source's history and patterns of compliance for the
purpose of determining appropriate levels of enforcement action.
12-1
-------
Comment: One commenter felt that the recordkeeping and
reporting requirements associated with tank truck vapor tightness
should be the responsibility of the vehicle owner, and not of the
terminal owner who does not own every truck that loads at the
terminal (IV-D-7, p. 11). Another commenter stated that the
terminal should be required to hold only the current certificate
indicating compliance with the tightness requirements, and that
the owner of operator of the tank truck should maintain records
for previous years (IV-D-20, p. 15).
Response: The owner or operator of an affected major source
bulk terminal is responsible for ensuring that tank trucks and
railcars loading gasoline and potentially emitting HAP's are
certified as vapor-tight through an annual pressure-vacuum test.
This includes both "branded" trucks and those operated by "for-
hire" tank truck firms, as well as any railcars that may load
gasoline at the facility. To carry out this responsibility,
records of all gasoline cargo tanks loading at the facility
together with the test records for those tanks need to be
collected at a central location, logically the affected facility.
For their part, cargo tank owners (whether terminal owners or
outside firms) who wish to load specific tanks at affected
terminals have the responsibility of equipping those tanks with
compatible vapor collection and loading hardware, having the
annual test performed, and supplying the test results for each
tank to the facility. (Cargo tanks are also subject to being
tested for vapor tightness at any time under a continuous
performance requirement; see Section 7.2.) The terminal owner
then need only cross-check to ensure that there is up-to-date
vapor tightness documentation for each tank loaded, within 2
weeks after the loading occurs. Finally, the terminal owner must
notify the cargo tank owner within 3 weeks of any loading that
was made into a noncertified tank, and then take steps to ensure
that any such tanks are not reloaded until the test requirement
has been met and the documentation provided.
12-2
-------
This is the same approach that has been in use under 40 CFR
part 60, subpart XX, the NSPS for tank trucks at new bulk
terminals since 1983. The EPA believes that these requirements
are reasonable and necessary to reduce a potentially large
emission source at bulk terminals. Further, it is reasonable for
the owner/operator of the affected source to bear the
responsibility for ensuring that loadings at the source are
performed using cargo tanks subject to the maximum control level.
Comment: Several commenters objected to the requirement to
store all records at the facility site, especially at unmanned
facilities. These commenters pointed out that space at these
facilities is often at a premium, and requested that some
alternate place of business in the same geographic area be
allowed to store the records and make them available within a
reasonable time period upon request (IV-D-18, p. 9; IV-D-20, p.
14; IV-D-22, p. 82; IV-D-34, p. 26).
Response: The concern of these commenters is addressed by
§63.10(b)(1) of the General Provisions, which apply to these MACT
standards. At a minimum, the most recent 2 years of data must be
retained on site; i.e., at the bulk terminal or pipeline breakout
station. The remaining 3 years of data may be retained off site.
Such files may be maintained on microfilm, on a computer (hard
disk), on computer floppy disks, on magnetic tape, or on
microfiche. However, they must be "recorded in a form suitable
and readily available for expeditious inspection and review."
12.2 REPORTING REQUIREMENTS
Comment: Several commenters felt that 45 days is much too
short a time period for facilities to provide initial
notification of applicability for existing facilities or of
construction for new facilities [per §63.9(b) of the General
Provisions]. They suggested a period of 1 year to make this
notification, or at least the 120 days provided in the final
General Provisions. One commenter said this extension is
12-3
-------
necessary because: (1) revisions to RFG and oxygenated fuel
mandates are still underway and may change HAP emission
calculations over the next 2 years, especially if States opt into
oxygenated fuel requirements; (2) EPA may still be reviewing the
revised emission factors for fugitive equipment leaks for these
sources and, unless completed by the final rule as requested,
additional time will be necessary for EPA to complete that review
and to allow States to adopt those factors as revised, and to
allow facilities to utilize those factors for major source
determination; and (3) the emissions audits that will be
necessary for many sources could require several months to
contract and conduct, once the gasoline makeup and emission
factor issues are fully resolved (IV-D-4, p. 3; IV-D-7, p. 10;
IV-D-15, p. 10; IV-D-18, p. 5; IV-D-22, p. 83). Two of these
commenters also recommended that this initial notification not be
irreversible and subject to substantial additional documentation,
and that the notifier not be subject to EPA enforcement action
for reversing their notification (IV-D-18, p. 5; IV-D-22, p. 84).
Response: The preamble to the proposed MACT standards
stated that, as outlined in §63.5 of the proposed General
Provisions, existing sources would have to submit their required
initial notifications within 45 days after promulgation of the
final rule. Subsequent to the preparation of the MACT proposal,
and after EPA had considered public comments on the proposed
General Provisions, the final General Provisions allowing a 120-
day period for submitting the initial notifications were
published. In addition, EPA has considered the comments and
concerns of commenters on the MACT proposal, and agrees that
there are factors that would make it difficult for many
facilities to submit this notification report even within 120
days. The principal issue (in addition to points made by the
commenters) concerns the collocation of marketing facilities with
refineries, which may necessitate more time for facilities to
assess the combined effects of the refinery and gasoline
12-4
-------
distribution MACT rules. Therefore, EPA is allowing up to 1 year
for existing facilities to submit the initial notification,
although many facilities will likely be able to make the
submittal within 120 days.
Since this extra time has been granted, the Agency feels
that the facility should be confident of the accuracy of the
information it has provided. Therefore, this notification will
be considered a source's final decision.
Comment: Two commenters believed that the required amount
of recordkeeping represents an excessive burden, and that all
reporting requirements in the rule should be consolidated into a
single annual report (IV-D-18, p. 9; IV-D-36, p. 5). The second
commenter stated that a quarterly report could be submitted for
periods of noncompliance.
Response: The proposed rule in §63.428(j) provided that
most of the required reports be consolidated into a periodic
report that must be submitted semiannually, except that
exceedances of the monitored operating parameter and failures to
maintain a vacuum in the vapor collection system during loading
were to be reported quarterly. The Administrator in proposed
§63.428(k) retained the authority to request more frequent
reporting of monitored operating parameter data if the source was
found to be frequently out of compliance or if the monitoring
system was out of service excessively. This latter provision has
been determined to be unnecessary because the part 63 General
Provisions already address more frequent reporting, as discussed
below.
These semiannual reporting requirements are consistent with
the part 70 operating permit program as well as the General
Provisions, §63.10(e)(3); furthermore, EPA believes that these
reporting requirements are reasonable for gasoline distribution
facilities. Therefore, the requirement for a semiannual report
has been retained in the final rule.
12-5
-------
All reporting is to be performed in accordance with
§63.10(e)(3)(i). As indicated under §63.10(e)(3)(i)(C), once a
source reports excess emissions, the source must follow a
quarterly reporting format until a request to reduce reporting
frequency is approved. Thus, the final rule follows the General
Provisions with regard to the frequency of reporting.
12.3 OTHER
Comment: One commenter stated that the requirements of
§63.428(c)(3) should be clarified to note their applicability
only to those flares that must comply with §63.11 of the General
Provisions. The commenter felt that the language suggests that
data must be recorded and reported for all flares (including
enclosed flares) in accordance with the requirements of §63.11
(IV-D-38, p. 3).
Response: The EPA agrees that the requirements of proposed
§63.428(c)(3) were intended to apply only to flares that cannot
be tested directly to determine their compliance with the
emission limit (as referred to under proposed §63.425(a)). The
language of §63.428(c) (3) [now §63.428(c) (2) (ii)] has been
revised in the final rule to reflect this intent.
12-6
-------
13.0 COSTS/ECONOMIC IMPACTS
13.1 STORAGE VESSEL CONTROL COSTS
Comment: Two commenters said that EPA's estimates of the
costs to comply with the proposed storage vessel requirements are
too low. One commenter estimated that it would cost $33,500 to
install an internal floating roof in a 50-foot diameter tank,
rather than EPA's estimate of $20,000 (IV-D-5, p. 4). Another
commenter estimated that it would cost approximately $50,000
(including the cost of cleaning, degassing, and sludge disposal)
to install a secondary seal on an existing external floating roof
tank, rather than EPA's estimate of $31,000 (IV-D-17, p. 3).
This commenter also provided a cost for degassing a tank at
$20,000-30,000 when the capital cost of additional tankage is
required (IV-D-17, p. 5).
Response: Storage vessel equipment costs can vary
significantly depending on the storage vessel contractor and
other site-specific factors. Retrofit costs for installing
secondary seals and gasketed fittings were provided to EPA by
industry (IV-D-44). These costs were collected from three
manufacturers and varied by as much as 100 to 1,300 percent.
To estimate storage tank equipment costs, EPA used the
document "Alternative Control Techniques (ACT) Document:
Volatile Organic Liquid Storage in Floating and Fixed Roof Tanks"
(IV-A-3). The average costs presented in the ACT document were
obtained from four contractors. The EPA believes that these
13-1
-------
costs are reasonable estimates and has used them in the final
cost analysis.
The EPA revised its degassing and cleaning costs assumed at
proposal using the estimates provided in the ACT document.
Assuming a sludge depth of 2 inches, a sludge disposal cost of
$5/gal, and a cleaning cost of approximately $150/foot-diameter,
the new revised degassing and cleaning costs for a 50 ft., 78
ft., and 100 ft. tank are $19,730, $41,470, and $63,930,
respectively. The degassing and cleaning costs are discussed
further in Appendix B.
Comment: Several commenters claimed that EPA's cost
analysis does not include an assessment of the costs associated
with the fitting controls required under 40 CFR 60, subpart Kb
and incorporated by reference into the proposed rule (IV-D-7, p.
7; IV-D-8, p. 4; IV-D-15, p. 9; IV-D-16, p. 2; IV-D-22, p. 30;
IV-D-31, p. 3; IV-F-1, p. 9). Another commenter estimated the
cost effectiveness of roof fitting control requirements at well
over $40,000 per Mg of HAP's reduced, which does not consider the
costs associated with tank degassing (IV-D-17, p. 5).
Response: In BID, Volume I, EPA stated that the cost for
installing controlled fittings on a new internal floating roof
tank was $200 (page 7-3) and the cost for installing controlled
fittings on an existing external floating roof tank was $680
(page 7-4). These costs were obtained from docket item II-A-24
and were restated in EPA's ACT document. Industry provided
information collected from three manufacturers regarding the cost
of installing gasketed fittings (IV-D-44). For the purposes of
this rulemaking, EPA used the lowest of the three cost estimates
provided by industry to determine the cost effectiveness of
fitting controls because the range of the costs was so large (100
to 1,300 percent). The estimated cost to install controlled
fittings (excluding degassing/cleaning cost) is $1,225 for a 50
ft. tank, $2,175 for a 78 ft. tank, and $2,800 for a 100 ft.
13-2
-------
tank. Based on these costs, a net savings will be realized for
external floating roof tanks that have controlled fittings
installed while they are being retrofitted with the required deck
rim seals. Controlled fitting costs are discussed in detail in
Appendix B.
13.2 OTHER
Comment: One commenter stated that EPA's estimates of
nationwide annual gasoline savings resulting from the proposed
standards (about 14.3 million gallons saved) would be accurate
only if incineration were banned as a control technology. The
commenter suggested that energy impacts be analyzed and presented
on a control technology basis. This will indicate that recovery
units save approximately 1.5 gallons/1,000 gallons loaded, while
incineration systems have negative savings (IV-D-28, p. 10).
Response: The EPA's calculations of net energy savings due
to various control alternatives were summarized in Table 6-6 of
BID, Volume I. As stated in footnote "b" of Table 6-6, the
assumption was that 25 percent of the control systems would be
recovery type devices, while the remainder would be destruction
(incineration) type devices. Therefore, for bulk terminals only
25 percent of the 12,400 Mg/yr emission reduction at loading
racks was used to derive the 1.2 million gallons per year
savings. The remaining 10.8 million gallons/yr saved at
terminals is due to controls on storage tanks, tank truck vapor
leakage, and equipment component leaks. If all control systems
were assumed to recover product, the loading rack gasoline
savings would have been calculated as 4.9 million gallons per
year.
In reanalyzing baseline emissions and regulatory impacts, as
well as the likely population of control technologies selected to
comply with the standards, EPA has calculated new energy impacts
13-3
-------
for the final regulation. These results are summarized in
Section 1.2.5.
Comment: Three commenters said that EPA has overestimated
the price of recovered product in the calculation of recovery
credits by using the average retail price of gasoline (which
includes taxes). In fact, the tax portion of the gasoline price
cannot be realized by a terminal owner. Two of the commenters
estimated the value of recovered product to be $0.50 to $0.60 per
gallon in 1990 dollars, plus any real dollar increase in price up
to the 1998 baseline (IV-D-2, p. 7; IV-D-5, p. 4; IV-D-17, p. 7).
Response: At proposal, EPA did overestimate the price of
recovered product because it included the value of taxes in the
price of gasoline. The price of gasoline currently used to
estimate recovery credits is the wholesale price at the point in
the distribution chain where the recovery occurs. The base year
for costing this standard is 1990. The EPA is using estimated
1990 wholesale prices of $0.79 per gallon at bulk terminals and
$0.77 per gallon at pipeline breakout stations in its
reestimation of recovery credits. These values were derived by
first estimating the price of gasoline at service stations and
subtracting an average State gas tax of $0,178 per gallon and an
average Federal tax of $0.14 per gallon (IV-J-1). The wholesale
price of gasoline at bulk terminals and breakout stations was
estimated using the pricing margins stated in Table 8-20 of BID,
Volume I. The Agency anticipates that the real price of motor
vehicle gasoline will increase from base year levels to 1998
impact year levels. Therefore, the value of recovered product
will be worth more in real terms in 1998 than it is worth today.
In order to be conservative, the 1990 wholesale prices above do
not reflect this real increase in value.
Comment: One commenter said that EPA should consider the
competitive effect that the proposed regulation will have on the
independent sector of the petroleum market. Independent
13-4
-------
marketers are generally smaller than the major integrated oil
companies. The cost of environmental compliance will have a
greater impact on the financial viability of the independent
marketer than of the major oil companies (IV-D-20, p. 16).
Response: The EPA did not separately address the potential
impact of the proposed rules on independent distributors because
the Agency did not collect or receive financial information on
any of the affected companies. Company size and diversification,
considered alone, do not affect a company's relative performance.
The Agency believes that the cost of the controls in this NESHAP
can be passed on to the end users without significant increases
in prices or significant changes in end-user demand. The Agency
does not possess information necessary to conclude that the
financial viability of the independent companies is more or less
threatened by the proposed rule than major integrated companies,
or that independent companies will be unable to obtain the
required capital funding to comply with the rule.
