v>EPA
United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
EPA 450/3-80-0346
May 1983
Air
Benzene Emissions
from Benzene
Storage Tanks -
Background
Information
for Promulgated
Standards
Final
EIS
Preliminary Draft
-------
NOTICE
This document has not been formally released by EPA and should not now be construed to represent Agency policy.
It is being circulated for comment on its technical accuracy and policy implications.
Benzene Emissions from Benzene
Storage Tanks -
Background Information for
Promulgated Standards
Preliminary Draft
Emissions Standards and Engineering Division
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air, Noise, and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
May 1983
-------
TABLE OF CONTENTS
Section Page
1 SUMMARY 1-1
1.1 Summary of Changes Since Proposal 1-1
1.2 Summary of Health, Environmental, Energy and Economic
Impacts of the Promulgated Action 1-5
2 SUMMARY OF PUBLIC COMMENTS 2-1
2.1 Selection of the Level of the Standard 2-1
2.2 Applicability 2-66
2.3 Risk Analysis and Health Impacts 2-76
2.4 Inspection, Reporting and Repair Requirements 2-90
2.5 General Issues 2-99
APPENDICES
C C-l
iii
-------
LIST OF TABLES
Table Page
1-1 Summary of NESHAP for Benzene Storage Vessels 1-3
2-1 List of Commenters on the Proposed Standards of
Performance for Metallic Mineral Processing Plants .... 2-2
2-2 Comparison of Emissions as Calculated from the EPA
Series and the 2519/2517 Series 2-11
2-3 Internal Floating Roof Tank Emissions by Source 2-14
2-4 Emissions from a Typical Benzene Storage Vessel 2-15
2-5 Estimated Installed Capital Cost of a Bolted Internal
Floating Roof for New Construction 2-19
2-6 Estimated Installed Cost of a Welded Contact Internal
Floating Roof with Secondary Seals for New Construction. . 2-21
2-7 Comparison of Degassing Capital Cost at Proposal and
, Current Estimated Cost 2-22
2-8 Additional Costs for Retrofit Considerations . 2-23
2-9 Lifetimes of Control Equipment 2-25
2-10 Cost Annualizing Assumptions 2-25
2-11 Capital and Annualized Costs of Control Equipment for
Equipment for Typical Benzene Tank . 2-26
2-12 Emissions and Absolute Cost Effectiveness of Retrofitting
an Existing Fixed Roof Tank with an Internal Floating
Roof 2-27
2-13 Incremental Cost Effectiveness of Seal Conversion with a
Fixed Roof Tank (New or Existing) as the Baseline 2-29
2-14 Absolute Cost Effectiveness of Seal Conversions with an
Existing Internal Floating Roof Tank as a Baseline .... 2-30
2-15 Incremental Cost Effectiveness of Controlling Deck Seam
Emission with a Fixed Roof Tank (New or Existing) as
Baseline 2-31
2-16 Cost Effectiveness of Controlling Fitting Emissions with
an Existing Internal Floating Roof Tank as the Baseline. . 2-31
2-17 Absolute Cost Effectiveness of Seal Conversions on
External Floating Roof Tanks 2-34
2-18 Comparison of Conventive and Permeability Losses from
Internal Floating Roof Seal Systems in the Model Tank. . . 2-36
IV
-------
LIST OF TABLES (Continued)
Table Page
2-19 Model Tank Emissions (Mg/yr) from a Fixed Roof Tank and a
Typical Internal Floating Roof Tank 2-37
2-20 Regulatory Alternatives and Incremental Cost Effectiveness
between Regulation Alternatives for Existing Benzene
Storage Vessels 2-45
2-21 Regulatory Alternatives and Incremental Cost Effectiveness
between Regulatory Alternatives for New Benzene Storage
Vessels 2-46
2-22 Emissions from Existing Model Plants 2-47
2-23 Emissions from New Model Plants 2-48
2-24 1983 Nationwide Emissions from Existing Benzene Storage
Tanks 2-49
2-25 Nationwide 1988 (fifth-year) Emissions on New Benzene
Storage Tanks 2-50
2-26 Capital Costs for Existing Model Plants 2-51
2-27 Total Annualized Cost (Without Product Recovery Credits)
for Existing Model Plants 2-52
2-28 Total Annualized Cost (With Product Recovery Credits)
for Existing Model Plants 2-53
2-29 Capital Costs for New Model Plants 2-54
2-30 Annualized Cost (Without Product Recovery Credits) for
New Model Plants 2-55
2-31 Annualized Cost (With Product Recovery Credits) for New
Model Plants 2-56
2-32 Incremental Cost Effectiveness for Existing Model Plants . 2-57
2-33 Incremental Cost Effectiveness for New Model Plants. . . . 2-58
2-34 Nationwide Capital Costs for Existing Model Plants .... 2-59
2-35 Nationwide Annualized Cost With Product Recovery Credits
for Existing Model Plants 2-60
2-36 Nationwide 1988 (fifth-year) Capital Costs for New Model
Plants 2-61
2-37 Nationwide 1988 (fifth-year) Annualized Costs (With
Product Recovery Credits) for New Model Plants 2-62
2-38 Vessels Containing Mixtures That May be More Than
10 Percent Benzene 2-71
2-39 Nationwide Cutoff .Impacts Analysis 2-74
-------
LIST OF TABLES (Concluded)
Page
Plants and Locations for Benzene Storage Tanks C-5
Model Inputs for Each Type of Model Plant C-ll
Estimated Maximum Concentration and Dosage for Benzene
Storage Tanks C-17
C-4 Estimated Nationwide Health Impacts for Benzene Storage
Tanks C-23
VI
-------
1. SUMMARY
On December 19, 1980, the U.S. Environmental Protection Agency
(EPA) proposed National Emissions Standards for Hazardous Air Pollutants
(NESHAP) for benzene storage vessels under the authority of Section 112
of the Clean Air Act. The proposed standards were published in the
Federal Register (45 FR 83952) with a request for public comment. A
public hearing was held on June 9, 1981. Six individuals representing
three organizations made presentations. A total of 22 comments from
industry, two trade associations, and an environmental group were
submitted during the comment period. Their comments and EPA's responses
are summarized in this document. The summary of comments and responses
serves as the basis for the revisions that have been made to the proposed
standards.
1.1 SUMMARY OF CHANGES SINCE PROPOSAL
In response to the public comments, and as a result of new emissions
test data, fundamental changes have been made in the proposed standards.
These changes are divided into the following three major categories:
1. Level of the Standard (controls);
2. Inspections; and
3. Applicability of Controls.
Changes in the selected control technologies have resulted in
changes to the inspection procedures, and the capacity at which vessels
must be equipped with controls. These changes are discussed in the
sections that follow.
1.1.1 Level of the Standard (controls)
The promulgated standard requires use of certain kinds of equipment
on each type of benzene storage vessel. The alternate control options
-------
differ from those considered at proposal, and are considered in
Section 2.1 of this document. Table 1-1 lists the requirements of the
promulgated standard.
The promulgated standard differs from the proposed standard in that
it does not require contact internal floating roofs or secondary seals
on internal floating roofs, but it does require control of roof fittings;
and for new tanks constructed after the promulgation date, control of
emissions from column wells. The promulgated standard also differs from
the proposed standard in that it allows external floating roofs which
meet certain specifications. The changes in requirements are based on
extensive new test data gathered by the American Petroleum Institute (API)
in close association with EPA.
The net effect of the promulgated standard is that most existing
benzene storage vessels will require no, or very few, additional controls
to meet the standard. This is due to compliance with one of two new
source performance standards that apply to storage vessels greater than
151.416 cubic-meters (40,000,gallons) and constructed since June 11,
1973, or due to similar State requirements for vessels constructed prior
to that date. The primary effect of the promulgated standard is that it
requires the remaining tanks to be controlled. For example, most fixed
roof benzene storage vessels already have internal floating roofs in
compliance with existing regulations; but a small proportion do not
because they are not affected by those regulations, and the promulgated
benzene standard requires that those tanks be upgraded to include the
internal roof. The promulgated standard also requires that when the
internal floating roof is added to an existing fixed roof tank, that a
liquid-mounted rather than a vapor-mounted seal be used with the roof
and that fittings on the roof be gasketed. These controls are in addition
to those required by the new source performance standard, but the recent
API studies have shown them to be effective in reducing emissions, and
they are relatively inexpensive to install when the internal floating
roof 1s being added to a tank. Existing fixed roof tanks that already
have internal floating roofs are not required to have their vapor-mounted
seals replaced with liquid-mounted seals, although they are required to
1-2
-------
Table 1-1. SUMMARY OF NESHAP FOR BENZENE STORAGE VESSELS
Tank size and time of
construction
Requirements
1. Fixed roof Internal floating roof tank
a. > 38 ms, commenced construction
prior to December 19, I960, and
had no Internal floating roof as
of Oecenber, 19, 1980.
Internal floating roof with liquid-mounted primary teal and
gasketed roof fittings
or
Internal floating roof with liquid-mounted primary seal and a
continuous secondary seal.
b. > 38 n3, commenced construction
prior to December 19, 1980, and
had an Internal floating roof as
of Oecenber 19, 1980.
Internal.floating roof with any type of seal and gasketed roof
fittings1
or
Internal floating roof with liquid-Mounted primary seal and
continuous secondary seal.
c. > 38 n3 and comenced construction Internal floating roof with liquid-mounted primary seal and
after December 19, 1980, and before gasketed roof fittings
date of promulgation, or
Internal floating roof with liquid-mounted primary seal and
continuous secondary seal.
d. > 38 m3 and comenced construction
on or after date of promulgation.
Internal floating roof with liquid-mounted primary seal,
gasketed roof fittings and pipe column with flexible fabric
sleeve.
2. External floating roof tank
2 3
a. > 38 m3 and commenced construction Liquid-mounted primary seal and a continuous secondary seal. '
prior to December 19, 1980.
b. £ 38 m3 and commenced construction
on or after December 19, 1980, and
before date of promulgation.
Liquid-mounted primary seal and a continuous secondary seal."
c. > 38 m3 and commenced construction
on or after date of promulgation.
Liquid-mounted primary seal and a continuous secondary seal.
Basketing of roof fittings Is required the first time tank 1s degassed.
'if external floating roof Is already equipped with liquid-mounted primary seal, the secondary seal is
required to be added the first time the tank is degassed.
Mechanical-shoe primary seal is also allowed, provided that the tank 1s also equipped with a continuous
secondary teal.
1-3
-------
constructed with.the same controls as are required for existing tanks
with no internal roof, i.e., with an internal floating roof, a liquid-
mounted primary seal and controlled roof fittings, and are also required
to have pipe columns equipped with a flexible fabric sleeve seal. New
storage vessels for which construction commenced before the promulgation
date will not be required to install pipe columns with flexible fabric
sleeves. Note that if tanks were equipped with a secondary seal in
accordance with the proposed standard, gasketed fittings will not be
required; the two control techniques achieve the same emission reduction.
Owners of existing and new external floating roof tanks will have
to install liquid-mounted primary seals (or mechanical-shoe seals) and
continuous secondary seals meeting certain gap requirements. Existing
external floating roof tanks already equipped with a liquid-mounted
primary seal, however, are not required to add the secondary seal until
the first degassing of the tank.
1.1.2 Inspections
The inspection requirements have also been changed since proposal
as the result of the new API datla and the changes in the equipment
requirements. Each internal floating roof vessel is to be inspected
from inside prior to the filling of the vessel (if it is emptied to
install control equipment) and at least once every 10 years. An internal
floating roof having defects, or a seal having holes or tears, is to be
repaired before the storage vessel is filled with benzene. The promulgated
standard also requires that the internal floating roof and its seal be
inspected visually through hatches on the fixed roof at least once
annually. Any major defects such as roof sinking or primary seal detachment
as viewed through the roof hatches must be repaired within 30 days or
the storage vessel emptied.
The promulgated standard also requires that, for external floating
roof tanks, the primary seal and secondary seal gaps be measured initially
and at least once every 5 years for the primary seal, and at least once
annually for the secondary seal.
1.1.3 Applicability of Controls
The promulgated standard differs from the proposed standard in
the capacity at which vessels must be controlled. The promulgated
1-4
-------
standard requires that each vessel with a capacity greater than or
equal to 38 cubic meters (m ) or 10,000 gallons be equipped with
controls. The proposed standards required that each vessel with a
capacity greater than or equal to 4 m (1,000 gallons) be equipped
with controls.
1.2 SUMMARY OF HEALTH, ENVIRONMENTAL, ENERGY AND ECONOMIC IMPACTS
OF THE PROMULGATED ACTION
The promulgated standard affects approximately 600 existing benzene
storage vessels. Emissions from existing benzene storage vessels will
be reduced from 620 megagrams per year (Mg/yr) to about 510 Mg/yr. As a
result of this emission reduction the estimated lifetime risk to the
most exposed population will be reduced from a range of 1.38 x 10 to
9.48 x 10"5 to a range of 1.11 x 10"5 to 7.62 x 10"5. The estimated
incidence of excess leukemia cases resulting from exposure to benzene
emissions from existing benzene storage vessels will be reduced from a
range of .017 to 0.113 leukemia cases per year to a range of .014 to
0.094 leukemia cases per year for the people living within 20 kilometers
of existing benzene storage vessels. Due to the assumptions that are
made in calculating these risk and incidence numbers, there is considerable
uncertainty associated with them beyond the range presented. The
uncertainties associated with these risk numbers are explained in the
Section 2.3.2.
The promulgated standard will also reduce the emissions from new
benzene storage vessels; By 1988 there will be an estimated 170 new
benzene storage vessels in use. Implementation of the promulgated
standard will reduce the 1988 emissions from new storage vessels from
about 150 megagrams (Mg) to about 90 Mg. The effect of the standard on
reduction of emissions and risks from new storage tanks will depend on
the location of the tanks, the number and distribution of people living
within the vicinity of the new sources, and the level of control which
would be employed in the absence of the standard. These factors cannot
be quantified at this time. The 1988 emissions from both new and
currently existing benzene storage vessels will be reduced from 770 Mg
to 600 Mg with the implementation of the promulgated standard for
benzene storage vessels.
1-5
-------
The promulgated standard will have no adverse Impacts on other
aspects of the environment. In addition, there will be no adverse
energy impacts associated with the promulgated standard.
The capital investment required for an existing model plant (with
multiple tanks) to comply with the promulgated standard will range from
about $12,000 to zero. The net annualized cost, taking into account the
value of benzene saved, will range from about $2,000 to about zero. The
total national capital and annual 1 zed costs for existing facilities will
be about $660,000 and $110,000, respectively. The price of benzene may
increase by as much as 0.02 percent as a result of the promulgated
standard. No plants are projected to close as a result of implementing
the promulgated standard.
The capital cost for a new model plant (with multiple tanks) to
comply with the promulgated standard will range from about $250 to about
$4,600. The net annualized cost will range from about $210 to a credit.
The total national capital and annualized costs for new facilities
constructed through 1988 to comply with the promulgated standard will be
approximately $67,000 and a credit, respectively.
The health, environmental, energy and economic impacts are discussed
in greater detail in Chapter 2.
1-6
-------
2. SUMMARY OF PUBLIC COMMENTS
A list of commenters, their affiliations, and the EPA docket entry
number assigned to each comment are shown in Table 2-1. Twenty-two
letters commenting on the proposed standard and the Background Information
Document (BID) for the proposed standard were received. Significant
comments have been combined into the following five categories:
2.1 Selection of the Level of the Standard
2.2 Applicability
2.3 Risk Analysis and Health Impacts
2.4 Inspection, Reporting and Repair Requirements
2.5 General Issues
!
Comments, issues, and their responses are discussed in the following
sections of this chapter. Changes to the regulations are summarized in
Subsection 1.2 of Chapter 1.
2.1 SELECTION OF THE LEVELS OF THE STANDARD
2.1.1 Background
The proposed standards which were based on Best Available Technology
(hereafter referred to as Best Demonstrated Technology or BDT), required
the use of a fixed roof in combination with an internal floating roof.
The proposed standards also required that the internal floating roof be
in contact with the liquid surface and be equipped with a liquid-mounted
primary seal and a continuous secondary seal.
Many commenters suggested that the EPA delay the development of the
final standards until the effectiveness of BDT equipment relative to
other equipment types could be reevaluated using data from the American
Petroleum Institute (API) 2519 Task Group testing program. The results
of this testing program have been received and evaluated by the EPA.
Comments were also received on other aspects of BDT, such as control
equipment costs. Because these comments are interrelated with each
other and with the comments on the API testing program, the Agency
2-1
-------
Table 2-1. LIST OF COMMENTERS ON THE PROPOSED STANDARDS
OF PERFORMANCE FOR METALLIC MINERAL PROCESSING PLANTS
Docket entry number3 Commenter/affiliation
IV-D-1 Edward W. Warren
Kirkland and Ellis
1776 K Street, Northwest
Washington, D.C. 20006
Counsel for the American Petroleum
Institute
IV-D-2 R. W. Bogan
GATX Terminals Corporation
120 South Riverside Plaza
Chicago, Illinois 60606
IV-D-3 Edward W. Warren
Kirkland and Ellis
1776 K Street, Northwest
Washington, D.C. 20006
Counsel for the American Petroleum
Institute
IV-D-4 John T. Barr
Air Products and Chemicals, Inc.
Box 538
Allentown, Pennsylvania 18105
IV-D-5 John Heinz
Unites States Senate
Committee on Energy and Natural
Resources
Washington, D.C. 20510
With attachment from Sun Petroleum
Production Company
IV-D-6 J. C. Pullen
Celanese Fibers Company
Box 32414
Charlotte, North Carolina 28232
IV-D-7 Herman A. Fritscher
Cities Service Company
Box 300
Tulsa, Oklahoma 74102
IV-D-8 E. M. Vancura
Union Oil Company of California
Box 7600
Los Angeles, California 90051
(continued)
2-2
-------
Table 2-1. Continued
Docket entry number3 Commenter/affillation
IV-D-9 D. P. Martin
Gulf Oil Company
Post Office Box 2001
Houston, Texas 77001
IV-D-10 Geraldine V. Cox
Chemical Manufacturers Association
2501 M Street, Northwest
Washington, D.C. 20037
IV-D-lOa Lance S. Granger
Chemical Manufacturers Association
2501 M Street, Northwest
Washington, D.C. 20037
Attachment to docket entry IV-D-10
IV-D-11 Paul J. Sienknecht
The Dow Chemical Company
Midland, Michigan 48640
IV-D-12 , Alfred G. Hoerrner
Merck Chemical Manufacturing Division
Post Office Box 2000
Rahway, New Jersey 07065
IV-D-13 Richard K. Meyers
Texaco, Incorporated
Post Office Box 509
Beacon, New York 12308
IV-D-14 .- F. M. Parker
Chevron U.S.A., Incorporated
575 Market Street
San Francisco, California 94105
IV-D-15 R. J. Samel son
PPG Industries, Incorporated
One Gateway Center
Pittsburgh, Pennsylvania 15222
IV-D-16 Daniel B. Rathbun
American Petroleum Institute
2101 L Street, Northwest
Washington, D.C. 20037
IV-D-17 John J. Moon
Phillips Petroleum Company
Bartlesville, Oklahoma 74004
(continued)
2-3
-------
Table 2-1. Concluded
Docket entry number Commenter/affillation
IV-D-18 Dennis L. Gehlhausen
Eli Lilly and Company
307 East McCarty Street
Indianapolis, Indiana 46285
IV-D-19 David D. Doniger
Natural Resources Defense Council,
Incorporated
1725 I Street, Northwest
Suite 600
Washington, D.C. 20006
IV-D-20 Wells Eddleman
General Energy Consulting
Route 1, Box 183
Durham, North Carolina 27705
IV-D-21 C. D. Mallach
Monsanto Company
800 North Lindbergh Boulevard
St. Louis, Missouri 63166
IV-H-1 T. L. Hurst
Kerr-McGee Corporation
Kerr-McGee Center
Oklahoma City, Oklahoma 73125
IV-F-1 National Air Pollution Control
Techniques Advisory Committee
Transcript of Meeting for National
Emission Standards for Hazardous
Air Pollutants from Benzene
Storage Vessels
U.S. Environmental Protection Agency
Office of Administration
Research Triangle Park, NC 27711
aThese designations represent docket entry numbers for Docket
No. A-80-14. These documents are available for public inspection at:
U.S. Environmental Protection Agency, Central Docket Section, West
Tower Lobby, Gallery 1, Waterside Mall, 401 M Street, Washington, D.C.
20460.
2-4
-------
decided to summarize all comments that have been received concerning BDT
and then follow this with one response in which BDT is reevaluated.
2.1.2 Selection of the Level of the Standard
2.1.2.1 Emission Data Base. Seven commenters suggested that the
emissions data base used in selection of the BDT at proposal was erroneous
and that the Agency should await the completion of a new API testing
program before selecting BDT prior to promulgation (IV-D-1, IV-D-2,
IV-D-3, IV-D-8, IV-D-10, IV-D-10a, IV-D-14).
2.1.2.2 BDT Equipment. Several commenters stated that other types
of equipment should be allowed as BDT. Two commenters stated that the
final standard should allow external floating roof tanks (IV-D-8, IV-D-6).
Six commenters stated that the final standard should not distinguish
between contact and noncontact roofs, and that both roof types be allowed
with several different types of seal systems (IV-D-21, IV-D-17, IV-D-8,
IV-D-6, IV-D-9, IV-D-2).
One commenter stated that because of the small price impacts,
thermal oxidation should have been selected as BDT (IV-D-19).
2.1.2.3 Cost of Controls. Four commenters (IV-D-9, IV-D-14,
IV-D-15, IV-D-20) supplied very specific criticisms of cost estimates.
Two (IV-D-9, IV-D-14) discussed the vessel degassing costs. Commenter
IV-D-9 stated that degassing is slower and more hazardous for vessels
with secondary seals, and, consequently, the economic analyses must
account for the more rigorous and expensive degassing procedures for
vessels with secondary seals. No estimates, however, were provided.
The other commenter (IV^D-W) stated that the EPA's estimate of $6,150
for degassing/cleaning/inspecting a storage vessel is much lower than
the industry estimate of $30,000/vessel. This commenter also noted that
the cost of vapor recovery or inert gas systems to prevent air/benzene
mixtures between the floating and fixed roofs, will add $50,000-
100,000/vessel to the capital costs. In addition, he noted that the
capital costs estimate did not include the costs of extra vessels for
use during vessel maintenance and their retrofit of the required technology.
No estimates were supplied of this cost. Another commenter (IV-D-15)
stated that not all benzene storage facilities are located in areas
2-5
-------
where volatile organic compounds (VOC) are a problem. These facilities,
therefore, may not have existing controls, but the economic impacts
calculation assumes that VOC controls are already fitted and need only
be upgraded. It has been assumed that only these upgrading costs will
have an economic impact. In addition, vessels with the "wrong" type of
roof or seals may require 100% replacement. These costs have not been
included. Another commenter (IV-D-20) suggested that vapor-mounted
seals could be repaired more easily and quickly than assumed. He
additionally was concerned that the most recent price of benzene be used
in the cost estimations.
2.1.3 Selection of BDT
The Agency has received and evaluated the results of the API testing
program. Because of this data and the above comments, the Agency has
reevaluated several controls options for BDT and added some new ones.
2.1.3.1 Selection of Emissions Data Base. There are four potential
emission data bases from which emission calculations could be developed.
These are:
1. A test series done by Chicago Bridge and Iron (CBI) for an
internal floating roof vendor. This series measured emissions
from a bolted, noncontact internal floating roof equipped with
wiper-type, vapor-mounted primary seals; and a welded contact
internal floating roof. The welded roof was equipped with a
liquid-mounted primary seal and in some instances a secondary
seal. All the tests were performed in a propane/octane binary
»«
mixture. This data base is hereafter referred to as the
Vendor report or series.
2. A large number of tests done on various external floating
roofs with propane/octane as the stored liquid. These tests
were also performed by CBI. The primary emphasis of this work
was to categorize emissions from various types of primary and
secondary seals and was used to update API bulletin 2517,
which is used in estimating emissions from external floating
roof tanks. It was also used in the 1981 revision of EPA
publication AP-42. This work is referred to as the 2517 series
or report.
2-6
-------
3. A test series done by CBI for the EPA using benzene as the
test liquid. This program tested a bolted noncontact internal
floating roof with vapor-mounted, shingled, primary and secondary
seals; a welded contact type internal floating roof equipped
with a liquid-mounted primary seal and in some instances, a
secondary seal; and an external floating roof equipped with a
mechanical shoe primary seal and in some instances a secondary
seal. This data base will be referred to as the EPA report or
series.
4. A test program done by CBI for API on emissions from internal
floating roofs. This program tested three roof types (non-
contact, bolted contact, welded contact), three primary seal
types (vapor-mounted wiper; vapor-mounted, foam-filled resilient
seal; liquid-mounted seal) with and without secondary seals,
in three different liquids (propane/octane, hexane, and octane).
Additional work was done on emissions from the components of
an internal floating roof. This consisted of deck fitting
emission tests, laboratory evaporation tests, laboratory
permeability tests, and bench permeability tests. This data
base will be referred to as the 2519 report or series.
Each of the above test series was performed in the CBI 20 foot
diameter pilot test tank. The first three were completed prior to the
development of the proposed standard, but the 2519 series was completed
after the date of proposal.
In evaluating the emissions data for internal and external floating
roofs prior to proposal it was noted that emissions from the EPA series
were significantly higher than those measured in either the Vendor or
the 2517 series when tests on similar equipment were normalized to the
same vapor pressure and molecular-weight. The Agency believed that the
difference in emissions resulted from a difference in liquids, namely
multicomponent liquids such as propane/octane and single component
liquids such as benzene. The reason for this difference was believed to
be due to that fact that in a mixed product (e.g. the propane/octane
mixture) the emission rate depends upon the ability of the component
2-7
-------
with the highest partial pressure (e.g. propane) to migrate through the
liquid to the liquid surface and replenish the component that is lost
through evaporation at the liquid surface. In a single component product
(e.g. benzene), however, the liquid surface does not tend to become
depleted of light ends at the liquid surface during the evaporation
process. Thus, a mixed product of the same vapor pressure as a single
component product was expected to have a lower evaporation rate due to
this phenomenon. Therefore, in selecting BDT at proposal only the EPA
test series was used because it was believed that the previous
propane/octane test work was not representative of single component
emissions such as a vessel storing benzene.
Industry representatives commented that the higher emissions were a
result of the test procedures and did not necessarily result from a
difference in evaporative properties. The 2519 test series shows that
when normalized to a common vapor pressure and molecular weight, there
is no significant emission difference between hexane, octane, and the
propane/octane binary mixture. Based on these results the Agency now
•
agrees that there is no evidence of evaporative difference between
single and multicomponent liquids stored in floating roof tanks, and
this is not a reason for the higher emissions measured in the EPA series.
One cause of at least a portion of the higher emissions from the
EPA series is that during certain internal roof tests done for the EPA
series, the roof fittings had openings that would not normally exist in
the field and were sealed with polyurethane film which, as previously
discussed, is permeable to benzene. This would lead to artificially
higher emissions being measured during the EPA series than would normally
be expected from a typical field tank. During tests done on the same
roof for the 2519 series the roof fitting openings that would not normally
exist in the field were sealed by welded metal seals not permeable to
benzene. This procedure would yield measured emissions more representative
of emissions from a typical field located tank.
Also during the EPA series, the bolted noncontact internal floating
roof was tested with shingled (i.e., noncontiguous) primary and secondary
seals, which are not as effective in reducing emissions as the more
typical continuous wiper or'foam-filled resillient seals. This again
would lead to higher emissions being measured during the EPA series.