Comment: One commenter suggested that EPA conduct more
definitive research pertaining to the number of tank trucks
potentially impacted by the standards, and reevaluate the
associated economic impacts. They stated that it is unclear in
the proposal how many tank trucks are estimated to be impacted by
the regulation. The commenter believes that retrofitting a four-
compartment tank truck to meet the mandated "closed loop" vapor
recovery will cost approximately $3,200 per unit (not including
downtime). As a result, the compliance cost for the tank truck
industry would be in excess of $32 million assuming 10,000
retrofits (IV-D-23, p. 3).
Response: At proposal, EPA estimated that approximately
2,500 tank trucks (including terminal owned and independent tank
trucks) would need to be retrofitted to comply with the standard
at a cost of $3,500 per truck (see page 7-26 of BID, Volume I)
resulting in a nationwide cost of approximately $8.8 million.
13-5
-------
The estimation of the number of trucks impacted by the regulation
was determined as follows:
1) The EPA estimated that 81,300 tank trucks (including new
and existing tank trucks) will be used in 1998 to distribute
motor vehicle (non-aviation) gasoline. Fifty-four percent of
these trucks, or 43,900 trucks, are estimated to be used in 1998
at bulk gasoline terminals. Of these trucks, 7,200 are estimated
to be owned by bulk terminal facilities and 36,700 are estimated
to be independent (see BID, Volume I Table 5-6).
2) The Agency assumes that 72.1 percent of the 1998 tank
truck population comprises existing tank trucks. The Agency
further assumes that only 29 percent of all existing terminal-
dedicated tank trucks loading at bulk terminals (B.T.'s) will
require retrofit because most tank trucks are regulated by bulk
terminal NSPS and tank truck CTG requirements. Finally, EPA is
estimating that approximately 27 percent of all bulk terminals
will be classified as major sources. Using the assumptions
stated above, the number of tank trucks that will be impacted by
this regulation is estimated as follows:
43,900 tank trucks x 72.1 % = 31,680 existing tank trucks.
31,680 T.T.'s x 29 % need to retrofit x 27 % are at major B.T.
= 2,500 tank trucks to be retrofitted.
Based on some changes in the analysis (e.g., assumption of a
lower RVP), EPA has estimated that only 24 percent of the bulk
terminals will be classified as major sources resulting in
approximately 2,200 tank trucks being impacted by this
regulation. As a result, the revised nationwide cost impact to
the tank truck industry is estimated at $7.6 million.
13-6
-------
The Agency's economic impact assessment, which uses a worst-
case assumption, indicates that up to 51 fewer jobs would be
required to satisfy post-regulatory demand for tank truck
services. This reduction is equivalent to 5 small for-hire
trucking firms, which employ 10 workers each.
Comment: One commenter suggested that EPA conduct the cost
analysis on both a per-facility and a per-control technology
basis. Based on EPA's estimated capital cost of $93 million and
that approximately 245 facilities will be impacted, the average
capital cost to comply with the 10 mg/liter standard is $380,000
per facility (excluding the additional cost of vacuum assist) .
The commenter stated that the cheapest approach (a direct-mode
incinerator) would cost $430,000 for an 800,000 gallon-per-day
facility, which does not include the cost of vacuum assist
equipment (IV-D-28, p. 10). The commenter suggested that EPA
survey the States that have had equipment installed in the last 5
years to obtain accurate cost information.
Response: The EPA did base its costs on both a per-facility
and a per-control technology basis. Section 7.1.3 of BID, Volume
I, discusses the loading rack conversion costs for the model
plant bulk terminals to meet the 10 mg/liter standard using
either carbon adsorption, refrigeration condenser, or thermal
oxidation units to process the vapors. The model plant costs
were based on various vendor quotes and represent 1990 dollars.
The commenter incorrectly calculated the average capital
cost of $380,000 per facility to comply with the 10 mg/liter
standard by including the number of breakout stations in the
calculation. The EPA currently estimates that 176 existing
terminals and 67 new terminals will be impacted by this
regulation at a nationwide capital cost of $55.2 million and $7.7
million, respectively. Therefore, the average capital cost per
facility is approximately $314,000 for existing facilities and
approximately $115,000 for new facilities. These costs do not
13-7
-------
represent the average cost of a new vapor processing unit but the
average cost to retrofit a unit to meet the 10 mg/liter standard.
For example, new facilities would already be subject to 40 CFR
part 60, subpart XX which requires loading racks at bulk
terminals to be controlled to 35 mg/liter. For the purposes of
calculating the cost impacts of this MACT regulation, EPA assumes
the additional cost to retrofit the vapor processing unit to meet
10 mg/liter. Likewise, EPA assumed that most of the existing
units could be retrofitted to meet the 10 mg/liter standard and
would not require a new control device. The costs for
retrofitting the control device to meet the 10 mg/liter standard
were obtained from various vendors and have been peer reviewed.
The EPA does not agree that further changes to the costing data
are required.
13.3 REFERENCES
II-A-24 Control of Volatile Organic Compound
Emissions from Volatile Organic Liquid
Storage in Floating and Fixed Roof Tanks -
Guideline Series. Draft. U.S. Environmental
Protection Agency, Research Triangle Park,
NC. September 1991.
IV-A-3 Alternative Control Techniques Document:
Volatile Organic Liquid Storage in Floating
and Fixed-Roof Tanks. U.S. Environmental
Protection Agency, Research Triangle Park,
NC. EPA 453/R-94-001. January 1994.
IV-D-44 Letter from Ferry, R., TGB Partnership, to
Shedd, S., Chemicals and Petroleum Branch,
U.S. Environmental Protection Agency,
Research Triangle Park, NC. June 3, 1994.
Cost basis for IFRT fitting controls.
IV-J-1 National Petroleum News. 1991 Fact Book
Issue. Volume 83, No. 7.
13-8
-------
APPENDIX A.
CARGO TANK VAPOR LEAKAGE
The purpose of this appendix is to explain the methodology
used to estimate the average amount of vapor that leaks from
loading cargo tanks subject to an annual vapor tightness test
allowing either a 1-inch of water pressure decay (in 5 minutes)
as required in California or a 3-inch pressure decay as discussed
in the control techniques guideline (CTG) for tank trucks (II-A-
9) and required under the bulk terminal NSPS, 40 CFR part 60,
subpart XX. This appendix also discusses the cost of annually
testing a cargo tank at the 1-inch pressure decay limit versus
the 3-inch limit.
A.1 VAPOR LEAKAGE FROM CARGO TANKS
A. 1.1 California Decay Limit of 1-inch H20
In order to calculate the cargo tank leakage factor for a
cargo tank just passing the 1-inch pressure decay limit required
by California, it was estimated that the allowable leakage from
cargo tanks passing the pressure decay test (change of less than
1 inch of water from an initial pressure of 18 inches, over 5
minutes) is approximately 0.3 percent. This value was derived
from the following equation located in the tank truck CTG:
VL = (0.5V) (T/tp) (1 - Pf/PJ ,
where VL = allowable volume of leakage (liters)
V = capacity volume of tank (liters)
T = total time for loading (minutes)
A-1
-------
tp = time limit for pressure test (minutes)
Pf = final pressure for test (inches H20 absolute)
P± = initial pressure for test (inches H20 absolute).
Using the bulk terminal model plant parameters shown in Chapter 5
of BID, Volume I (III-B-1),
V = 32,200 liters, T = 32,200 liters/2,270 1pm = 14 minutes.
Therefore, VL = (0.5) (32, 200) (14/5) (1 - 424/425)
= 106 liters.
The percentage leakage with a 1-inch pressure decay is then
106/32,200 = 0.0033 = 0.33 percent. Using the same methodology,
a 3-inch pressure decay results in vapor leakage of 1 percent
(see Section A.1.2).
Second, EPA analyzed the cargo tank failure rates from data
submitted by four commenters (IV-D-14, IV-D-22, IV-D-28, IV-D-
38). One of the data sets (IV-D-14) summarized failure rates for
tank trucks surveyed by the Bay Area Air Quality Management
District (BAAQMD) (5 percent failure rate) and the John F. Jordan
Company (2.5 percent failure rate). The BAAQMD data (1986-1992)
were identified as representing tank trucks that failed the
annual tightness tests based on not sustaining adequate pressure
(exceeded the 1-inch allowable pressure decay). The data
collected by the John F. Jordan Company were listed by State and
did not identify any California tank trucks. As a result, these
data represented tank trucks failing the CTG and 40 CFR part 60,
subpart XX requirements (3-inch pressure decay). Two of the
commenters (IV-D-28, IV-D-38) supplied leakage data from only the
BAAQMD, which were identical to the data discussed above. One
commenter (IV-D-22) supplied over 10 years of data from the
BAAQMD (which were identical to the BAAQMD data supplied in IV-D-
14), four oil companies, and the Jordan Company (which was
approximately the same information supplied by IV-D-14). From
these data, it was determined that the data collected from both
A-2
-------
the oil companies and John F. Jordan Company represented tank
trucks failing the CTG and 40 CFR part 60, subpart XX
requirements (3-inch pressure decay). Consequently, from all the
leakage data received, EPA determined that only the BAAQMD data
represent cargo tanks failing a 1-inch pressure decay limit. The
EPA elected to use the most current data (for the years 1988
through 1992) because it would be more representative of current
conditions. The BAAQMD data are summarized in Table A-l.
TABLE A-l. 1988 TO 1992 BAAQMD TANK TRUCK LEAKAGE DATA
(Source: IV-D-14, IV-D-22, IV-D-28, IV-D-38)
Year
#TT Tested
1988
1989
1990
1991
1992
772
1, 003
785
606
442
TOTALS
3, 608
#Leakers
Failed
43
45
29
14
5
5.6
4.5
3.7
2 . 3
1 . 1
136
Avg. =3.8
As seen in Table A-l, a total of 3,608 tank trucks were
tested, with 136 failing the initial test (before repairs).
Thus, the failure rate was 136/3,608 = 3.8 percent for this
sample of tanks. The percentage passing the test was 100 - 3.8 =
96.2 percent.
To estimate the leakage occurring from cargo tanks that do
not pass the annual vapor tightness testing on the first attempt,
EPA relied on information in the EPA document "Bulk Gasoline
Terminals--Background Information for Proposed Standards" (II-A-
14). In Table C-4 of that document, tank truck leakage data were
separated into two groups, those tanks that would pass and those
tanks that would fail the current certification standards. The
data from Table C-4 showed that 16 of the trucks tested leaked
A-3
-------
over a range of 1.9 to 35.8 percent, with an overall average of
12.1 percent. For the purposes of this analysis, it was assumed
that trucks failing either annual certification test (1-inch
pressure decay or 3-inch pressure decay) would leak 12.1 percent.
The average leakage and failure rate data were combined to
determine the average leakage from all tank trucks subject to the
annual test. The average leakage from all tank trucks is
calculated as follows:
(0.038) (0.121) + (0.962) (0.0033) = 0.0078 = 0.8 percent.
The leakage emission factor for cargo tanks subject to the
1-inch decay limit is (0.8 percent)(1,014 mg/liter) = 8 mg/liter.
A.1.2 Tank Truck CTG Decay Limit of 3 inches H20
Using the same methodology discussed above, VL = 318 liters
when assuming a 3-inch pressure decay. The percentage vapor
leakage with a 3-inch pressure decay is 318/32,200 = 0.0099 = 1
percent. As discussed above, EPA determined that the test
failure rate data collected from the four oil companies and the
John F. Jordan Company represented tank trucks subject to an
annual vapor tightness test with a 3-inch vapor decay limit.
Although the commenters supplied over 10 years worth of data, EPA
elected to use the more current data for the years 1989 through
1994 since it would be more representative of current conditions.
The data supplied by the four oil companies and the John F.
Jordan Company for the years 1989 through 1984 are summarized in
Table A-2.
TABLE A-2. 1989 TO 1994 TANK TRUCK LEAKAGE DATA
(DATA FROM 4 OIL COMPANIES AND JORDAN CO.)
Year
#TT Tested
#Leakers
% Failed
1989
1990
212
395
24
16
11 . 3
4 . 1
A-4
-------
1991
1992
1993
1994
392
3,386
75
388
6
10
91
1
1.5
2 . 5
2.7
1 .3
TOTALS
4, 848
148
Avg. =3.1%
As seen in Table A-2, a total of 4,848 tank trucks were
tested, with 148 failing the initial test (before repairs).
Thus, the failure rate was 148/4,848 = 3.1 percent for this
sample of tanks, which is a slightly lower failure rate than for
the California trucks which are required to pass at the 1-inch
criterion (3.8 percent). The percentage passing the test was 100
- 3.1 = 96.9 percent.
Using the same methodology as discussed above, the average
leakage from cargo tanks required to pass a 3-inch annual vapor
tightness test is 1.3 percent, which results in a leakage
emission factor of 13 mg/liter.
A.2 COST OF 1-INCH PRESSURE DECAY TEST
As discussed in Section 7.1.3 of BID, Volume I, the cost of
the EPA Method 27 pressure/vacuum test was estimated at
approximately $350 per cargo tank. The cost impact of the
proposed regulation, which required a 3-inch pressure test, was
only $150 per cargo tank because all trucks were already required
to pass an annual DOT pressure test estimated to cost $200 per
cargo tank.
The final rule requires annual pressure and vacuum tests at
the 1-inch decay limit. To estimate the cost of the 1-inch test,
EPA used data supplied by one of the commenters (IV-D-38). The
commenter estimated that the cost to perform the 1-inch pressure
test ranged from $300 to $1,200 per cargo tank, depending on the
amount of repair required to pass the test. Assuming that the
trucks which pass the annual test incur the cost of $300 and the
trucks which fail the annual test incur the cost of $1,200, the
A-5
-------
average cost for a tank truck to pass the annual test is
calculated as:
(0.038) ($1, 200) + (0.962) ($300) = $334 per cargo tank.
Comparing the proposal cost estimate for the annual vapor
tightness test of $350 to the estimated cost of $334 for the 1-
inch test from a second source, EPA cannot discern a cost
difference in certifying tanks to the 1-inch limit versus the 3-
inch limit. As a result, the $150/tank cost impact of tank truck
testing at the 3-inch pressure decay limit is assumed to be a
reasonable estimate for the 1-inch pressure decay test.
A- 6
-------
Control of Volatile Organic Compound Leaks
from Gasoline Tank Trucks and Vapor
Collection Systems. U.S. Environmental
Protection Agency, Research Triangle Park,
NC. E PA-450/2-78-051. December 1978.
Bulk Gasoline Terminals--Background
Information for Proposed Standards. U.S.