2-8
-------
Either wiper or foam-filled resillient seals were tested during the
other test series.
Other physical mechanisms that could explain the higher emissions
in the EPA series were sought. The permeability results in the 2519 series
were examined to ascertain if permeation of the seal system could be
responsible for the higher benzene emission. As detailed in Appendix A,
the permeation rate of benzene through a typical seal fabric (polyurethane)
was significantly higher than the rates at which hexane or propane/octane
permeate. Because there are no direct measurements of benzene permeation
rates through an entire seal system, theoretical models were developed.
The most reasonable model of permeation through a liquid mounted seal
predicts emissions of 0.0102 pound moles per day in the test tank (see
Appendix A). While permeation and equipment differences may explain
some of the emission differences between the benzene test work and the
other test work, it is not sufficient to account for the total difference.
Another explanation of the higher emissions from floating roof
tanks shown by the EPA series has to do with the test procedures used.
The vendor series and the 2517 series used the same test procedure as
the EPA series, that is, a floating roof and seal system is installed in
the pilot tank, and air is blown over the floating roof. The air is
collected and analyzed for hydrocarbon content. In the vendor, 2517,
and EPA series test work, the temperature of the air being blown across
the roof was uncontrolled. During periods when the air is cold (such as
during the winter), the benzene vapor being emitted will condense during
periods when actual tests are not being run. When a test is then begun,
the benzene vapor that condensed will be measured during the test when
it was actually emitted before the test run began. In the case of
benzene (EPA series) this could lead to artificially higher results. In
the case of the propane/octane mixture, the uncontrolled air temperature
is not as important to the results since this mixture is less likely to
condense in the cold air. In the 2519 series, the air temperature was
controlled, and no emissions differences were observed between the three
tested liquids.
As just explained, because the 2519 series test conditions were
more controlled than during the EPA series and because of the equipment
tested (continuous versus shingled seals), this test series resulted in
2-9
-------
more representative emission measurements. The 2519 series was also
structured to make it possible to better ascertain the relative
contributions to emissions of the various emission points (e.g., seals,
roof seams and roof fittings). Also, the data obtained from the 2519
series are similar to the vendor series which tested similar roofs and
seals and used a propane/octane binary mixture.
The higher permeability of benzene, the difference in equipment
tested and the differences in test procedures explain most of, but not
all, the higher emissions from floating roof tanks measured during the
EPA series. Currently, however, there is no explanation beyond what has
already been discussed as to why benzene emissions would be any higher
than the hexane and octane emissions measured during the tests done
during the 2519 series.
Since there is no reason (other than possibly permeability which is
addressed later) for benzene emissions (normalized for vapor pressure
and molecular weight) to be higher than hexane and octane emissions
during the 2519 tests, and since the 2519 series was conducted with more
refined procedures and more thoroughly evaluated the emission sources
and control techniques for each source, the Agency has decided to use
the data from this series to evaluate the emission reduction potential
for various control technologies applied to fixed roof and internal
floating roof tanks. For similar reasons, the Agency has selected the
2517 series as the data base for evaluating controls for external floating
roof vessels. The 2517 tests are more extensive in terms of equipment
tested and, for the same reasons as the 2519 series, have measured
emissions more representative of emissions from a typical external
floating roof.
Table 2-2 compares emissions from selected floating roof tank types
as calculated using data from the EPA series and as calculated using
data from the 2519 and 2517 series. It should be noted that because of
differences in tested equipment and test procedures, the emissions are
not strictly comparable. However, it can be seen that the sharp difference
in emissions (particularly in terms of mass rather than percentage)
between the equipment configurations vanished in the 2519 and 2517 test
series. Making the decision that the 2517 and 2519 test series are
superior to the EPA test series meant that it was then necessary to
2-10
-------
Table 2-2. COMPARISON OF EMISSIONS AS CALCULATED FROM THE EPA
SERIES AND THE 2519/2517 SERIES
Tank type/equipment
Test series
EPA 2517/2519
emissions emissions
(Mg/yr)
(Mg/yr)
I. Internal Floating Roof
A. Bolted deck with vapor-mounted
primary and secondary seals
B. Welded deck with liquid-mounted
primary seal
C. Welded deck with liquid-mounted
primary and secondary seals
II. External Floating Roof with
Mechanical Shoe Primary Seal
A. Primary seal only
B. With rim-mounted secondary
3.56J
1.15
0.67
6.99
2.63
0.42'
0.38
0.34
1.11
0.087
Both primary and secondary seals were shingle design.
'All seals were continuous.
-------
reexamine control options for benzene storage vessels and reconsider
which options should be selected as BDT, and to reevaluate an alternative
more stringent than BDT to determine if the risks remaining after the
application of BDT are unreasonable.
2.1.3.2 Emissions. In selecting BDT, the Agency examined the
emission points from possible baseline tank types and possible control
technologies. As explained in the Volume I BID there are four types of
tanks that could be used to store benzene. These are:
1. Fixed roof tanks;
2. Noncontact internal floating roof tanks;
3. Contact internal floating roof tanks; and
4. External floating roof tanks.
Based on the 2519 test series, there is no inherent difference between
contact and noncontact deck types. Analysis of the data concluded that
deck seams emit at the same rate if they are in contact with the liquid
or saturated vapor. Contact decks may be welded (no deck seams) or
bolted (e.g., mechanically connected panels or sections that have seams).
A bolted contact deck would 'have deck seam emissions at the same rate
per foot of deck seam as a noncontact deck. Because of this, for the
purpose of developing regulatory alternatives the two types of internal
floating roofs were merged into the general classification of internal
floating roof. This procedure reduced the basic starting cases to three
tank types: fixed roof, internal floating roof (bolted deck assumed),
and external floating roof.
The mechanisms of 'fixed roof tank and external floating roof tank
emissions have been fully discussed in the Volume I BID. Although the
external floating roof tank emission factors have changed based on the
2517 series, the emission mechanisms are still the same. Fixed roof
tank emissions have not changed since proposal.
The 2519 series allows for a more detailed breakdown of internal
floating roof tank emissions into:
1. Standing storage losses, consisting of:
a. Rim seal emissions;
b. Fitting losses; and
c. Deck seam emissions
2. Working losses.
2-12
-------
Table 2-3 presents losses from a model benzene storage vessel by point
of loss, and Table 2-4 compares emissions from various selected tank
configurations. The model tank used in these and all subsequent
calculations in this section has a volume of 606 m3 (160,000 gallons), a
diameter of 9.1 m (30 feet), and undergoes 50 turnovers per year.
Internal floating roofs are typically bolted decks equipped with
vapor-mounted seals and Case A fittings (defined below). In the model
tank, emissions from the vapor-mounted seal are about 35 percent of
total emissions. Emissions from the vapor-mounted seal could be reduced
through the use of a liquid-mounted primary seal, a secondary seal, or
both. A liquid-mounted seal reduces emissions from the vapor-mounted
primary seal by about 55 percent. The addition of a secondary seal to
the vapor-mounted primary seal would reduce emissions by about 63 percent.
The addition of a secondary seal to a liquid-mounted primary seal reduces
emissions by about 46 percent over the liquid-mounted primary seal
alone. Converting a vapor-mounted primary seal system to a liquid-mounted
primary seal with a secondary seal reduces emissions from the seal area
by about 76 percent over the vapor-mounted primary seal alone.
The next major source of internal floating roof tank emissions are
losses from fittings. Fittings in general are ancillary equipment such
as hatches or column wells that penetrate the deck. Such penetrations
will emit benzene. Typical fittings are: (1) access hatch, with
ungasketed, unbolted cover; (2) automatic gauge float well, with ungasketed,
unbolted cover; (3) built-up column wells, with ungasketed sliding
cover; (4) ladder well, with ungasketed sliding cover; (5) adjustable
roof legs; (6) sample well with slit fabric (10% open area); (7) 1-inch
diameter stub drains; and (8) vacuum breaker with, gasketed weighted
mechanical actuation. This equipment is referred to as "Case A". In
the model tank, emissions from Case A fittings account for about 48 percent
of total emissions. Emissions from Case A type fittings could be reduced
through the use of gaskets, bolting covers, and constructing pipe columns
with flexible fabric sleeve seals on the column well in place of built-up
columns equipped with ungasketed sliding covers in the column wells.
This configuration of fittings is referred to as Case B, and is the
level of control that could be obtained in new benzene storage vessels
equipped with internal floating roofs. Specifically, "Case B" is defined
2-13
-------
Table 2-3. INTERNAL FLOATING ROOF TANK EMISSIONS BY SOURCE'
ro
t-*
f>
Seal
Type
Vapor- mounted
Liquid-mounted
losses
Emission
(Mg/yr)
0.19
0.085
Fitting
Case
.. A2
B3
losses
Emission
(Mg/yr)
0.26
0.16
Deck
Roof type
Bolted
Welded
losses
Emission
(Mg/yr)
0.06
0.0
Working losses
Emission
(Mg/yr)
0.03
Vapor-mounted
with secondary 0.071
Liquid-mounted
with secondary 0.046
0.19
Tank Parameters: Volume = 160,000 gallons
Diameter = 30 ft. diameter
Turnovers = 50 turnovers per year
y
'Case A assumes: (1) access hatch, with ungasketed, unbolted cover; (2) automatic gauge float well,
with ungasketed, unbolted cover; (3) built-up column wells, with ungasketed sliding cover; (4) ladder
well, with ungasketed sliding cover; (5) adjustable roof legs; (6) sample well with slit fabric (10%
open area); (7) 1-inch diameter stub drains; and (8) vacuum breaker with, gasketed weighted mechanical
actuation.
Case B assumes: (1) access hatch, with gasketed, bolted cover; (2) automatic gauge float well, with
gasketed, bolted cover; (3) pipe column with flexible fabric sleeve seal; (4) ladder well, with
gasketed sliding cover; (5) adjustable roof legs; (6) sample well with slit fabric (10% open area);
(7) 1-inch diameter stub drains; and (8) vacuum breaker, with gasketed weighted mechanical actuation.
Case C is identical to Case B except that built-up columns with gasketed sliding covers are assumed
instead of pipe columns.
-------
Table 2-4. EMISSIONS FROM A TYPICAL BENZENE STORAGE VESSEL
Tank type/equipment Emissions (Mg/yr)
I. Fixed Roof 9.2
II. Internal Floating Roof
A. Bolted deck, vapor-mounted 0.54
seal, Case A fittings
B. Bolted deck, liquid-mounted 0.44
Case A fittings
C. Bolted deck, liquid-mounted 0.34
seal, Case B fittings
D. Bolted deck, liquid-mounted 0.30
primary with secondary, Case B
fittings
E. Welded deck, liquid-mounted 0.24
primary with secondary, Case B
fittings
III. External Floating Roof
A. Mechanical Shoe Primary Seal 1.11
1. Primary seal only
2. With rim-mounted secondary 0.087
B. Vapor-mounted Primary Seal 6.9
1. Primary seal only
2. With rim-mounted secondary 2.31
C. Liquid-mounted Primary Seal 0.36
1. Primary seal only
2. With rim-mounted secondary 0.080
2-15
-------
as: (1) access hatch, with gasketed, bolted cover; (2) automatic gauge
float well, with gasketed, bolted cover; (3) pipe column wells with
flexible fabric sleeve seal; (4) ladder well, with gasketed sliding
cover; (5) adjustable roof legs; (6) sample well with slit fabric (10%
open area); (7) 1-inch diameter stub drains; and (8) vacuum breaker,
with gasketed weighted mechanical actuation. Case B fittings would
reduce emissions from the typical fittings Case A by about 38 percent.
Existing internal floating roof benzene tanks typically use built-up
columns to support the fixed roof. Such vessels could not be equipped
with pipe columns without replacing the columns. In most instances,
this would be equivalent to requiring the construction of a new tank.
Therefore, an intermediate control strategy was sought. Emissions from
built-up column wells could be controlled by gasketing the sliding
cover. This strategy is referred to as "Case C" and represents the
level of fitting control available in existing internal floating roof
tanks.
Specifically, Case C is identical to Case B except that built-up
columns with gasketed sliding covers are assumed instead of pipe columns.
Case C fittings would provide about a 27 percent emission reduction over
Case A fittings. Because most existing fixed roof benzene tanks are
equipped with built-up columns, Case C represents the level of control
of fitting emissions generally available for existing tanks.
The next source of internal floating roof tank emissions are deck
seams. Decks that are constructed of sections bolted together have
emissions along the seam. As discussed previously, seams emit at the
same rate if they are in contact with the liquid surface (contact deck)
or contain a saturated vapor on one side (noncontact deck). Because of
this fact, the distinction between contact and noncontact decks has been
dropped, and these decks are now referred to as "bolted" for emission
purposes. Emissions from the deck seams in the model tank are about
11 percent of total emissions.
Deck seam emissions could be controlled by installing decks that
have no seams. Such decks are generally made out of steel sections
welded together. These decks are generally in contact with the liquid
surface, and are referred to as "welded" for emission purposes.
2-16
-------
The last emission type in an internal floating roof tank is the
working loss. These losses are fully discussed in the Volume I BID, and
account for about 6 percent of typical losses. No controls for working
losses are available.
As Table 2-4 shows, fixed roof tank emissions could be reduced by
about 94 percent by the installation of internal floating roofs. Emissions
could be further reduced through the use of additional controls on
seals, fittings, and deck seams.
Emissions from external floating roofs could be reduced by the
addition of a secondary seal over the primary seal. In the case of the
mechanical shoe primary seal, this would reduce emissions by about
92 percent. Emissions from vapor-mounted primary seals could be reduced
by replacing these seals with mechanical shoe seals or liquid-mounted
primary seals alone or further reduced with secondary seals.
In restructuring the control options based on the 2519 and 2517
test series, EPA applied increasingly effective control technologies to
each type of storage vessel. Many of the options were the same as the
ones considered at proposal; however, the emission reductions for these
options based on the 2519 and 2517 test series are quite different than
those calculated at proposal. For example, the 2519 series showed that
the control effectiveness of an internal floating roof (of any type) in
a fixed roof tank is much more effective in reducing emissions than was
believed at proposal, based on the EPA test series. On the other hand,
the 2519 test series showed that a secondary seal in an internal floating
roof tank is much less effective in reducing emissions than believed at
proposal based on the EPA test series. This is to be expected because
the internal floating roof is more effective than believed previously
and as a result, there are less residual emissions to be controlled by
the secondary seal. One option which was considered at proposal, but
which was rejected in the final rulemaking is the use of a contact, as
opposed to a noncontact, internal floating roof; the 2519 test series
showed these two roof types are equally effective in reducing emissions.
The 2519 test series also showed that liquid-mounted seals are more
effective in reducing emissions than vapor-mounted seals. This type of
seal can be used with both contact and noncontact roofs, and was considered
as a control option in the final rulemaking. Furthermore, the 2519
2-17
-------
series showed that control of roof fittings, column wells, and roof deck
seams does reduce emissions, so these control options were added to the
ones considered at proposal. Using the 2517 and 2519 data in combination
shows that external floating roofs can, when used with effective seals,
reduce emissions as effectively as internal floating roofs. This also
affected structuring of the control options.
In selecting the level of the final standard, both the effectiveness
of control and the cost of control were examined. Emissions and controls
have been discussed above. The cost of control will be discussed below.
2.1.3.3 Cost of Controls. As discussed previously, some commenters
stated that the costs of the equipment selected as BDT at proposal were
underestimated and should be reevaluated. While the commenters1 focus
was on the equipment proposed as BDT, had other equipment been selected
as BDT, the Agency may have received similar comments. Additionally,
the new control options resulting from the new test data were not costed
at the time of proposal. All of these factors led to a decision to
completely update the costing for the purposes of the Volume II BID.
The cost analysis methodology has not changed from the Volume I
BID. However, there are three significant differences in costs. First,
an additional allowance has been made to retrofit existing fixed roof
tanks with internal floating roofs. This additional allowance accounts
for the costs of cutting vents in the tank shell and other smaller
ancillary work. Second, based on information supplied by vendors and
commenters, the equipment lifetime for primary seals and floating roofs
was extended from 10 years to 20 years. Third, the costs of degassing a
storage vessel is now estimated to be significantly higher than at
proposal. This is in part due to the listing of benzene contaminated
materials as a hazardous waste. It is assumed that any water used to
wash the tank and any debris (rust, seal system, etc.) resulting from
the cleaning would have to be disposed of as a solid waste subject to
those regulations authorized by the Resource Conservation and Recovery
Act. This cost was not accounted for at the time of proposal.
Additional equipment costs were obtained from a large number of
vendors. Table 2-5 presents installed, new construction costs for a
noncontact internal floating roof of various configurations. This type
of roof was examined because it is the least costly and most typical
2-18
-------
Table 2-5. ESTIMATED INSTALLED CAPITAL COST OF A BOLTED INTERNAL FLOATING ROOF
FOR NEW CONSTRUCTION (fourth-quarter 1982 dollars)
ro
i
vr>
Tank
diameter,
(m)
5
10
15
20
25
30
Basic
roof
cost3
($)
6,280
11,600
17,000
22,300
27,700
33,000
Cost with
liquid-mounted
primary seal
($)
6,320
11,680
17,120
22,460
27,900
33,250
Cost with
liquid-mounted
primary seal
and Case C
($)
6,320
11,680
17,120
22,460
27,900
33,250
Cost with
liquid-mounted
primary seal ,
controlled deck
fittings and .
secondary seal
($)
7,620
14,300
21,000
27,600
34,400
41,000
Estimated from the equation: Cost ($) = 1,069 D + 939; where, D = tank diameter in meters; with the
correlation coefficient r2 = 0.889. This correlation generates installed cost estimates for an aluminum
noncontact internal floating roof with a vapor mounted wiper type seal.
The additional cost of the liquid-mounted primary seal over an elastomeric wiper seal is estimated to
be $2.60 per linear meter of circumference based on quotes from one vendor.
°Case C deck fittings include a gasketed, bolted cover, access hatch; a gasketed, bolted cover automatic
gauge float well; pipe columns with a flexible fabric sleeve seals; a gasketed sliding cover for the
ladder well; a sample well with a split fabric seal and 10% open area; and a weighted, mechanical
actuation, gasketed, vacuum breaker. Based on vendor estimates, the additional cost of controlled
fittings over the cost of the normally installed deck fittings is negligible.
Cost of secondary seal is estimated to be $85 per linear meter of tank circumference. This is the
average price of 13 seals from 8 different vendors.
-------
type of internal floating roof. Table 2-6 presents installed new
construction costs for a welded, steel, contact internal floating roof.
The regression equations that appear in the table notes are for new
construction and were used in the cost analysis. As noted in Table 2-4,
the cost difference of constructing a liquid-mounted primary seal rather
than a vapor-mounted primary seal is $2.60 per linear meter. The additional
cost of Case B or Case C fittings over Case A is estimated to be negligible.
The cost of adding a secondary seal is estimated to be $85 per linear
meter. The costs for liquid-mounted seals and secondary seals is an
additional cost that must be added to the basic roof cost in Table 2-5.
The costs in Table 2-6 include a liquid-mounted seal. Again, the
cost of Type B fittings over Type A fittings is negligible.
To make modifications needed to control emissions from existing
tanks, the first step is to clean and degass the storage vessel. Additional
costing data were obtained for this procedure. Limited data provided by
two vendors was used to develop the following relationship:
(tank caPacity in cubic nieters)0'5132
and
for tanks <52 m3 a $1,000 minimum cost should be assumed.
Table 2-7 compares these costs with values in the Volume I BID.
To account for retrofit costs (installing an internal floating roof
in an existing fixed roof tank) vendors indicated that the cost of
additional work (i.e. cutting roof vents, etc.) could be estimated as
5 percent of the installed capital cost of new construction. This would
apply to both bolted and welded decks. However, with a welded deck
there would be an additional charge to cut an opening in the tank shell
so that the steel sections may be put in the tank. This charge is
estimated to be $1,300.
To replace an existing bolted internal floating roof with a welded
roof, it would be necessary to remove the existing internal floating
roof. This cost is estimated to be about $4,300.
The retrofit costs to add secondary seals is unchanged from the
$85 per linear meter for new construction. However, the cost to retrofit
an existing vapor-mounted seal system with a liquid-mounted seal system
is $148.50 per linear meter. Table 2-8 lists retrofit costs.
2-20
-------
Table 2-6. ESTIMATED INSTALLED COST OF A WELDED CONTACT INTERNAL
FLOATING ROOF WITH SECONDARY SEALS FOR NEW CONSTRUCTION
(fourth-quarter 1982 dollars)
Tank
diameter cost
(m) ($)
5 15,900
10 30,000
15 44,000
20 58,100
25 72,100
30 86,100
aThe basic cost of the roof and primary seal is estimated from the
equation: cost ($1000) = 1.91 + 2.54D; where D equals the tank
diameter in meters with the correlation coefficient r2 = 0.883. The
additional cost of a secondary seal is estimated based on the factor,
$85 per linear meter of circumference. The secondary seal cost is the
average price of 13 seals from 8 different vendors.
2-21
-------
Table 2-7. COMPARISON OF DEGASSING CAPITAL COST AT PROPOSAL
AND CURRENT ESTIMATED COST
Tank volume
(M3)
603
1,990
2,617
4,156
8,588
Cost at proposal
($)
1,300
2,000
2,500
3,400
6,150
2
Current estimates
($)
3,500
6,450
7,250
9,400
13,700
•4980 dpllars.
21982 dollars.
2-22
-------
Table 2-8. ADDITIONAL COSTS FOR RETROFIT CONSIDERATIONS
(fourth-quarter 1982 dollars)
Item
Cost ($)
1. Degassing
2. Additional work
3. Door sheet opening
4. Removal of existing aluminum
internal floating roof
Cost = 130.8V0'5132; or
$1,000: whichever is greater
where V = tank volume in cubic
meters
5% of installed capital cost
for new roof construction
1,300
4,300
Needed for installation of welded, steel deck only.
2-23
-------
A retrofit example will be discussed to illustrate this procedure.
The cost of retrofitting an existing fixed roof storage vessel with a
noncontact, internal floating roof with liquid-mounted primary seals,
secondary seals and Case C fittings will be discussed. First the cost
of degassing the tank is calculated. This is added to 105 percent of
the basic roof cost calculated from Table 2-5. Added to that are charges
of $2.60 per linear meter for the additional cost of the liquid-mounted
seal and $85 per linear meter for the secondary seal. No cost is assumed
for the Case C fittings. The cost of the liquid-mounted seal is $2.60
per linear meter because the deck is "new". No existing vapor-mounted
seal needs to be removed and no modifications to the deck need to be
made to support the seal system. The $148.50 per linear meter applies
only to cases where an existing vapor-mounted primary seal is replaced
with a liquid-mounted seal.
Capital costs were annualized in the same manner as in the Volume I
BID. Tables 2-9 and 2-10 contain the assumptions made in converting
capital costs to annualized cost. Table 2-11 shows capital and total
annualized costs without product recovery credits for the model tank.
2.1.3.4 Cost Effectiveness. The information on emissions and
costs in the previous section was combined into two kinds of cost
effectiveness values for various controls and base cases. The absolute
(or average) cost effectiveness is defined as the total annualized cost
minus the value of product recovery credits (net annualized cost) of
requiring a particular control option, divided by the total emissions
reduction achieved by going from no control to that control option.
Incremental cost effectiveness is defined as the difference in net
annualized cost (including product recovery credits) between two control
options, divided by the difference in emissions reduction between the
same two options.
Table 2-12 presents the absolute cost effectiveness of retrofitting
an existing fixed roof tank with various type of internal floating
roofs. It should be noted that the cost effectiveness for all retrofits
is under $500 per Mg.
2-24
-------
Table 2-9. LIFETIMES OF CONTROL EQUIPMENT
Device
Tank and floating roof
Secondary seals
Carbon adsorber
Thermal oxidizer
Lifetime (yrs)
20
10
10
10
Capital
recovery
factor
0.11746
0.16275
0.16275
0.16275
Capital recovery factor determined by the equation:
CRF = i(l + 1)"/[(! + i)n - 1];
where i = the annual interest rate and,
n = the equipment lifetime.
Table 2-10. COST ANNUALIZING ASSUMPTIONS
Item
Charge
Tax, insurance, and administration
Maintenance
Inspection
Interest rate
Recovery credits
4% of capital cost
5% of capital cost
1% of capital cost
10%
$360 per Mg
2-25
-------
Table 2-11. CAPITAL AND ANNUALIZED COSTS OF CONTROL EQUIPMENT FOR TYPICAL BENZENE TANKJ
ro
i
ro
1.
2.
3.
4.
5.
Item
Bolted deck
Welded deck
Secondary seal for internal or external
floating roof tank
Liquid-mounted seal
A. Retrofit from a vapor-mounted seal
B. "New"
Retrofit costs
A. Additional work for internal floating
roof tanks
1. bolted deck
2. welded deck
B. Degassing
Capital cost
($)
10,700
25,100
2,440
4,270
75
535
2,555
3,500
Annual i zed cost
($)
2,327
5,458
641
928
16
116
555
761
-------
Table 2-12. EMISSION AND ABSOLUTE COST EFFECTIVENESS OF RETROFITTING
AN EXISTING FIXED ROOF TANK WITH AN INTERNAL FLOATING ROOF
Tank type/equipment
Emissions Cost effectiveness
(Mg/yr) ($/Mg)
I. Fixed Roof
II. Internal Floating Roof
A. Bolted deck, vapor-
mounted seal, Case A
fittings
B. Bolted deck, liquid-
mounted seal, Case A
fittings
C. Bolted deck, liquid-
mounted seal, Case C
fittings
D. Bolted deck, liquid-
mounted primary with
secondary, Case C
fittings
E. Welded deck, liquid-
mounted primary with
secondary, Case C
fittings
9.2
0.54
0.44
0.37
0.33
0.27
9.97
7.61
4.7
75
470
2-27
-------
Table 2-13 presents the incremental cost effectiveness of seal
conversions with the fixed roof tank as the base case. Because these
are incremental values between seal type control options, the cost of
degassing does not appear as part of the cost of the conversion. Because
the tank would have already been degassed to install the internal floating
roof, no degassing costs are ascribed in calculating the incremental
value for seals. Also because the fixed roof tank was the base case,
the liquid-mounted seal was costed at the $2.60 per linear meter rate.
Therefore, Table 2-13 also represents the incremental cost effectiveness
of seal conversions for new construction. The cost effectiveness with
one exception has a high value as opposed to the absolute values in
Table 2-12. Also the emission reduction in the absolute case with the
fixed roof tank baseline is relatively large (=8.8 Mg). This large
emission reduction provides substantial recovery credits that help to
finance the cost of the control equipment, thus reducing the net annualized
cost. However, the incremental values have no large emissions reduction
and the recovery credits are correspondingly small (Table 2-2).
The liquid-mounted seal, however, does generate sufficient recovery
credits to provide a net credit. This is because the annualized cost
(Table 2-11) is small. As seen in Table 2-2 the emissions from the
vapor-mounted primary seal with a secondary seal are slightly less than
those from the liquid-mounted seal, but the large difference in cost
makes the liquid-mounted seal more cost effective.
Table 2-14 presents the absolute cost effectiveness of controlling
seal emissions with an existing internal floating roof tank as a baseline.
In this case, because the internal floating roof exists, the costs of
degassing are included. Also as discussed previously, because there is
an existing vapor-mounted seal which must be removed, the cost of the
liquid-mounted seal is charged at the $145.50 per linear meter value.
This fact makes all of the absolute values in Table 2-14 relatively
large.
As presented in Table 2-15, the incremental cost effectiveness of
controlling deck seam emissions from a bolted deck by the construction
of a welded deck is $51,800 per Mg. Again, because the tank would have
already been degassed to install the internal floating roof, no degassing
2-28
-------
ro
i
ro
10
Table 2-13. INCREMENTAL COST EFFECTIVENESS OF SEAL CONVERSION WITH A
FIXED ROOF TANK (NEW OR EXISTING) AS THE BASELINE
Base case
Vapor-mounted
Liquid-mounted
Vapor-mounted with secondary
($/Mg)
5,030
—
End Case
Liquid-mounted
($/Mg)
credit
-1
Liquid-mounted with secondary
($/Mg)
4,200
16,100
Not applicable.
-------
Table 2-14. ABSOLUTE COST EFFECTIVENESS OF SEAL CONVERSIONS WITH AN
EXISTING INTERNAL FLOATING ROOF TANK AS A BASELINE
End Case
Base case Vapor-mounted with secondary Liquid-mounted Liquid-mounted with secondary
ro
i
CO
o
"($/Mg) ($/Mg)
Vapor-mounted 11,400 15,700
Liquid-mounted — —
($/Mg)
15,800
35,600
-------
Table 2-15. INCREMENTAL COST EFFECTIVENESS OF CONTROLLING
DECK SEAM EMISSION WITH A FIXED ROOF TANK
(NEW OR EXISTING) AS BASELINE
Incremental
cost effectiveness
Base case End case ($/Mg)
Bolted deck
Welded deck
51,800
Table 2-16. COST EFFECTIVENESS OF CONTROLLING FITTING EMISSIONS WITH
AN EXISTING INTERNAL FLOATING ROOF TANK AS THE BASELINE
Incremental
cost effectiveness
Base case End case ($/Mg)
Case A Case C 9,900
2-31
-------
costs are ascribed to the incremental value. While the cost effectiveness
of controlling deck seam emissions from an existing internal floating
roof is not presented, it is substantially higher than the $51,806 per
Mg presented in Table 2-15.
Table 2-16 presents the absolute cost effectiveness of controlling
fitting emissions (Case C fittings) with an existing internal floating
roof tank as the base case (Case A fittings). The only costs ascribed
were those of degassing. No costs were ascribed to the installation of
the Case C fittings. The recovery credit due to the emission reduction
of 0.07 Mg, however, is not large enough to offset the cost of degassing;
therefore, the incremental cost effectiveness for this lowest cost
estimate for existing internal floating roofs is $9,900. Because the
cost of installing Case C fittings in an existing internal floating roof
tank is unlikely to be zero, actual costs would be higher.
On the other hand, if a small additional capital cost, estimated to
be about $200 at a maximum for gaskets and other required equipment, is
assumed for Case B or Case C fittings over Case A fittings, the incremental
cost effectiveness would be about $300 per Mg excluding the cost of
degassing. As discussed before, in the case of new construction or in
the case of a fixed roof tank as baseline, the cost effectiveness of
controlling fittings would be about $0 per Mg because the cost of the
Case B or Case C fittings is assumed to be zero. This is because the
controls are part of the basic design of the internal floating roof, and
degassing costs (if any) are ascribed to the internal floating roof.
It should be noted that in examining the cost effectiveness of
controlling emissions from an existing internal floating roof tank,
degassing cost has been ascribed to each control option. That is, when
seal conversions were considered (Table 2-14) the degassing costs were
included in the calculation of cost effectiveness. When controls for
fittings were examined, the cost of degassing was included in that
calculation. Controlling several emission sources with only one degassing
would spread the degassing costs over several control options, resulting
in an overall cost effectiveness that is lower than any single control
option. Therefore, the simultaneous application of control options at
one degassing was examined. The cost effectiveness of converting an
internal floating roof with vapor-mounted primary seals and Case A
2-32
-------
fittings to an internal floating roof with a vapor-mounted primary seal,
a secondary seal, and Case C fittings is about $7,000 per Mg. In this
calculation, it was assumed that the cost of the fitting conversions is
zero.
The cost effectiveness of adding secondary seals to existing external
floating roofs with mechanical primary shoe seals is $1,010 per Mg.
This includes the cost of degassing and a charge of $85 per linear meter
for the secondary seal. The cost effectiveness of retrofitting a vapor-
mounted primary and a secondary seal system to a liquid-mounted primary
and secondary seal system was estimated to be $680 per Mg. This includes
the cost of degassing, the cost of the liquid-mounted primary seal at
$145.50 per linear meter, and the cost of the secondary seal at $85 per
linear meter. The value of product recovery credits, and the large
emission reduction makes this conversion more cost effective than adding
a secondary seal over a mechanical shoe primary seal.
Table 2-17 presents the cost effectiveness of primary seal transitions
and secondary seals for external floating roof tanks. In reading the
table it is important to note that mechanical shoe seals are the most
common type of primary seal. When equipped with a rim-mounted secondary
seals, these seals are equivalent to a liquid-mounted primary seal with
rim-mounted secondary seals.
Also, there is no reason why new benzene storage vessels must be
built with vapor-mounted primary seals rather than mechanical shoe or
liquid-mounted primary seals.
2.1.3.5 Uncertainty of Emission Estimates. As briefly mentioned
previously, there is an additional source of emissions that has not been
fully considered up to this point. This is the permeability of seal
systems and gaskets to benzene. The 2519 series and the open literature
point to the fact that aromatics such as benzene have higher permeability
rates through polymers than some other types of compounds. Because no
direct measurements of seal permeability are available, the Agency
examined this emission source by developing theoretical models.
These models represented:
1. A foam-filled liquid-mounted seal;
2. A wiper type, vapor-mounted primary seal; and
3. Each of the above with a wiper type secondary seal.
2-33
-------
Table 2-17. ABSOLUTE COST EFFECTIVENESS OF SEAL CONVERSIONS
ON EXTERNAL FLOATING ROOF TANKS
Base case
Liquid-mounted
End case
rim-mounted secondary
Liquid-mounted with secondary
Cost effectiveness
($/Mg)
Retrofit New
Vapor- mounted
Vapor-mounted
Vapor-mounted
Mechanical shoe
Vapor-mounted with secondary
Liquid-mounted
Liquid-mounted with secondary
Mechanical shoe with
credit
credit
680
1,010
credit
credit
credit
270
5,000
2,100
2-34
-------
Each seal consists of two parts:
1. Two layers (top and bottom) of seal fabric; and
2. Open cell foam situated between the fabric layers.
In selecting the fabric layers for modeling, it was discovered that
there was little data on what fabrics are actually in use, and little
data on measured fabric permeability rates. Because the Agency had
permeability measurements on 0.037 inch thick polyurethane-coated nylon
fabric from the 2519 tests, and because this material is currently in
use in field tanks, the Agency decided to use this material as the
fabric in the models.
In modeling the open cell foam it was assumed that the foam presented
no permeability barrier. Transport between the fabric layers was assumed
to be diffusion (it was assumed that the foam did not allow convective
transport). These models done on a 20 foot diameter tank are contained
in Docket Item _. For the purpose of comparability to the model tank
(30 foot diameter) emissions the results have been extroplated to the
model tank.
Table 2-18 compares the convective losses presented in Table 2-3
with the calculated permeability losses. It is seen that permeation may
account for more than 50 percent of seal losses if:
1. The permeation rates are correct; and
2. The models realistically represent actual systems.
Table 2-19 examines how this effects the overall effectiveness of
controls compared to a fixed roof tank. The reduction in overall
effectiveness is less than 3 percent.
However, the Agency examined how permeability emissions may be
controlled. These emissions could be controlled by a seal permeability
specification. Such a specification would limit permeability emissions
to a specified limit per unit area of seal. -However, the variation in
measured values in open literature indicate that such measurements would
be difficult to make reliably. Seal materials must withstand abrasion
and flexing as the floating roof moves. At this point in time, the
Agency is aware of no materials or laminar composites that would have
both the necessary characteristics of material strength and permeation
2-35
-------
Table 2-18. COMPARISON OF CONVECTIVE AND PERMEABILITY LOSSES
FROM INTERNAL FLOATING ROOF SEAL SYSTEMS IN THE MODEL TANK
Seal Type
Emissions (Mg/yr)
Convective
Modeled
Permeation
Possible
Total
Losses
Vapor-mounted
Liquid-mounted
Vapor-mounted
with secondary
Liquid-mounted
with secondary
0.19
0.085
0:071
0.046
0.21
0.20
0.11
0.10
0.40
0.285
0.181
0.146
2^36
-------
Table 2-19. MODEL TANK EMISSIONS (Mg/yr) FROM A FIXED ROOF TANK
AND A TYPICAL INTERNAL FLOATING ROOF TANK
Tank Type
Emission
Percent Control
Fixed roof
Internal floating roof with
bolted deck, Case A fittings,
vapor-mounted primary seal
only, no permeability
Internal floating roof with
bolted deck, Case A fittings,
vapor-mounted primary seal,
permeability
9.2
0.54
94.1
0.75
91.8
2-37
-------
rates lower than the modeled fabric. Such a specification could be made
with additional research on materials.
2.1.3.6 Best Demonstrated Technology. Best demonstrated technology
for new and existing sources is that technology which, in the judgement
of the Administrator, is the most effective level of control considering
the economic, energy, and environmental impacts and any technological
problems associated with the retrofitting of existing sources.
The costs of the storage vessel control techniques are small relative
to the capital and operating costs of the process units where the tanks
are located. As a consequence, none of these control techniques impact
the ability of an owner or operator to raise capital or do they measurably
impact product prices. Therefore, EPA selected BDT based primarily on a
comparison of incremental costs and emission reductions associated with
each alternative control technique. Incremental costs and emission
reductions are calculated by taking the difference between the emissions
and annualized costs of one control option and the next less stringent
control option. The control options considered were arranged in order
of increasing incremental cost effectiveness. In selecting BDT, EPA
selected control techniques that achieve the most emission reduction
with reasonable incremental control costs per megagram of emission
reduction.
• Existing Fixed Roof Tank
The first level of control that can be installed in a fixed roof
tank is an internal floating roof with a vapor-mounted primary seal.
This has a cost effectiveness of about $10 per megagram. Another control
option which can be applied is to use a liquid-mounted rather than a
vapor-mounted primary seal. This results in a credit over the vapor-mounted
primary seal case. The next step is to control fittings on the roof.
The incremental cost effectiveness of this control is approximately
zero. Therefore, the internal floating roof, liquid-mounted primary
seal, and controlled fittings have reasonable costs and thus, were
selected as BDT for existing fixed roof tanks. However, the other
control options, which consist of installation of a secondary seal and
welding of roof deck seams, have incremental cost effectivenesses ranging
2-38
-------
from $16,100/Mg to $51,800/Mg. These costs were judged to be unreasonable
and thus these control techniques were not selected as BDT for fixed
roof tanks. A control technique applicable to a new fixed roof or
internal floating roof tank that was rejected for existing tanks is
control of column wells. Column well controls have to be constructed at
the time the original tank is constructed. Essentially complete
reconstruction would be required to retrofit these.
• Existing Internal Floating Roof Tank With Vapor-mounted Primary Seal
The incremental cost effectiveness of the control techniques applicable
to internal floating roof tanks with vapor-mounted primary seals range
from $10,000/Mg (immediate control of fitting emissions) to $112,000/Mg
(control of deck seam emissions). All of these incremental cost effective-
nesses were judged to be unreasonably high and BDT for these tanks was
determined to be no additional control. It may be noted that the
incremental cost effectiveness of installing liquid-mounted primary
seals and gasketing fittings were judged to be reasonable for fixed roof
tanks, but not for internal floating tanks that already have vapor-mounted
seals. The reason is due to retrofit cost differences. These controls
can be applied with little cost to a fixed roof tank when the tank is
being degassed to install an internal floating roof. When a tank already
has an internal floating roof, it would still have to be degassed to
replace the seal and to gasket fittings. The only emission reduction
that would be achieved for the cost of degassing is that from controlling
fittings and replacing the primary seal. This is a smaller emission
reduction than that associated with installing an internal roof, and
thus a larger cost effectiveness results. When a tank is degassed for
other reasons, the capital cost of controlling fittings is estimated to
be about $200, and the incremental cost-effectiveness of gasketing
fittings at this time is about $300/Mg. This cost is judged to be
reasonable, and therefore, BDT for these tanks consists of gasketing
fittings at the first degassing.
However, the incremental cost effectiveness of controlling seal
emissions is still at least $5,000 per Mg even if the tank is degassed
for other reasons. This was judged to be unreasonable, and control of
2-39
-------
seal emissions was not selected as BDT for existing internal floating
roof tanks.
• Existing Internal Floating Roof Tank with Liquid-mounted Primary Seal
The incremental cost effectiveness of initially installing fitting
or roof control techniques applicable to internal floating roof tanks
that already have liquid-mounted primary seals range from $10,000/Mg to
$112,000/Mg. All these cost effectivenesses were judged to be unreasonably
high. However, as with the internal floating roof tanks with vapor-mounted
primary seals, the incremental cost-effectiveness of installing gaskets
on roof fittings at the first degassing is reasonable and was selected
as BDT for these tanks.
• Existing External Floating Roof Tank with Vapor-mounted Primary Seal
Two control options were considered for these tanks. These are
adding a secondary seal and replacing the vapor-mounted primary seal
with a liquid-mounted seal. The incremental cost effectiveness of both
of these options, which ranged from a credit to a cost of $680/Mg, were
judged to be reasonable and were both selected as components of BDT for
these tanks.
• Existing External Floating Roof Tank with Liquid-mounted Primary Seal
Only one control option is applicable to this tank, the addition of
a secondary seal. The incremental cost effectiveness of adding a secondary
seal to this type of tank ($5,000/Mg) is much higher than for an external
floating roof tank with a vapor-mounted primary seal (credit). In both
cases, degassing of the tank is a major portion of the cost necessary to
add the seal. However, the emissions from a liquid-mounted primary seal
are much lower than from a vapor-mounted seal. This means there are
less emissions available for control by the secondary seal, and a higher
incremental cost effectiveness results. The cost effectiveness of
degassing a tank with a liquid-mounted primary seal and installing a
secondary seal ($5,000/Mg) is judged to be unreasonable. However, the
cost effectiveness of adding the secondary seal when degassing of the
tank occurs for other reasons ($2,100/Mg) is considered reasonable.
Therefore, BDT for these tanks consist of adding a secondary seal when
the tank is degassed for other purposes.
2-40
-------
• New Fixed Roof Tank
Controls can be applied to new tanks with much less cost, because
the controls are part of the basic construction and there are no degassing
or other retrofit costs.
New fixed roof tanks can be built as new internal floating roof
tanks. This level of control is already required for many tanks by the
new source performance standard (NSPS) for petroleum liquid storage
vessels, which requires, as the minimum level of control for fixed roof
tanks greater than 40,000 gallons, the installation of internal floating
roofs with vapor-mounted primary seals. The cost of this control option
is a credit rather than a cost. The second level of control which can
be applied is to use a liquid-mounted primary seal rather than a vapor-
mounted primary seal. This also results in a savings rather than a
cost. The next step is to control fitting losses. The incremental cost
of controlled fittings over uncontrolled fittings is a net credit. All
of these controls have reasonable costs and were selected as BDT for new
fixed roof tanks.
The next option to consider is the addition of a secondary seal
over the liquid-mounted primary seal. The incremental cost effectiveness
of this would be about $16,100 per Mg. Another control option would be
to control deck seam emissions by requiring welded decks. The incremental
cost effectiveness of this would be about $51,000 per Mg. These costs
were judged to be unreasonable, and thus, these control options were not
selected as BDT for new fixed roof tanks.
;•
• New External Floating Roof Tank
New external floating roof tanks with vapor-mounted primary seals
only, could be built as external floating roof tanks equipped with
vapor-mounted primary seals and continuous secondary seals. This is the
least stringent requirement of the NSPS for petroleum liquid storage
vessels, and as such, this level of control is already required for many
tanks. This level of control results in a savings rather than a cost.
The next control option to consider is a liquid-mounted primary seal
with a continuous secondary seal rather than the vapor-mounted primary
seal with a continuous secondary seal. The incremental cost effectiveness
2-41
-------
of requiring this control option over the base case (vapor-mounted
primary seal with a secondary seal) is a credit. A mechanical-shoe
primary seal with a continuous secondary seal has similar costs and
achieves similar emission reductions. There is no control option that
will achieve more emission reduction than these two seal systems.
Therefore, BDT for new external floating roofs is a liquid-mounted
primary seal with a continuous secondary seal or a mechanical-shoe
primary seal with a continuous secondary seal.
In summary, the level of control selected as BDT includes the
following controls for benzene storage vessels:
(1) Existing Fixed Roof Tank - installation of internal floating
roof with a liquid-mounted primary seal and gasketing of
fittings;
(2) Existing Internal Floating Roof Tank with Vapor or Liquid-Mounted
Primary Seal - gasketing of fittings at first degassing;
(3) Existing External Floating Roof Tank with Vapor-mounted Primary
Seal - replacement of vapor-mounted seal with a liquid-mounted
seal and installation of a secondary seal;
(4) Existing External Floating Roof Tank with Liquid-mounted
Primary Seal - addition of secondary seal at first degassing;
(5) New Internal Floating Roof Tank - use of liquid-mounted primary
seal, gasketing of fittings, and control of column wells; and
(6) New External Floating Roof Tank - use of liquid-mounted primary
seal with a continuous secondary seal or a mechanical-shoe
primary seal with a continuous secondary seal.
2.1.3.7 Consideration of Permeability in Selection of BDT. As
previously discussed, seals have emissions resulting from permeation
that may be as significant as convective losses from seals from both
internal and external floating roof tanks. Before these emissions could
be regulated by a materials specification, a long research program would
be needed to develop fundamental information on the permeability rates
and the durability of different seal materials.
In the absence of a long research program, the Agency examined the
use of secondary seals in internal floating roof tanks to control
2-42
-------
permeability emissions from seals. The cost effectiveness of adding a
secondary seal to a new liquid-mounted primary seal, including consideration
of permeability emissions, is $4,250 per Mg in the model tank. The cost
of requiring existing internal floating roof tank to retrofit with
secondary seals is $6,040 per Mg if permeability emissions are considered.
Because BDT for most external floating roof tanks already includes
a secondary seal, the only additional control would be the immediate
addition of a secondary seal on those existing tanks equipped with
liquid-mounted primary seals only. The cost of requiring this control
option is $3,300 per Mg if permeability emissions are considered.
Because of uncertainty in the models and basic information, the
cost effectiveness cited above could be higher or lower. The above
values, which represent the current best estimate, are judged to be
unreasonable and the Agency decided not to require controls for permeability
as part of BDT.
The EPA plans to investigate this issue further, and control of
these emissions may be required at a future time. However, in the
interim period, the Agency has decided not to delay promulgation until
this issue has been resolved. Therefore, no BDT controls are specified
for permeability in the promulgated standards.
2.1.3.8 Environmental and Cost Impacts.
2.1.3.8.1 Regulatory alternatives. BDT was tentatively selected
by examining the incremental cost-effectiveness of adding different
control techniques to different individual model tanks. The next step
was to examine the nationwide environmental and economic impacts of the
tentatively selected BDT to determine if there were any unreasonable
adverse impacts that would affect the selection of BDT. The number and
distribution of the different types of tanks have to be taken into
consideration when calculating nationwide cost, emission reduction, and
economic impacts. To do this, regulatory alternatives were established
that included different control levels on the different types of tanks
combined. Several regulatory alternaties were examined because the
level that would be tentatively selected as BDT based on incremental
cost-effectiveness was unknown at the time. The distribution of tank
2-43
-------
types was based on the model plants examined in the BID Volume I. Each
model plant contains multiple tanks. Tables 2-20 and 2-21 show the
regulatory alternatives for existing and new facilities, respectively.
In both cases, Alternative 0 is simply the baseline and will be
used to illustrate the impacts of no additional regulatory action by the
Agency. Alternatives I and II show the impacts of each portion of BDT;
further control of seal emissions and fitting emissions for new and
existing fixed roof tanks, and further control of seal emissions for
existing external floating roof tanks.
Alternative III represents BDT. The difference between
Alternative 0 and III represents the impacts of BDT. To examine the
impact of controls more stringent than BDT, Alternative IV was
developed. Alternative IV represents the next most stringent control
option.
2.1.3.8.2 Emissions. Emissions for each regulatory alternative
were calculated for each of the 4 model plants developed in the Volume I
BID. Tables 2-22 and 2-23 present emissions from the existing facility
model plants and new facility model plants respectively. This emissions
information was used to develop nationwide impacts for each regulatory
alternative. Tables 2-24 and 2-25 present the nationwide emissions
impacts for existing and new storage vessels, respectively.
2.1.3.8.3 Costs. Capital and annualized costs were calculated for
the model plants. This information is presented in Tables 2-26
through 2-31. The costing and annualizing procedures are those
described previously. Where no additional equipment over baseline was
required, the cost assigned to the regulatory alternative was $0. As
previously discussed, recovery credits were valued at $360 per Mg.
Tables 2-32 and 2-33 present the incremental cost effectiveness of the
regulatory alternatives.
Tables 2-34 through 2-37 present the nationwide capital and
annualized costs (with product recovery), of the regulatory alternatives
for existing and new model plants.
2-44
-------
Table 2-20. REGULATORY ALTERNATIVES AND INCREMENTAL COST EFFECTIVENESS BETWEEN REGULATORY ALTERNATIVES
FOR EXISTING BENZENE STORAGE VESSELS1
Regulatory Alternative
Type of Existing
Benzene Storage
Vessel
FR
II
III
IV
Incremental cost
effectiveness
- 9.97
Incremental cost
effectiveness
credit
Incremental cost
effectiveness
credit
Incremental cost
effectiveness
16,100
en
IFR -2 — 2 -2
blhKVM,A
EFR — 2 1,010 EFRss EFR$S
— 2 7,000 b1FRVH ss C
EFRss EFRss
'Notation is as follows:
FR = fixed roof tank
IFR = Internal floating roof tank
EFR = external floating roof tank
b = bolted deck
LM = liquid-mounted primary seal
VN = vapor-mounted primary seal
ss = secondary seal
ps = primary seal (mechanical shoe assumed)
A = Case A fittings
C = Case C fittings
j
No controls required.
-------
Table 2-21. REGULATORY ALTERNATIVES AND INCREMENTAL COST EFFECTIVENESS BETWEEN REGULATORY ALTERNATIVES
FOR NEW BENZENE STORAGE VESSELS1
Regulatory Alternative
FR
bIFRVM.A
EFRss
— credit
2
,,1FR^ credit
II
, A
bIFRLM,A
credit
Cred1t
III
,,„ 16'100
bIFRLM>B 16.100
2
IV
en Notation Is as follows:
'FR - fixed roof tank
IFR = internal floating roof tank
EFR = external floating roof tank
b = bolted deck
LM = 11quid-mounted primary seal
VM = vapor-mounted primary seal
ss = secondary seal
ps = primary seal (mechanical shoe assumed)
A = Case A fittings
C = Case C fittings
No controls required.
-------
Table 2-22. EMISSIONS FROM EXISTING MODEL PLANTS
Regulatory Alternative
Tank dimensions ,
(meters x meters)
Large benzene producer
12 x 9
18 x 12
8x5
9x9
13 x 13
24 x 9
27 x 15
Total
Small benzene producer
3 x 11
13 x 13
8 x 11
32 x 7
Total
Benzene consumer
12 x 11
18 x 15
Total
Bulk storage terminal
12 x 11
18 x 15
Total
0
0.72
2.19
0.48
0.59
0.68
1.36
1.82
7.84
1.27
0.68
0.50
2.17
4.61
0.64
0.97
1.61
0.64
0.97
1.61
I
0.72
0.13
0.45
0.59
0.68
1.36
1.82
5.78
0.32
0.68
0.50
2.17
3.66
0.64
0.97
1.61
0.64
0.97
1.61
II
Emissions (Mg/y)
0.72
0.13
0.48
0.59
0.68
1.36
1.82
5.78
0.28
0.68
0.50
2.17
3.62
0.64
0.97
1.61
0.64
0.97
1.61
III
0.72
0.13
0.48
0.59
0.68
1.36
1.82
5.78
0.21
0.68
0.50
2.17
3.56
0.64
0.97
1.61
0.64
0.97
1.61
IV
0.56
0.13
0.38
0.47
0.52
1.04
1.46
4.55
0.20
0.51
0.40
1.76
2.87
0.48
0.73
1.21
0.48
0.73
1.21
Diameter x height.
2-47
-------
Table 2-23. EMISSIONS FROM NEW MODEL PLANTS
Regulatory Alternative
Tank dimensions ,
(meters x meters)
Large benzene producer
12 x 9
18 x 12
8x5
9x9
13 x 13
24 x 9
27 x 15
Total
Small benzene producer
3 x 11
13 x 13
8 x 11
32 x 7
Total
Benzene consumer
12 x 11
18 x 15
Total
Bulk storage terminal
12 x 11
18 x 15
Total
0
0.72
0.13
0.48
0.59
0.68
1.36
1.82
5.78
1.27
0.68
0.50
2.17
4.61
0.64
0.97
1.61
0.64
0.97
1.61
I
Emi
0.72
0.13
0.48
0.59
0.68
1.36
1.82
5.78
0.32
0.68
0.50
2.17
3.66
0.64
0.97
1.61
0.64
0.97
1.61
II
ssions (Mg/y)
0.58
0.13
0.39
0.48
0.54
1.07
1.50
4.70
0.28
0.53
0.41
1.80
3.02
0.50
0.76
1.26
0.50
0.76
1.26
III
0.49
0.13
0.30
0.38
0.44
0.98
1.23
3.93
0.18
0.43
0.31
1.49
2.41
0.40
0.65
1.06
0.40
0.66
1.06
IV
0.43
0.13
0.26
0.34
0.38
0.87
1.11
3.52
0.17
0.37
0.28
1.36
2.17
0.35
0.58
0.93
0.35
0.58
0.93
Diameter x height.
2-48
-------
Table 2-24. 1983 NATIONWIDE EMISSIONS FROM
EXISTING BENZENE STORAGE TANKS
Model plant
Regulatory Alternative
II
III
Emissions (Mg/y)
IV
Large benzene producer 269
Small benzene producer 192
Benzene consumer 152
Bulk storage terminal ' 8
Total 621
198
152
152
8
510
198
151
152
8
509
198
148
152
8
506
156
120
114
6
396
2-49
-------
Table 2-25. NATIONWIDE 1988 (fifth-year) EMISSIONS
ON NEW BENZENE STORAGE TANKS
Regulatory Alternative
Model plant
Large benzene producer
Small benzene producer
Benzene consumer
Bulk storage terminal
Total
0
55
53
42
2
152
I
55
42
42
2
141
II
Emissions
45
35
33
2
115
III
(Mg/y)
37
28
28
1
94
IV
33
25
24
1
83
2-50
-------
Table 2-26. CAPITAL COSTS FOR EXISTING MODEL PLANTS
Regulatory Alternative
Tank dimensions -,
(meters x meters)
Large benzene
producer
12 x 9
18 x 12
8x5
9x9
13 x 13
24 x 9
27 x 15
Total
Small benzene
producer
3 x 11
13 x 13
8 x 11
32 x 7
Total
Benzene consumer
12 x 11
18 x 15
Total
Bulk storage
terminal
12 x 11
18 x 15
Total
0
0
0
0
0
0
0
0
0
0
0
0
0
6
0
0
0
0
0
0
I
••
0
12,470
0
0
0
0
0
12,470
5,570
0
0
0
5,570
0
0
0
0
0
0
II
Costs ($)
0
12,470
0
0
0
0
0
12,470
5,590
0
0
0
5,590
0
0
0
0
0
0
III
0
12,470
0
0
0
0
0
12,470
5,590
0
0
0
5,590
0
0
0
0
0
0
IV
7,590
12,470
4,280
5,990
9,170
15,390
20,510
75,400
6,930
9,190
5,220
18,910
40,250
8,150
13,090
21,240
8,150
13,090
21,240
Diameter x height.