Environmental Protection Agency, Research
Triangle Park, NC. EPA-450/3-80-038a.
December 1980.
Gasoline Distribution Industry (Stage I) -
Background Information for Proposed Standards
(NESHAP) . E PA-453/R-94-002a. U.S.
Environmental Protection Agency, Office of
Air Quality Planning and Standards. Research
Triangle Park, NC. January 1994.
A-7
-------
APPENDIX B.
ENVIRONMENTAL AND COST IMPACTS
OF STORAGE VESSEL CONTROLS
The purpose of this appendix is to present the environmental
and economic impacts of imposing 40 CFR part 60, subpart Kb (NSPS
subpart Kb) rim seal and controlled (or gasketed) fitting
requirements for storage vessels. Section B.l discusses the
methodology used to calculate emissions generated from the
cleaning and degassing of gasoline storage vessels. Section B.2
discusses the balancing of degassing and cleaning emissions
versus the emission reductions achieved by installing rim seals
and controlled deck fittings on various types and sizes of
storage vessels. Section B.3 discusses the cost of cleaning and
degassing storage vessels as well as the costs for adding the
controls specified under NSPS subpart Kb.
B.l DEGASSING AND CLEANING EMISSIONS FACTOR DEVELOPMENT
For the purposes of this rulemaking, EPA has estimated that
tank degassing and cleaning emissions are generated from both the
release of gasoline vapors that exist beneath the floating roof
prior to cleaning (degassing emissions) and the expulsion of
vapors from beneath the roof during the initial refilling after
cleaning (refilling emissions). Degassing and refilling
emissions are assumed to be equivalent because the volume of
vapors beneath the roof during degassing and refilling are equal.
These emissions were estimated by calculating the working losses
from a hypothetical fixed-roof tank (FRT) assuming a height of 6
feet and 1 turnover (the typical leg heights for floating roofs
B-1
-------
are 6 feet) . These working losses, Lw, were calculated using
equation 1-23 in Chapter 12 of AP-42 (IV-A-2) shown below.
Lw = 0.0010 Mv Pva Q Kn Kp
where,
Mv = vapor molecular weight (65.8 lb/lb-mole)
Pva = vapor pressure at daily liquid surface temperature (5.6
psia)
Q = annual net throughput, bbl/yr (assuming 1 turnover and
6' roof height)
Kn = turnover factor, dimensionless (1)
Kp = working loss product factor, dimensionless (1).
Degassing and cleaning emissions are shown for various size
storage vessels in Table B-l. Clingage emissions (vapors
evaporating from the exposed tank wall) were also calculated but
were found to be insignificant from only one turnover. Also, due
to lack of data, emissions were not estimated from the handling
and disposal of the sludge removed from the tank bottom.
The EPA considered the approach of one commenter (IV-D-22)
who used the factor of 1 Mg of VOC generated from cleaning a
200,000 gallon tank, generating 2,000 gallons of sludge. The EPA
could not establish the support for this emission factor. For
the purposes of this analysis, therefore, EPA estimated degassing
and cleaning emissions for various size tanks using the method
described above.
TABLE B-l. STORAGE VESSEL DEGASSING AND REFILLING EMISSIONS
Tank
Diameter 30 100 150 180 240
(ft. )
Degassing
Emissions 0.35 Mg 1.40 Mg 3.15 Mg 4.55 Mg 8.1 Mg
B-2
-------
Refilling
Emissions 0.35 Mcr 1.40 Mcr 3.15 Mcr 4.55 Mct 8.1 Mcr
Total
Emissions 0.70 Mg 2.8 Mg 6.3 Mg 9.1 Mg 16.2 Mg
B. 2 BALANCE OF EMISSIONS AND EMISSION REDUCTIONS
The "TANKS" program cited in Chapter 12 of EPA's AP-42
document (IV-A-2) was used to calculate evaporative emissions
from gasoline storage vessels. The following methodology was
then used to determine if degassing and cleaning emissions exceed
the emission reductions achieved by adding NSPS subpart Kb rim
seals and controlled fittings within the statutory requirement of
3 years as suggested by commenters.
First, the annual evaporative emissions were calculated for
storage vessels meeting 40 CFR part 60, subpart K or Ka
requirements [e.g., internal floating roof tanks (IFRT's) with
only a vapor-mounted primary seal or external floating roof tanks
(EFRT's) with only a primary seal]. Second, annual evaporative
emissions were calculated for the same size storage vessels
meeting the NSPS subpart Kb rim seal or the full NSPS subpart Kb
(rim seal and controlled fitting) requirements. Third, emission
reductions resulting from the installation of NSPS subpart Kb rim
seals or NSPS subpart Kb rim seals and controlled fittings were
calculated by taking the difference between the annual NSPS
subpart Kb losses and the annual NSPS subpart K or Ka losses.
Fourth, the degassing and cleaning emissions, which were
calculated using the methodology described in Section B.l, were
divided by the annual emission reductions to determine the number
of years in which the emission reductions achieved by adding the
controls are equivalent to the degassing and cleaning emissions
("years to balance").
Shown in Tables B-2 through B-9 are the "years to balance"
for various types of storage vessels currently meeting the NSPS
B-3
-------
subpart K or Ka level of control when retrofitted to meet either
the NSPS subpart Kb rim seal or the full NSPS subpart Kb control
level.
Tables B-2 through B-4 show the "years to balance" for
storage vessels currently meeting NSPS subpart K or Ka
requirements when retrofitted to meet the NSPS subpart Kb rim
seal requirements. In general, the larger the diameter of the
storage vessel, the longer it takes for the emission reductions
TABLE B-2. EMISSIONS BALANCING—INSTALL SECONDARY SEAL
(SUBPART Kb RIM SEALS) ON AN IFRT WITH A
VAPOR-MOUNTED PRIMARY SEAL
Tank Diameter
(Capacity)
Annual
Emissions at
Ka level of
control
(Mg/yr)
Annual
Emissions
at Kb Rim
Seals
(Mg/yr)
Annual
Emission
Reduction
(Mg/yr)
Degassin
g and
Cleaning
Emission
s
(Mg)
Years
to
Balance
30 ft. (190,360 gal)
1. 93
1.48
0 .45
0.70
1. 6
100 ft. (2,100,000
gal)
7 . 6
6.1
1.5
2 . 8
1. 9
150 ft. (5,287,300
gal)
13.5
11.33
2 .17
6.3
2 . 9
180 ft. (7,613,700
gal)
19.2
16.5
2.7
9.1
3 . 4
TABLE B-3. EMISSIONS BALANCING—INSTALL SECONDARY SEAL
(SUBPART Kb RIM SEALS) ON AN EFRT WITH A
MECHANICAL-SHOE PRIMARY SEAL
Tank Diameter
(Capacity)
Annual
Emissions at
K level of
control
(Mg/yr)
Annual
Emissions
at Kb Rim
Seals
(Mg/yr)
Annual
Emission
Reduction
(Mg/yr)
Degassin
g and
Cleaning
Emission
s
(Mg)
Years
to
Balance
30 ft. (190,360 gal)
6.66
3 . 95
2.71
0.70
0.26
150 ft. (5,287,300
gal)
23.3
9.8
13.5
6.3
0.5
240 ft. (13,535,500
gal)
35 . 9
14.24
21.7
16.2
0.75
NOTE: The environmental impacts for less-controlled
external floating roof tanks (i.e., EFRT's equipped with a
B-4
-------
vapor-mounted primary seal) will be less than shown for the
EFRT's equipped with a mechanical-shoe primary seal.
TABLE B-4. EMISSIONS BALANCING—INSTALL SECONDARY SEAL
(SUBPART Kb RIM SEALS) ON AN EFRT WITH A
LIQUID-MOUNTED PRIMARY SEAL
Tank Diameter
(Capacity)
Annual
Emissions at
K level of
control
(Mg/yr)
Annual
Emissions
at Kb Rim
Seals
(Mg/yr)
Annual
Emission
Reduction
(Mg/yr)
Degassin
g and
Cleaning
Emission
s
(Mg)
Years
to
Balance
30 ft. (190,360 gal)
3 . 8
2 . 8
1
0.70
0.7
150 FT. (5,287,300
gal)
9.25
4.3
4 . 95
6.3
1.3
240 ft. (13,535,500
gal)
12.79
CO
-J
7 . 92
16.2
2 . 0
to balance the degassing and cleaning emissions. As shown in
Table B-2, the "years to balance" for a 150-ft. internal floating
roof tank adding a secondary seal is 2.9 years. The "years to
balance" for a 180-ft. internal floater adding a secondary seal
is 3.4 years, which is longer than the statutory period for
compliance mandated by the Act (3 years). As shown in Tables B-3
and B-4, the "years to balance" for external floating roof tanks
retrofitting to meet NSPS subpart Kb rim seals is well within the
statutory compliance period of 3 years.
Tables B-5 through B-7 show the "years to balance" for
storage vessels currently meeting NSPS subpart K or Ka
requirements when retrofitted to meet the full NSPS subpart Kb
requirements [both rim seals and controlled (gasketed) fittings].
As shown in these tables, the emission reductions achieved by any
storage vessel retrofitting both rim seals and gasketed fittings
balances the degassing and cleaning emissions well within the
statutory compliance period of 3 years.
Tables B-8 and B-9 show the "years to balance" for storage
vessels currently meeting NSPS subpart Kb rim seals to retrofit
to meet the NSPS subpart Kb controlled (gasketed) fitting
B-5
-------
requirements. As shown in Table B-8, the emission reductions
achieved by adding controlled fittings to a 100-ft. internal
floating roof tank meeting NSPS subpart Kb rim seals is slightly
longer than the statutory compliance period of 3 years. Also as
shown in Table B-9, the emission reductions achieved by adding
controlled fittings to a 150-ft. external floating roof is
slightly less than the statutory compliance period of 3 years.
B.3 STORAGE TANK COSTS/ECONOMIC IMPACTS
As stated in Section 13.1, EPA believes that the
installation/equipment costs for storage vessel controls
calculated at proposal are reasonable. However, EPA was
questioned by commenters and has reconsidered the costs for
storage vessel degassing and cleaning, as well as for installing
the emission controls required under NSPS subpart Kb.
B-6
-------
TABLE B-5. EMISSIONS BALANCING—INSTALL SECONDARY SEAL
AND CONTROLLED FITTINGS (SUBPART Kb) ON AN IFRT WITH A
VAPOR-MOUNTED PRIMARY SEAL
Tank Diameter
(Capacity)
Annual
Emissions at
Ka level of
control
(Mg/yr)
Annual
Emissions
at Kb level
of control
(Mg/yr)
Annual
Emission
Reduction
(Mg/yr)
Degassin
g and
Cleaning
Emission
s
(Mg)
Years
to
Balance
30 ft. (190,360 gal)
1. 93
1.16
0.77
0.70
0 . 9
150 ft. (5,287,300
gal)
13.5
9.8
3.7
6.3
1.7
180 ft. (7,613,700
gal)
19.2
14.2
5
9.1
1. 8
240 ft. (13,535,500
gal)
32 . 0
24 .35
7 . 65
16.2
2.2
TABLE B-6. EMISSIONS BALANCING—INSTALL SECONDARY SEAL
AND CONTROLLED FITTINGS (SUBPART Kb) ON AN EFRT WITH A
MECHANICAL-SHOE PRIMARY SEAL
Tank Diameter
(Capacity)
Annual
Emissions at
K level of
control
(Mg/yr)
Annual
Emissions
at Kb level
of control
(Mg/yr)
Annual
Emission
Reduction
(Mg/yr)
Degassin
g and
Cleaning
Emission
s
(Mg)
Years
to
Balance
30 ft (190,360 gal)
6.66
1.71
4 . 95
0.70
0 .14
240 ft (13,535,500
gal)
35 . 9
12
23 . 9
16.2
0 . 68
NOTE: The environmental impacts for less-controlled
external floating roof tanks (i.e., EFRT's equipped with a
vapor-mounted primary seal) will be less than shown for the
EFRT's equipped with a mechanical-shoe primary seal.
TABLE B-7. EMISSIONS BALANCING—INSTALL SECONDARY SEAL
AND CONTROLLED FITTINGS (SUBPART Kb) ON AN EFRT WITH A
LIQUID-MOUNTED PRIMARY SEAL
B-7
-------
Tank Diameter
(Capacity)
Annual
Emissions at
K level of
control
(Mg/yr)
Annual
Emissions
at Kb level
of control
(Mg/yr)
Annual
Emission
Reduction
(Mg/yr)
Degassin
g and
Cleaning
Emission
s
(Mg)
Years
to
Balance
30 ft (190,360 gal)
3.76
0.54
3.22
0.70
0.2
240 ft (13,535,500
gal)
12.79
2 . 64
10 .1
16.2
1. 60
B-8
-------
TABLE B-8. EMISSIONS BALANCING—INSTALL CONTROLLED FITTINGS
(SUBPART Kb) ON AN IFRT WITH A VAPOR-MOUNTED PRIMARY SEAL
AND SECONDARY SEAL (SUBPART Kb RIM SEALS)
Tank Diameter
(Capacity)
Annual
Emissions at
Kb Rim Seal
(Mg/yr)
Annual
Emissions
at Kb level
of control
(Mg/yr)
Annual
Emission
Reduction
(Mg/yr)
Degassin
g and
Cleaning
Emission
s
(Mg)
Years
to
Balance
30 ft. (190,360 gal)
1.48
1.16
0.32
0.70
2 .1
100 ft. (2,100,000
gal)
6.1
5.22
CO
CO
o
2 . 8
3.2
180 ft. (7,613,700
gal)
8.86
6.55
2.31
9.1
3 . 9
240 ft. (13,535,500
gal)
14 . 65
10.59
4.06
16.2
4 . 0
TABLE B-9. EMISSIONS BALANCING—INSTALL CONTROLLED FITTINGS
(SUBPART Kb) ON AN EFRT WITH A MECHANICAL-SHOE PRIMARY SEAL
AND SECONDARY SEAL (SUBPART Kb RIM SEALS)
Tank Diameter
(Capacity)
Annual
Emissions at
Kb Rim Seals
(Mg/yr)
Annual
Emissions
at Kb level
of control
(Mg/yr)
Annual
Emission
Reduction
(Mg/yr)
Degassin
g and
Cleaning
Emission
s
(Mg)
Years
to
Balance
30 ft. (190,360 gal)
3 . 94
1.71
2.23
0.70
0.3
100 ft. (2,100,000
gal)
7.34
5 .11
2.23
2 . 8
1.3
150 ft. (5,287,300
gal)
9.78
7.54
2.24
6.3
2 . 8
180 ft. (7,613,700
gal)
11.28
9.05
2.23
9.1
4 .1
240 ft. (13,535,500
gal)
14 .25
12 .01
2.24
16.2
7.2
NOTE: The environmental impacts for less-controlled
external floating roof tanks (i.e., EFRT's equipped with a
vapor-mounted primary seal) will be less than shown for the
EFRT's equipped with a mechanical-shoe primary seal.
B-9
-------
B.3.1 Degassing and Cleaning Costs
Degassing and cleaning costs were reconsidered using
information from the EPA ACT document (IV-A-3). According to the
ACT document, cleaning and degassing costs can be divided into
two separate costs: 1) cleaning and 2) hazardous waste disposal.
In the ACT document, it is estimated that the cost to clean and
dispose of the hazardous waste generated by a 200,000 gallon tank
is approximately $18,000 to $20,000 (2,000 gallons of sludge and
1,000 gallons of rinseate are generated from the tank). The cost
to dispose of the hazardous waste is approximately $15,000 (@ $5
per gallon of waste) leaving $3,000 to $5,000 for cleaning the
tank. The dimensions of the vessel are not shown in the document
(only the capacity of the tank is given); however, assuming that
the tank is approximately 33 feet in diameter, the cleaning cost
of $5,000 represents a cost of about $150 per foot-diameter. For
the purposes of this analysis, if one assumes a tank cleaning
cost of $150 per foot-diameter, the costs to clean a 50-ft., 78-
ft., and 100-ft. tank are $7,500, $11,700, and $15,000,
respectively.
To calculate sludge disposal costs, the EPA ACT document
stated that 2,000 gallons of sludge would be generated from a
200,000 gallon tank. Assuming that the tank is 33 feet in
diameter, the sludge depth in the tank was calculated at
approximately 3.75 inches. It was assumed that this tank was
storing crude oil, because crude oil is one of the model liquids
discussed in the document. However, for the purposes of this
analysis, it was assumed that gasoline, being a refined petroleum
product, would contain less sludge-forming materials than crude
oil. Therefore, it was assumed that the average sludge depth in
a gasoline storage tank is 2 inches. It was also assumed that no
rinseate would be generated in cleaning the tank since gasoline
is a less viscous fluid than crude oil. Therefore, assuming a
sludge depth of 2 inches and a disposal cost of $5 per gallon,
B-10
-------
the sludge disposal costs for a 50-ft., 78-ft., and 100-ft. tank
are approximately $12,230, $29,770, and $48,930, respectively.
B.3.2 Controlled (or Gasketed) Fitting Costs
As stated in Section 13.1, EPA reconsidered the cost to
install gaskets on existing uncontrolled fittings. Industry
provided estimates obtained from three manufacturers for various
tank sizes (IV-D-44) which varied significantly (100 to 1,300
percent difference). It is uncertain why the cost estimates
ranged as much as 1,300 percent, but for the purposes of this
analysis, EPA believes that installing gaskets should be
relatively inexpensive when performed while degassing and
cleaning the tank. The EPA used the lowest of the three cost
estimates provided by industry because it was closer to the cost
documented in the ACT document. The EPA interpolated between
these reported costs for the given tank sizes to determine the
capital cost to add controlled (gasketed) fittings for the
gasoline distribution model tank sizes. The capital costs for
adding controls to the fittings on the model tanks are $1,225 for
a 50-ft. tank, $2,175 for a 78-ft. tank, and $2,800 for a 100-ft.
tank. The cost effectiveness for installing controlled fittings
(with and without degassing costs) on floating roof tanks is
summarized in Table B-10. As shown in the table, the cost
effectiveness for adding controlled (gasketed) fittings is
significantly higher when the tank degassing and cleaning cost is
included in the compliance costs. However, as also shown in the
table, the approximate increase in the price of gasoline when
considering the cost of installing fittings on the storage tanks
is relatively insignificant (well below 0.1 cent per gallon) with
or without the tank degassing and cleaning costs.
Tables B-ll through B-20 are similar to Tables 7-1 and 7-2
of BID, Volume I and show the costs for model storage tanks of
different control levels to meet the NSPS subpart Kb rim seal
B-ll
-------
requirements only, as well as the full NSPS subpart Kb rim seal
and controlled fitting requirements.
Tables B-21 and B-22 show the model tank and nationwide
costs for bulk terminals and breakout stations to meet only the
NSPS subpart Kb rim seal requirements for storage vessels.
Tables B-23 and B-24 show the model tank and nationwide costs for
bulk terminals and breakout stations to control all their tanks
to meet the rim seal and controlled fitting requirements of NSPS
subpart Kb.
B-12
-------
TABLE B-10. COST AND EMISSIONS IMPACT OF REQUIRING CONTROLLED FITTINGS
ON ALL GASOLINE STORAGE TANKS
Wait Until Next
Costs)
Degassing,
or 10 years
(No Degassing and
Cleaning
Degas
Tank (Adds
Degassing and Cleaning
Costs)
Tank
Type
E .R.
@ 5%
HAP
(Mg/yr
)
E .R.
@ 16%
HAP
(Mg/yr
)
Capital
Cost
($)
Net
Ann.
Cost
($/yr)
C.E.
@ 5%
HAP
($/Mg)
C.E.
@ 16%
HAP
($/Mg
)
A
increas
e in
gasolin
e price
K/gal.
)
E .R.
@ 5%
HAP
(Mg/yr
)
E .R.
@ 16%
HAP
(Mg/yr
)
Capital
Cost
($)
Net
Ann.
Cost
($/yr)
C.E.
@ 5%
HAP
($/Mg)
C.E.
@ 16%
HAP
($/Mg)
A
increas
e in
gasolin
e price
K/gal.
)
PER
TANK
B.T.
EFR
0 .11
0.352
$2,175
($211)
($1,900
)
($600
)
(0.001)
0 .11
0.352
$43,645
$8,813
$80,000
$25,00
0
0 .04
IFR
. 02
.064
$1,225
$142
$7,100
$2,20
0
0.00002
. 02
0 . 064
$20,955
$4,435
$220,00
0
$69,00
0
0 . 05
B.O
EFR
0 .11
.352
$2,800
($57)
$(520)
($160
)
(0.0000
1)
0 .11
0.352
$66,730
$13,854
$126,00
0
$39,00
0
0 .004
IFR
.045
0.144
$2,800
$337
$7,500
$2,30
0
0.0001
.045
0.144
$66,730
$14,248
$317,00
0
$99,00
0
0 .005
NATIO
N-
WIDE
B.T.
EFR
90
$1,600,00
0
($153,000
)
(1,
530)
(0.001)
90
$31,642,0
00
$6,389,00
0
71,
000
0 .04
IFR
15
$780,000
$93,000
6,
200
0 .002
15
$13,774,0
00
$2,916,00
0
194
000
0 . 05
B.O.
EFR
25
$420,000
($49, 000)
(340)
(0.0000
1)
25
$9,440,00
0
$1,960,00
0
78,
000
0 .004
B-13
-------
IFR
$200,000 $24,000
4,800
0.0001
$4,545,00 $970,000
0
194,000
0 .005
B-14
-------
TABLE B-ll. COSTS TO INSTALL FLOATING DECK, LIQUID-MOUNTED
PRIMARY SEAL, AND CONTROLLED FITTINGS ON A 50' FRT
(THIRD QUARTER 1990 DOLLARS)
Tank Capacity = 2,680 m3
Tank Diameter = 15.2 m
Tank Height = 14.6 m
13 Turnovers
Deck + Deck + Seals and
Seal Onlv Controlled
Fittings
Capital Cost & Installation
Degassing, Cleaning, & Waste $19,730 $19,730
Disposal® $19,900 $19,900
Roof with Liquid-Mounted Seal13 $0 $1,225
Controlled Deck Fittings0
Total Capital Cost $39,630 $40,885
Annualized Costs ($/yr)
Maintenance (5%)b $1,982 $2,043
Taxes, Insurance, G&A (4%)b $1,585 $1,634
Inspections (l%)b $369 $409
Annual Capital Charges (11.76%, $4,660 $4,805
20 yrs. @ 10%)c
Total Annualized Cost $8,596 $8,890
Product Recovery Credit $13,743d $13,867e
Net Annualized Cost ($/yr)f ($5,147) ($4,977)
Emission Reduction
Deck + Seal = 4 4.2 Mg VOC
Deck + Seal + Fittings = 44.6 Mg
VOC
Cost Effectiveness @ 5% HAP ($/Mg) (2,329) ($2,232)
Cost Effectiveness @ 16% HAP ($/Mg) ($728) ($697)
B-15
-------
FOOTNOTES FOR TABLE B-ll
a Assuming sludge depth of 2 inches, with a disposal cost of $5
per gallon of sludge. Assumed cleaning cost is approximately
$150 per foot-diameter.
b Docket item IV-A-3.
c Docket item IV-D-44.
d Based on a calculation which subtracts annual evaporative
losses from an internal floating roof tank equipped with a
liquid-mounted primary seal (subpart Kb rim seal requirements)
from uncontrolled evaporative losses from a fixed-roof tank,
and a cost of gasoline at bulk terminals of $0.79/gal.
0 Based on a calculation which subtracts annual evaporative
losses from an internal floating roof tank equipped with a
liquid-mounted primary seal and controlled deck fittings (full
subpart Kb requirements) from uncontrolled evaporative losses
from a fixed-roof tank, and a cost of gasoline at bulk
terminals of $0.79/gal.
f Parentheses indicate a net savings.
B-16
-------
TABLE B-12. COSTS TO INSTALL A FLOATING DECK, LIQUID-MOUNTED
PRIMARY SEAL, AND CONTROLLED FITTINGS ON A 100' FRT
(THIRD QUARTER 1990 DOLLARS)
Tank Capacity = 8,000 m3
Tank Diameter = 30 m
Tank Height = 12 m
150 Turnovers
Deck + Deck + Seals and
Seal Onlv Controlled
Fittings
Capital Cost & Installation
Degassing, Cleaning, & Waste Disposal® $63,930 $63,930
Roof with Liquid-Mounted Seal13 $41, 550 $41,550
Controlled Deck Fittings0 $0 $2,800
Total Capital Cost $105,480 $108,280
Annualized Costs ($/yr)
Maintenance (5%)b $5,274 $5,414
Taxes, Insurance, G&A (4%)b $4,219 $4,331
Inspections (l%)b $1,055 $1,083
Annual Capital Charges (11.76%, 20 yrs. $12,404 $12,734
@ 10%)
Total Annualized Cost $22,952 $23,562
Product Recovery Credit $158,646d $158,9190
Net Annualized Cost ($/yr)f ($135,693 ($135,358)
)
Emission Reduction
Deck + Seal = 522.6 Mg VOC
Deck + Seal + Fittings = 523.5 Mg VOC
Cost Effectiveness @ 5% HAP ($/Mg) ($5,192) ($5,162)
Cost Effectiveness @ 16% HAP ($/Mg) ($1,623) ($1,613)
B-17
-------
FOOTNOTES FOR TABLE B-12
a Assuming sludge depth of 2 inches, with a disposal cost of
$5/gallon of sludge. Assumed cleaning cost is approximately
$150 per foot-diameter.
b Docket item IV-A-3.
c Docket item IV-D-44.
d Based on a calculation which subtracts annual evaporative
losses from an internal floating roof tank equipped with a
liquid-mounted primary seal (subpart Kb rim seal requirements)
from uncontrolled evaporative losses from a fixed-roof tank,
and a cost of gasoline at breakout stations of $0.77/gal.
0 Based on a calculation which subtracts annual evaporative
losses from an internal floating roof tank equipped with a
liquid-mounted primary seal and controlled deck fittings (full
subpart Kb requirements) from uncontrolled evaporative losses
from a fixed-roof tank, and a cost of gasoline at breakout
stations of $0.77/gal.
f Parentheses indicate a net savings.
B-18
-------
TABLE B-13. COSTS TO INSTALL SECONDARY SEAL AND CONTROLLED
FITTINGS ON 78' EFRT WITH A MECHANICAL-SHOE PRIMARY SEAL
(THIRD QUARTER 1990 DOLLARS)
Tank Capacity = 5,7 60 m3
Tank Diameter = 23.8 m
Tank Height = 12 m
13 Turnovers
Capital Cost & Installation
Degassing, Cleaning, & Waste Disposal®
Secondary Seal Cost13
Controlled Deck Fittings0
Total Capital Cost
Annualized Costs ($/yr)
Maintenance (5%)b
Taxes, Insurance, G&A (4%)b
Inspections (l%)b
Annual Capital Charges (11.76%, 20 yrs.
@ 10%)
Secondary
Seal Onlv
$41,470
$13,200
$0
$54,670
$2,734
$2,187
$547
$6,429
Secondary Seal
and
Controlled
Fittings
$41,470
$13,200
$2,175
$56,845
$2,842
$2,274
$568
$6,685
Total Annualized Cost
Product Recovery Credit
$11,897
$3, 12 6d
$12,369
$3,824e
Net Annualized Cost ($/yr) $8,770
Emission Reduction
Secondary Seal = 10.1 Mg VOC
Secondary Seal + Fittings = 12.3 Mg VOC
Cost Effectiveness @ 5% HAP ($/Mg) $17,340
Cost Effectiveness @ 16% HAP ($/Mg) $5,419
$8,545
$13,895
$4,342
B-19
-------
FOOTNOTES FOR TABLE B-13
a Assuming sludge depth of 2 inches, with a disposal cost of $5
per gallon of sludge. Assumed cleaning cost is approximately
$150 per foot-diameter.
b Docket item IV-A-3.
c Docket item IV-D-44.
d Based on a calculation which subtracts annual evaporative
losses from an external floating roof tank equipped with a
mechanical-shoe primary seal and secondary seal (subpart Kb r
seal requirements) from losses from an external floating roof
tank equipped with a mechanical-shoe primary seal (subpart K
requirements), and a cost of gasoline at bulk terminals of
$0.79/gal.
0 Based on a calculation which subtracts annual evaporative
losses from an external floating roof tank equipped with a
secondary seal and controlled deck fittings (full subpart Kb
requirements) from losses from an external floating roof tank
equipped with a mechanical-shoe primary seal (subpart K
requirements), and a cost of gasoline at bulk terminals of
$ 0.7 9/gal.