2--51
-------
Table 2-27. TOTAL ANNUALIZED COST (WITHOUT PRODUCT RECOVERY CREDITS)
FOR EXISTING MODEL PLANTS
Regulatory Alternative
Tank dimensions ,
(meters x meters)
Large benzene
producer
12 x 9
18 x 12
8x5
9x9
13 x 13
24 x 9
27 x 15
Total
Small benzene
producer
3 x 11
13 x 13
8 x 11
32 x 7
Total
Benzene consumer
12 x 11
18 x 15
Total
Bulk storage
terminal
12 x 11
18 x 15
Total
0
0
0
0
0
0
0
0
0
0
0
0
•o
0
0
0
0
0
0
0
I
0
2,960
0
0
0
0
0
2,960 '
1,210
0
0
0
1,210
0
0
0
0
0
0
II
Costs ($)
0
2,960
0
0
0
0
0
2,960
1,220
0
0
0
1,220
0
0
0
0
0
0
III
0
2,960
0
0
0
0
0
2,960
1,220
0
0
0
1,220
0
0
0
0
0
0
IV
1,810
2,960
1,040
1,420
1,270
3,360
4,830
16,690
1,430
1,220
1,240
4,540
8,430
1,940
2,660
4,600
1,940
2,660
4,600
Diameter x height.
2-52
-------
Table 2-28.
TOTAL ANNUALIZED COST (WITH PRODUCT RECOVERY CREDITS)
FOR EXISTING MODEL PLANTS
Regulatory Alternative
Tank dimensions ,
(meters x meters)
Large benzene
producer
12 x 9
18 x 12
8x5
9x9
13 x 13
24 x 9
27 x 15
Total
Small benzene
producer
3 x 11
13 x 13
8 x 11
32 x 7
Total
Benzene consumer
12 x 11
18 x 15
Total
Bulk storage
terminal
12 x 11
18 x 15
Total
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
I
0
2,220
0
0
0
0
0
2,220
870
0
0
0
870
0
0
0
0
0
0
II
Costs ($)
0
2,220
0
0
0
0
0
2,220
860
0
0
0
860
0
0
0
0
0
0
III
0
2,220
0
0
0
0
0
2,220
840
0
0
0
840
0
0
0
0
0
0
IV
1,750
2,220
1,000
1,380
1,210
3,240
4,700
15,500
1,040
1,150
1,200
4,390
7,780
1,880
2,570
4,450
1,880
2,570
4,450
Diameter x height.
2-53
-------
Table 2-29. CAPITAL COSTS FOR NEW MODEL PLANTS
Regulatory Alternative
Tank dimensions ,
(meters x meters)
Large benzene
producer
12 x 9
18 x 12
8x5
9x9
13 x 13
24 x 9
27 x 15
Total
Small benzene
producer
3 x 11
13 x 13
8 x 11
32 x 7
Total
Benzene consumer
12 x 11
18 x 15
Total
Bulk storage
terminal
12 x 11
18 x 15
Total
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
I
0
0
0
0
0
0
0
0
4,150
0
0
0
4,150
0
0
0
0
0
0
II
Costs ($)
100
0
65
75
105
195
220
760
4,170
105
65
260
4,600
100
150
250
100
150
250
III
100
0
65
75
105
195
220
760
4,170
105
65
260
4,600
100
150
250
100
150
250
IV
3,300
0
2,200
2,475
3,575
6,605
7,430
25,585
4,970
3,575
2,200
8,805
19,550
3,300
4,955
8,255
3,300
4,955
8,255
Diameter x height.
2-54
-------
Table 2-30. ANNUALIZED COST (WITHOUT PRODUCT RECOVERY CREDITS)
FOR NEW MODEL PLANTS
Regulatory Alternative
Tank dimensions ..
(meters x meters)
Large benzene
producer
12 x 9
18 x 12
8x5
9x9
13 x 13-
24 x 9
27 x 15
Total
Small benzene
producer
3 x 11
13 x 13
8 x 11
32 x 7
Total
Benzene consumer
12 x 11
18 x 15
Total
Bulk storage
terminal
12 x 11
18 x 15
Total
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
o
0
I
0
0
0
0
0
0
0
1 o
900
0
0
0
900
0
0
0
0
0
0
II
Costs ($)
20
0
15
15
25
45
50
170
910
25
15
55
1,005
20
30
50
20
30
50
III
20
0
15
15
25
45
50
170
910
25
15
55
1,005
20
30
50
20
30
50
IV
860
0
575
650
935
1,725
1,940
6,685
1,125
935
575
2,302
4,937
860
1,295
2,155
860
1,295
2,155
Diameter x height.
2-55
-------
Table 2-31. ANNUALIZED COST (WITH PRODUCT RECOVERY
CREDITS) FOR NEW MODEL PLANTS
Tank dimensions ,
(meters x meters)
Large benzene
producer
12 x 9
18 x 12
8x5
9x9
13 x 13
24 x 9
27 x 15
Total
Small benzene
producer
3 x 11
13 x 13
8 x 11
32 x 7
Total
Benzene consumer
12 x 11
18 x 15
Total
Bulk storage
terminal
12 x 11
18 x 15
Total
0
0
0
0
0
0
0
0
0
0
0
0
0
o'
0
0
0
0
0
0
I
0
0
0
0
0
0
0
.- 0
560
0
0
0
560
0
0
0
0
0
0
Regulatory
II
Costs
credit
0
credit
credit
credit
credit
credit
credit
535
credit
credit
credit
430
credit
credit
credit
credit
credit
credit
Alternative
III
($)
credit
0
credit
credit
credit
credit
credit
credit
520
credit
credit
credit
210
credit
credit
credit
credit
credit
credit
IV
775
0
655
560
830
1,550
1,685
6,055
730
825
495
2,010
4,060
755
1,125
1,180
755
1,125
1,880
LDiameter x height.
2-56
-------
Table 2-32. INCREMENTAL COST EFFECTIVENESS FOR EXISTING MODEL PLANTS
Regulatory Alternative
Tank Dimensions,
(meters x meters)
Large benzene
producer
12 x 9
18 x 12
8x5
9x9
13 x 13
24 x 9
27 x 15
Small benzene
producer
3 x 11
13 x 13
8 x 11
32 x 7
Benzene consumer
12 x 11
18 x 15
Bulk storage
terminal
12 x 11
18 x 15
0
0
0
0
0
0
0
0
0
0
0
0-
0
0
0
0
I
0
1,080
0
0
0
0
0
915
0
0
0
0
0
0
0
II
($/Mg)
0
0
0
0
0
0
0
credit
0
0
0
0
0
0
0
III
0
0
0
0
0
0
0
credit
0
0
0
0
0
0
0
IV
10,900
0
42,800
11,500
7,560
10,130
13,060
20,000
6,760
12,000
10,710
11,750
64,250
11,750
64,250
Diameter x height.
2-57
-------
Table 2-33. INCREMENTAL COST EFFECTIVENESS FOR NEW MODEL PLANTS
Regulatory Alternative
0
(meters x meters)
Large benzene
producer
12 x 9 —
18 x 12 —
8x5 —
9x9 —
13 x 13 —
24 x 9 —
27 x 15 —
Small benzene
producer
3 x 11 —
13 x 13 —
8 x 11 —
32 x 7 — •
Benzene consumer
12 x 11 —
18 x 15 —
Bulk storage
terminal
12 x 11 —
18 x 15 —
I II
($/Mg)
— credit
— —
— credit
— credit
— credit
— credit
— credit
590 credit
— credit
— credit
— credit
— credit
— credit
— credit
— credit
III
credit
—
credit
credit
credit
credit
credit
credit
credit
credit
credit
credit
credit
credit
credit
IV
16,800
—
17,630
15,500
14,800
14,900
15,400
21,000
14,800
18,300
16,900
16,400
17,300
16,400
17,300
Diameter x height.
2-58
-------
Table 2-34. NATIONWIDE CAPITAL COSTS
FOR EXISTING MODEL PLANTS
Model Plant
Regulatory Alternative
I
II
III
Costs ($106)
IV
Large benzene producer 0 0.43
Small benzene producer 0 0.23
Benzene consumer 0 0
Bulk storage terminal 0 0
Total 0 0.66
0.43
0.23
0
0
0.66
0.43
0.23
0
0
0.66
2.6
1.7
2.0
1.0
7.3
2-59
-------
Table 2-35. NATIONWIDE ANNUALIZED COST WITH PRODUCT RECOVERY
CREDITS FOR EXISTING MODEL PLANTS
r
Regulatory Alternative
Model Plant
Large benzene producer
Small benzene producer
Benzene consumer
Bulk storage terminal
Total
0
0
0
0
0
0
I
76
36
0
0
112
II
Costs ($103/yr)
76
36
0
0
112
III
76
35
0
0
111
IV
530
330
420
22
1,300
2-60
-------
Table 2-36. NATIONWIDE 1988 (fifth-year) CAPITAL COSTS
FOR NEW MODEL PLANTS
Model Plant
Large benzene producer
Small benzene producer
Benzene consumer
Bulk storage terminal
Total
0 I II
Costs ($103)
-a -b 7.2
— a 48.0 53.0
-a -b 6.5
-a -b 0.3
^-a 48.0 67.0
III
7.2
53.0
6.5
0.3
67.0
IV
243
225
215
11
694
a
Baseline.
3No change from previous regulatory alternative.
2-61
-------
Table 2-37. NATIONWIDE 1988 (fifth-year) ANNUALIZED COSTS
(WITH PRODUCT RECOVERY CREDITS) FOR NEW MODEL PLANTS
0
Model Plant
Large benzene producer —
Small benzene producer — a
Benzene consumer — a
Bulk storage terminal — a
Total — a
I
__b
6,500
_b
_b
6,500
II
Costs ($/yr)
credit
5,000
credit
credit
2,700
III
credit
2,400
credit
credit
credit
IV
57,000
47,000
49,000
2,500
155,000
a
Baseline.
JNo change from previous regulatory alternative.
2-62
-------
It should be noted that the cost effectiveness values which appear
on the tables in this section are higher than those calculated for the
model tank. This is because the fixed roof tank (1st tank in the small
producer) is much smaller than the model tank (21,000 gallon as compared
to 160,000 gallon). The decisions for BDT were appropriately based on
an average or model tank size. A size cutoff was selected at a later
step.
2.1.3.9 Economic Nonair Environmental and Energy Impact Considerations
of BDT. The nationwide capital and annualized costs of the controls
tentatively selected as BDT were calculated, using the current number
and distribution of existing tanks. These costs are $727,000 and $105,000
per year, respectively. These costs are judged to be reasonable. They
are considerably smaller than they were at proposal, both on a per plant
and nationwide basis. No unreasonable adverse economic impacts were
identified at proposal, based on the projected costs at proposal. Since
the costs for BDT selected at promulgation are considerably smaller than
those projected at proposal, it was reasonable to conclude that these
costs would also not cause any unreasonable adverse economic impacts.
There are no adverse nonair environmental or energy impacts associated
with the control techniques selected as BDT based on incremental cost
effectiveness. Therefore, these control techniques continued to be
selected as BDT after consideration of factors other than cost
effectiveness, i.e., nonair environmental impacts, energy and economic
impacts.
2.1.3.10 Beyond BDT Consideration. After identifying certain
control techniques as BDT (Regulatory Alternative III), the EPA made a
judgment concerning the unreasonableness of the health risks remaining
after the application of BDT by examining the health risks reductions,
costs, and other impacts that would result from application of more
stringent control techniques (Regulatory Alternative IV).
The health and cost impacts of this beyond-BDT regulatory alternative
were calculated. Due to the assumptions used in calculating the health
numbers, there is considerable uncertainty associated with them beyond
the ranges presented here. The uncertainties associated with these risk
2-63
-------
numbers are explained in Section 2.3.2. Requiring the beyond-BDT regulatory
alternative instead of BDT would reduce the estimated leukemia incidence
within 20 kilometers of plants with benzene storage vessels from a range
of 0.017 to 0.113 cases per year to a range of 0.011 to 0.074 cases per
year for existing sources. The incidence reduction for new sources
would depend on the location of those sources, but assuming the same
distribution as for existing sources, would be less than a third of the
figure for existing sources. It would reduce estimated maximum lifetime
risk at the point of maximum exposure caused by benzene storage vessels
from a range of 1.38 x 10 to 9.48 x 10 to a range of 8.66 x 10 to
5.95 x 10 . Requiring this level of control rather than BDT would
increase capital cost from $660,000 to $7.3 million for existing sources
and from $67,000 to $694,000 for new sources. Total annualized cost
would increase from $111,000 to $1.3 million for existing sources and
from a credit to $155,000 for new sources. Because of the relatively
small health benefits to be gained with the additional costs for requiring
the level of control more stringent than BDT, the EPA considers the
risks remaining after application of BDT to existing and new sources not
to be unreasonable. The EPA considers this level of emission reduction
to provide an ample margin of safety, and consequently, no more stringent
emission controls than BDT for existing and new sources are required.
2.1.3.11 New Sources Constructed Between Proposal and Promulgation.
A new source is defined as one for which construction or modification
commenced after the date the standard was proposed. The controls selected
for new internal floating roof storage vessels include liquid-mounted
primary seals and control of fittings and column wells. The requirement
for liquid-mounted primary seals was included in the proposal. The
requirements for control of fittings and column wells were not. Therefore,
the owner of a new storage vessel constructed between proposal and
promulgation would not have known to control fittings or column wells.
Controlling column wells after a tank has already been constructed would
require major reconstruction of the entire tank. Requiring retrofit of
this control on new tanks for which construction commenced before
promulgation would be unreasonable. Therefore, the standard requires
2-64
-------
that only tanks for which construction commences after the promulgation
date have to control column wells. Even though the owner of a new
storage vessel constructed before promulgation was not aware before
promulgation that control of fittings would be required, he was put on
notice that secondary seals would be required. Continuous secondary
seals achieve as much control as does controlling fittings. Therefore,
if any storage vessel is equipped with a continuous secondary seal,
control of fittings is not required. If a tank constructed between
proposal and promulgation has a continuous secondary seal, therefore,
that tank will be considered in compliance with the part of the final
rule that requires control of fittings. All of the other requirements
in the final rule are less stringent than in the proposed rule. For
example, although there are requirements for external floating roofs in
the final rule which were not in the proposed rule, external floating
roofs were not allowed by the proposed rule. Therefore, anyone constructing
a new storage vessel between proposal and promulgation in accordance
with the proposed rule would not have constructed an external floating
roof, so the requirements in the final rule have no retrofit effect on
new tanks.
As discussed earlier, the promulgated standard also requires all
external floating roofs to install liquid-mounted primary seals and
secondary seals. However, data show that a mechanical shoe primary seal
combined with a continuous secondary seal is equivalent in reducing
emissions to a liquid-mounted primary seal combined with a secondary
seal system. Therefore, the promulgated standard considers new and
existing external floating roofs equipped with a mechanical shoe seal
and secondary seal to be in compliance with the standard.
The promulgated standard also allows existing external floating
roofs already equipped with a liquid-mounted seal to wait to install the
secondary seal until the first degassing of the tank. This is allowed
because the incremental cost effectiveness of adding the secondary seal
initially is $5,000 per Mg, which is judged to be unreasonable. The
cost of initially adding a secondary seal to an existing external floating
roof equipped with a mechanical shoe primary seal is $l,000/Mg, which is
2-65
-------
judged to be reasonable. The promulgated standard, therefore, requires
the Initial addition of a secondary seal to an existing external floating
roof equipped with a mechanical shoe primary seal.
2.2 APPLICABILITY AND APPLICABILITY OF CONTROL TECHNOLOGIES
2.2.1 Storage Vessels Attached to Moving Vehicles
Comment: One commenter (IV-D-7) stated that definition of storage
vessels should exclude storage vessels attached to mobile vehicles, such
as tankers, barges and tank trucks.
Response: The control technologies that would be necessary to
control benzene emissions from storage vessels attached to mobile vehicles,
such as tankers, barges, or tank trucks, are completely different from
those that are appropriate for other storage vessels. Additionally,
data collection on tanks, barges, and tank trucks was not part of the
survey performed by the EPA to develop a data base to support the Benzene
Storage Vessel NESHAP. For these reasons, it was never the intent of
the EPA to consider these types of benzene storage vessels as designated
sources under this NESHAP.
However, the comment indicates that the scope of the applicability
must need clarification. Therefore, the EPA agrees with the commenter
and has clarified the applicability of the NESHAP. To accomplish this,
storage vessels attached to mobile vehicles have been specifically
excluded in §61.120d.
2.2.2 Coke Oven Byproduct Vessels
Comment: One commenter (IV-D-15) stated that the regulations
should not exempt tanks at coke oven by-product facilities. He felt
that if benzene is hazardous, all facilities should be required to meet
uniform control requirements.
Response: A separate NESHAP is currently being developed for coke
oven by-product facilities. Vessels at coke oven by-product facilities
were not incorporated into the NESHAP for Benzene Storage Vessels because
the applicable control techniques are different than the ones considered
for this NESHAP. This is a function of the nature of the coking and
byproduct processes. For this reason, the Agency has decided that a
separate standard for vessels at coke oven by-product facilities is
appropriate and that such vessels should not be incorporated into these
standards.
2-66
-------
Comment: Two commenters (IV-D-6, IV-D-12) questioned the applicability
of the control requirements for underground storage tanks. One commenter
stated that the EPA did not evaluate the application of floating roof
technology to underground storage tanks in terms of feasibility, effective-
ness, or costs. The other commenter recommended that underground storage
tanks be exempted from regulation because these types of tanks have
virtually no breathing losses caused by diurnal temperature changes.
One commenter (IV-D-12) stated that the required floating roof
technology was inappropriate for either small or horizontal tanks. The
commenter indicated that the design and structure of both small and
horizontal tanks would prohibit the use of floating roofs. The commenter
believes that control costs relative to the emissions reduction would be
higher for these tanks than for larger conventional tanks.
Response: It is true that in the development of the proposed
standards the Administrator did not specifically evaluate the installation
of internal floating roofs in underground or horizontal benzene storage
vessels at the time of proposal. The Agency is currently aware of
14 existing underground benzene'storage vessels. Five of these underground
vessels are horizontal. The volumes of the 14 vessels ranged from a
high of 190 m3to a low of 75 m3 with an average volume of 150 m3.
It is not technically feasible to equip a horizontal storage vessel
with an internal floating roof because, unlike a vertical storage vessel,
the liquid surface area is not constant with changes in liquid level.
The internal floating roof in a horizontal storage vessel would be
required to vary its surface area as the liquid surface area varied. No
such internal floating roofs exist.
The installation of internal floating roofs in existing underground
storage vessels involves a number of significant issues. (1) It would
be necessary to disinter the vessel to install the control equipment.
This is a cost that was not considered in the proposal. (2) Most
underground vessels do not have accessible entrance ports in the fixed
roof. Thus, as currently operated, it would not be possible to conduct
the required visual inspection of the internal floating roof and seal
without disintering the vessel. (3) To conduct the internal inspection,
2-67
-------
it would be necessary to again disinter the storage vessel. While no
quoted costs for the dis- and reinterment of an underground vessel are
available, the cost may exceed the capital cost of the control equipment.
Thus, it may not be economically practicable to install internal floating
roofs in underground storage vessels.
Based on the problems cited above, the Agency concurs with the
commenters assertions that the controls selected as BDT (internal floating
roofs) are impracticable for both underground and horizontal benzene
storage vessels. Therefore, the Administrator specifically examined the
impacts associated with requiring that benzene emissions from underground
and horizontal storage vessels be controlled with vapor collection and
recovery or disposal systems.
The capital cost of vapor collection and recovery system (carbon
adsorber) was estimated to be $231,000; the capital cost of a vapor
collection and disposal system (thermal oxidizer) is about $188,000 (all
figures are first-quarter 1979 dollars from the Volume I BID). The
application of these control techniques to a single typical underground
storage vessel (150 m3 tank undergoing 10 turnovers a year) results in
annualized operating costs of $71,900 for carbon adsorption or $62,100 for
thermal oxidation. Uncontrolled emissions from this vessel are about
0.867 megagrams per year. The cost effectiveness in these cases is
about $83,100 per megagram controlled for the carbon adsorption case and
about $72,200 per megagram controlled if thermal oxidation is assumed.
Based on this analysis, the Agency concluded that the costs of
requiring vapor collection and recovery or disposal systems for underground
or horizontal storage vessels are unreasonably high for the emissions
reduction obtained. Because there are no other techniques by which to
control benzene emissions from these vessels, the Administrator has
decided to exempt each existing underground and each existing horizontal
benzene storage vessel from the promulgated standards.
There is, however, no technical reason why industry must construct
new underground or horizontal benzene storage vessels. Underground
storage vessels may have minor advantages over above-ground vessels in
freezing climates because of the insulating nature of the earth. But
2-68
-------
these advantages may be offset by other factors such as limited access.
It may be possible to locate a horizontal vessel within a plant where a
vertical vessel would not fit (e.g., under a pipe rack). However, this
is simply a matter of operator convenience. Therefore, the Administrator
has decided not to extend the above exemption for existing underground
and horizontal vessels to new benzene storage vessels. While an owner
or operator may construct a new underground or horizontal benzene storage
vessel, emissions from such a vessel must be controlled in accordance
with the final regulation. Such control would probably consist of a
vapor recovery or vapor disposal system.
2.2.3 Benzene Mixtures
Comment: One commenter (IV-D-19) pointed out that the proposed
standards would apply only to vessels that store pure benzene. He asked
if vessels that store mixtures of benzene and other substances existed;
and if so, why such vessels were not affected by these standards.
Response: It is true that vessels storing mixtures of benzene and
other chemicals exist, but such vessels were never intended to be part
of this source category. In part, this is because many vessels storing
mixtures, such as those associated with coke oven byproduct processes,
have different control options than those affected by the NESHAP. The
controls and impacts of control strategies for vessels storing mixtures
would have to be examined as part of a separate NESHAP. For this reason,
the Agency decided not to extend the applicability of this NESHAP to
vessels storing benzene mixtures.
However, some information is currently available on vessels storing
benzene mixtures, which for completeness will be presented here. There
are three general classes of stored liquids that are composed of benzene
that would not be affected by the proposed standards. These are:
1. Liquids such as gasoline, which are stored in large quantities,
but do not, on a fractional basis, contain more than 10 percent
benzene;
2. Mixtures in which benzene may be more than 10 percent; and
3. Benzene that does not meet the specific gravity specification
for industrial grade benzene (crude benzene).
2-69
-------
New vessels storing gasoline (~2 percent benzene), are affected
facilities under Standards of Performance for Petroleum Liquid Storage
Vessels (40CFR60: Subpart K(a)). These standards discussed above
require controls that are almost identical to those that have been
selected as BDT for new benzene storage vessels in the NESHAP final rule
requirements. Many state implementation plans (SIPs) require that
existing gasoline storage vessels be controlled to some extent, although
not as stringently as BDT for existing benzene storage vessels.
Data were gathered on vessels storing liquids of the second class
(Table 2-38). This data was obtained from a data base of 4,054 vessels
associated with the Synthetic Organic Chemical Manufacturing
Industry (SOCMI). Fifteen (15) were thought to possibly contain more
than 10 percent of benzene. The total volume of these vessels is about
2.7 million gallons. This can be compared to the estimated 500 vessels
with a total volume of about 308 million gallons that stored industrial
grade benzene in 1979. The total tank volume (tankage) devoted to the
storage of this type of benzene mixtures is less than one percent of the
!t
tankage devoted to benzene. Because vessels storing mixtures will have
reduced amounts of benzene in them, the true amount of benzene stored
may be significantly reduced from the above two million gallons.
The last class of liquids consists of unfinished (crude) benzene or
off specification benzene. Most such liquids are petroleum liquids and
many are affected facilities under Subpart K(a) or the SIPs and as such,
would be controlled to some extent. There was only one such tank in the
data base.
2.2.4 Tank Size Cutoff
Comment: Three commenters questioned the 4 m3 (1,000 gallon) limit
of applicability. One commenter stated that based upon information
presented in the draft Volume I BID "VOC Emissions from Volatile Organic
Liquid Storage Tanks" the proposed control technology did not obtain a
demonstrated emissions reduction in tank volumes smaller than 150 m3.
He concluded that the proposed controls would not provide an emission
reduction in small tanks (IV-D-10, lOa). The same commenter stated that
internal floating roof vendors indicated that the installation of internal
2-70
-------
Table 2-38. VESSELS CONTAINING MIXTURES THAT MAY BE
MORE THAN 10 PERCENT BENZENE3
Vessel contents
Volume (1000's of gallons)
Benzene Caprolactum
Benzene Lactum
Benzene/Toluene
EA, Benzene, Water
EA, Benzene
Light Aromatic Distillate
4.4
7
37.8, 237, 42, 8.8, 17.0, 1272.7l
2, 2, 2b
4.75
515, 515, 63.5b
Including crude benzene.
"'Multiple vessels with same contents.
2-71
-------
floating roofs in tanks less than 10 feet in diameters would generally
be impracticable; two other vendors he had contacted indicated that such
an installation would require a site specific design and, therefore, be
unreasonably expensive.
The second commenter stated that applying the standards in such
small vessels would encourage users to handle benzene in environmentally
unsound ways, such as drumming. This could increase spills and benzene
emissions. This commenter recommended that the applicability limit be
set at 38 m3 (10,000 gallons).
The third commenter stated that internal floating roofs for 4 m3
tanks would have to be custom built, and therefore, because of costs the
applicability should apply to tanks with volumes greater than or equal
to 19,800 gallons (IV-D-21).
Response: Because of the above comments, and because the equipment
selected for the standard has changed from that specified at proposal,
the Agency has reevaluated the tank size cutoff. Three potential size
cutoffs were examined in terms of nationwide and individual tank impacts.
They were 1,000 gallons, 10,000 gallons, and 20,000 gallons. In the
analyses, costs were generated from the regression equations and techniques
presented in Section 2.1.3.3, except where the tank diameter was less
than 10 feet. To account for custom fittings, the costs of equipment
for tanks less than 10 feet in diameter were not allowed to be less than
those for 10 foot diameter tanks. The regressed capital cost for BDT
equipment in a 10 foot tank is about $4,200. However, some vendors made
actual quotes as low as" $2,400. The $1,800 margin between the lowest
quote and the regressed cost, combined with the assumption that there is
no reduction in cost below 10 feet in diameter should be sufficient to
account for any customizing costs in small vessels.
In performing the analysis, both the cost effectiveness of controlling
an individual tank and nationwide emissions reductions were considered.
However, cost effectiveness of controlling an individual tank was the
primary basis for decision making, while nationwide impacts were considered
mainly as ancillary information that was used to provide a general
examination of nationwide emissions as a function of vessel size. A
2-72
-------
different number of vessels would change neither the decision making
process or the final decision that follows below. The nationwide impacts
of the various cutoffs will be presented first followed by the more
important cost effectiveness information.
To examine nationwide impacts, the Agency sought additional data on
tanks smaller than 20,000 gallons. A data base developed for another
project (VOC Emissions from Volatile Organic Liquid (VOL) Storage Vessels)
contains information on approximately 4,000 fixed and floating roof
storage vessels. Nineteen of these contain benzene and have volumes
less than 20,000 gallons. Sixty-one contain benzene and have volumes
greater than or equal to 20,000 gallons.
Most of the plants surveyed for VOL would be considered consumers.