B-20
-------
TABLE B-14. COSTS TO INSTALL SECONDARY SEAL AND CONTROLLED
FITTINGS ON 100' EFRT WITH A MECHANICAL-SHOE PRIMARY SEAL
(THIRD QUARTER 1990 DOLLARS)
Tank Capacity = 8,000 m3
Tank Diameter = 30 m
Tank Height = 12 m
150 Turnovers
Capital Cost & Installation
Degassing, Cleaning, & Waste Disposal®
Secondary Seal Cost13
Controlled Deck Fittings0
Total Capital Cost
Annualized Costs ($/yr)
Maintenance (5%)b
Taxes, Insurance, G&A (4%)b
Inspections (l%)b
Annual Capital Charges (11.76%, 20 yrs.
@ 10%)
Secondary
Seal Onlv
$63,930
$16,960
$0
$80,890
$4,405
$3,236
$809
$9,513
Secondary Seal
and
Controlled
Fittings
$63,930
$16,960
$2,800
$83,690
$4,185
$3,348
$837
$9,842
Total Annualized Cost
Product Recovery Credit
$17,963
$3, 87 9d
$18,211
$4, 546e
Net Annualized Cost ($/yr) $14,084
Emission Reduction
Secondary Seal = 12.8 Mg VOC
Secondary Seal + Fittings = 15.0 Mg VOC
Cost Effectiveness @ 5% HAP ($/Mg) $22,006
Cost Effectiveness @ 16% HAP ($/Mg) $6,877
$13,665
$18,220
$5,694
B-21
-------
FOOTNOTES FOR TABLE B-14
a Assuming sludge depth of 2 inches, with a disposal cost of $5
per gallon of sludge. Assumed cleaning cost is approximately
$150 per foot-diameter.
b Docket item IV-A-3.
c Docket item IV-D-44.
d Based on a calculation which subtracts annual evaporative
losses from an external floating roof tank equipped with a
mechanical-shoe primary and secondary seal (subpart Kb rim seal
requirements) from an external floating roof tank equipped with
a mechanical-shoe primary seal (subpart K requirements), and a
cost of gasoline at breakout stations of $0.77/gal.
0 Based on a calculation which subtracts annual evaporative
losses from an external floating roof tank equipped with a
primary and secondary seal and controlled deck fittings (full
subpart Kb requirements) from losses from an external floating
roof tank equipped with a mechanical-shoe primary seal (subpart
K requirements), and a cost of gasoline at breakout stations of
$0.77/gal.
B-22
-------
TABLE B-15. COSTS TO INSTALL SECONDARY SEAL AND CONTROLLED
FITTINGS ON 50' IFRT WITH A VAPOR-MOUNTED PRIMARY SEAL
(THIRD QUARTER 1990 DOLLARS)
Tank Capacity = 2,680 m3
Tank Diameter = 23.8 m
Tank Height = 12 m
13 Turnovers
Capital Cost & Installation
Degassing, Cleaning, & Waste Disposal®
Secondary Seal Cost13
Controlled Deck Fittings0
Total Capital Cost
Annualized Costs ($/yr)
Maintenance (5%)b
Taxes, Insurance, G&A (4%)b
Inspections (l%)b
Annual Capital Charges (11.76%, 20 yrs.
@ 10%)
Secondary
Seal Onlv
$19,730
$4,080
$0
$23,810
$1,191
$952
$238
$2,800
Secondary Seal
and
Controlled
Fittings
$19,730
$4,080
$1,225
$25,035
$1,252
$1,001
$250
$2,944
Total Annualized Cost
Product Recovery Credit
$5,181
$218d
$5,448
$342e
Net Annualized Cost ($/yr) $4,963
Emission Reduction
Secondary Seal = 0.7 Mg VOC
Secondary Seal + Fittings = 1.1 Mg VOC
Cost Effectiveness @ 5% HAP ($/Mg) $141,800
Cost Effectiveness @ 16% HAP ($/Mg) $44,313
$5,106
$92,829
$29,009
B-23
-------
FOOTNOTES FOR TABLE B-15
a Assuming sludge depth of 2 inches, with a disposal cost of $5
per gallon of sludge. Assumed cleaning cost is approximately
$150 per foot-diameter.
b Docket item IV-A-3
c Docket item IV-D-44.
d Based on a calculation which subtracts annual evaporative
losses from an external floating roof tank equipped with a
vapor-mounted primary seal and secondary seal (subpart Kb rim
seal requirements) from losses from an external floating roof
tank equipped with only a vapor-mounted primary seal (subpart K
requirements), and a cost of gasoline at bulk terminals of
$ 0.7 9/gal.
0 Based on a calculation which subtracts annual evaporative
losses from an external floating roof tank equipped with a
vapor-mounted primary seal, secondary seal, and controlled
fittings (full subpart Kb requirements) from losses from an
external floating roof tank equipped with only a vapor-mounted
primary seal (subpart K requirements), and a cost of gasoline
at bulk terminals of $0.79/gal.
B-24
-------
TABLE B-16. COSTS TO INSTALL SECONDARY SEAL AND CONTROLLED
FITTINGS ON 100' IFRT WITH A VAPOR-MOUNTED PRIMARY SEAL
(THIRD QUARTER 1990 DOLLARS)
Tank Capacity = 8,000 m3
Tank Diameter = 30 m
Tank Height = 12 m
150 Turnovers
Capital Cost & Installation
Degassing, Cleaning, & Waste
Disposal®
Secondary Seal Cost13
Controlled Deck Fittings0
Total Capital Cost
Annualized Costs ($/yr)
Maintenance (5%)b
Taxes, Insurance, G&A (4%)b
Inspections (l%)b
Annual Capital Charges (11.76%,
yrs. @ 10%)
Secondary
Seal Onlv
$63,930
$8,170
$0
$72,100
$3,605
$2,884
$721
$8,479
Secondary Seal
and Controlled
Fittings
$63,930
$8,170
$2,800
$74,900
$3,745
$2,996
$749
$8,808
Total Annualized Cost $15,689 $16,298
Product Recovery Credit $454d $727e
Net Annualized Cost ($/yr) $15,235 $15,571
Emission Reduction
Secondary Seal = 1.5 Mg VOC
Secondary Seal + Fittings = 2.4 Mg
VOC
Cost Effectiveness @ 5% HAP ($/Mg) $203,133 $129,758
Cost Effectiveness @ 16% HAP ($/Mg) $63,479 $40,549
B-25
-------
FOOTNOTES FOR TABLE B-16
a Assuming sludge depth of 2 inches, with a disposal cost of $5
per gallon of sludge. Assumed cleaning cost is approximately
$150 per foot-diameter.
b Docket item IV-A-3.
c Docket item IV-D-44.
d Based on a calculation which subtracts annual evaporative
losses from an internal floating roof tank equipped with a
vapor-mounted primary seal and secondary seal (subpart Kb rim
seal requirements) from losses from an internal floating roof
tank equipped with a vapor-mounted primary seal (subpart K
requirements), and a cost of gasoline at breakout stations of
$0.77/gal.
0 Based on a calculation which subtracts annual evaporative
losses from an internal floating roof tank equipped with a
vapor-mounted primary seal, secondary seal, and controlled deck
fittings (full subpart Kb requirements) from losses from an
internal floating roof tank equipped with only a vapor-mounted
primary seal (subpart K requirements), and a cost of gasoline
at breakout stations of $0.77/gal.
B-26
-------
TABLE B-17. COSTS TO INSTALL CONTROLLED FITTINGS ON
50' IFRT WITH A LIQUID OR MECHANICAL PRIMARY SEAL—
WITH AND WITHOUT DEGASSING COSTS
(THIRD QUARTER 1990 DOLLARS)
Tank Capacity = 2,680 m3
Tank Diameter = 15.2 m
Tank Height = 14.6 m
13 Turnovers
With Degassing Without
Costs Degassing
Costs
Capital Cost & Installation
Degassing, Cleaning, & Waste Disposal® $19,730 $0
Roof with Liquid-Mounted Seal13 $0 $0
Controlled Deck Fittings0 $1,2250 $1,225
Total Capital Cost $20,955 $1,225
Annualized Costs ($/yr)
Maintenance (5%)b $1,048 $61
Taxes, Insurance, G&A (4%)b $838 $49
Inspections (l%)b $210 $12
Annual Capital Charges (11.76%, 20 yrs. $2,464 $144
@ 10%)a
Total Annualized Cost $4,560 $267
Product Recovery Credit"1 $124 $124
Net Annualized Cost ($/yr) $4,435 $142
Emission Reduction
(controlled fittings) = 0.4 Mg VOC
Cost Effectiveness @ 5% HAP ($/Mg) $221,722 $7,110
Cost Effectiveness @ 16% HAP ($/Mg) $69,304 $2,222
a Assuming sludge depth of 2 inches, with a disposal cost of $5
per gallon of sludge. Assumed cleaning cost is approximately
$150 per foot-diameter.
b Docket item IV-A-3
c Docket item IV-D-44.
d Based on a calculation which subtracts annual evaporative
losses from an internal floating roof tank equipped with either
a liquid-mounted or mechanical-shoe primary seal and controlled
deck fittings (full subpart Kb requirements) from losses from
an internal floating roof tank equipped with only a liquid-
mounted primary seal (subpart Ka requirements), and a cost of
gasoline at bulk terminals of $0.79/gal.
B-27
-------
TABLE B-18. COSTS TO INSTALL CONTROLLED FITTINGS ON
100' IFRT WITH A LIQUID OR MECHANICAL PRIMARY SEAL—
WITH AND WITHOUT DEGASSING COSTS
(THIRD QUARTER 1990 DOLLARS)
Tank Capacity = 8,000 m3
Tank Diameter = 30 m
Tank Height = 12 m
150 Turnovers
Capital Cost & Installation
Degassing, Cleaning, & Waste Disposal®
Roof with Liquid-Mounted Seal13
Controlled Deck Fittings0
Total Capital Cost
Annualized Costs ($/yr)
Maintenance (5%)b
Taxes, Insurance, G&A (4%)b
Inspections (l%)b
Annual Capital Charges (11.76%, 20 yrs.
@ 10%)
Total Annualized Cost
Product Recovery Credit"1
With Degassing
Costs
$63,930
$0
$2,800
$66,730
$3,337
$2,669
$667
$7,847
$14,520
$273
Without
Degassing
Costs
$0
$0
$2,800
$2,800
$140
$112
$28
$329
$609
$273
Net Annualized Cost ($/yr)
$14,248
$337
Emission Reduction
(controlled fittings) =
0.9 Mg VOC
Cost Effectiveness @ 5% HAP ($/Mg)
Cost Effectiveness @ 16% HAP ($/Mg)
$316,616
$98,942
$7,749
$2,337
a Assuming sludge depth of 2 inches, with a disposal cost of
$5/gallon of sludge. Assumed cleaning cost is approximately
$150 per foot-diameter.
b Docket item IV-A-3.
c Docket item IV-D-44.
d Based on a calculation which subtracts annual evaporative
losses from an internal floating roof tank equipped with either
a liquid-mounted or mechanical-shoe primary seal and controlled
deck fittings (full subpart Kb requirements) from losses from
an internal floating roof tank with only a liquid-mounted or
mechanical-shoe primary seal (subpart Ka requirements), and a
cost of gasoline at breakout stations of $0.77/gal.
B-28
-------
B-29
-------
TABLE B-19. COSTS TO INSTALL CONTROLLED FITTINGS ON A
78' EFRT WITH A MECHANICAL-SHOE PRIMARY SEAL AND SECONDARY SEAL—
WITH AND WITHOUT DEGASSING COSTS
(THIRD QUARTER 1990 DOLLARS)
Tank Capacity = 5,7 60 m3
Tank Diameter = 23.8 m
Tank Height = 12 m
13 Turnovers
With degassing Without
Costs degassing costs
Capital Cost & Installation
Degassing, Cleaning, & Waste Disposal®
Secondary Seal Cost13
Controlled Deck Fittings0
Total Capital Cost
Annualized Costs ($/yr)
Maintenance (5%)b
Taxes, Insurance, G&A (4%)b
Inspections (l%)b
$41,470
$0
$2,175
$43,645
$2,182
$1,746
$436
$0
$0
$2,175
$2,175
$109
$87
$22
Annual Capital Charges (11.76%, 20 yrs.
@ 10%)
$5,133
$256
Total Annualized Cost $9,497 $473
Product Recovery Credit"1 $684 $684
Net Annualized Cost ($/yr)e $8,813 ($211)
Emission Reduction
(controlled fittings) = 2.2 Mg VOC
Cost Effectiveness @ 5% HAP ($/Mg) $80,199 ($1,196)
Cost Effectiveness @ 16% HAP ($/Mg) $25,037 ($599)
B-30
-------
FOOTNOTES FOR TABLE B-19
a Assuming sludge depth of 2 inches, with a disposal cost of $5
per gallon of sludge. Assumed cleaning cost is approximately
$150 per foot-diameter.
b Docket item IV-A-3.
c Docket item IV-D-44.
d Based on a calculation which subtracts annual evaporative
losses from an external floating roof tank equipped with a
mechanical-shoe primary seal, secondary seal, and controlled
fittings (full subpart Kb requirements) from losses from an
external floating roof tank equipped with only a mechanical-
shoe primary seal and secondary seal (subpart Ka requirements),
and a cost of gasoline at bulk terminals of $0.79/gal.
e Parentheses indicate a net savings.
B-31
-------
TABLE B-20. COSTS TO INSTALL CONTROLLED FITTINGS ON
100' EFR WITH A MECHANICAL-SHOE PRIMARY SEAL AND SECONDARY SEAL—
WITH AND WITHOUT DEGASSING COSTS
(THIRD QUARTER 1990 DOLLARS)
Tank Capacity = 8,000 m3
Tank Diameter = 30 m
Tank Height = 12 m
150 Turnovers
With Degassing Without
Costs Degassing Costs
Capital Cost & Installation
Degassing, Cleaning, & Waste Disposal®
Secondary Seal Cost13
Controlled Deck Fittings0
Total Capital Cost
Annualized Costs ($/yr)
Maintenance (5%)b
Taxes, Insurance, G&A (4%)b
Inspections (l%)b
$63,930
$0
$2,800
$66,730
$3,337
$2,669
$667
$0
$0
$2,800
$2,800
$140
$112
$28
Annual Capital Charges (11.76%, 20 yrs.
@ 10%)
$7,847
$329
Total Annualized Cost
Product Recovery Credit"3
Net Annualized Cost ($/yr)G
Emission Reduction
(controlled fittings) = 2.;
Mg VOC
Cost Effectiveness @ 5% HAP ($/Mg)
Cost Effectiveness @ 16% HAP ($/Mg)
$14,520
$667
$13,854
$125,943
$39,357
$609
$667
($57)
($522)
($1,163)
B-32
-------
FOOTNOTES FOR TABLE B-2 0
a Assuming sludge depth of 2 inches, with a disposal cost of $5
per gallon of sludge. Assumed cleaning cost is approximately
$150 per foot-diameter.
b Docket item IV-A-3.
c Docket item IV-D-44.
d Based on a calculation which subtracts annual evaporative
losses from an external floating roof tank equipped with a
mechanical-shoe primary seal, secondary seal, and controlled
fittings (full subpart Kb requirements) from losses from an
external floating roof tank equipped with only a mechanical-
shoe primary seal and secondary seal (subpart Ka requirements),
and a cost of gasoline at breakout stations of $0.77/gal.
e Parentheses indicate a net savings.