The Benzene Storage data base estimates that there are 167 tanks storing
benzene with volumes greater than 20,000 gallons at consumers. This
compares to 61 such tanks in the VOL data base. On this basis, a sealer
of 2.74 was developed. Applying this sealer to the 19 vessels with
volumes less than 20,000 gallons results in a total of 52 vessels with
volumes less than 20,000 gallons. The VOL data shows that overall ratio
of vessels between 10 and 20 thousand gallons should be 1. The Agency
is not aware of any reason why small benzene storage vessels should have
a different volume distribution than VOL small storage vessels. Therefore,
in the calculations of impacts, it was assumed that there were 26 vessels
with volumes between 10,000 and 20,000 gallons, and 26 vessels with
volumes less than 10,000 gallons.
The emissions analysis of a 1,000 gallon tank demonstrated that no
emission reduction is obtained by applying BDT to such a tank. This is
because fixed roof tank emissions are generally proportional to tank
volume. As the tank volume becomes small, baseline emissions decrease
rapidly. Emissions from the 1,000 gallon fixed roof tank are calculated
as 0.08 Mg/yr, while emissions from a 1,000 gallon tank equipped with
BDT are the same or slightly larger. For this reason, 1,000 gallons was
rejected as the size cutoff.
The analysis then focused on the 10,000 gallon cutoff (Table 2-39).
This cutoff would obtain an additional emission reduction of 18.2 Mg
2-73
-------
Table 2-39 NATIONWIDE CUTOFF IMPACTS ANALYSIS
ro
i
Volume range Number of
(gallons) tanks
1, 000-10, 000a 26
10,000-20,000a 26
Baseline
emission
in range
(Mg)
7.9
21.8
BDT emissions
in range
(Mg)
3.1
3.6
Emissions
reduction
in range
(Mg)
4.8
18.2
Annual i zed
cost in
range
(Mg)
30,700
34,900
Cost effectiveness
($/Mg)
6,000
1,550
Exclusive.
-------
over a 20,000 gallon cutoff, but would lose only 4.8 Mg of potential
emission reduction by not regulating smaller vessels. Additionally, the
average cost effectiveness of controlling tanks with volumes between
10,000 and 20,000 gallons is generally less than $1,600 per Mg as compared
to the smaller range where the average cost effectiveness is about
$6,000/Mg. The cost effectiveness of controlling a tank in the 10 to
20 thousand gallon range varies from about $100 per Mg to about $2,300
per Mg, which is judged to be reasonable.
Additionally, the requirement of controls at 10,000 gallons would
not result in tanks smaller than 10 feet in diameter being equipped with
internal floating roofs. Therefore, based on nationwide and individual
impacts, the Agency selected 10,000 gallons as the tank size cutoff.
2.2.5 Feasibility of BDT in Cold Climates
Comment: One commenter (IV-D-10) stated that the proposed BDT was
not feasible at any location where temperatures fall below the melting
point (5.5°C) of benzene because freezing benzene will result in technical
and safety problems. The commenter indicated that while some freezing
could be tolerated in fixed roof tanks, freezing in floating roof tanks
could cause problems with the roof and seal system. Freezing could
interfere with the seal system, the ability of the roof to slide on the
columns, and other problems. The commenter stated that heating the
walls of the tank to prevent freezing creates as many problems as it
solves. Heating the seal area increases emissions making it more likely
that an explosive mixture will develop between the floating and fixed
tank roof. The commenter stated that because of these problems, an
informal survey could find no floating roof applications to benzene
storage in the middle or northern states.
Response: A formal survey of benzene storage vessels undertaken
under the authority of Section 114 of the Clean Air Act, located many
vessels in northern states that employ technologies very similar to
those specified as BDT in the final rule. Both internal and external
floating roof tanks are used to store benzene in such states as Illinois,
Pennsylvania, Ohio, and other locations where the temperature may be
below the freezing point of benzene for months at a time.
2-75
-------
In evaluating the comment, the Agency concluded that the application
of BDT would cause no technical or safety problems that are not routinely
faced by many operators. Therefore, the Agency concludes that BDT is
feasible for affected facilities in cold climates.
2.3 RISK ANALYSIS AND HEALTH IMPACTS
2.3.1 Benzene Listing
A number of commenters stated that benzene should not have been
listed as a hazardous air pollutant under Section 112 of the Clean Air
Act, and that, therefore, the proposed benzene storage NESHAP should be
withdrawn. Various reasons were commonly cited to support this contention.
These comments are summarized below.
Comment. Some commenters (IV-D-lOa, IV-D-13, IV-D-16, IV-D-21,
IV-F-1) felt that benzene does not constitute the kind of risk deemed
hazardous by the courts or under Section 112 of the Clean Air Act.
Response: Response to this comment can be found in "Response to
Public Comments on EPA's Listing of Benzene Under Section 112 and Relevant
Procedures for the Regulation of Hazardous Air Pollutants" (EPA-450/5-
82-003), which was prepared to address the listing of benzene under
Section 112 of the Clean Air Act.
Comment: A number of commenters stated that there exists little or
no evidence to substantiate risk from emissions of benzene. Some commenters
(IV-D-2, IV-D-13, IV-D-14, IV-D-21) stated that there is no evidence.
Others (IV-D-4, IV-D-9, IV-D-15, IV-D-16) stated that there is insufficient,
suspect, or tenuous evidence.
Response: Response to this comment can be found in EPA-450/5-82-003,
which was prepared to address the listing of benzene under Section 112
of the Clean Air Act.
Comment: Two commenters (IV-D-4, IV-D-11) felt that the listing
and rulemaking proceedings for benzene are premature since they are
based on a draft EPA policy regarding airborne carcinogens. One of the
commenters (IV-D-4) felt that to proceed before the airborne carcinogen
policy is finalized is a violation of Section 307 of the Clean Air Act
and of Section 533 of the Administrator Procedure Act. One commenter
(IV-D-11) felt that EPA exceeded its legal authority and offended good
scientific practice in utilizing the airborne carcinogen policy to list
benzene.
2-76
-------
Response: Response to this comment can be found in EPA-450/5-82-003,
which was prepared to address the listing of benzene under Section 112
of the Clean Air Act.
2.3.2 Need for the Standard
Several commenters contended that the proposed benzene storage
emissions standard is not needed and, therefore, should be withdrawn.
These comments address the following: (1) significant and relative
proportion of risk associated with benzene storage emissions; (2) dupli-
cation of federal and State regulations and guidelines; (3) information
indicating that risks are smaller than estimated in the preamble to the
proposed standards; (4) acceptable residual risk; and (5) lack of data
to demonstrate risk.
When the standard for benzene storage vessels was proposed on
December 19, 1980, the Administrator made the judgment that "benzene
emissions from benzene storage vessels create a significant risk of
cancer and require the establishment of a national emissions standard
under Section 112" (45 FR 83954).
The data base used to calculate emissions from storage vessels has
changed since the standard was proposed. This change is based on new
test data acquired since proposal. This data base and the reasons for
using it have been described previously. Based on these new data, the
emission estimates for fixed roof tanks (totally uncontrolled tanks)
remains unchanged. The emission estimates for internal floating roof
tanks and external floating roof tanks are lower than at proposal.
Since a substantial proportion of existing tanks have internal or external
floating roofs, this change resulted in a substantial reduction in the
estimate of nationwide emissions from these tanks. For this reason, the
Administrator reevaluated the need to establish Section 112 standards
for benzene storage vessels.
While benzene storage vessels are currently neither the sole cause
of benzene emissions to the atmosphere nor the largest source emitter,
they contribute significantly to nationwide benzene emissions. In the
absence of a standard, storage vessels would still emit about 620 Mg per
year.
2-77
-------
Benzene Is causally related to leukemia incidence in humans. The
estimated 20 to 30 million people living within 20 kilometers of the
143 existing facilities having benzene storage vessels are exposed to
higher levels of benzene than if they lived at greater distances. These
people are exposed not only to emissions from benzene storage vessels,
but also benzene fugitive emission sources that exist at the same plants.
Some of the people are also exposed to emissions from process vents at
maleic anhydride plants and ethylbenzene/styrene plants. These plants
have process vents, storage vessels, and fugitive emission sources that
all have benzene emissions. Because no known threshold exists for
benzene's carcinogenic effects, the people living in the vicinity of
benzene storage vessels not only incur a higher benzene exposure but
also run greater risk of contracting leukemia due to that exposure.
Using the new emission data and a new modeling approach adopted
since proposal, EPA estimated leukemia cases and maximum lifetime risks
that occur due to exposure from storage vessels. There is considerable
uncertainty associated with these estimates. A small portion of that
uncertainty has been considered by calculating ranges. The ranges
presented represent uncertainty in estimates of benzene concentrations
to which workers were exposed in occupational studies of Infante, Aksoy,
and Ott that serves as the basis for developing the benzene unit risk
factor (Part I Docket Item II-A-31). Ranges are based on a 95-percent
confidence interval that assumes estimated benzene concentrations to
which workers were exposed are within a factor of two of actual
concentrations.
Other uncertainties associated with estimating health impacts are
not quantified here. Maximum lifetime risk and leukemia incidence were
calculated based on a no-threshold linear extrapolation of leukemia risk
associated with a presumably healthy white male cohort of workers exposed
to benzene concentrations in the parts per million to the general
population, which includes men, women, children, nonwhites, the aged,
and the unhealthy, who are exposed to concentrations in the parts per
billion range. These widely diverse population segments may or may not
have differing susceptibility to leukemia than do workers in the studies.
2-78
-------
In addition, the exposed population is assumed to be immobile, remaining
at the same location 24 hours per day, 365 days per year, for a lifetime
(70 years). A counterbalancing assumption, particularly in the calculation
of leukemia incidence is that no one moves into the area, either as a
permanent resident or as a transient. Assumptions that must be made in
order to estimate ambient concentrations by dispersion modeling and
exposed populations by census tract also introduce uncertainties into
the risk estimates. Modeled ambient benzene concentrations depend upon:
(1) plant configuration, which is difficult to determine for more than a
few plants; (2) emission point characteristics, which can be different
from plant to plant and are difficult to obtain in more than a few
plants; (3) emission rates, which may vary over time and from plant to
plant; and (4) meteorology, which is seldom available for a specific
plant. The particular dispersion model used can also influence the
numbers. The best model to use (ISC) is usually too resource intensive
for modeling a large number of sources. Less complex models introduce
further uncertainty through a greater number of generalizing assumptions.
Dispersion models also assume that the terrain in the vicinity of the
source is flat. For sources located in complex terrain, the maximum
annual concentrations could be underestimated by several-fold due to
this assumption.
Furthermore, leukemia incidence is the only benzene health effect
considered. Other health effects, such as aplastic anemia and chromosomal
aberrations, are not as.easily quantifiable and are not reflected in the
risk estimates. Although these other health effects have been observed
at occupational levels, it is not clear if they occur at ambient exposure
levels. Overlapping benzene exposures from other source categories are
also not included in the estimates. Finally, these estimates do not
include cumulative or synergistic effects of concurrent exposure to
benzene and other substances. As a result of these uncertainties, the
number of leukemia cases and the maximum lifetime risk calculated around
benzene storage vessels could be overestimated. However, they could
just as likely be underestimated for the same reasons.
2-79
-------
In the absence of a standard, estimated maximum lifetime risk would
range from 1.38 x 10 to 9.48 x 10 to the most exposed individuals.
Maximum lifetime risk is the probability of someone within the assumed
exposed population contracting leukemia who is exposed to the highest
maximum annual average benzene concentration during an entire lifetime
(70 years). Based on an estimate of the number of people exposed and
the associated level of exposure, a range of 0.017 to 0.113 leukemia
cases per year due to benzene emissions from existing benzene storage
vessels is estimated.
Based on the human carcinogencity of benzene, the amount of benzene
currently emitted from storage vessels (as revised based on the new test
data), the number of people exposed to benzene emissions from storage
vessels, and the uncertainties associated with the estimates of maximum
risks and leukemia incidence, the Administrator has concluded that
benzene emissions from benzene storage vessels pose a significant risk
of leukemia to the general public.
Several other factors were considered in the Administrator's
determination that a standard for benzene storage vessels is needed.
Although many existing storage vessels are controlled, at least to some
extent, there are in existence several uncontrolled tanks. Control
technologies are available to reduce emissions from these tanks and add
controls to other tanks at a reasonable cost. In fact some controls
would result in a net savings rather than a cost. In the broader
perspective of prudent public health policy, it is reasonable to require
that all tanks be controlled to at least the degree that already occurs
at many locations and to have controls that recent API studies have
shown to have no or relatively low costs for the emission reductions
achieved.
Comment: Three commenters stated that the EPA has not demonstrated
that benzene storage emissions, relative to other benzene source categories,
pose a significant risk that merits the adoption of a benzene storage
standard (IV-D-lOa, IV-D-16, IV-F-1). One of these commenters (IV-D-10a)
stated that Section 112 requires that a NESHAP be established at the
level which in the Administrator's judgement provides "an ample margin
2-80
-------
of safety to protect the public health from such hazardous air pollutant."
According to the commenter the Supreme Court has held that, absent a
"clear mandate" from Congress to eliminate all risk, the statutory term
"safe" (regarding exposure levels), rather than meaning "absolutely
risk-free," means a level that protects against a "significant risk of
harm." The commenter noted that risk levels that the EPA has calculated
are not "significant" as that term has been used by the court.
Two commenters (IV-D-16, IV-D-21) felt that the risk from exposure
to benzene emissions is insignificant compared to other commonly accepted
societal risks. Two commenters (IV-D-13, IV-D-16) noted that the risk
from benzene storage emissions is insignificant in comparison to the
background leukemia incidence risk.
Two commenters (IV-D-4, IV-D-13) further compared the risk from
benzene storage emissions to other government determinations of risk
acceptability and noted that, under these determinations, the risk from
exposure to benzene storage emission sources would have been considered
not worthy of regulation.
Response: Response to these comments can be found in EPA-450/5-82-003.
Comment: Several commenters stated that regulation of benzene
storage emissions duplicates existing Federal and State regulations and
guidelines. Several of these commenters asserted that emissions of
benzene will be reduced under other air pollution programs required by
the Clean Air Act. One commenter (IV-D-13) stated that SIP's proposed
involving National Ambient Air Quality Standards (NAAQS) for ozone
parallel the EPA's proposed provisions benzene storage monitoring,
recordkeeping, and reporting. He stated that 82% of all man-made sources
of benzene are controlled by the hydrocarbon and ozone NAAQS. Other
comments (IV-D-lOa, IV-F-1) supported the assertion that regulation of
benzene emissions under authorities other than Section 112 would be more
logical. For example, since more than 80 percent of benzene emissions
come from mobile sources not subject to Section 112, it would seem
appropriate to regulate benzene under the NAAQS.
Response: Response to these comments can be found in EPA-450/5-82-003.
2-81
-------
2.3.3 Standard is Overdue
Comment: One commenter (IV-D-19) supported the prompt issuance of
final standards, noting that the standard is long overdue since it
should have been set within a year of the June 1977 listing of benzene
as a hazardous air pollutant.
Response: Response to this comment can be found in EPA-450/5-82-003,
which was prepared to address the listing of benzene under Section 112
of the Clean Air Act.
2.3.4 Clean Air Act Authority for EPA to Set Control Requirements
Comment: One commenter (IV-D-19) felt that EPA lacks the authority
to "water down" the control requirements of Section 112. The commenter
agreed with the EPA's contention that since no threshold exposure can be
defined for carcinogens, no level of exposure can be considered safe.
Since there is no safe level of exposure, the commenter noted, Section 112
of the Clean Air Act establishes a goal of eliminating benzene emissions.
According to the commenter, best demonstrated technology is inadequate
for this task and should be replaced in a technology-forcing fashion by
"best performing technology," including transfer technology.
Response: Response to this comment can be found in EPA-450/5-82-003,
which was prepared to address the listing of benzene under Section 112
of the Clean Air Act.
2.3.5 Authority for Risk-Benefit Analysis
Comment: One commenter (IV-D-19) asserted that EPA lacks the
authority to engage in risk-benefit analyses as it has in the benzene
proceedings. According to the commenter, Congress has consistently
rejected risk-benefit analysis as having no place in the Clean Air Act;
furthermore, there is no place in the Act where risk-benefit analysis is
less consistent with the statutory language and intent than Section 112,
with its directive to set standards that protect health with "an ample
margin of safety".
Response: Response to this comment can be found in EPA-450/5-82-003,
which was prepared to address the listing of benzene under Section 112
of the Clean Air Act.
2.3.6 Suspension of Benzene Proceedings
Comment: A number of commenters (IV-D-1, IV-D-3, IV-D-4, IV-D-11,
IV-D-13, IV-D-16, IV-D-21, IV-F-1) maintained that the EPA should temporarily
2-82
-------
suspend or postpone proceedings on the benzene source-specific standards
until EPA makes final decisions on the listing of benzene. One commenter
(IV-D-1) noted that a brief suspension of the proceedings will not
result in undue delay.
Response: EPA believed that a suspension or delay of the standard
was unnecessary. The standard development process takes enough time
that any issues arising with respect to the listing of benzene have
ample time to be reviewed and resolved before the standard is promulgated.
In addition, Section 112 requires that EPA promptly establish national
emission standards for hazardous air pollutants.
2.3.7 Dose/Response Analysis
Comment: Two commenters (IV-D-1, IV-D-2, IV-D-3, IV-D-4, IV-D-9,
IV-D-lOa, IV-D-13, IV-D-16, IV-D-21) stated that the EPA's assumption
that leukemia risk can be extrapolated from high doses to very low doses
is not justified by available direct evidence. Two commenters (IV-D-lOa,
IV-D-13) believed that the linear dose-response model is the most
conservative method that could be applied to the data, and it results in
an upper-limit estimate of the leukemia risk for benzene. These commenters
contended that available empirical evidence suggests the absence of
health effects below 10 parts per million. According to the commenters,
the EPA notes that benzene has been connected with other adverse health
effects, such as pancytopenia, aplastic anemia, chromosome changes, and
reproductive effects; but the commenter contends that these effects
result only from exposures in excess of 10 parts per million. Moreover,
they add there is no direct evidence that benzene is carcinogenic or
leukemogenic at levels below 100 parts per million. The commenters
noted that the EPA estimated that maximum benzene exposures would be
only in the very low parts per billion range, and average exposure
within 20 kilometers of the source would be only 19 parts per trillion.
They feel that no leukemia risk can be substantiated at these low levels
of exposure.
Response: Response to this comment can be found in EPA-450/5-82-003,
which was prepared to address the listing of benzene under Section 112
of the Clean Air Act.
2-83
-------
Comment: Three commenters (IV-D-4, IV-D-13, IV-D-21) noted that,
in its ruling on the Occupational Safety and Health Administration's
(OSHA) reduction of allowable occupational exposure to benzene from
10 ppm to 1 ppm, the U.S. Supreme Court in 1980 determined that OSHA
"made no finding that the Dow study, or any other evidence, or any
opinion testimony demonstrated that exposure to benzene at or below the
20 ppm level had ever in fact caused leukemia". According to the commenter,
the EPA has based its evaluation of public exposure to benzene storage
vessels on an estimated maximum out-of-plant concentration in a parts-per-
billion range, significantly below the levels addressed in the OSHA
proceedings.
Commenters (IV-D-lOa, IV-D-21) acknowledged that there is some
epidemiologic data that appear to support an association between leukemia
and benzene at high concentrations. However, he added that the leukemogenic
action of benzene at these high levels is preceded by blood changes,
such as cytopenia and pacytopenia, and that these pre-leukemic changes
do not occur at levels below about 35 ppm. According to the commenter,
one study of benzene-exposed pliofilm rubber workers that allegedly
found excess leukemia at low levels turned out to have underestimated
the exposure, which in fact substantially exceeded 100 ppm. In two
other epidemiologic studies, one on petroleum workers exposed to benzene
and one on benzene-exposed chemical workers, he noted that no excess
leukemia was found. As the commenter stated, the exposures in the
latter two groups, although not precisely quantifiable, were clearly
much greater than the nonoccupational exposures in the community. Other
comments (IV-F-1) made similar general statements.
Response: Response to this comment can be found in EPA-450/5-82-003,
which was prepared to address the listing of benzene under Section 112
of the Clean Air Act.
Comment: Two commenters (IV-D-lOa, IV-D-21) observed that the CAG
risk factor was overestimated because CAG misinterpreted the results of
the three epidemiological studies used in the analysis. According to
the commenter, one of the studies showed no statistically significant
increase in leukemia deaths, and those leukemias that did occur were
2-84
-------
only doubtfully related to benzene exposure. The commenter asserted
that CAG's analysis of the other two studies overstated the relative
leukemia risk by overcounting leukemia incidence in the study population,
underestimating leukemia incidence in control populations, and under-
estimating the concentration of benzene to which study groups were
exposed. As a result, the commenter concluded that the CAG risk factor
overstated the exposure risk by at least an order of magnitude.
Response: Response to this comment can be found in EPA-450/5-82-003,
which was prepared to address the listing of benzene under Section 112
of the Clean Air Act.
2.3.8 Exposure Assessment
A number of commenters (IV-D-1, IV-D-3, IV-D-4, IV-D-10, IV-D-lOa,
IV-D-13, IV-D-21, IV-F-1) stated that the model plant methodology used
by EPA overestimates risk from benzene exposure. The commenters suggested
that a more realistic and accurate risk estimate would be obtained using
actual plant emission data, actual population data, and available plant-
specific emission data.
One commenter (IV-D-10a) maintained that the Agency's benzene
emissions exposure analysis relied upon incomplete and inaccurate
meteorologic data. Rather than use site-specific climatological data as
required by the Agency guidelines, the commenter remarked that the
analysis relies entirely on conditions at the Gulf Coast to apply to
storage vessels throughout the nation.
According to the commenter, the EPA concedes that this assumption
causes an overstatement of estimated exposure, noting its data were
"representative of poor dispersion conditions in the area in order to
develop a potential worst-case situation". He concluded that since
climatological data for approximately 300 sites throughout the U.S. are
available in the Agency archives, the EPA's total reliance on Gulf Coast
meteorology was not justified. The commenter also stated that EPA
arbitrarily oriented the benzene storage vessels of a hypothetical
facility in order to maximize the ambient concentrations at the plant
boundary. According to the commenter, this was done despite the fact
that the actual storage vessels are not usually arranged in a straight
line configuration.
2-85
-------
The commenter further asserted that the EPA failed to validate the
results of its air quality modeling as a check on its accuracy, as
required by Agency guidelines. According to the commenter, in this
exposure analysis, the EPA repeatedly has relied upon unsupported
assumptions about emissions, meteorology, population distribution, and
other factors, even though accurate data were readily available. The
approach taken results in an unacceptably high degress of uncertainty in
the Agency's exposure estimates; in some instances, the exposure estimate
may be off by a factor of 100 or more.
Response: The EPA has revised its original risk assessment for
benzene storage vessels: the unit risk factor has been recalculated;
new emission estimates have been developed based on the new API data;
the meteorology of the area where each plant is located has been used;
and an improved population model (Human Exposure Model) has been used.
The EPA considered the option of using plant-specific data for all
parameters in order to run an exposure model for each plant. The EPA
compared the uncertainty that would result using the plant-specific data
approach with the uncertainty that would result using the model plant
and extrapolation approach. The EPA also compared the level of effort
that would be required to complete the two options.
The plant-specific approach probably would not improve the precision
or accuracy of the results enough to justify the level of effort required
to gather the input data. A plant-specific approach would entail using
"Section 114" letters to gather plant information on emissions, meteorology,
and plant configuration from about 130 plants. This would require
substantial effort from plant owners as well as from the EPA. The
dispersion and exposure models would then have to be run about 390 times,
at least three times for each plant. The resultant increase in precision
and accuracy would probably be small compared to the uncertainty still
remaining that is inherent in the dispersion and exposure models and in
the input data used. Both the Industrial Source Complex Long Term
computer model (ISCLT) and the Human Exposure Model, even with perfect
input data, are subject to substantial uncertainty. (The ISCLT model,
even with state-of-the-art input data, is estimated to have a 95 percent
2-86
-------
confidence interval of plus or minus a factor of two.) The plant-specific
input data would also exhibit wide variability and thus introduce
uncertainty in the results of the study.
The increase in precision and accuracy would not substantially
improve upon the precision and accuracy of the model plant extrapolation
method being used by the EPA. The Agency has not exaggerated the precision
of the results of the model plant extrapolation method, nor has the EPA
attempted to refine the results of the model plant extrapolation method
any more than is warranted by the quality of the data and the modeling
technique. Uncertainties are clearly delineated. The results are
presented in highly aggregate, nonspecific terms, in a fashion that
exhibits much less uncertainty than if the EPA tried to obtain more
detailed, refined results from the extrapolation. Using the model plant
extrapolation method, inaccurate deviations in the results for specific
plants tend to average out when the total national incidence is computed.
Because a plant-specific approach would be very costly and would not
substantially improve upon the precision and accuracy already achieved
by the model plant extrapolation" approach, the EPA has elected to use
the model plant extrapolation approach.
Comment: One commenter (IV-D-lOa) added that deficiencies exist in
the population concentration estimates contained in the exposure analysis.
According to the commenter, the EPA assumed that population is distributed
uniformly in all directions at each site, which introduces an uncertainty
factor of 10 to 100 into the overall exposure estimate.
Response: Response to the comment can be found in EPA-450/3-80-032b
and EPA-450/5-82-003.
Comment: A commenter (IV-D-lOa) stated the analysis failed to
account for population activity patterns and population mobility, thereby
overestimating exposure levels for persons residing in the affected area
surrounding these plants. He further criticized the EPA's estimate of
"maximum individual lifetime risk" by noting that the Agency has no
evidence that any individual ever lives an entire lifetime 0.1 kilometers
from the plant at a point of maximum benzene concentration.
2-87
-------
Response: Response to this comment can be found in EPA-450/5-82-003,
which was prepared to address the listing of benzene under Section 112
of the Clean Air Act and EPA-450/3-80-032b.
Comment: One commenter (IV-D-19) felt that EPA had understated the
risk of exposure to benzene storage emissions. According to one commenter,
the scientific knowledge necessary for reasonably reliable and precise
estimates of human cancer risks simply is not available. The commenter
felt that, given interactions and synergisms, it is much more likely
that exposure to multiple chemicals will have an additive or multiplicative
effect than that such chemicals will cancel each other out. This commenter
cited many sources of uncertainty in the risk assessment and concluded
that the EPA may have drastically understated the real leukemia risk
associated with benzene. According to the commenter, the estimates
given by the EPA may well underestimate the health benefits of the
increment between the proposed requirements and use of vapor recovery or
thermal destruction of emissions. He added that it is unacceptable that
the noncarcinogenic effects of benzene exposure have virtually dropped
out of the EPA's analysis due to the fact that they cannot be readily
quantified. According to the commenter, the proposal makes no efforts
to see that these effects get appropriate weight in the decision to stop
short of more stringent regulatory alternatives.
Response: Response to this comment can be found in EPA-450/5-82-003,
which was prepared to address the listing of benzene under Section 112
of the Clean Air Act and EPA-450/3-80-032b.
Comment: A commenter (IV-D-19) noted that EPA assumed that many
benzene-emitting facilities have a life expectancy of 50 years or more.
Yet the quantifications of risk used to compare the proposed approach
with a more protective one assume a 20-year lifetime. According to the
commenter, this understates the number of benzene victims for such
facilities by two and one-half times or more, and reconsideration of the
decision not to adopt Alternatives IV or V with the appropriate health
effects timeframe may lead to a different decision (IV-D-31:5-6).
Response: Twenty years is an average figure for the lifetime of a
plant. Some plants have a life expectancy of 50 years; some have lifetimes
2-88
-------
shorter than 20 years. Roughly, a plant will have a lifetime of 20 years.