B-33
-------
TABLE B-21. BULK TERMINAL MODEL TANK AND NATIONWIDE COSTS
TO MEET SUBPART Kb RIM SEAL REQUIREMENTS (FLOOR)
Tank Type
No . of
Tanks
Model Tank
Capital
Cost ($)
Model
Tank
Annual
Cost ($)
Nationwide
Capital
Costs
($)
Nationwide
Annual Costs
($/year)
External floater w/seca
309
$0
$0
$0
$0
External floater
w/primaryb
416
$56,670
$8,770
$22,742,720
$3,648,320
Internal floater
w/vapor-mounted
primaryb
154
$23,810
$4,963
$3,666,740
$764,302
Internal floater
w/liquid-mounted
primary3
314
$0
$0
$0
$0
Fixed-roofc
184
$39,630
($5,147)
$7,291,920
($947,048)
TOTAL
$33,701,380
$3,465,574
a Currently meeting subpart Kb rim seal requirements and incurs
no cost to meet the floor.
b Needs a secondary seal to meet subpart Kb rim seal
requirements (floor).
c Needs a floating deck with a liquid-mounted primary seal to
meet subpart Kb rim seal requirements (floor).
TABLE B-22. BREAKOUT STATION MODEL TANK AND NATIONWIDE COSTS
TO MEET SUBPART Kb RIM SEAL REQUIREMENTS (FLOOR)
Tank Type
No . of
Tanks
Model Tank
Capital
Cost
($)
Model Tank
Annual
Cost
($/year)
Nationwide
Capital
Costs
($)
Nationwide
Annual
Costs
($/year)
External floater
w/ seca
54
$0
$0
$0
$0
External floater
w/primaryb
95
$80,890
$14,084
$7,684,550
$1,337,980
Internal floater
w/vapor-mounted
primaryb
15
$72,100
$15,235
$1,081,500
$228,525
Internal floater
w/liquid-mounted
primary3
$0
$0
$0
$0
Fixed-roofc
28
$105,480
($135,693)
$2,953,440
($3,799,404
)
TOTAL
$11,719,490
($2,232,899
)
B-34
-------
a Currently meeting subpart Kb rim seal requirements and incurs
no cost to meet the floor.
b Needs a secondary seal to meet subpart Kb rim seal
requirements (floor).
c Needs a floating deck with a liquid-mounted primary seal to
meet subpart Kb rim seal requirements (floor).
B-35
-------
TABLE B-23. BULK TERMINAL MODEL TANK AND NATIONWIDE COSTS
TO MEET FULL SUBPART Kb REQUIREMENTS FOR ALL TANKS
Tank Type
No . of
Tanks
Model Tank
Capital
Cost ($)
Model
Tank
Annual
Cost ($)
Nationwide
Capital
Costs
($)
Nationwide
Annual Costs
($/year)
External floater w/seca
309
$2,175
(211)
$672,075
($65,200)
External floater
w/primaryb
416
$56,845
$8,545
$23,647,520
$3,555,000
Internal floater
w/vapor-mounted
primaryb
154
$25,035
$5,106
$3,855,390
$786,300
Internal floater
w/liquid-mounted
primary3
314
$1,225
$142
$384,650
$44,600
Fixed-roofc
184
$40,855
($4,977)
$7,517,320
($915,800)
TOTAL
$36,076,955
$3,405,000
Currently meeting NSPS subpart Kb rim seal requirements and
needs only the controlled fittings.
Needs a secondary seal and controlled fittings.
Needs a floating deck, liquid-mounted primary seal, and
controlled fittings.
TABLE B-24. BREAKOUT STATION MODEL TANK AND NATIONWIDE COSTS
TO MEET FULL SUBPART Kb REQUIREMENTS FOR ALL TANKS
Tank Type
No . of
Tanks
Model Tank
Capital
Cost
($)
Model Tank
Annual
Cost
($/year)
Nationwide
Capital
Costs
($)
Nationwide
Annual
Costs
($/year)
External floater
w/ seca
54
$2,800
($57)
$151,200
($3,078)
External floater
w/primaryb
95
$83,690
$13,665
$7,950,550
$1,298,175
Internal floater
w/vapor-mounted
primaryb
15
$74,900
$15,571
$1,123,500
$233,565
Internal floater
w/liquid-mounted
primary3
29
$2,800
$337
$81,200
$9,773
Fixed-roofc
28
$108,280
($135,358)
$3,031,840
($3,790,024
)
TOTAL
$12,338,290
($2,251,589
)
B-36
-------
a Currently meeting NSPS subpart Kb rim seal requirements and
needs only the controlled fittings.
b Needs a secondary seal and controlled fittings.
c Needs a floating deck, liquid-mounted primary seal, and
controlled fittings.
B-37
-------
B.4 REFERENCES
IV-A-2
IV-A-3
IV-D-22
IV-D-44
B-38
Compilation of Air Pollutant Emission Factors
(AP-42). Chapter 12, Storage of Organic
Liquids. October 1992. U.S. Environmental
Protection Agency, Research Triangle Park,
NC.
Alternative Control Techniques Document:
Volatile Organic Liquid Storage in Floating
and Fixed Roof Tanks. EPA-453/R-94-001.
U.S. Environmental Protection Agency,
Research Triangle Park, NC. January 1994.
API Storage Tank Analysis on Net Emission
Increases from Premature Degassing and
Cleaning of Gasoline Storage Tanks, and Cost
Analysis. March 29, 1994. Appendix H of
API's comments.
Memorandum. Ferry, R., The TGB Partnership,
to Shedd, S., U.S. Environmental Protection
Agency. June 3, 1994. Cost Basis for IFRT
Fitting Controls Gasoline Distribution MACT
for Storage Tanks.
-------
APPENDIX C.
EMISSIONS FROM EQUIPMENT COMPONENTS AT BULK TERMINALS
AND PIPELINE BREAKOUT STATIONS
The purpose of this appendix is to discuss the methodology
used to recalculate the emissions occurring from equipment
components at bulk terminals and pipeline breakout stations in
the gasoline distribution (Stage I) industry. Two major elements
have been changed in the analysis prepared before proposal--the
emission factors and the equipment counts at each model facility.
Section C.l discusses the new emission factors, Section C.2
discusses the revised equipment component populations, and
Section C.3 summarizes the revised baseline equipment emissions
occurring at bulk terminals and pipeline breakout stations.
C.l EMISSION FACTORS
As discussed in Section 9.2, at proposal EPA used refinery
equipment leak emission factors to calculate the baseline
equipment emissions for both bulk terminals and pipeline breakout
stations. However, the American Petroleum Institute (API)
provided mass emissions data [bagging of leaks at four terminals
and equipment leak frequency data from 74 terminals and pipeline
facilities (IV-D-22)] which EPA evaluated and then used to
recalculate emissions from leaking equipment. Based on the API
data, EPA determined that emissions occurring from equipment at
bulk gasoline terminals and pipeline breakout stations are much
less than was previously estimated through use of the refinery
emission factors.
C-l
-------
The EPA issued a draft report for public review, discussing
the preliminary results of the comparison and analysis of the
1993 refinery data, the 1993 marketing terminal data, and the
1980 refinery data. It is expected that new EPA emission factors
for equipment at gasoline distribution facilities will be
developed in the near future. For this analysis, API's suggested
emission factors have been used.
Table C-l compares the refinery emission factors used at
proposal to the new API emission factors. The API's new emission
factor for pump seals is approximately 99.8 percent lower, and
the new factor for valves is 99.6 percent lower, than the
respective refinery emission factors for these components.
At proposal, emission factors were not available for connectors,
loading arm valves, open-ended lines, and other miscellaneous
components at bulk terminals and pipeline breakout stations and,
therefore, EPA did not attempt to estimate emissions from such
equipment.
C.2 EQUIPMENT POPULATIONS
The API's new emission factors for equipment leaks included
factors for connectors, open-ended lines, loading arm valves, and
other miscellaneous components (IV-D-22, Appendices A and B).
Also included in the data were equipment component counts for
both bulk terminals and pipeline breakout stations. In order to
estimate emissions from equipment leaks using the new API
factors, EPA estimated the equipment component populations for
both bulk terminals and pipeline breakout stations using the
equipment count data submitted by API.
C.2.1 Equipment Populations for Bulk Terminals
The API provided selected characteristics (e.g., number of
components and throughputs) for four petroleum marketing
terminals in Table 2-1 of Appendix A of their comments (IV-D-22).
Table C-2 below summarizes the throughput and component count
data extracted from that table. Using these data, EPA
C-2
-------
reevaluated the equipment count populations for the four "model
plants" in BID, Volume I (III-B-1). As a starting point, EPA
assumed that Terminals A and D shown in Table C-2 were comparable
to Model Plant 4 in BID, Volume I on the basis of throughput.
TABLE C—1. REVISED EMISSION FACTORS FOR EQUIPMENT LEAKS
Equipment Type
Emission Factor Used
at Proposal
(lbs/hr)a
Revised
Emission
Factor
(lbs/hr)b
Pump Seals
liquid service
Valves
Connectors
Loading Arm Valves
gas service
liquid service
Open-Ended Lines
gas service
liquid service
Other
0.25
0 . 024
Not Available
Not Available
Not Available
Not Available
Not Available
Not Available
0.00053
9.2E-05
8.4E-05
0 . 045
0 . 00087
0 . 0067
0 . 0065
0.0003
Docket item II-A-17.
Docket item IV-D-22. (The EPA expects to develop
its own emission factors in the near future.)
TABLE C-2. TERMINAL THROUGHPUTS AND COMPONENT COUNTS3
Terminal Terminal Terminal Terminal Terminal
Description A B CD
Avg. >900,000 >2,500,000 >1,300,000 >650,000
throughput
(gal/day)
C-3
-------
# of light 3,600 6,800 5,800 4,300
liquid and
gas equipment
components
a Appendix A of docket item IV-D-22. Extract from
Table 2-1, Comparison of Selected Petroleum Marketing
Terminals Characteristics.
Consequently, it was assumed that the average of the number of
light liquid and gas components at terminals A and D
(approximately 4,000 components) represent the Model Plant 4
facility.
Also provided in the API data was information that
approximated the percentage of each component type and service
occurring at a bulk gasoline terminal. These data are summarized
in Table C-3.
Based on the assumption made at proposal that Model Plant 4
has 12 loading arms (3 loading arms per rack), it was assumed
that there are 12 loading arm valves (LAV's) in liquid service
and 4 LAV's in gas vapor service (1 vapor return line per loading
rack). All other equipment component types were estimated using
the distribution of component types (by percent) shown in Table
C-3 (excluding components in heavy liquid service). The total
equipment populations for bulk terminal Model Plants 1, 2, and 3
were estimated by scaling down Model Plant 4 using the ratio of
the number of components for each model plant shown in Table 5-3
of BID, Volume 1. For example, at proposal it was estimated that
Model Plant 1 has 100 components and Model Plant 4 has 170
components. To re-estimate the total number of equipment
components for Model Plant 1, the ratio of 100/170 was applied to
the total number of components in Model Plant 4 (approximately
4,000). The total equipment populations for Model Plants 1, 2,
and 3 were distributed into the various component types using the
same methodology as used for Model Plant 4. The distribution of
C-4
-------
the total number of components for each of the bulk terminal
model plants is shown in Table C-4.
C.2.1 Equipment Populations for Breakout Stations
The API provided equipment counts for 36 pipeline breakout
stations in Appendix B of their comments (IV-D-22). These data
are summarized in Table C-5. As shown in Table C-5, the range of
the number of component types varies significantly among breakout
stations. For example, the number of fitting/connectors ranges
from 15 components at the code-named "Plum" facility to 169,587
components at the "Zucchini" facility. Also, there was no
TABLE C-3. OCCURRENCE OF COMPONENT TYPES AT
BULK TERMINALS3
Component Type and Service Percent of
Total
Connector:
Gas Vapor 5.15
Heavy Liquid 5.22
Light Liquid 74.86
Valves:
Gas Vapor 0.57
Heavy Liquid 1.3
Light Liquid 10.48
Loading Arm Valves:
Gas Vapor 0.2
Heavy Liquid 0.04
Light Liquid 0.49
Open-Ended Lines:
Gas Vapor 0.08
Heavy Liquid 0.08
Light Liquid 0.44
Pump Seals:
Gas Vapor NA
Heavy Liquid 0.05
Light Liquid 0.41
"Other"
Gas Vapor 0.09
Heavy Liquid 0.03
Light Liquid 0.51
C-5
-------
Total
100.00
a Appendix A of docket item IV-D-22.
Extract from Table 5-13, Summary of
Components Studied.
discussion regarding the throughput characteristics for any of
the breakout stations.
At proposal, EPA assumed two model plant sizes for pipeline
breakout stations (see Table 5-2 of BID, Volume 1). The average
number of components shown in Table C-5 was distributed among the
two model plants by assuming that Model Plant 1 has 50 percent of
the average number of each component type and that Model Plant 2
has 150 percent of the average number of each component type.
TABLE C-4. EQUIPMENT COMPONENT POPULATIONS AT
BULK TERMINALS
MP1
# Comp.
MP2
# Comp.
MP3
# Comp.
MP4
# Comp.
Pump seals
LLa
10
15
15
20
Valves
290
350
400
500
Connectors
2, 000
2,400
2,700
3,400
Loading Arm Valves
gasb
2
3
3
4
LLa
6
9
9
12
Open-Ended Lines
gasb
7
8
8
12
LLa
7
8
8
12
Other0
15
17
20
25
# Comp.
2,337
2, 810
3, 163
3, 985
# Facilities
410
230
280
100
LL = light liquid service,
gas = gas vapor service.
C-6
-------
c "Other" - as stated in Appendix A of API's comments,
includes components such as hatches, covers, manholes,
thermal wells, and pressure relief valves.
The revised number of components for breakout stations is shown
in Table C-6.
C.3 REVISED NATIONWIDE BASELINE EQUIPMENT EMISSIONS
The revised equipment component populations shown in Tables
C-4 and C-6 and the new emission factors shown in Table C-l were
used to recalculate the nationwide baseline emissions occurring
from equipment leaks. The basic methodology used to recalculate
the new baseline emissions (per model plant) is shown below.