Since there are little data available that estimate plant lifetimes, the
EPA considers 20 years to be a reasonable estimate. However, EPA agrees
there is uncertainty associated with this number.
2.3.9 Risk Methodology Consistency
Comment: A number of commenters (IV-D-lOa, IV-D-4, IV-F-1) stated
that there should be some consistency in risk assessment methodologies
between the four current benzene proposals. One commenter stated that
if benzene is to be regulated by a NESHAP standard, the emission concerns
and risk/benefit analysis should be completed for all types of emissions
(e.g., process emissions, storage tank emissions, fugitive emissions,
etc.) simultaneously. This integrated analysis, the commenter maintained,
would prevent duplication of effort, errors, or inconsistencies and
result in an overall analysis of the risk/benefit of a product. According
to another commenter's (IV-D-21) review of the four current benzene
proposals, a great deal of duplication has occurred with little or no
health benefit to the public.
Response: The risk assessment methodologies used for the four
benzene standards have been made more consistent. The only area in
which they are different is that the affected industries voluntarily
submitted detailed plant-specific information on the maleic anhydride
and ethyl benzene plants. Obtaining this kind of information for the
143 or more plants that have benzene fugitive and storage sources would
be too costly (considering the uncertainty of the final results either
way) for the industry or the EPA to obtain. Because of the detailed
information available on maleic anhydride and ethylbenzene/styrene
plants and the relatively small number of these plants, the more precise
ISC dispersion model was used for all those plants and the SAI model was
used for the benzene storage and benzene fugitive plants. However, the
ISC model was used for a few plants with benzene storage and fugitive
emissions to compare the results of that model with the SAI model. For
plants containing multiple sources, the same meteorological and population
data were used for each plant each time the risks were calculated for
one of the sources in that plant.
2-89
-------
2.4 INSPECTION, REPORTING AND REPAIR REQUIREMENTS
2.4.1 Inspection Requirements
2.4.1.1 Quarterly Visual Inspection.
Comment: Six commenters (IV-D-5, IV-D-6, IV-D-10, IV-D-15, IV-D-19,
IV-D-21) questioned the 3-month inspection interval. One commenter
(IV-D-6) stated that inspections on a 6-month rather than a 3-month
interval should be adequate because of the expected slow deterioration
rate of roof seals. Other commenters (IV-D-6, IV-D-7, IV-D-lOa, IV-D-14)
stated that a 3-month inspection interval was unnecessary or of questionable
effectiveness but offered no rationale for this opinion. Three commenters
(IV-D-5, IV-D-15, IV-D-21) suggested that annual inspections would be
adequate for detection of major failures. One of these three commenters
(IV-D-15) indicated that the annual inspection interval could be shortened
if numerous storage vessel problems are observed.
Response: The intent of the proposed 3-month inspections was to
detect and repair seal gaps as well as major failures (e.g., roof sinking
or hangups, seal detachment) as rapidly as possible. Seal gaps, however,
are the most likely problem and will occur more frequently than the
major failures. In order to correct any of these problems, the storage
vessel must be drained and degassed to permit vessel entry and repair.
Emissions occur during this degassing/repair operation. The degassing
emissions resulting from the correction of a seal gap may exceed the
emissions resulting from a seal gap. The degassing emissions for repair
of a gap of 3-square inches per foot of diameter for a typical vessel
(based upon American Petroleum Institute data) are equivalent to
approximately 7 or 10 years of unrepaired gap emissions for liquid-mounted
or vapor-mounted seals, respectively. The emissions for major failures,
however, are much larger than those for seal gaps. In addition, the
frequency of such major failures is much smaller than the frequency of
seal gaps.
The promulgated standard requires immediate repair of major failures
as at proposal. Because degassing and repair of seal gaps causes more
emissions than the seal gaps themselves, however, the repair of seal
gaps will be made only during internal inspections or during repair of
2-90
-------
major failures when tanks are being degassed for other reasons. Inspections
as frequently as quarterly would not reduce emissions due to the detection
of seal gaps and would be unnecessarily frequent for the less likely
major failures. Industry experience tends to indicate that such major
failures may occur on a time frame of years rather than months.
Consequently, for existing storage vessels equipped with internal
floating roofs, the initial visual inspection is to be performed within
90 days of the effective date, and visual inspections are to be made
annually thereafter. For new storage vessels, or for existing storage
vessels without an internal floating roof, an initial visual inspection
of the newly installed internal floating roof will be made prior to
filling the storage vessel with benzene. This inspection will ensure
that the controls have been properly installed, that the seals are free
of defects such as holes or tears, that the internal floating roof has
no defects, and that the fittings meet BDT specifications. Visual
inspection of the controls from the fixed roof will be made annually
thereafter.
Comment: Three commenters '(IV-D-5, IV-D-lOa, IV-D-15) questioned
whether the required visual inspections of the floating roofs and seals
could be performed safely. These commenters believed that the proposed
requirement to open a manhole to visually inspect the internal floating
roof and seals was unreasonable because this could only be done safely
when the VOC concentration inside the tank was well below the lower
explosive limit. They doubted that the pan-type, ventilated and covered,
internal floating roof technology proposed by the EPA would accomplish
this at all times because of its tendency to sink.
Response: Neither the proposed or the promulgated standards require
steel pan type internal floating roofs. The promulgated standards allow
any type of internal floating roof (noncontact or contact). Noncontact
internal floating roofs have an inherent buoyancy that makes them difficult
to sink. Several types of contact internal floating roofs are more
resistant to sinking than the steel pan. These designs include honeycomb
panels, double deck, pontoon, and bulkhead type decks. These designs
are all difficult to sink.
2-91
-------
Use of any of these roof types will prevent a free liquid surface
from occurring inside the tank and thus prevent explosive concentrations
from occurring inside the tank.
Comment: Three commenters (IV-D-8, IV-D-lOa, IV-D-15) also noted
that the visual inspections, to be effective, would require vessel entry
and consequently the use of full-face masks, auxiliary lights, and
visual aids. They noted that not only would the lights cause additional
safety problems, but the need for masks and OSHA regulations, which
prevent sticking one's head in the tank, would limit the useful information
gained.
One commenter (IV-D-8) stated that the requirement for an inspector
to enter the tank for visual inspections was unsafe. He further stated
that even though fully suited with protective equipment, the risk involved
in having an inspector descend onto a floating roof was unwarranted. He
felt that visual inspections from the top of the tank would be appropriate
and that EPA should delete requirements to enter the tank. Another
commenter (IV-D-10) recommended that in-service inspections be limited
to those that would not require inspectors to enter the tank but rather
to observe through open hatches.
Response: The information gained by looking into a storage vessel
from outside of a hatch is admittedly limited, but it is adequate for
ascertaining the occurrence of a catastrophic failure such as the sinking
or hangup of a roof or the detachment of a seal. The use of auxiliary
lights, if employed outside rather than inside the vessel, are anticipated
to pose no safety problem.
Both the proposed and final regulations do not require tank entry
for the visual inspections. Observation is to be conducted externally
from the fixed roof hatch. Correctly performed from outside the vessel,
the visual inspection will provide the necessary information on catastrophic
failure at no occupational hazard to personnel.
Comment: One commenter (IV-D-19) stated that the 3-month visual
inspection was solely based up industry practice and should be required
more frequently. He suggested that an initial inspection be required
"soon" after the storage vessel has been filled and a second time "after
2-92
-------
the roof has had a chance to rise and fall a couple of times." In
addition, the use of visual aids, e.g. binoculars, bright lights, should
be required.
Response: The visual inspection is intended to identify the occurrence
of major failures. The frequency of such an event is low. In a documented
example, a benzene storage vessel incurred no event that required vessel
entry for 12 years. This is indicative at least of the order of magnitude
of the frequency for such major failures. The final rule will require a
reasonable inspection interval consistent with this information: initial
visual inspection within three months and subsequent annual visual
inspections. Because a typical storage vessel will have approximately
12 turnovers in a 3-month period, ample time is made available to observe
any obvious problems in the first inspection. The use of such equipment
as bright lights and binoculars are unnecessary to observe such catastrophic
failures as are of interest in the visual inspection.
2.4.1.2 Five Year Internal Inspection.
Comment: Five commenters (IV-D-5, IV-D-8, IV-D-lOa, IV-D-17,
IV-D-19) questioned the requirement for a five-year internal inspection
of the required control equipment. Another commenter (IV-D-21), however,
agreed that the internal inspection interval should be five years. One
commenter (IV-D-5) suggested elimination of the 5-year inspection without
a supporting rationale. One commenter (IV-D-8) stated that internal
inspections are necessary only when sludge accumulates and/or mechanical
repairs are made and that tanks in benzene service do not experience
enough mechanical work or sludge build-up to warrant an internal inspection
once every five years. This commenter suggested that a 10-year inspection
interval would be more appropriate. Another commenter (IV-D-17) stated
that the internal inspection interval should be extended to once every
10 years because of the difficulties inherent in taking tanks out of
service and switching products to other tanks.
A commenter (IV-D-lOa) stated that seal lifetimes were longer than
those estimated by the EPA. The commenter stated that seal failures are
caused by ultraviolet degradation of the seal material, the presence of
debris that punctures the seal (usually due to poor maintenance of the
2-93
-------
weather shields), abrasion of the seal due to out-of-roundness, and
product incompatibility. He further stated that the proposed requirement
of a fixed roof over a floating roof would extend seal lifetime because
the presence of a fixed roof will minimize ultraviolet degradation,
because the fixed roof will minimize weather-induced rust (the main
source of debris causing seal damage), and because the presence of a
fixed-roof will improve the roundness of the tank as compared to an
external floating roof tank. The commenter concluded that the final
rule should require the internal inspection whenever the tank is emptied
and degassed, but at least once every ten years.
The final commenter felt that the five-year period was based on
current industry practice and was unreasonably long for large facilities
(IV-D-19). The commenter recommended that large facilities could have
one centralized stand-by tank connected by piping to the storage tanks.
He felt that this configuration would permit transfer of benzene to the
stand-by tank and would allow more frequent inspections, at perhaps
annual intervals.
Response: The main purpose for the 5-year internal inspection was
to allow for the inspection of the primary seal that could not be inspected
during the annual visual inspection due to the presence of the required
secondary seal. The promulgated standard, however, does not require a
secondary seal, and therefore, the primary seal can be inspected during
the annual inspections. As already explained, the storage vessel will
be required to be degassed only when major failures occur. However,
internal inspections do have other advantages other than the opportunity
to repair major failures. Internal inspections also allow for the
inspection of the underside of the primary seal for deterioration, the
inspection of the portion of the primary seal farthest from the roof
hatch that might be difficult to see during the annual inspection, and
of course, for repair of seal gaps. Commenters have indicated that
tanks are generally degassed on the average of once every ten years for
inspection as standard practice. Therefore, if owners or operators were
required to perform internal inspections on their tanks at least once
every ten years, this requirement would cause no additional degassings
2-94
-------
of the tank, and hence no additional emissions. Consequently, since
there are advantages to performing internal inspections on an internal
floating roof storage vessel and since they are inspected routinely on
the average of every ten years, the promulgated standard requires an
internal inspection of all internal floating roof storage vessels at
least once every ten years. If a vessel is drained and degassed to
repair a catastrophic failure, this can be substituted for the 10-year
inspection and another internal inspection will not be required for
another ten years.
2.4.2 Reporting Requirements
2.4.2.1 Quarterly Reports Submission.
Comment: Six commenters (IV-F-1, IV-D-5, IV-D-6, IV-D-7, IV-D-12,
IV-D-15) questioned whether the quarterly reporting requirements in the
proposed standards were relevant. An additional commenter (IV-D-lOa)
stated that the requirement appears to mandate reporting merely for the
sake of reporting. He suggested that the EPA is requiring quarterly
reports either for ease of enforcement or for mere data collecting
purposes. Four commenters (IV-D-5, IV-D-7, IV-D-lOa, IV-D-12) recommended
that the EPA revise the proposed standards by only requiring that the
owner or operator of each tank maintain a report of each inspection on
file for a specific period of time. Under such a procedure, industry
would only be required to report excursions from the standard and to
certify, under penalty of law, compliance with the provisions. If a
source complied with the 30-day repair period procedure, no reporting
would be required to certify compliance. Only in the event of a delay
of repairs in excess of 30 days would a report be required stating
(1) the reason or reasons for the delay, and (2) when the repair or
repairs will be completed.
Two commenters (IV-D-6, IV-D-15) stated that quarterly reports
should not be required because annual reports would be fully adequate to
report the results of inspections.
Response: Because the visual inspection in the final standards are
required annually, reporting the results of the visual inspection will
also be on an annual, rather than a quarterly, basis.
2-95
-------
Reports provide documentation that the inspection was performed and
thus are an important method of determining compliance with the standards.
Also, the reporting provisions are not burdensome (about 2 labor hours
per year), especially when compared with recordkeeping (1 labor hour per
year). Therefore, reports are required after each annual visual inspection.
2.4.2.2 30-day Notice Prior to Refill.
Comment: Two commenters (IV-D-5, IV-D-21) stated the 30-day notice
prior to refilling the vessel with benzene was an unnecessary additional
regulatory burden.
Response: An internal inspection assures correct installation,
repair, and maintenance of the required controls by providing the only
directly observable means of enforcement verification for the Agency.
The proposed requirement for a 30-day notice before refilling the vessel
is intended to ensure the opportunity for an internal inspection by the
EPA during those infrequent times when the vessel may be entered,
i.e., following initial installation, catastrophic events repair
(unscheduled internal inspections), or scheduled 10-year internal
!
inspections.
2.4.2.3 30-day Notice of Internal Inspection.
Comment: One commenter (IV-D-10) stated that the 30-day notice
prior to refilling the vessel is adequate for any scheduled inspection
but inadequate for unscheduled inspections (e.g., low inventory due to
plant shutdowns, shipping delays, or strikes). He suggested that this
requirement is punative and should be deleted.
Response: As previously explained, the 30-day notice prior to
refilling the vessel with benzene will afford the Administrator the
opportunity to have an observer present to ascertain the condition of
the control equipment before the vessel is refilled with benzene. This
is reasonable when the internal inspection takes place as scheduled in
the tenth year.
However, the EPA agrees that there are instances in which the
internal inspection may take place early. An unplanned plant shutdown
or other event may provide the owner or operator a convenient time to
2-96
-------
inspect and, if necessary, to repair. To require that the vessel remain
empty for possible 30 days prior to refill in these situations may deter
owners and operators from inspecting the control equipment. This is
because in many cases, requiring the tank to remain empty may require
that the plant be shutdown for the 30 days. The costs of this would be
so punitive as that owners or operators would elect not to empty and
degass.
However, the inspection and repair of control equipment whenever
possible is desirable. To alleviate the burden of the 30-day notice of
refill in those unscheduled cases, the promulgated standards allow for a
shorter notification period. If the internal inspection is performed
prior to the tenth year, the owner and operator shall notify the
Administrator in writing and by express mail at least seven days prior
to refill and include documentation demonstrating why 30-days notice
could not have been given.
2.4.3 Repair Requirements
2.4.3.1 30-day Repair or Empty Requirement.
Comment: Two commenters (IV-D-7, IV-D-14) expressed the opinion
that a very minor defect would create fewer emissions than the repair of
the defect. In addition, one of these (IV-D-14) stated that the shutdowns
and startups of the 30-day repair requirement would incur significant
(but unenumerated) incremental energy costs. The other of these (IV-D-7)
suggests some discretionary latitude for operators. Another commenter
(IV-D-10) suggested repairs for major defects only.
Response: The EPA agrees with the commenters that the correction
of a minor defect, such as a seal gap, may incur significantly larger
emissions in vessel degassing and defect repair than in the unrepaired
defect. The intent of the final standard is to identify and correct
immediately only major failures, where failure emissions are larger than
those of degassing/repair. The magnitude of emissions from major failures
are so large as to require immediate correction. The startup/shutdown
emissions incurred in this repair are small when compared to the emissions
due to a major failure. The requirement to visually inspect the control
equipment may detect major failures (such as seal detachment) that would
2-97
-------
have gone undetected in the absence of standards. Therefore, the 30-day
empty or repair requirement will cause the loss of the benzene that is
emitted during the emptying and degassing process. This will be a cost
(in energy or dollars) to the owner or operator. However, that cost
will be more than offset by the benzene emissions saved through the use
of controls that has not undergone a major failure. Therefore, the next
effect of the final standard is a savings (in terms of energy or cost)
through the repair of major failures.
In the final regulations, if a major failure is detected in the
annual visual inspection, the repair and the 10-year internal inspection
may be completed at the same time. The next internal inspection can
then be rescheduled for 10 years from the repair date.
Comment: Two commenters (IV-D-14, IV-D-15) stated that spare
storage vessels are unavailable. They assert that the 30-day repair or
empty requirement will increase the number of shutdowns and the necessary
number of acceptable standby storage vessels. One commenter (IV-D-15)
suggested that a variance be granted where the "normal rate of removal"
will not empty the storage vessel in 30 days or less. Another commenter
(IV-D-10) recommends that 90 days is a more reasonable time period to
find alternate storage vessels and to clean the vessel before repair.
Yet another commenter (IV-D-9) stated that 45 to 60 days was "reasonable,"
but he did not specify the reasons for this assessment.
Commenter IV-D-19 states that at large facilities spare storage is
generally available, and, therefore, the 30-day repair requirement
should be reduced to 5 days for large facilities. If no spare storage
is available, the facility could request an extension or certify and
report to the EPA that alternate storage is unavailable.
Response: A survey of benzene storage facilities (Docket Number 11-67
through 11-70) indicates that most facilities could empty a storage
vessel within 30 days. The daily emissions due to an unrepaired
catastrophic event are large. Because, as shown in the cited survey,
most operators are capable of emptying the benzene storage vessel within
30 days, there is no reason to permit additional days of large emissions
beyond 30 days. Because a repair or empty period of less than 30 days
2-98
-------
may be unachievable by many operators, the final regulations maintain
the 30-day requirement. In special circumstances where alternate storage
cannot be found, the operator must certify the unavailability of alternate
storage and may apply to the Administrator for a waiver from the 30-day
repair or empty requirement.
2.5 GENERAL ISSUES
2.5.1 90-Day Compliance
Comment: Several commenters (IV-D-6, II-D-7, II-D-9) characterized
the proposed standards requirement for compliance within 90 days as
impossible to achieve within the specified time and as unreasonable.
Two commenters (IV-D-6, II-D-17) stated that procurement and installation
of retrofit equipment requires 9 to 12 months; they suggested the Agency
allow 12 months for compliance before a waiver must be obtained. Another
commenter (IV-D-9), citing the same difficulties, suggested 2 years as a
compliance period. A third commenter (IV-D-7) noted that taking each
tank out of service was economically infeasible and incurred difficulties
in maintaining continuous production. He suggests compliance over
3 years as vessels are taken out of service for maintenance and repairs.
A commenter (IV-D-15) cited the compliance time as "generally
inadequate" and suggested that explicit rules for obtaining a variance
(waiver) be included in the preamble and that the EPA be prepared to act
swiftly on variance requests. Another commenter (IV-D-12) noted that it
is apparent that a majority of the facilities subject to the standards
will be required to apply for a waiver, and such a procedure is
inappropriate where large numbers of facilities are not in compliance.
He suggested that the 90-day compliance requirement be deleted and that
the facilities be required instead to submit compliance schedules to the
EPA for review and approval within a specified time.
Finally, a commenter (IV-H-1) stated that sources that had installed
controls for the Occupational Safety and Health Act or State regulations
should be given a "grace period". He recommends that compliance with
the benzene storage standards be achieved within 10 years of the existing
controls installation.
2-99
-------
Response: The commenters are correct in asserting that many existing
facilities would not have been able to install the controls required by
the proposed standards within the 90-day time period. However, the
90-day compliance period for existing sources does not result from a
determination made by the Administrator, but is stipulated as a provision
of the Clean Air Act, as amended (1977) (hereafter referred to as CAA).
Section 112(c)(l)(B) of the CAA states:
"(B) no air pollutant to which such standard applies
may be emitted from any stationary source in violation
of such standard, except that it the case of an
existing source-
(i) such standard shall not apply until
90 days after its effective date, and
(ii) the Administrator may grant a waiver
permitting such source a period of up to
two years after the effective date of a
standard to comply with the standard, if
he finds that such period is necessary for
the installation of controls and that
steps will be taken during the period of
the waiver to assure that the health of
persons will be protected from imminent
endangerment."
A waiver may be obtained by the submission of a written request to
the Director of the appropriate EPA Region Enforcement Division. The
EPA will, within the limits of its resources, process and act on requests
for waivers as quickly as possible. The information to be included in a
waiver request is contained in the General Provisions to Part 61.
Many of the sources already have the controls required by the
standard. A waiver is not needed for those sources and would not be
appropriate. The time required to install controls on the other sources
depends on the extent of control they already have and plant-specific
characteristics. A blanket waiver with one set time period would not be
appropriate for these sources.
Furthermore, the number of existing facilities requiring a waiver
to achieve compliance with the final standards is significantly smaller
than the number that would require a waiver under the proposed standards.
2-100
-------
Under the proposed standards, almost all of the 500 existing facilities
would have required a waiver. The estimated number of existing facilities
requiring a retrofit/waiver under the final standards is 65-100 facilities.
Thus, while the compliance period by statute may not be lengthened and
waivers will very probably be necessary, the number of the waiver
applications under the final standards will be much smaller than those
under the proposed standards.
2.5.2 Flares, Vapor Recovery
2.5.2.1 Flares.
Comment: One commenter (IV-D-lOa) objected to the EPA's determination
that flares are only 60% efficient. He asserted that flare efficiency
is higher and should, therefore, be adequate as a control technology for
benzene emissions. No data were presented to support this assertion.
Response: A study of flares (completed after the benzene NESHAP
proposal) by the Chemical Manufacturers Association (CMA) in collaboration
with the EPA indicates that the commenter is correct in his assessment
of flare efficiency: properly operated flares meeting certain
specifications achieve an efficiency of 98% (docket item no. ).
The Agency has, therefore, revised the standard to allow flares meeting
certain specifications.
2.5.2.2 Vapor Recovery.
Comment: The commenter (IV-D-6) has pointed out that in a closed
vent system and control device that handles vapors from facilities
affected by the NESHAP and other emission sources, it may not be possible
to ascertain if the closed vent system and control device is 95 percent
efficient in reducing emissions from the affected facilities (benzene
storage tanks).
Response: The final standards allow the owner or operator to
control emissions through the use of a closed vent system and 95 percent
effective, by weight, control device. The final standards are design
standards and, as such, do not require measurement of the 95 percent
control. The 95 percent control is to be documented through design and
operating specifications.
2-101
-------
The control effectiveness of many control devices is dependent upon
the specific vapors entering the control device. For example, a condenser
that is 95 percent effective in reducing perc chloroethylene emissions
may not be 95 percent effective in reducing benzene emissions.
Additionally, the fact that a control device receives vapors from other
sources may affect the ability of the control device to reduce benzene
emissions by 95 percent. For example, a condenser receiving vapors from
more than one source may have a reduced efficiency as far as benzene is
concerned because of increased loading. Because of these factors, the
final standards require that benzene emissions from designated sources
be reduced by 95 percent in light of the fact that the control device
receives vapors from nondesignated sources. Therefore, if the control
device receives vapors from nondesignated sources, the owner or operator
must demonstrate to the Administrator's satisfaction that the control
will achieve 95 percent control of benzene emissions from benzene storage
vessels despite increases in flow and other factors.
However, thermal oxidation devices are known with specifications
that will allow reduction of all VOC vapors by 98 percent and that
would, therefore, meet the 95 percent control of benzene emissions
required by these standards even in combined vent cases. Recent tests
have demonstrated smokeless steam-assisted, air-assisted and nonassisted
flares can achieve 98 percent control over a broad range of vapor types
if the heat content of the flared gas was maintained above 7.45 MJ/scm
(200 Btu/scf) or 11.2 MJ/scm (300 Btu/scf) (depending upon flare design),
and if the exit velocity of the flare is less than 18 m/sec. Monitoring
provisions for flares are provided in the regulation. An enclosed
combustion device with a minimum residence time of 0.75 seconds and a
minimum temperature of 816°C will also provide 98 percent control.
Documentation that these conditions exist is sufficient to meet the
requirements of these standards.
Comment: One commenter (IV-F-1) requested that the EPA eliminate
the "prejudice" against closed vent systems and control devices by
allowing downtime for both malfunction and preventive maintenance of the
system. With regard to the abatement level, he requested that the
2-102
-------
percent emissions reduction of a system be calculated on an annual
average basis. This, he suggested, would allow a company to "bank"
downtime hours by operating at a higher efficiency when the system was
functioning to affect those times when the system was being repaired.
Such a strategy could achieve an 95 percent efficiency on an annual
basis without requiring two totally redundant vapor recovery systems.
Response: The Agency agrees that vapor recovery systems will
experience unavoidable failures (malfunctions). Unlike many processes
that can be shut down during malfunction, emissions from a storage
vessel cannot be "turned off." The storage vessel will continue to emit
regardless of the availability of the control technology, in this case a
vapor recovery unit. Transferring the benzene from a vessel attached to
a malfunctioning vapor recovery unit will also result in emissions.
Additionally, in the case of integrated plant wide vapor recovery system
that handles vapors from more than one tank, there may be no controlled
tanks in which to store the benzene.
The Agency examined requiring redundant systems. Such a requirement
would require owners and operators to construct two totally independent
vapor recovery systems. This would double the cost of vapor recovery
systems to control only those emissions during malfunction of one system.
At proposal, the Agency determined that the costs of requiring vapor
recovery was not reasonable and has determined that requiring owners and
operators of existing systems to build another system to control excess
emissions would also have unreasonable costs. The incremental cost
effectiveness of this requirement would vary according to the number of
malfunctions but would generally exceed several hundreds of thousands of
dollars per Mg. This is far in excess of the beyond-BDT requirements
that were rejected previously and, therefore, has been rejected as a
strategy to control excess emissions. Because there is no other means
of controlling excess emissions and because excess emissions cannot be
avoided, the Agency has decided to allow these emissions. Provisions
have been added to the final rule that allow excess emissions during
malfunctions. The owner will have to demonstrate that the control
system failure is unavoidable, i.e. that it meets the definition of
malfunction. In addition, the malfunction must be repaired as soon as
possible, and emissions must be minimized during the malfunction.
2-103
-------
Unlike a malfunction, routine maintenance on a control system is
planned ahead of time. Control system maintenance can be planned so
that it occurs when the plant is shut down for maintenance. Even then,
however, the storage vessels can not be expected to contain no benzene,
and storage vessels containing benzene emit benzene whether the plant is
operating or not. Since tanks controlled by a control device are usually
tied into one device, there would be nowhere else to put the benzene
where the emissions would be controlled during the maintenance operation.
Therefore, the EPA decided that a provision is needed to allow storage
vessels to be uncontrolled during maintenance. However, the provision
is structured in such a way to minimize emissions.
First, the regulation requires that an operation and maintenance
plan be submitted for approval along with other information for plants
that are meeting the standard with a control device. That operation and
maintenance plan will specifiy the number of days a year the device will
be down for maintenance. The regulation requires that the plant owner
adhere to the operation and maintenance plan.
Secondly, the regulation requires that a static level of benzene be
maintained in the tank during the maintenance. Emissions from storage
vessels at a given point are heavily dependent upon changes of the
liquid level in the tank. For example, consider a large (1 million gallon)
fixed roof benzene storage tank. If the liquid level is held static,
losses from breathing during a 24-hour period would be about 2.5 kg. If
the liquid level is raised by adding 100,000 gallons (10 percent of the
tank volume), losses during that period would be about 130 kg. As this
emissions comparison points out, in most fixed roof tanks breathing
losses are very small compared to working losses. As long as the liquid
level in the tank is not raised, emissions will be small. To assure
compliance, the final rule requires that during periods of downtime, a
record of the liquid level in the tank be kept.
2.5.2.3 Equivalence.
Comment: Three commenters (IV-D-lOa, IV-D-12, IV-D-21) discussed
the equivalence procedures. One (IV-D-lOa) cited them as unreasonable.
All three commenters suggested that the EPA produce a comprehensive
2-104
-------
listing of proven technologies that are equivalent controls for the
benzene storage vessel NESHAP. An owner/operator would then only notify
the EPA of the intended use of a "listed" technology. The commenter
noted that this would prevent "reinventing the wheel" on each equivalence
determination. He further objects to the use of 95% efficiency as the
benchmark for choice of equivalent technology. This is premised upon
his comment that vapor recovery efficiency varies with vessel size.
Another commenter (IV-D-21) further requests a simplification of
the demonstration of equivalence procedure to permit utilization of
laboratory results together with engineering evaluations. This is
intended to promote innovative technology development.
A third commenter (IV-D-12) states that the requirement for approval
of alternate technologies for each facility will impose unnecessary
administrative burdens on the EPA and will cause intolerable delays in
industry. Agreeing with the other commenters concerning the need for a
comprehensive list of equivalent technologies, he further notes that
these technologies should include vapor recovery systems, vapor
condensation, and carbon adsorption, and that any of these should be
implementable without prior EPA approval.
Response: Where possible, the EPA will evaluate the technical
parameters of control technologies rather than individual trade-name
equipment for equivalence. Thus, any equipment possessing the same
technical parameters (no matter who produces the equipment) would be
considered equivalent. The benzene standards' listing of control
technologies is comprehensive in that all floating roof and seal control
technologies currently known to the EPA are listed. In the event that
innovative control technologies are developed, however, only one equivalence
determination for each new technology is required provided that the
technology will operate the same regardless of the conditions at a
particular source. Upon the EPA's finding of equivalence, the new
technology is added to standards list and is, therefore, automatically
available for use by other owners/operators without additional (and
redundant) equivalence determinations.
2-105
-------
Vapor recovery systems can vary widely in both emissions reduction
efficiency and design. The technologies mentioned by the commenter can
achieve the required efficiency if properly designed and operated.
However, if these systems are not properly designed to handle the maximum
anticipated flow rate or other key parameters, they will not be capable
of achieving the requisite 95 percent emissions reduction. Also, the
system must be properly operated. For example, carbon beds must be
desorbed before reaching breakthrough or they will not continue to
reduce emissions. For these reasons unlike floating roof and seal
technologies, the Agency cannot issue a blanket equivalency determination
for vapor recovery units.
2.5.3 Economic and Cost/Benefit Analyses
Comment: Four commenters (IV-D-1, IV-14, IV-D-15, and IV-D-21)
were concerned about the accuracy of the economic analysis completed by
the EPA. One (IV-D-1) simply stated that the accuracy of the cost
estimates and their economic impacts were questionable. The remaining
three stated that the true costs for their benzene storage vessels were
underestimated; each commenter Suggested a different cost underestimation
factor: 3-4 (IV-D-21), 5 (IV-D-14), and 7.5 (IV-D-15). Another commenter
(IV-D-20) found the EPA economic analysis methodology basically sound.
One commenter (IV-D-2) stated that the standards required expensive
replacements for no air quality improvement. Another commenter (IV-D-4)
claimed that the EPA failed (1) to conduct an adequate costs/benefit
analysis and (2) to choose the most cost effective alternative, as
required by Executive Order 12291.
A commenter (IV-D-18) that operates a number of benzene storage
vessels informs the EPA that his company was not listed as a benzene
consuming facility. He implies that at the 1,000 gallon cutoff (4 m3)
there are many more than the EPA-estimated 77 consuming facilities on a
nationwide basis and concludes, therefore, that the EPA nationwide
economic impacts are underestimated.
Response: As discussed in the section on the cost of controls, the
latest EPA cost estimates (fourth quarter 1982) are based upon vendor
data collected from 15 companies. The Agency has no reason to believe
2-106
-------
that these underlying vendor data or the cost estimates based upon them,
are underestimated.
It is estimated that 100 vessels or less will be required to retrofit
control technology under the final standards. Thus, the number of
replacements will be significantly fewer than the 500 retrofits that
would have been required under the proposed standards. In deciding what
controls to require for each type of existing tank, the EPA considered
the retrofit costs of the controls. In selecting BDT, the EPA examined
the incremental costs and emission reduction for each type of control
applied to each type of tank. In each case, the EPA selected for BDT
that control option that gets the most emission reduction for a reasonable
cost. The cost and health benefits of a level of control more stringent
than BDT were also examined, but the final standard is based on BDT.
The Agency acknowledges that indeed there may be more benzene
storage vessels that were not included in nationwide estimate of affected
facilities. However, the decisions about the level of standard (control
equipment required) and the size cut-off were not based on nationwide
impact numbers, but on individual tank considerations. The decisions
about the standard would not change regardless of the actual number of
tanks.
2-107
-------
APPENDIX C
METHODOLOGY FOR ESTIMATING LEUKEMIA INCIDENCE AND
MAXIMUM LIFETIME RISK FROM EXPOSURE TO
BENZENE STORAGE TANKS
C-l
-------
APPENDIX C
METHODOLOGY FOR ESTIMATING LEUKEMIA INCIDENCE AND MAXIMUM
LIFETIME RISK FROM EXPOSURE TO BENZENE STORAGE TANKS
C.I INTRODUCTION
The purpose of this appendix is to describe the methodology and to
provide the information used to estimate leukemia incidence and maximum
lifetime risk from population exposure to benzene emissions from benzene
storage tanks. The methodology consists of four major components: estimation
of annual average concentration patterns of benzene in the region surrounding
each plant, estimation of the population exposed to each computed concentration,
calculation of dosage by summing the products of the concentrations and
associated populations, and calculation of annual leukemia incidence and
maximum lifetime risk from the concentration and dosage estimates and a
health effects estimate represented by a unit risk factor. Due to the
assumptions made in each of these four steps of the methodology, there is
considerable uncertainty associated with the lifetime individual risk and
leukemia incidence numbers calculated in this appendix. These uncertainties
are explained in Section C.6 of this appendix. A description of the health
effects and derivation of the unit risk factor for benzene is not included
in this appendix; however, they are discussed in EPA docket number OAQPS 79-3
and Response to Public Comments on EPA's Listing of Benzene Under Section 112
and Relevant Procedures for the Regulation of Hazardous Air Pollutants.
EPA-450/5-82-003.
C-2
-------
C.2 ATMOSPHERIC DISPERSION MODELING AND PLANT EMISSION RATES
The Human Exposure Model (HEM) was used to estimate concentrations of
benzene around approximately 126 plants that contain benzene storage tanks.
The HEM estimates the annual average ground-level concentrations resulting
from emissions from point sources. The dispersion model within the HEM is
a Gaussian model that uses the same basic dispersion algorithm as the
2
climatological form of EPA's Climatological Dispersion Model. Gaussian
concentration files are used in conjunction with STAR data and emissions
data to estimate annual average concentrations. Details on this aspect of
the HEM can be found in Reference 1.
Seasonal or annual stability array (STAR) summaries are principal
meteorological input to the HEM dispersion model. STAR data are standard
climatological frequence-of-occurrence summaries formulated for use in EPA
!
models and available for major U.S. sites from the National Climatic Center,
Asheville, N.C. A STAR summary is a joint frequency of occurrence of wind
speed stability and wind direction categories, classified according to the
Pasquill stability categories. For this modeling analysis, annual STAR
summaries were used.
The model receptor grid consists of 10 downwind distances located
along 16 radials. The radials are separated by 22.5-degree intervals
beginning with 0.0 degrees and proceeding clockwise to 337.5 degrees. The
10 downwind distances for each radial are 0.2, 0.3, 0.5, 0.7, 1.0, 2.0,
5.0, 10.0, 15.0, and 20.0 kilometers. The center of the receptor grid for
each plant was assumed to be the center as determined by review of maps.
C-3
-------
Inputs to the dispersion model include the geographical coordinates
for each plant, and the emission rates, dimensions and plume characteristics
for each storage tank in each plant. The latitudes and longitude for each
plant, used in selecting the STAR site, are listed in Table C-l. Four
model units representing the different types of plants that would have
benzene storage tanks were developed: large producers of benzene, small
producers of benzene, benzene consumers, and bulk storage terminals. The
model units were assigned to each plant according to the uses of benzene
within the plant. Where a plant had two model units assigned to it
(e.g., a plant may be both a producer and consumer of benzene), emissions
from both model units were used in calculating the concentration pattern
around the plant. The model units assigned to each plant are listed in
Table C-l.
Each model unit consists of a set of benzene storage tanks with specified
dimensions, roof types, turnovers, and emission rates. The tank parameters
used in the dispersion model are the same for benzene consumers and bulk
storage terminals; therefore, no differentiation was made between them for
modeling purposes. Table C-2 shows, for each model unit, the height and
vertical cross-sectional area (used in downwash calculations) of each tank.
The table also shows the emissions from each tank for the three levels of
control: baseline (current level), best demonstrated technology (BDT), and
a more stringent level (beyond BDT). Because the emissions vary among the
three alternatives according to changes in the type of roof seal, the roof
types have also been listed in Table C-2. Emissions from all the tanks
were assumed to be at ambient temperature, which the model assigns as
293°Kelvin. Because the gas exit velocity is negligible, it was assumed
C-4
-------
lable U-i. HLANIb ANU LUtMllUNb hUK BtN^tNb blUKAUt
TANKS
Coordinates
Plant
Region II
1. American Cyanamid
2. DuPont
3. Exxon
4. Standard Chlorine
5. Texaco
6. Ashland Oil
7. ICC Industries
8. Commonwealth Oil
9. Phillips Puerto Rico
10. Puerto Rico Olefins
11. Union Carbide
12. Amerada Hess
Region III
13. Getty
14. Standard Chlorine
15. Sun-Olin
16. Continental Oil
17. Atlantic Richfield
18. Gordon Terminals
19. Gulf Oil
20. Standard Oil
(Ohio)/BP Oil
Location
Boundbrook, NJ
Gibbstown, NJ
Linden, NJ
Kearny, NJ
Westville, NJ
North Tonawanda, NY
Niagara Falls, NY
Penuelas, PR
Guyama, PR
Penuelas, PR
Penuelas, PR
St. Croix, VI
Delaware City, DE
Delaware City, DE
Claymont, DE
Baltimore, MD
Beaver Valley, PA
McKees, PA
Philadelphia, PA
Marcus Hook, PA
Longitude
74°06'04"
75017'50M
74°12'49"
74°06'39"
75008'42"
78°55'27"
79°00'55"
66°42'00"
66°07'00"
66°42'00"
66°42'00"
64°44'00"
75°37'45"
75°38'47"
76°25'40"
77°34'02"
80021'20"
80°03'10"
75°12I31"
75°37'45"
Latitude
40°33'25"
39°50'25"
40°38'10"
40°45'03"
39°52'05"
42°59'45"
43°03'33"
18°04'00"
17°59'00"
18°04'00"
18°04'00"
17°45'00"
39°35'15"
39°33'54"
39°48'20"
39°14'19"
40°39'2r'
40°28'22"
39°54'18"
39°35'15"
Model Plant
Type3
C/T
C/T
SP,C/T
C/T
LP,C/T
SP
C/T
LP.C/T
LP,C/T
C/T
C/T
LP
SP
C/T
C/T
C/T
C/T
C/T
LP.C/T
SP
C-5
-------
Table C-l. PLANTS AND LOCATIONS FOR BENZENE STORAGE
TANKS (continued)
Coordinates
Plant
Region III (concluded)
21. Sun Oil
22. U.S. Steel
23. Allied Chemical
24. American Cyanamid
25. Mobay Chemical
26. PPG
27. Union Carbide
Region IV
28. Jim Walter Resources
29. Reichhold Chemicals
30. Ashland Oil
31. B.F. Goodrich
32. GAP
33. 01 in Corporation
34. Chevron
35. First Chemical
Region V
36. Clark Oil
37. Core-Lube
38. Monsanto
39. National Distillers
(U.S.I.)
40. Northern
Petrochemicals
41. Shell Oil
Location
Marcus Hook, PA
Neville Island, PA
Moundsville, WV
Willow Island, WV
New Martinsville, WV
Natrium, WV
Institute, WV
Birmingham, AL
Tuscaloosa, AL
Ashland, KY
Calvert City, KY
Calvert City, KY
Brandenburg, KY
Pascagoula, MS
Pascagoula, MS
Blue Island, IL
Danville, IL
Sauget, IL
Tuscola, IL
Morris, IL
Wood River, IL
Longitude
75024'51"
80°05'00"
80°48'04"
81°19'08"
80°49'50"
80°52'14"
81047'05"
86°47'30"
87028'21"
82°36'32"
88019'51"
88°24'48"
86°07'15"
88°28'37"
88°29'45"
87°42'07"
87°32'30"
90°10'11"
88°21'00"
88°25'42"
90°04'24"
Latitude
39°48'45"
40°30'00"
39°55'00"
39°21'45"
39°43'30"
39044 '46"
38°22'40"
33035'30"
33°15'06"
38°22'30"
37°03'19"
37°02'51"
38°00'30"
30°19'04"
30°20'57"
41°39'19"
40°07'10"
38°36'06"
39047'53"
41°21'28"
38°50'26"
Model Plant
Type3
SP
C/T
C/T
C/T
C/T
C/T
C/T
C/T
C/T
LP
C/T
C/T
C/T
SP
C/T
C/T
C/T
C/T
C/T
C/T
LP
C-6
-------
Table C-l. PLANTS AND LOCATIONS FOR BENZENE STORAGE
TANKS (Continued)
Plant
Region V (concluded)
42. Union Oil
(California)
43. Dow Chemical
44. Dow Chemical
45. Sun Oil
Region VI
46. Vertac/Transvaal
47. Allied Chemical
48. American Hoechst
49. Cities Service
50. Continental Oil
51. Cos-Mar, Inc.
52. Dow Chemical
53. Exxon
54. Gulf Coast Olefins
55. Gulf Oil
56. Gulf Oil
57. Pennzoil United
(Atlas Processing)
58. Rubicon
59. Shell Oil
60. Tenneco
61. Union Carbide
62. Sun Oil
63. Amerada Hess
64. American Hoechst
65. American Petrofina
of Texas
Location
Lemont, IL
Bay City, MI
Midland, MI
Toledo, OH
Jacksonville, AR
Geismar, LA
Baton Rouge, LA
Lake Charles, LA
Lake Charles, LA
Carrville, LA
Plaquemine, LA
Baton Rouge, LA
Taft, LA
Alliance, LA
Donaldsonville, LA
Shreveport, LA
Geismar, LA
Norco, LA
Chalmette, LA
Taft, LA
Tulsa, OK
Houston, TX
Bayport, TX
Port Arthur, TX
Coordi
Longitude
88°00'10"
89°52'22"
84°12'18"
83°31'40"
92°04'56"
91°03'12"
91°12'40"
93019'0r'
93°16'35"
91°04I09"
91°14'30"
91°10'17"
90°26'23"
89°58'26"
90°55'19"
93°46'13"
91°00'37"
90°27'35"
89°58'19"
90°27'15"
96°01'15"
95014'15"
95°01'15"
93°53'20"
nates
Latitude
41°40'20"
43°37'21"
43°35'42"
41°36'52"
34°55'36"
30012'55"
30°33'03"
30°10'58"
30°14'30"
30°14'16M
30°19'50"
30°29'14"
29°59'16"
29°41'00"
30°05'44"
32°28'12"
30011'06"
29°59'42"
29°55'56"
29059'17"
36008'25"
29°41'39"
29036'10"
29°57'30"
Model Plant
Type3
SP
LP.C/T
C/T
LP
C/T
C/T
C/T
SP
C/T
C/T
LP
LP
C/T
LP
C/T
LP
C/T
C/T
SP
LP
C/T.SP
C/T
C/T
SP
C-7
-------
Table C-l. PLANTS AND LOCATIONS FOR BENZENE STORAGE
TANKS (Continued)
Coordinates
Plant
Region VI (continued)
66. American Petrofina
(Cosden Oil)
67. American Petrofina/
Union Oil of
California
68. Atlantic Richfield
69. Atlantic Richfield
(ARCO/Polymers)
70. Atlantic Richfield
(ARCO/Polymers)
71. Celanese
72. Charter
International
73. Coastal States Gas
74. Corpus Christi
Petrochemicals
75. Cosden Oil
76. Crown Central
77. Dow Chemical (A)
78. Dow Chemical (B)
79. Dow Chemical
80. DuPont
81. DuPont
82. Eastman Kodak
83. El Paso Natural Gas
84. El Paso Products/
Location
Big Spring, TX
Beaumont, TX
Channel view, TX
Houston, TX
Port Arthur, TX
Pampa, TX
Houston, TX
Corpus Christi, TX
Corpus Christi , TX
Groves, TX
Pasadena, TX
Freeport, TX
Freeport, TX
Orange, TX
Beaumont, TX
Orange, TX
Longview, TX
Odessa, TX
Longitude
101°24'55"
93°58'45"
95°07'30"
95°13'54"
93°58'15"
100°57'47"
95°15'09"
97°26'44"
97031'21"
93°52'58"
95010'30"
95°19'55"
95°24'09"
93°45'14"
94°01'40"
93044-44"
94°41'24"
102019'29"
Latitude
32°16'ir'
30°00'00"
29°50'00"
29°43'10"
29°51'24"
35032'07"
29°40'17"
27°48'42"
27°50'02"
29°57'46"
29°44'40"
28°57'23"
28°59'17"
30°03'20"
30°00'51"
30°03'24"
32026'17"
31°49'27"
Model Plant
Type3
LP.C/T
SP.C/T
LP
LP
C/T
C/T
SP
LP.C/T
SP.C/T
C/T
SP
LP.C/T
LP.C/T
C/T
C/T
C/T
C/T .
C/T
(Rexene Polyolefins)
Odessa, TX
102°20'00"
31°49'22'
C/T
C-8
-------
Table C-l. PLANTS AND LOCATIONS FOR BENZENE STORAGE
TANKS (Continued)
Coordinates
Plant
Region VI (continued)
85. Exxon
86. GATX Terminal Group
87. Georgia-Pacific Corp.
88. Goodyear Tire and
Rubber
89. Gulf Oil Chemicals
90. Gulf Oil Chemicals
91. Hercules
92. Howe 11
93. Independent Refining
Corp.
94. Kerr-McGee Corp.
(Southwestern)
95. Marathon Oil
96. Mobil Oil
97. Monsanto
98. Monsanto
99. Oxirane
100. Petrounited Terminal
Services
101. Phillips Petroleum
102. Phillips Petroleum
103. Phillips Petroleum
104. Quintana-Howell
105. Shell Chemical
106. Shell Oil
107. Shell Oil
Location
Baytown, TX
Houston, TX
Houston, TX
Bayport, TX
Cedar Bayou, TX
Port Arthur, TX
McGregor, TX
San Antonio, TX
Winnie, TX
Corpus Christi, TX
Texas City, TX
Beaumont, TX
Alvin (Choco-
late Bayou)
Texas City, TX
Channel view, TX
Houston, TX
Borger, TX
Pasadena, TX
Sweeny, TX
Corpus Christi, TX
Houston, TX
Deer Park, TX
Odessa, TX
Longitude
95°01'04"
95013'29"
95°03'00"
95002 -44"
94°55'10"
93°58'30"
97°16'30"
98°27'36"
94°20'28"
97°25'24"
94054'47"
94°03'30"
95°12'44"
94°53'40"
95°06'29"
95°01'23"
101022'05"
95010'53"
95°45'10"
97°27'30"
95°01'45"
95°07'33"
102°19'20"
Latitude
29°44'50"
29°43'17"
29°37'20"
29°39'43"
29°49'29"
29°51'30"
31°30'15"
29°20'51"
29050'04"
27°48'16"
29022'21"
30°04'00"
29°15'09"
29°22'44M
29°50'00"
29°33'51"
35°42'05"
29043.59..
29°04'24"
27°48'30"
29°38'15"
29042 -55"
31°49'05"
Model Plant
Type3
LP.C/T
C/T
C/T
C/T
C/T
LP.C/T
C/T
SP
SP
SP
SP.C/T
LP,C/T
LP.C/T
LP.C/T
C/T
C/T
SP
C/T
SP.C/T
SP
C/T
LP
SP
C-9
-------
Table C-l. PLANTS AND LOCATIONS FOR BENZENE STORAGE
TANKS (Concluded)
Coordinates
Plant
Region VI (concluded)
108. Standard Oil
(Indiana)
109. Standard Oil
(Indiana)XAmoco
110. Sun Oil
111. Texaco
112. Texaco/Jefferson
Chemical
113. Union Carbide
114. Union Carbide
115. USS Chemicals
Region VII
116. Chemplex
117. Getty Oil
118. Monsanto
Region IX
119. Atlantic Richfield
120. Chevron
121. Specialty Organics
122. Standard Oil of
California (Chevron
Chemical)
123. Union Carbide
124. Witco Chemical
125. Montrose Chemical
126. Stauffer Chemical
Location
Alvin, TX
Texas City, TX
Corpus Christi, TX
Port Arthur, TX
Port Neches, TX
Seadrift, TX
Texas City, TX
Houston, TX
Clinton, 10
El Dorado, KA
St. Louis, MO
Wilmington, CA
Richmond, CA
Irwindale, CA
El Segundo, CA
Torrance, CA
Carson, CA
Henderson, NV
Henderson, NV
Longitude
95°11'55"
94°55'45"
97°31'38"
93°54'43"
93°56'00"
96°45'59"
94°56'33"
95°15'06"
96°17'29"
96°52'00"
90°12'00"
118°14'30
122°23'36"
117°55'56"
118°24'4r'
118°20'50M
118°14'13"
115°00'40"
115°00'40"
Latitude
29°13'06"
29°21'58"
27049.57..
29°52'00"
29°57'50
28°30'38"
29°22'27"
29°42'18"
41°48'24"
37°47'10"
38°35'00"
33043.49..
37°56'12"
34°06'18"
33°54'39"
33°51'11"
33°49'18"
36°02'28"
36°02'28"
Model Plant
Type3
C/T
LP,C/T
LP.C/T
LP.C/T
C/T
C/T
C/T
C/T
C/T
SP.C/T
C/T
SP
SP.C/T
C/T
SP.C/T
C/T
C/T
C/T
C/T
aC/T represents a benzene consumer or bulk storage terminal; LP represents a large
producer of benzene; SP represents a small producer of benzene.
C-10
-------
TABLE C - 2. MODEL INPUTS FOR EACH TYPE OF MODEL PLANT
o
Type of
Model
Plant and
Tank Number
Benzene Producer:
Large. Facility
(throughput of
224. & x 108
Liters/year)
1
2
3
4
5
6
7
Tank Dimensions Baseline
Height Vertical Cross- "oofa Emissions
(m) Sectional Area Type (kg/yr)
(m2)
9
12
5
9
13
9
15
108
216
40
81
169
216
405
ncIFR
EFRps
cIFRps
cIFRps
ncIFR
ncIFR
ncIFR
720
2,190
480
590
680
1,360
1,820
BDT
Roof Emissions
Type3 (kg/yr)
ncIFR
EFRss
cIFRps
cIFRps
ncIFR
ncIFR
ncIFR
720
130
480
590
680
1,360
1,820
Beyond
Roof
Type3
ncIFRss
EFRss
cIFRss
cIFRss
ncIFRss
ncIFRss
ncIFRss
BDT
Emissions
(kg/yr)
560
130
380
470
520
1,040
1,460
Benzene Producer:
Small Facility
(throughput of
46.3 x 106 liters/yr)
1
2
3
4
11
13
11
7
33
169
88
224
FR
ncIFR
ncIFR
cIFRps
1,270
680
500
2,170
ncIFRlm
ncIFR
ncIFR
cIFRps
280
680
500
2,170
ncIFRlmss
ncIFRss
ncIFRss
cIFRss
270
510
400
1,760
Benzene Consumer or
Bulk Storage Terminal
1
2
11
15
132
270
ncIFR
cIFRps
640
970
ncIFR
cIFRps
640
970
ncIFRss
cIFRss
480
730
FR - Fixed-roof tank, IFR - internal floating-roof tank, ERF - external floating-roof tank, c - contact roof,
nc - noncontact roof, ps - primary seal, ss - secondary seal, 1m - liquid-mounted seal.
C-ll
-------
to be 0 m/s. The model was run in the nonurban mode. More information on
the development of model plants and emission rates can be found in Chapter 2
of this document.
C.3 POPULATION AROUND PLANTS CONTAINING BENZENE STORAGE TANKS
The HEM was used to estimate the population that resides in the vicinity
of each receptor coordinate surrounding each plant containing benzene
storage tanks. A slightly modified version of the "Master Enumeration
District List—Extended" (MED-X) data base is contained in the HEM and used
for population pattern estimation. This data base is broken down into
enumeration district/block group (ED/BG) values. MED-X contains the population
centroid coordinates (latitude and longitude) and the 1970 population of
each ED/BG in the United States (50 States plus the District of Columbia).
For human exposure estimations, MED-X has been reduced from its complete
form (including descriptive and summary data) to produce a randomly accessible
computer file of only the data necessary for the estimation. A separate
file of county-level growth factors, based on the 1970 to 1980 growth
factor at the county level, has also been created for use in estimating
1980 population figures for each ED/BG. The population "at risk" to benzene
exposure was considered to be persons residing within 20 km of plants
containing benzene storage tanks. The population around each plant was
identified by specifying the geographical coordinates of that plant. The
geographical coordinates are shown for each plant in Table C-l.
C.4 POPULATION DOSAGE METHODOLOGY
C.4.1 Dosage Methodology
The HEM uses benzene atmospheric concentration patterns (see Section C.2)
together with population information (see Section C.3) to calculate population
C-12
-------
dosage. For each receptor coordinate, the concentration of benzene and the
population estimated by the HEM to be exposed to that particular concentration
are identified. The HEM multiplies these two numbers to produce population
dosage estimates and sums these products for each plant. A two-level
scheme has been adopted in order to pair concentrations and populations
prior to the computation of dosage. The two-level approach is used because
the concentrations are defined on a radius-azimuth (polar) grid pattern
with nonuniform spacing. At small radii, the grid cells are much smaller
than ED/BG's; at large radii, the grid cells are much larger than ED/BG's.
The area surrounding the source is divided into two regions, and each ED/BG
is classified by the region in which its centroid lies. Population dosage
is calculated differently for the ED/BG's located within each region.
For ED/BG centroids located between 0.1 km and 2.8 km from the emission
source, populations are divided between neighboring concentration grid
points. There are 96 (6 x 16) polar grid points within this range. Each
grid point has a polar sector defined by two concentric arcs and two wind
direction radials. Each of these grid points is assigned to the nearest
ED/BG centroid identified from MED-X. The population associated with the
ED/BG centroid is then divided among all concentration grid points assigned
to it. The exact land area within each polar sector is considered in the
apportionment.
For population centroids between 2.8 km and 20 km from the source, a
concentration grid cell, the area approximating a rectangular shape bounded
by four receptors, is much larger than the area of a typical ED/BG (usually
1 km in diameter). Since there is a linear relationship between the logarithm
of concentration and the logarithm of distance for receptors more than 2 km
C-13
-------
from the source, the entire population of the ED/BG is assumed to be exposed
to the concentration that is geometrically interpolated radially and azimuthally
from the four receptors bounding the grid cell. Concentration estimates
for 80 (5 x 16) grid cell receptors at 2.0, 5.0, 10.0, 15.0, and 20.0 km
from the source along each of 16 wind directions are used as reference
points for this interpolation.