# of Facilities x # components x Emission Factor x 24 hrs/day x # operating davs x Ma
(per model plant) (by type) (lbs/hr) yr 2,204
lbs
The total emissions for each model plant are added together
over all model plant sizes to derive the nationwide totals. It
TABLE C-5. BREAKOUT STATION COMPONENT DATA (IV-D-22, App. A)
Station ID
ifittings/
conn.
Others
PRD's *
Pumps
Valves
Total
Avocado
343
140
129
110
1, 397
2,119
Brussels
Sprout
4, 602
1, 059
1, 539
351
17,406
24,957
Cabbage
20,534
497
—
95
7, 194
28,320
Cashew
2, 550
239
53
40
830
3, 712
Cauliflower
20
10
—
—
75
105
Chive
806
277
13
6 6
820
1, 982
Cilantro
203
126
63
—
602
994
Date
6, 322
253
96
137
4, 829
11,637
Fig
24,481
792
48
7
7, 137
32,465
Grape
855
19
10
6
220
1,110
Grapefruit
19,025
2, 823
110
301
9, 764
32,023
Kumquat
3,210
1, 176
677
399
7, 352
12,814
Leek
104
166
73
32
900
1, 275
Lemon
15
10
50
75
C-l
-------
Lettuce
12,
. 918
189
602
75
5,
. 971
19,
.755
Nutmeg
15
15
—
5
185
220
Okra
2,
. 555
397
65
34
1,
. 117
4,
.168
Olive
3,
.309
246
76
26
1,
.416
5,
. 073
Papaya
2,
.500
79
32
19
794
3,
. 424
Peanut
11,
. 827
1,
, 664
288
100
6,
. 028
19,
. 907
Pear
4,
.969
170
52
52
1,
.167
6,
.410
Pickle
10,
. 437
500
2,442
253
8,
. 117
21,
.749
Pistachio
756
13
36
24
435
1,
.264
Plum
15
—
—
4
85
104
Poi
3,
.394
121
286
77
1,
.760
5,
. 638
Prune
68
20
—
12
232
332
Radish
470
245
10
50
2,
. 350
3,
. 125
Raisin
24,
.898
3,
, 333
687
99
6,
.768
35,
.785
Rice
1,
. 103
72
12
36
598
1,
. 821
Spinach
130
56
10
18
460
674
Squash
6,
. 425
622
600
109
2,
. 971
10,
.727
Tangerine
671
20
1
4
161
857
Truffle
12,
.464
590
—
63
4,
. 227
17,
.344
Turnip
128
112
34
—
396
670
Y am
955
43
69
11
373
1,
. 451
Zucchini
169,
.587
8,
,589
2.729
1.358
85,
. 948
268,
.211
Average
9,
.796
686
301
110
5,
.282
16,
. 175
*PRD's = Pressure Relief Devices.
C-8
-------
is assumed that bulk terminals operate 340 days per year and
breakout stations operate 365 days per year). The nationwide
emissions are summarized in Table C-7.
C-9
-------
TABLE C-6. REVISED EQUIPMENT COMPONENT POPULATIONS
FOR BREAKOUT STATIONS
MP1
# Comp,
MP2
# Comp,
Emission
Factor
(lbs/hr)
Pump seals
(2/pump)
LL
Valves
-avg.
Connectors
-avg.
Other
-avg.
# Comp.
# Facilities
110
2, 640
4, 900
340
7, 990
150
330
7, 920
14,700
1, 030
23,980
120
0.00053
9.2E-05
8.4E-05
0 . 0003
TABLE C-7. NATIONWIDE BASELINE EQUIPMENT EMISSIONS
FOR BULK TERMINALS AND BREAKOUT STATIONS
Baseline
Emissions
Bulk Terminal3
Equipment
Emissions
(Mg/yr)
Breakout Station13
Equipment
Emissions
(Mg/yr)
HAP
VOC
HAP
VOC
Existing
facilities
100
1,310
110
1,500
Newc
facilities
40.
510
10
150
Total
140
1, 820
120
1, 650
a Facility operates 24 hours per day, 340 days per year.
b Facility operates 24 hours per day, 365 days per year.
c Approximately 24 percent of bulk terminal facilities are
new and approximately 9 percent of breakout station
facilities are new.
C-10
-------
Compilation of Air Pollutant Emission
Factors, Fourth Edition (AP-42). U.S.
Environmental Protection Agency, Research
Triangle Park, NC. Section 9.1.3, Fugitive
Emissions and Controls at Petroleum
Refineries. September 1985.
Gasoline Distribution Industry (Stage I) -
Background Information for Proposed Standards
(NESHAP) . E PA-453/R-94-002a. U.S.
Environmental Protection Agency, Office of
Air Quality Planning and Standards. Research
Triangle Park, NC. January 1994.
Development of Fugitive Emission Factors and
Emission Profiles for Petroleum Marketing
Terminals, API Report Number 4588. March
1993. Appendix A of API's comments.
Marketing Terminal and Pipeline Breakout
Station Data. Prepared by Hal Taback Co.,
Phillips Ranch, CA. April 5, 1994. Appendix
B of API's comments.
C-ll
-------
APPENDIX D.
REVISIONS TO THE
NATIONWIDE EMISSION REDUCTION AND COST ESTIMATES
The purpose of this appendix is to discuss the methodology
used and to present the results of the calculation of the
nationwide emission reductions and costs of the final regulatory
action. Section D.l discusses the emission reductions and
Section D.2 discusses the compliance cost estimates that have
been revised since proposal.
As discussed in Section 1.1, several regulatory requirements
have been changed since proposal and this has affected both the
HAP emission reductions and the cost impacts of this rulemaking.
Table D-l summarizes the revised nationwide emission reductions
and compliance costs resulting from the final rule. Tables D-2
and D-3 summarize the revised baseline emissions resulting from
the new assumptions (e.g., revisions to the equipment leak and
cargo tank leakage emission factors).
D.l NATIONWIDE EMISSION REDUCTIONS
As discussed in Section 10.2, EPA's estimate at proposal of
a nationwide average gasoline Reid vapor pressure (RVP) of 11.4
psia has been reduced based on recent EPA analyses. The new
value of 10.4 psia has the effect of lowering the emission
factors used to calculate baseline emissions. Since these
emission factors are lower, the number of facilities estimated to
satisfy the major source definition (10 tpy of a single HAP or 25
tpy of a combination of HAP's) is now lower than at proposal.
Table D-4 presents the estimated number of bulk terminals and
D-l
-------
pipeline breakout stations that will be affected sources under
TABLE D-l. SUMMARY OF NATIONWIDE EMISSION REDUCTIONS
AND COST IMPACTS OF THE FINAL RULE
NEW FACILITIES
Promulgated
Action
HAP
voc
Cap.
Ann.
HAP
VOC
Red.
Red.
Cost
Cost
C/E
C/E
(Mg/yr
)
(Mg/yr)
($1000)
($1000
)
($/Mg)
($/Mg)
Breakout Stations
Storage Tanks (NSPS Kb)
0
0
0
0
Equipment Leaks
Monthly Visual
UD
UD
0
<1
0
0
Quarterly Visual
UD
UD
0
<1
0
0
Recordkeeping & Reporting
NA
NA
0
4
NA
NA
Bulk Terminals
Loading Racks (@ 10 mg/1)
55
950
7,700
1,800
32,700
1, 900
Storage Tanks (NSPS Kb)
0
0
0
0
T.T. Leakage (annual
vapor tightness @ 1
10
150
0
0
inch)
Equipment Leaks
UD
UD
0
11
0
0
Monthly Visual
UD
UD
0
3
0
0
Quarterly Visual
NA
NA
0
672
NA
NA
Recordkeeping & Reporting
TOTALS*
65
1,100
$7,700
$2,490
$38,300
$2,260
w/o Recordkeeping and
65
1,100
$7,700
$1,818
$28,000
$1,650
Reporting
EXISTING FACILITIES
Promulgated Action
HAP
VOC
Cap.
Ann.
HAP
VOC
Inc .
Red.
Red.
Cost
Cost
C/E
C/E
HAP
(Mg/yr)
(Mg/yr)
($1000)
($1000)
($/Mg)
($/Mg
C/E
)
($/Mg)
Breakout Stations
Storage Tanks
Floor
430
5,700
11,720
(2, 249)
(5,190)
(390)
+ EFRT Fittings
446
5, 910
11,986
(2,244)
(5,000)
(380)
(340)
+ IFRT Fittings
449
5, 950
12,106
(2,258)
(5,030)
(370)
4,800
Equipment Leaks
Monthly Visual
UD
UD
0
4
0
0
Quarterly Visual
UD
UD
0
<1
0
0
Record & Reporting
NA
NA
0
37
NA
NA
Bulk Terminals
Loading Racks
710
12,500
55,200
8, 150
11,500
2,700
(@ 10 mg/1)
Storage Tanks
Floor
600
10,800
33,700
3,466
5,780
320
+ EFRT Fittings
650
11,715
34,606
3, 378
5,200
290
(1,750
+ IFRT Fittings
658
11,850
35,020
3, 425
5,210
290
)
T.T. Leakage
6, 000
(annual vapor
tightness @ 1
240
4,500
7, 600
2,000
10,500
540
inch)
Equipment Leaks
UD
UD
0
28
0
0
Monthly Visual
UD
UD
0
7
0
0
Quarterly Visual
NA
NA
0
1, 660
NA
NA
Record & Reporting
D-2
-------
TOTALS**
w/o Record &
Reporting
2,046 34,625
2,046 34,625
$109,39 $13,013
2
$11,316
$109,39
2
$6,360 375
$5,530 330
TOTAL NEW & EXISTING
2,111 35,725
$117,09 $15,503
$7,340 $430
w/o Record and
2,111 35,725
$13,134
$6,200 $370
Report
$117,09
2
* Totals include monthly visual leak inspections.
** Totals assume that storage tanks must meet NSPS subpart Kb rim seal requirement and
EFRT's adding rim seals must also add controlled fittings. Also assumes monthly
visual leak inspections.
UD = undetermined.
D-3
-------
TABLE D-2. BASELINE EMISSIONS FROM
PIPELINE BREAKOUT STATIONS
Facilities
Storage Tank
Emissions (Mg/yr)
Equipment Leak
Emissions
(Mg/yr)
Total Emissions
(Mg/yr)
HAP
VOC
HAP
VOC
HAP
VOC
Existing
6, 320
83,460
110
1,500
6, 430
84,960
New
50
630
10
150
60
780
TOTAL
6, 370
84,090
120
1, 650
6, 490
85,740
TABLE D-3. BASELINE EMISSIONS FROM BULK TERMINALS
Loading Rack
Tank Truck Leakage
Equipment Leak
Storage Tank
Total
Emissions
Emissions
Emissions
Emissions
Emissions
Facilities
(Mg/yr)
(Mg/yr)
(Mg/yr)
(Mg/yr)
(Mg/yr)
HAP
VOC
HAP
VOC
HAP
VOC
HAP
VOC
HAP
VOC
Existing
3,210
59,450
1,180
21,270
100
1,310
4, 390
70,100
8,880
152,130
New
250
4, 070
90
1, 610
40
510
540
9, 050
920
15,240
TOTAL
3, 460
63,520
1, 270
22,880
140
1, 820
4, 930
79,150
9, 800
167,370
D-4
-------
TABLE D-4. NUMBER OF AFFECTED SOURCE BULK
TERMINALS AND PIPELINE BREAKOUT STATIONS
Facility
Type
1
Model
2
Plant
3
4
Total
Bulk
Terminals
0
15
166
61
242
Breakout
Stations
0
20
—
—
20
the final rule for each model plant size. There may be
additional terminals and pipeline facilities affected by this
rule as a result of being collocated on plant sites with other
HAP-emitting sources. However, data are not available to
distinguish these facilities from those included in these
estimates. Since both groups are likely to be composed of large
sources, the difference is not considered significant.
Table D-5 shows the potential total HAP emissions occurring
at pipeline breakout stations and at gasoline bulk terminals that
have uncontrolled loading racks. The HAP emissions were
calculated assuming HAP contents of 4.8 percent (representing the
arithmetic average HAP content of normal gasoline), 11.0 percent
(representing the maximum HAP content of normal gasoline), and
16.0 percent (representing the average HAP content of
reformulated/oxygenated gasoline). Since Table D-5 estimates the
total potential HAP emissions occurring at each of the model
plant facilities, potential major sources occur where the total
HAP emissions exceed 25 tons per year (tpy). As can be seen in
the table, no breakout stations are classified as major sources
under the 25 tpy criteria. As can also be seen from the table,
either the larger bulk terminals or those terminals that handle
reformulated or oxygenated gasoline may be considered major
sources.
D-5
-------
Table D-6 shows the potential maximum single HAP emissions
occurring at both pipeline breakout stations and bulk gasoline
terminals assuming a HAP content of 11.0 percent (represents the
MTBE content of reformulated/oxygenated gasoline). Since Table
D-6
-------
TABLE D-5. MODEL PLANT POTENTIAL TOTAL HAP EMISSIONS
POTENTIAL EMISSIONS FROM MODEL PLANTS
(tons/year)
Model Plant
HAP/VOC%
1
Model Plant
HAP/VOC%
2
Model Plant
HAP/VOC%
3
Model Plant 4
HAP/VOC%
4 . 8
11. 0
16.0
4 . 8
11. 0
16.0
4 . 8
11. 0
16.0
4 . 8
11.0 16.0
BREAKOUT STATIONS
Storage Vessels
Fugitive Emissions
Total Emissions
3 . 4
0.2
3 . 6
7 . 8
0 . 4
8.2
11. 4
0 . 6
12 . 0
4.3
0.5
4 . 8
9.8
1.2
11. 0
14.3
1.7
16.0
BULK TERMINALS
Truck Loading Racks
Storage Vessels
Fugitive Emissions
4 .1
1. 9
0 .1
9.5
4.3
0.2
13 . 9
6.1
0.3
10 . 4
2.5
0 .1
23 . 8
5.7
0.2
34.7
8.3
0.3
20 . 8
3 .1
0 .1
47 . 6
7 .1
0.2
69.3
10.3
0.3
41.
5
95.3 138.9
8.5 12.4
0.3 0.5
Total Emissions
6.1
14 . 0
20.3
13 . 0
29.7
43.3
24 . 0
49.9
79.9
3.7
104. 151.8
1
0 .1
45 .
3
TABLE D-6. MODEL PLANT MAXIMUM INDIVIDUAL HAP EMISSIONS
POTENTIAL EMISSIONS FROM MODEL PLANTS
USING REFORMULATED AND OXYGENATED GASOLINE WITH MTBE
ASSUMING HAP/VOC% OF 11. 9a
(tons/year)
Model Plant
1
MTBE
Emissions
Model Plant
2
MTBE
Emissions
Model Plant
3
MTBE
Emissions
Model Plant
4
MTBE
Emissions
BREAKOUT STATIONS
Storage Vessels
Fugitive Emissions
Total Emissions
BULK TERMINALS
8.4 10.6
0 . 4 1.3
8.8 11.9
D-7
-------
Truck Loading Racks 10.3 25.7 51.5 103.1
Storage Vessels 4.7 6.2 7.7 9.2
Fugitive Emissions 0 ¦ 2 0¦2 0¦2 0¦3
Total Emissions 15.2 32.1 59.4 112.6
MTBE is 74 percent of total HAP emissions for this category.
D-8
-------
D-6 estimates the maximum single HAP emissions occurring at each
of the model plant facilities, potential major sources occur
where the total HAP emissions exceed 10 tpy. As can be seen from
the table, model plant 2 pipeline breakout stations would be
classified as major source facilities and all bulk gasoline
terminals model plant facilities would be classified as major
sources. Based on this analysis, either large facilities or
facilities which handle gasoline reformulated/oxygenated with
MTBE will classify as major sources. Such facilities are
expected to occur in attainment areas and in ozone and carbon
monoxide nonattainment areas.
D.l.l Fugitive Leaks
As presented and discussed in Appendix C, the equipment leak
emission factors and equipment count/populations have been
revised since proposal. National emission estimates for
equipment leaks presented in Tables D-2 and D-3 are taken from
Table C-5 in Appendix C. The EPA has determined that the factors
supplied by industry represent a controlled situation because the
data on which the factors are based are comprised predominantly
of facilities that utilize a leak detection and repair (LDAR)
program or a visual inspection program. The EPA does not have
any data regarding current uncontrolled emission factors and so
cannot determine the emission reductions generated by the monthly
visual inspection program. As seen in Table D-l, emission
reductions resulting from the visual inspection program are
undetermined (UD) for both new and existing facilities.
D.1.2 Storage Vessels
As discussed in Sections 10.1 and 10.2, the storage tank
emission factors have been recalculated using the latest
estimation techniques described in Chapter 12 of AP-42 (IV-A-2)
and the estimated nationwide RVP of gasoline has been reduced
from 11.4 psia to 10.4 psia. The revised storage tank emission
factors are summarized in Tables D-7 and D-8. At proposal, the
emission reductions from existing storage vessels were calculated
including only the emissions reduced due to rim seal retrofits.
D-9
-------
However, the revised emission reductions include the reductions
achieved by installing controlled fittings on all external
TABLE D-7. EMISSION FACTORS FOR PIPELINE BREAKOUT STATION
STORAGE VESSELSa,b
Type of Emission
VOC Factor
NonWinter Winter
Units
Fixed-Roof Uncontrolled
Breathing losses
Working losses
Internal Floating Roof
Vapor-mounted rim seal losses
Liquid-mounted rim seal
losses
Vapor primary and secondary
seal
Uncontrolled fitting losses0
Controlled fitting losses'1
Deck seam losses
Working losses
External Floating Roof
Standing Storage losses
Primary seal0
Secondary seal1
Primary + secondary +
fittingsg
Working losses
33.5
433.3
2 . 05
0 . 92
0.76
47.2
564
2 . 87
1.29
1. 07
2.36 3.31
1.61 2.25
2.08 2.9
7.33 x 10~8
14 . 0
6.27
4 . 63
19. 61
8.79
6.38
4 . 61 x 10"
Mg VOC/yr/tank
Mg VOC/yr/tank
Mg VOC/yr/tank
Mg VOC/yr/tank
Mg VOC/yr/tank
Mg VOC/yr/tank
Mg VOC/yr/tank
Mg VOC/yr/tank
Mg VOC/bbl
throughput
Mg VOC/yr/tank
Mg VOC/yr/tank
Mg VOC/yr/tank
Mg VOC/bbl
throughput
a Emission factors calculated with equations from Chapter 12
of AP-42 (using the TANKS program) using a nonwinter RVP of
9.3 psi, a winter RVP of 12.8 psi, and a temperature of
60°F.
b Assumes storage vessels at pipeline breakout stations have a
capacity of 8,000 m3 (50,000 bbl), a diameter of 30 meters
(100 feet), and a height of 12 meters (40 feet).
c Calculated assuming the "typical" level of control in the
"TANKS" program.
d Calculated assuming the "controlled" level of control in the
"TANKS" program.
D-10
-------
Assumes that the EFRT is equipped with a primary metallic
shoe seal and uncontrolled fittings.
Assumes that the EFRT is equipped with a shoe-mounted
secondary seal and uncontrolled fittings.
Assumes that the EFRT is equipped with a primary metallic
shoe seal, a shoe-mounted secondary seal, and controlled
fittings.
TABLE D-8. EMISSION FACTORS FOR
BULK TERMINAL STORAGE VESSELS3
Type of Emission
VOC Factor
Nonwinter Winter
Units
Fixed-Roof Uncontrolled111
Breathing losses
Working losses
Internal Floating Roofb
Vapor-mounted rim seal losses
Liquid-mounted seal losses
Vapor primary and secondary
Uncontrolled fitting losses'1
Controlled fitting losses0
Deck seam losses
Working losses
External Floating Roofc
Standing Storage losses
Primary seal1
Secondary sealg
Primary + Secondary +
Working losses
8.55
32.3
1. 02
0.46
0.38
1.01
0 . 69
0.52
7 .33 x lO-
ll. 4
5.38
3.49
4 . 61 x 10-
12 . 0
42 .1
1.44
0 . 64
0.54
1.42
0 . 97
0.73
15 . 98
7.54
4 .88
Mg VOC/yr/tank
Mg VOC/yr/tank
Mg VOC/yr/tank
Mg VOC/yr/tank
Mg VOC/yr/tank
Mg VOC/yr/tank
Mg VOC/yr/tank
Mg VOC/yr/tank
Mg VOC/bbl
Mg VOC/yr/tank
Mg VOC/yr/tank
Mg VOC/yr/tank
Mg VOC/bbl
Emission factors calculated with equations from Chapter 4.3
of AP-42 (TANKS program) using a nonwinter RVP of 9.3 psia,
a winter RVP of 12.8 psia, and a temperature of 60°F.
Assumes fixed-roof and IFRT's at bulk terminals have a
capacity of 2,680 m3 (16,750 bbl), a diameter of 15.2 meters
(50 feet), and a height of 14.6 meters (48 feet).
Assumes EFRT's at bulk terminals have a capacity of 5,7 60 m3
(36,000 bbl), a diameter of 24.4 meters (78 feet), and a
height of 12.5 meters (40 feet).
Calculated assuming the "controlled" level of control in the
"TANKS" program.
D-ll
-------
0 Assumes that the EFRT is equipped with a primary metallic
shoe seal and uncontrolled fittings.
f Assumes that the EFRT is equipped with a primary metallic
shoe seal and uncontrolled fittings.
g Assumes that the EFRT is equipped with a shoe-mounted
secondary seal and uncontrolled fittings.
h Assumes that the EFRT is equipped with a primary metallic
shoe seal, a shoe-mounted sec. seal and uncontrolled
fittings.
floating roof tanks that are taken out of service to perform the
rim seal retrofits. As seen in Table D-l, the emission
reductions for existing storage tanks located at breakout
stations are estimated at 446 Mg of HAP per year (430 Mg reduced
by adding rim seals and 16 Mg reduced by adding controlled
fittings to EFRT's). Similarly, emission reductions for existing
storage tanks located at bulk terminals are estimated at 650 Mg
of HAP per year (600 Mg reduced by adding rim seals and 50 Mg
reduced by adding controlled fittings to EFRT's). As noted at
proposal and shown in Table D-l, new gasoline storage tanks will
be subject to NSPS subpart Kb requirements; therefore, the final
rule will not achieve any emission reductions from new tanks.
D.1.3 Cargo Tank Leakage
The EPA's estimates of nationwide baseline emissions and
emission reductions related to cargo tank leakage at new bulk
terminals have been revised significantly since proposal. First,
the proposed vacuum assist requirement has been deleted from the
rule and replaced with a multi-level cargo tank certification and
testing program similar to the one in California (see Sections
5.3 and 7.2). The estimated leakage emission factor for the
loading of cargo tanks that are certified at the 1-inch decay
limit is 8 mg/liter. Secondly, the controlled tank truck leakage
factor has been revised from 10 percent volume leakage loss to
1.3 percent for tank trucks operating in areas where an annual
vapor tightness testing program (at 3 inches allowable drop in
pressure) is implemented. The new leakage emission factor for
D-12
-------
cargo tanks in the 3-inch testing program is 13 mg/1, as
calculated in Section A. 2 of this document. A computational
error was discovered after proposal which led to an
underestimation of the HAP emission reductions from existing
cargo tank leakage. Correcting the error and considering the
additional benefits of the revised certification and testing
requirements, the nationwide HAP emission reduction for cargo
tank leakage at existing terminals is 240 Mg of HAP per year.
D.1.4 Loading Rack Emissions
Due to a computational error discovered after proposal, the
estimate of HAP emission reductions from loading racks has
increased from 670 to 710 Mg per year.
D.2 NATIONWIDE COSTS OF THE REGULATORY ACTION
D.2.1 Equipment Leaks
Industry indicated that the costs associated with performing
a visual inspection program to identify leaking equipment would
be minimal since most facilities were already implementing such a
program (IV-E-2). Using the data received from industry (IV-D-
22, Appendix B), EPA estimated that approximately 81 percent of
the industry is already implementing some type of visual
inspection program. Therefore, it was assumed that 19 percent of
the industry did not perform any type of visual inspection
program to identify equipment leaks. The EPA assumed that the
equipment at a bulk terminal can be visually inspected in
approximately 2 hours and the equipment at a breakout station can
be visually inspected in 3 hours. Assuming an average labor rate
of $35 per hour, the nationwide annual cost to implement a
monthly visual program at gasoline bulk terminals and pipeline
breakout stations is calculated as follows:
Bulk Terminals
0.19 x 242 major sources x 2 hrs/monitoring x 12 monitorings/year x $35/hr =
$38,60 0/yr.
D-13
-------
Breakout Stations
0.19 x 20 major sources x 3 hrs/monitoring x 12 monitorings/year x $35/hr =
$4,800/yr.
As shown above, the nationwide annual cost to implement a monthly
visual inspection program is relatively insignificant. The EPA
assumed that industry would not incur any capital costs to
initiate such a program.
D.2.2 Storage Vessels
As discussed in Appendix B, storage vessel degassing and
cleaning costs and the costs of controlled fittings were
reevaluated, which increased the overall compliance cost for
storage vessels. Tables B-21 and B-22 in Appendix B summarized
the model tank and nationwide costs to bring all storage vessels
at affected facilities to the NSPS subpart Kb rim seal
requirements. Tables B-23 and B-24 presented costs for all such
tanks to achieve compliance with the subpart Kb rim seal and
fitting control requirements. Table D-l combines this cost
information in presenting the compliance costs for controlling
all storage vessels to the level of the final rule. Existing
storage vessels already having the required rim seals are not
required to install controlled fittings under the final rule, so
these impacts were not evaluated. Also, since new storage
vessels at new and existing facilities are already required to
satisfy the full NSPS subpart Kb requirements, impacts on new
storage vessels were not calculated.
D.2.3 Cargo Tanks
As discussed in Section A.3 of Appendix A, EPA cannot
discern a cost difference in implementing the new annual cargo
tank certification and testing requirements at 1 inch pressure
decay versus the proposal 3-inch pressure decay test. As a
result, no additional cost impact was assumed to occur for owners
or operators to meet the final cargo tank testing requirements
discussed in Sections 5.3 and 7.2.
D.2.4 Loading Racks
D-14
-------
The model plant costs for complying with the 10 mg/liter
emission standard have not changed since proposal. However, the
nationwide cost impacts have been reduced because 33 fewer bulk
terminal sources are expected to be impacted due to the decrease
in the RVP of gasoline, which causes fewer terminals to be
considered major HAP sources. The number of pipeline breakout
stations affected by this rulemaking (20 facilities) has not
changed.
D.2.5 Reporting and Recordkeeping
The EPA's nationwide reporting and recordkeeping cost
estimates were reevaluated after proposal. The revised
nationwide annual cost for bulk terminal respondents to meet the
reporting and recordkeeping requirements is estimated at
$2,331,700, representing 66,620 hours of annual reporting and
recordkeeping. The revised nationwide annual cost for pipeline
breakout station respondents to meet the reporting and
recordkeeping requirements is estimated at $40,910, representing
1,169 hours of annual reporting and recordkeeping. The overall
cost for affected gasoline distribution facilities is 67,790
hours and $2,372,600 per year.
D-15
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-453/R-94-0 02b
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Gasoline Distribution Industry (Stage I) -
Background Information for Promulgated Standards
5. REPORT DATE
November 1994
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Office of Air Quality Planning and Standards
US Environmental Protection Agency
Research Triangle Park, North Carolina 27711
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-D1-0116
12. SPONSORING AGENCY NAME AND ADDRESS
Office of Air and Radiation
US Environmental Protection Agency
Washington, DC 20460
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA/200/04
15. SUPPLEMENTARY NOTES
16. ABSTRACT
National emission standards for hazardous air pollutants (NESHAP) are
promulgated for the gasoline distribution industry under authority of
Section 112(d) of the Clean Air Act as amended in 1990. This background
information document provides technical information and analyses used in
the development of the final NESHAP and Agency responses to public
comments on the proposed rule. The alternatives analyzed are to limit
emissions of hazardous air pollutants (HAPs) from existing and new Stage
I gasoline distribution facilities. Gasoline vapor emissions contain
about ten of the listed HAPs. Stage I sources include bulk gasoline
terminals and plants, pipeline facilities, and underground storage tanks
at service stations. Emissions of HAP's from these facilities occur
during gasoline tank truck and railcar loading, gasoline storage, and
from vapor leaks from tank trucks, pumps, valves, flanges and other
equipment in gasoline service.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
b. IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Air Pollution
Volatile Organic
Compounds Hazardous Air
Pollutants - Gasoline
Bulk Terminals Bulk
Plants Pipelines Service
Stations
Air Pollution Control
13 b
18. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS {TUsReporQ
Unclassified
21. NO. OF PAGES
203
20. SECURITY CLASS (TlllsPase)
Unclassified
22. PRICE
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION IS OBSOLETE
------- |