In summary, two approaches are used to arrive at coincident concentration/
population data points. For the 96 concentration points within 2.8 km of
the source, the pairing occurs at the polar grid points using an apportionment
of ED/BG population by land area. For the remaining portions of the grid,
pairing occurs at the ED/BG centroids themselves, through the use of log-log
linear interpolation. (For a more detailed discussion of the methodology
used to estimate dosages, see Reference 1.)
C.4.2 Total Dosage
3
Total dosage (persons-ug/m ) is the sum of the products of concentration
and population, computed as illustrated by the following equation:
N
Total dosage = I (P.C.) (1)
x X
where
P. = population associated with point i,
C. = annual average benzene concentration at point i, and
N = total number of polar grid points between 0 and 2.8 km
and ED/BG centroids between 2.8 and 20 km.
The computed total dosage is then used with the unit risk factor to
estimate leukemia incidence. This methodology and the derivation of maximum
lifetime risk are described in the following sections.
C-14
-------
C.5 LEUKEMIA INCIDENCE AND MAXIMUM LIFETIME RISK
C.5.1 Unit Risk Factor
_Q
The unit risk factor (URF) for benzene is 9.9 x 10 (cases per year)/
(ug/m -person years), as calculated by EPA's Carcinogen Assessment Group
(CAG). The derivation of the URF can be found in the CAG report on population
o
risk to ambient benzene exposure and updated in Response to Public Comments on
EPA's Listing of Benzene Under Section 112 and Relevant Procedures for the
Regulation of Hazardous Air Pollutants. EPA-450/5-82-003.
C.5.2 Annual Leukemia Incidence
Annual leukemia incidence (the number of leukemia cases per year)
associated with a given plant under a given regulatory alternative is the
3
product of the total dosage around that plant (in persons - ug/m ) and the
unit risk factor, 9.9 x 10 . Thus,
Cases per year = (total dosage) x (unit risk factor), (2)
where total dosage is calculated according to Equation 1 and the unit risk
factor equals 9.9 x 10"8.
C.5.3 Maximum Lifetime Risk
The populations in areas surrounding plants containing benzene storage
tanks have various risk levels of contracting leukemia from exposure to
benzene emissions. Using the maximum annual average concentration of
benzene to which any person is exposed, it is possible to calculate the
maximum lifetime risk of leukemia (lifetime probability of leukemia to any
person exposed to the highest concentration of benzene) attributable to
benzene emissions using the following equation:
C-15
-------
Maximum lifetime risk = C. m3V x (URF) x 70 years (3)
I 9 nlcLX
where
C. = the maximum concentration among all plants at any receptor
1 y iTlaX
location where exposed persons reside,
URF = the unit risk factor, 9.9 x 10"8, and
70 years = the average individual's life span.
C.5.4 Example Calculations
The following calculations illustrate how annual leukemia incidence
and maximum lifetime risk were calculated for specific plants listed in
Table C-l. Table C-3 presents the maximum annual average concentration and
the total dosage for each plant under the three control levels of baseline
(current level), best demonstrated technology (BDT), and the next more
stringent level beyond BDT (BBDT).
C.5.4.1 Annual Leukemia Incidence. As an example for calculating
annual leukemia incidence the Gulf Oil plant in Philadelphia, Pennsylvania,
is used. As shown in Table C-3, the total dosage under the current (baseline)
level of emission control is 3.30 x 10 persons-pg/m . Therefore, under
the baseline, the cases per year are computed according to Equation 2 as
follows:
Cases per year = 3.30 x 10 x 9.9 x 10"8
Cases per year = 0.003
C.5.4.2 Maximum Lifetime Risk. Plant numbers 73 (Coastal States and
Gas) and 117 (Sun Oil) had the highest maximum annual average benzene
concentration of 5.22 |jg/m . Using this maximum concentration and Equation 3,
maximum lifetime risk under the current (baseline) level of control is
calculated as follows:
C-16
-------
Baseline
TABLE C-3. ESTIMATED MAXIMUM CONCENTRATION
AND DOSAGE FOR BENZENE STORAGE TANKS
BDT
Beyond BDT
Plant
Number
Region
1
2
3
4
5
6
7
8
9
10
11
12
Region
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Maximum Annual
Average Benzene
Concentration
(pg/m3)
II
4.45X10"1
2.50X10"1
2.60x10°
l.OOxlO"1
3.03x10°
3.19x10°
7. 44x1 O"1
b
b
b
b
b
III
1.77x10°
2.50X10"1
8.67xlO"3
2.50X10"1
2.50X10"1
4.50X10"1
3.03x10°
1.77x10°
1.77x10°
4. 50x1 O"1
l.SOxlO"2
2.50X10"1
9.62xlO"3
5.00X10"1
4.87X10"1
Maximum Annual
Total Average Benzene
Dosage Concentration
(person pg/m3) (pg/m3)
5. 26x1 O3
2.27xl03
3.05xl04
1.44xl04
2. 32x1 O4
3.95xl03
7. 93x1 O2
b
b
b
b
b
1.21xl03
3. 20x1 O2
S.SOxlO1
l.SlxlO2 -
5.20xl02
3.75xl03
3. 30x1 O4
1.21xl03
4.50xl03
2.32xl03
5.39xl02
1.21xl02
7.24X101
l.OSxlO2
1.07xl03
4.45X10"1
2.50X10"1
2.24x10°
l.OOxlO"1
2.43x10°
2.50x10°
7. 44x1 O"1
b
b
b
b
b
1.39x10°
2.50X10"1
8.67x!0"3
2.50X10"1
2.50X10"1
4.50X10"1
2.43x10°
1.39x10°
1.39x10°
4.50X10"1
l.SOxlO"2
2.50X10"1
9.62xlO"3
5.00X10"1
4. 87x1 O"1
Total
Dosage
(person pg/m3)
5.26xl03
2. 27x1 O3
2.55xl04
1.44xl04
l.SlxlO4
3.10xl03
7.93xl02
b
b
b
b
b
9.45xl02
3. 20x1 O2
8.50X101
l.SlxlO2
5.20xl02
3.75xl03
2.59xl04
9.45xl02
3.53xl03
2.32xl03
5.39xl02
1.21xl02
7.24X101
1.03xl02
1.07xl03
Maximum Annual
Average Benzene Total
Concentration Dosage
(pg/m3) (person pg/m3;
3.34X10"1
l.OOxlO"1
1.79x10°
5.00xlO"2
1.89x10°
2.04x10°
5.59X10"1
b
b
b
b
b
1.13x10°
l.OOxlO"1
6.52X10"3
2.50X10"1
2.50X10"1
3.38X10"1
1.89x10°
1.13x10°
1.13x10°
3.38X10"1
1.36xlO"2
2.50X10"1
7.23xlO"3
2.50X10"1
3.66X10"1
3.95xl03
1.70xl03
2.03xl04
l.OSxlO4
1.42xl04
2.51xl03
5.96xl02
b
b
b
b
b
7.67xl02
2.41xl02
6.39X101
9. 87x1 01
3.91xl02
•2. 82x1 O3
2. 02x1 O4
7.67xl02
2.86xl03
1.75xl03
4.05xl02
9.06X101
5.44X101
7.76X10"1
8. 02x1 O2
C-17
-------
Baseline
TABLE C-3. ESTIMATED MAXIMUM CONCENTRATION
AND DOSAGE FOR BENZENE STORAGE TANKS
BDT
Beyond BDT
Maximum Annual
Average Benzene
Plant Concentration
Number (pg/m3)
Region IV
28
29
30
31
32
33
34
35
Region V
36
37
38
39
40
41
42
43
44
45
Region VI
46
47
48
49
50
-l
3.47x10 '
l.OOxlO"1
1.00x10°
1.04xlO"2
1.92xlO"2
1.53xlO"2
9.78xlO"3
S.OlxlO"1
3.70X10"1
2.50X10"1
4.11X10"1
6.05xlO"3
3.70X10"1
2.51x10°
1.64x10°
1.00x10°
3. 88x1 O"1
2.87x10°
l.OOxlO"1
2.50xlO~]
2.50X10"1
5. OOxl O"1
2.50X10"1
Maximum Annual
Total Average Benzene
Dosage Concentration
(person ug/m3) (pg/m3)
0
1.70xlOJ
4. 04x1 O2
2.57xl03
6.45X101
1.19xl02
1.41xl02
3.69xl02
3.77xl02
5.30xl03
1.96xl02
2.67xl03
4.75X101
4.91xl02
3.42xl03
3.43xl03
2.51xl02
4.81xl02
1.22xl04
1.63xl02
1.42xl02
6.57xl02
8. 88x1 O2
3.65xl02
-i
3.47x10 '
l.OOxlO"1
1.00x10°
1.04xlO"2
1.92xlO"2
1.53xlO"2
7.67xlO"3
S.OlxlO"1
3.70X10"1
2.50X10"1
4.11X10"1
6.05xlO"3
3.70X10"1
1.93x10°
1.32x10°
1.00x10°
3. 88x1 O"1
2.26x10°
l.OOxlO"1
2.50X10"1
2.50X10"1
5.00X10"1
2.50X10"1
Total
Dosage
(person pg/m3)
•3
1.70X1013
4. 04x1 O2
1.90xl03
6. 45x1 O1
1.19xl02
1.41xl02
2.90xl02
3. 77x1 O2
5.30xl03
1.96xl02
2. 67x1 O3
4.75X101
4.91xl02
2.54xl03
2.69xl03
1.96xl02
4.81xl02
9.10xl03
1.63xl02
1.42xl02
6.57xl02
6. 98x1 O2
3.65xl02
Maximum Annual
Average Benzene Total
Concentration Dosage
(pg/m3) (person pg/m3)
-n
2.60x10 '
l.OOxlO"1
1.00x10°
7.81xlO"3
1.45xlO"2
1.15xlO"2
6.22xlO"3
6.02X10"1
2. 78x1 O"1
2.50X10"1
3.09X10"1
4. 55x1 O"3
2. 78x1 O"1
1.52x10°
1.07x10°
1.00x10°
2.92X10"1
1.77x10°
l.OOxlO"1
l.OOxlO"1
2.50X10"1
5.00X10"1
2.50X10"1
0
1.28X10-3
3. 04x1 O2
l.SOxlO3
4.85X101
8.96X101
1.06X102
2.35xl02
2.83X102
3. 98x1 O3
1.47xl02
2.00xl03
3.57X101
3.69xl02
2.00xl03
2.19xl03
1.53xl02
3.61xl02
7.17xl03
1.22xl02
1.07xl02
4.94xl02
5.66X102
2.75xl02
C-18
-------
Baseline
TABLE C-3. ESTIMATED MAXIMUM CONCENTRATION
AND DOSAGE FOR BENZENE STORAGE TANKS
BDT Beyond BDT
Plant
Number
Region
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
Maximum Annual
Average Benzene
Concentration
(ug/m3)
VI Cont. ,
2.50x10 '
2.66x10°
1.00x10°
5.75X10"1
2.61xlO"2
2.50X10"1
3.72x10°
2.50X10"1
5.00X10"1
1.00x10°
2.50x10°
3.19x10°
4.39X10"1
8.13X10"1
1.00x10°
2.50x10°
1.00x10°
5.19x10°
1.00x10°
2. 50x1 O"1
3.25X10"1
2.01x10°
5.22x10°
3.78x10°
l.OOxlO"1
5.00X10"1
Maximum Annual
Total Average Benzene
Dosage Concentration
(person ug/m3) (ug/m3)
_
1 . 59.xlO
1.49xl03
l.OSxlO4
3.51xl02
2. 44x1 O2
2.01xl02
1.55xl04
1.35xl02
2.29xl02
l.OSxlO4
1.14xl03
5.29xl03
4.61xl03
7.39xl02
1.60xl03
7.50xl02
2.03xl03
4.34xl03 '
1.95xl04
3.43xl02
4.55xl02
1.54xl04
4. 89x1 O3
1.24xl03
3.59xl02
7.30xl03
,
2.50x10"'
2.10x10°
1.00x10°
5.75X10"1
1.93xlO~2
2.50X10"1
2.91x10°
2.50X10"1
5.00X10"1
1.00x10°
2.50x10°
2.68x10°
4.39X10"1
8.13X10"1
1.00x10°
1.00x10°
1.0x10°
4.02x10°
1.00x10°
2.50X10"1
3.25X10"1
1.64x10°
4.20x10°
3.14x10°
l.OOxlO"1
2.50X10"1
Total
Dosage
(person ug/m3)
9
1.59x10^
l.llxlO3
7.80xl03
3.51xl02
l.SOxlO2
2.01xl02
1.16xl04
1.35xl02
2.29xl02
8. 42x1 O3
8. 47x1 O2
4. 42x1 O3
4.61xl03
7.39xl02
1.25xl03
5. 88x1 O2
1.69xl03
3.21xl03
1.45xl04
3. 43x1 O2
4. 55x1 O2
1.21xl04
3. 83x1 O3
1.04xl03
3.59xl02
5.73xl03
Maximum Annual
Average Benzene Total
Concentration Dosage
(ug/m3) (person ug/m3)
_i
1.00x10 '
1.65x10°
1.00x10°
4.32X10"1
1.52xlO"2
2.50X10"1
2.29x10°
l.OOxlO"1
2.50X10"1
1.00x10°
1.00x10°
2.15x10°
3.30X10"1
e.iixio"1
1.00x10°
1.00x10°
1.00x10°
3.16x10°
1.00x10°
l.OOxlO"1
2. 44x1 O"1
1.33x10°
3.28x10°
2.51x10°
5.00xlO"2
2.50X10"1
9
1.19x10^
8.73xl02
6.15xl03
2. 64x1 O2
1.42xl02
1.51xl02
9.10xl03
l.OlxlO2
1.72xl02
6.83xl03
6.67xl02
3.52xl03
3.46xl03
5.55xl02
1.02xl03
4. 59x1 O2
1.35xl03
2.53xl03
1.14xl04
2.58xl02
3.42xl02
9.83xl03
2.99xl03
8. 24x1 O2
2.70xl02
4.65xl03
C-19
-------
Baseline
TABLE C-3. ESTIMATED MAXIMUM CONCENTRATION
AND DOSAGE FOR BENZENE STORAGE TANKS
BDT
Beyond BDT
Maximum Annual
Average Benzene
Plant Concentration
Number (ug/m3)
Region VI
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
Cont. n
3.10xlOu
1.00x10°
l.OOxlO"1
2.50X10"1
l.OOxlO"1
l.OOxlO"1
l.OOxlO"1
4.12X10'1
1.00x10°
4.39X10"1
8.13X10"1
2.50X10"1
2.50X10"1
1.00x10°
2.50X10"1
4.72x10°
1.56xlO"2
3.07x10°
2.26x10°
2.50x10°
1.00x10°
3.10x10°
5.00X10"1
8.13xlO"T
1.00x10°
8. ISxlO"1
1.00x10°
Maximum Annual
Total Average Benzene
Dosage Concentration
(person ug/m3) (ug/m3)
l.OSxlO3
1.21xl03
3.42xl02
3. 57x1 02
3. 42x1 02
3.40xl02
5.94xl02
7.13xl02
4.70xl03
3.62X103
7.71xl02
8. OOxl O2
2.07xl02
1.95xl03
1.32X102
1.36xl04
8.62X101
5.13xl03
2.16xl03
3. 24x1 03
1.71xl02
3.81xl03
6. 24x1 02
7.33xl02
2.72X102
2.60X103
2.79xl02
2.48x10°
1.00x10°
l.OOxlO"1
2.50X10"1
l.OOxlO"1
l.OOxlO'1
l.OOxlO"1
4. 12X10"1
1.00x10°
4.39X10"1
'8. 13X10"1
2.50X10"1
2.50X10"1
1.00x10°
2.50X10"1
3.85x10°
1.22xlO"2
2.43x10°
1.85x10°
1.00x10°
1.00x10°
2.48x10°
5.00X10"1
8.13X10"1
1.00x10°
8.13X10"1
l.OOxlO2
Total
Dosage
(person ug/m3)
8. 49x1 O2
9. 48x1 O2
3. 42x1 O2
3.57xl02
3. 42x1 02
3.40xl02
5. 94x1 O2
7.13xl02
3.68xl03
3.62xl03
7.71xl02
S.OOxlO2
2.07xl02
1.53xl03
1.32xl02
1.07xl04
6.76X101
4.00xl03
l.SlxlO3
2.53xl03
1.34xl02
2.99xl03
6.24X102
7.33xl02
2.12xl02
2.60xl03
2.33xl02
Maximum Annual
Average Benzene Total
Concentration Dosage
(ug/m3) (person ug/m3)
1 . 93x1 0°
1.00x10°
l.OOxlO"1
l.OOxlO"1
5.00xlO"2
l.OOxlO"1
l.OOxlO"1
S.lOxlO"1
5.00xlO"]
3.30X10"1
e.iixio"1
2.50X10"1
2.50X10"1
1.00x10°
l.OOxlO"1
3.13x10°
9.89xlO"3
1.98x10°
1.48x10°
1.00x10°
1.00x10°
1.93x10°
2.50X10"1
6-llxlO"1
5-OOxlO"1
6-llxlO"1
1.00x10°
6.63xl02
7.40xl02
2.57xl02
2.69xl02
2.57xl02
2.55xl02
4.47xl02
5.36xl02
2.87xl03
2.72xl03
5.79xl02
6.01xl02
1.56xl02
1.19xl03
9.95X101
8. 68x1 O3
5.48X101
3.25xl03
1.44X103
1.98xl03
1.05xl02
2.33xl03
4.69xl02
5. 50x1 O2
1.72xl02
1.96xl03
1.86xl02
C-20
-------
Baseline
TABLE C-3. ESTIMATED MAXIMUM CONCENTRATION
AND DOSAGE FOR BENZENE STORAGE TANKS
BDT
Beyond BDT
Plant
Number
Region
104
105
106
107
108
109
no
111
112
113
114
115
Region
116
117
118
Region
119
120
121
122
123
124
125
126
Maximum Annual
Average Benzene
Concentration
VI Cont.
3.07x10°
8.13X10"1
2.50x10°
1.85x10°
3.36xlO"3
3.10x10°
5.22x10°
2.50x10°
l.OOxlO"1
S.OOxlO"1
2.50X10"1
4.39X10"1
VII
6.11xlO"3
5-OOxlO"1
2.50X10"1
IX
2.85x10°
4.50x10°
5.61X10"1
n
3.85xlOu
7.33xlO"]
5.61X10"1
2.50X10"1
2.50X10"1
Maximum Annual
Total Average Benzene
Dosage Concentration
(person ug/m3) (ug/m3)
1.73xl03
8.31xl02
7.55xl03
1.76xl03
9.02x10°
2.66xl03
2.13xl03
3. 60x1 O3
6.22xl02
2.47X101
4. 88x1 O2
4. 55x1 O3
1.58X101
6. 08x1 O2
2. 82x1 O3
2. 44x1 O4
1.07xl04
6. 29x1 O3
A
2.40x10^
6.93xl03
8.47xl03
5.18xl02
5.18xl02
2.43x10°
S.lSxlO"1
1.00x10°
1.50x10°
3.36xlO"3
2.48x10°
4.20x10°
1.00x10°
l.OOxlO"1
5. OOxl O"1
2.50X10"1
4.39X10"1
6.11xlO"3
5.00X10"1
2.50X10"1
2.41x10°
3.76x10°
5.61X10"1
n
3.20xlOu
7.33X10"1
5.61X10"1
2.50X10"1
2.50X10"1
Total
Dosage
(person |jg/m3)
1.35xl03
8.31xl02
5.58xl03
1.38xl03
9.02x10°
2. 08x1 O3
1.67xl03
2.82xl03
6.22xl02
2.47X101
4. 88x1 O2
4.55xl03
1.58X101
5. 07x1 O2
2.82xl03
1.91xl04
8. 94x1 03
6.29xl03
A
2.01x10^
6.93xl02
8. 47x1 O3
5.18xl02
5.18xl02
Maximum Annual
Average Benzege Total
Concentration Dosage
(pg/m3) (person pg/m3!
1.98x10°
e.nxio"1
1.00x10°
1.22x10°
2.52xlO"3
1.93x10°
3.28x10°
1.00x10°
5.00xlO"2
2.50X10"1
2.50X10"1
3.30X10"1
4.59xlO"3
5.00X10"1
l.OOxlO"1
1.96x10°
3.02x10°
4.21X10"1
n
2.56xlOu
5.51X10"1
4.21X10"1
l.OOxlO"1
l.OOxlO"1
l.lOxlO3
6.24xl02
4.40xl03
1.12xl03
6.78x10°
1.63xl03
1.30xl03
2.20xl03
4.67xl02
1.85X101
3.66xl02
3.42xl03
1.19X101
4.04xl02
2.12xl03
1.55xl04
7.11xl03
4.73xl03
4
1.60xl(T
5.20xl03
6.36xl03
3.89xl02
3.89xl02
a This table lists the maximum annual average benzene concentration to which at least one"'
person is exposed.
b Population estimate is not included in the HEM for this plant.
C-21
-------
Maximum lifetime risk = 5.22 x 9.9 x 10"8 x 70
Maximum lifetime risk = 3.62 x 10
C.5.5 Summary of Impacts
Table C-4 summarizes the estimated nationwide impacts for the three
levels of emission control: baseline (current level), best demonstrated
technology (BDT) and a more stringent level (beyond BDT). The nationwide
annual leukemia incidence was calculated by summing the total dosages over
all the plants and multiplying by the unit risk factor. The maximum lifetime
risk for all three levels of control was calculated as shown in Section C.5.4.2.
C.6 UNCERTAINTIES
Estimates of both leukemia incidence and maximum lifetime risk are
primarily functions of estimated benzene concentrations, populations, the
unit risk factor, and the exposure model. The calculations of these variables
are subject to a number of uncertainties of various degrees. Some of the
major uncertainties are identified below.
C.6.1 Benzene Concentrations
Modeled ambient benzene concentrations depend upon: (1) plant configuration,
which is difficult to determine for more than a few plants; (2) emission
point characteristics, which can be different from plant to plant and are
difficult to obtain for more than a few plants; (3) emission rates, which
may vary over time and from plant to plant; and (4) meteorology, which is
seldom available for a specific plant. The particular dispersion modeling
used can also influence the numbers. The dispersion models also assume
that the terrain in the vicinity of the source is flat. For sources located
in complex terrain, the maximum annual concentration could be underestimated
by several fold due to this assumption. The best model to use (ISC) is
usually too resource intensive for modeling a large number of sources. The
C-22
-------
Table C-4 ESTIMATED NATIONWIDE HEALTH IMPACTS
FOR BENZENE STORAGE TANKS
Baseline
BDT
Beyond BDT
Max. Annual Average.,
Concentration (pg/m )
Maximum Lifetime Risk
Range for Maximum
Lifetime Risk3
Total Dosage 3
(persons-pg/m )
Incidence (cases/yr)
Range for Incidence
(cases/yr)
5.22
'5
3.6 x 10
1.38 x 10"
9.48 x 10
4.37 x 10£
0.043
0.017 to
0.113
4.20
2.91 x
1.11 x
7.62 x
3.64 x 10
0.036
0.014 to
0.094
-5 to
5
3.28
2.27 x 10
8.66 x 10
5.95 x 10
'5
-6
-5 to
2.85 x 10'
0.028
0.011 to
0.074
These ranges respresent the uncertainty of estimates concerning benzene
concentrations to which workers were exposed in the occupational studies
of Infante, Aksoy, and Ott that served as the basis for developing the
unit risk factor. They represent the 95 percent confidence interval that is
based on an assumption that actual benzene concentrations are within a factor
of 2 of the estimated concentrations.
C-23
-------
less complex model that was used for benzene storage tanks introduces
further uncertainty through a greater number of generalizing assumptions.
The dispersion coefficients used in modeling are based on empirical measurements
made within 10 kilometers of sources. These coefficients become less
applicable at long distances from the source, and the modeling results
become more uncertain. Assuming the inputs to the dispersion model are
accurate, the predicted benzene concentrations are considered to be accurate
to within a factor of 2. This uncertainty factor was not included in the
calculations in this analysis.
C.6.2 Exposed Populations
Several simplifying assumptions were made with respect to the assumed
exposed population. The exposed population was assumed to be immobile,
remaining at the same location 24 hours per day, 365 days per year, for a
lifetime (70 years). This assumption is counterbalanced to some extent (at
least in the calculation of incidence) by the assumption that no one moves
into the exposure area either permanently as a resident or temporarily as a
transient. The population "at risk" was assumed to reside within 20 km of
each plant regardless of the estimated concentration at that point. The
selection of 20 km is considered to be a practical modeling stop-point
considering the uncertainty of dispersion estimates beyond 10 km. The
results of dispersion modeling are felt to be reasonably accurate within
that distance (see above). The uncertainty of these assumptions has not
been quantified.
C.6.3 Unit Risk Factor
The unit risk factor contains the uncertainty of estimates concerning
benzene concentrations to which workers were exposed in the occupational
studies of Infante, Aksoy, and Ott, which serve as the basis for the unit
C-24
-------
risk factor. The range that is given for the unit risk factor represents
a 95 percent confidence interval and is based on an assumption that the
actual benzene concentrations to which the workers were exposed are within
a factor of 2 of the estimated concentrations. Other uncertainties regarding
the occupational studies and the workers exposed that may affect the unit
risk factor were raised during the public comment period and focus on
assumptions and inconclusive data contained in the studies. However, those
uncertainties have not been quantified.
C.6.4 Other Uncertainties
There are several uncertainties associated with estimating health
impacts. Maximum lifetime risk and annual leukemia incidence were calculated
using the unit risk factor, which is based on a no-threshold linear extrapolation
of leukemia risk and applies to a presumably healthy white male cohort of
workers exposed to benzene concentrations in the parts per million range.
It is uncertain whether the unit risk factor can be accurately applied to
the general population, which includes men, women, children, nonwhites, the
aged, and the unhealthy, who are exposed to concentrations in the parts per
billion range. It is uncertain whether these widely diverse segments of
the population may have susceptabilities to leukemia that differ from those
of workers in the studies. Furthermore, while leukemia is the only benzene
health effect considered in these calculations, it is not the only possible
health effect. Other health effects, such as aplastic anemia and chromosomal
aberrations, are not as easily quantifiable and are not reflected in the
risk estimates. Although these other health effects have been observed at
occupational levels, it is not clear if they can result from ambient benzene
exposure levels. Additionally, benefits that would affect the general
population as the result of indirect control of other organic emissions in
C-25
-------
the process of controlling benzene emissions from benzene storage tanks are
not quantified. Possible benzene exposures from other sources also are not
included in the estimate. For example, an individual living near a benzene
storage tank is also exposed to benzene emissions from automobiles. Finally,
these estimates do not include cumulative or synergistic effects of concurrent
exposure to benzene and other substances.
C-26
-------
C.8 REFERENCES
1. Systems Applications, Inc. Human Exposure to Atmospheric
Concentrations of Selected Chemicals. (Prepared for the U.S.
Environmental Protection Agency, Research Triangle Park, North
Carolina). Volume I, Publication Number EPA-2/250-1, and Volume II,
Publication Number EPA-2/250-2. May 1980.
2. Busse, A.D. and J.R. Zimmerman. User's Guide for the Climatological
Dispersion Model. (Prepared for the U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina.) Publication
Number EPA-R4-73-024. December 1973.
3. Albert, R. E. Carcinogen Assessment Group's Final Report on
Population Risk to Ambient Benzene Exposures. U.S. Environmental
Protection Agency. Publication No. EPA-450/5-80-004. Docket
Number A-79-27-II-A-28. January 1979.
C-27
------- |