EPA-450/3-80-032b
Benzene Fugitive Emissions—
Background Information
for Promulgated Standards
Emission 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
June 1982
-------
This report has been reviewed by the Emission Standards and Engineering Division of the Office of Air
Quality Planning and Standards, EPA, and approved for publication. Mention of trade names or commercial
products is not intended to constitute endorsement or recommendation for use. Copies of this report are
available through the Library Services Office (Mp-35), U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina 27711; or, for a fee, from the National Technical Information Services, 5285
Port Royal Road, Springfield, Virginia 22161.
-------
ENVIRONMENTAL PROTECTION AGENCY
Background Information
Final Environmental Impact Statement
Benzene Fugitive Emissions
Prepared by:
Jack R. Farmer
Director, Emission Standards and Engineering Division
U.Si Environmental Protection Agency
Research Triangle Park, North Carolina 27711
1.
The promulgated national emission standard will limit fugitive
emissions of benzene from existing and new equipment in benzene
service. The promulgated standard implements Section 112 of the
Clean Air Act and is based on the Administrator's determination of
June 8, 1977, (42 FR 29332) that benzene presents a significant risk
to human health as a result of air emissions from one or more
stationary source categories and is, therefore, a hazardous air pollutant
2. Copies of this document have been sent to the following Federal
Departments: Office of Management and Budget; Labor, Health and
Human Services, Defense, Transportation, Agriculture, Commerce,
Interior, and Energy; the National Science Foundation; the Council
on Environmental Quality; members of the State and Territorial Air
Pollution Program Administrators; the Association of Local Air
Pollution Control Officials; EPA Regional Administrators; and other
interested parties.
3. For additional information contact:
Mr. Gilbert Wood
Standards Development Branch (MD-13)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Telephone: (919) 541-5578
4. Copies of this document may be obtained from:
U.S. EPA Library (MD-35)
Research Triangle Park, NC 27711
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
m.
-------
-------
TABLE OF CONTENTS
Title
1.0 SUMMARY . . . - -.':,-• .
1.1 Introduction :. . .
1.2 Summary of Changes Since Proposal . . ... .
1.2.1 Standards for Pumps and Compressors. ... .
1.2.2 Requirements for Difficult-to-Monitor and
Unsafe-to-Monitor Valves .........
1.2.3 Alternative Standards for Valves .... .
1.2.4 Delay of Repair Provisions . .;
1.2.5 Requirements for Control Devices
1.2.6 Calibration Gas Requirements for Reference
Method 21
1.2.7 Alternative Requirements for Vendors and
Manufacturers
1.2.8 Reporting Requirements
1.2.9 Benzene Usage Cutoff ...
1.3 Summary of Impacts of Promul gated Action. . . . .
1.3.1 Alternatives to Promulgated Action ....
1.3.2 Environmental Impacts of Promulgated
• Action ............
1.3.3 Energy and Economic Impacts of Promulgated
Action .......
1.3.4 Health Impact of Promulgated Action. . . .
1.3.5 Other Considerations ...........
2.0 SUMMARY OF PUBLIC COMMENTS . . .
2.1 Legal Considerations and Need for the Standard .
2.1.1 Benzene Listing
2.1.2 Need for the Standard . . .
2.1.3 Standard is Overdue
2.1.4 Clean Air Act Authority for EPA to Set
Control Requirements . . . . . . . . ...
2.1.5 Authority for Risk-Benefit Analysis . . .
2.1.6 Ability to Quantify Margin of Safety . . .
2.1.7 Suspension of Benzene Proceedings ....
Page
1-1
1-1
1-1
1-2
1-3
1-4
1-5
1-7
1-9
1-10
1-10
1-10
1-10
1-10
1-11
1-12
1-12
1-13
2-1
2-1
2-1
2-4
2^18
2-18
2-18
2-19
2-19
-------
TABLE OF CONTENTS (Continued)
Title
2.2 Health Effects and Risk Assessment of Benzene
Exposure
2.2.1 Over-estimation of Risk
2.2.2 Underestimation of Risk
/ 2-2.3 Consistency in Methodologies
2.3 Selection of the Final Standard
2.4 Control Technology .
2.4.1 Equipment Specifications
2.4.2 Leak Detection and Repair Requirements
2.5 Recordkeeping and Reporting
2.5.1 Recordkeeping
2.5.2 Reporting
2.5.3 Miscellaneous Comments Addressing Both
Recordkeeping and Reporting
2.6 Costs
2.6.1 Impacts on Small Plants
2.6.2 Cost Effectiveness
2.6.3 Benefit-Cost Consideration
2.7 Test Methods and Procedures
2.8 General Issues
2.8.1 Applicability
2.8.2 Selection of Baseline
2.8.3 Format of the Standard
2.8.4 Emissions Data Base
2.8.5 Consistency with Other Regulations and
Guidelines
2.8.6 Incorporation of Other Comments or
Documents
2.8.7 Use of Alternative Control Systems . .
2.8.8 Siting Criteria
2.8.9 Withdrawal of the Standard
APPENDIX A ANNUALIZED COSTS AND COST EFFECTIVENESS FOR
BENZENE FUGITIVE EMISSION SOURCES
Page
2-19
2-19
2-28
2-30
2-31
2-44
2-44
2-64
2-82
2-82
2-83
2-87
2-89
2-89
2-91
2-97
2-99
2-103
2-103
2-110
2-111
2-114
2-124
2-126
2-128
2-129
2-129
A-l
VI
-------
TABLE OF CONTENTS (Concluded)
Title
APPENDIX B MODEL FOR EVALUATING THE EFFECTS OF LEAK
DETECTION AND REPAIR ON BENZENE FUGITIVE EMISSIONS
FROM CERTAIN VALVES AND PUMPS ....
B.I Introduction
B.2 Description of Model
B.3 References
APPENDIX C METHODOLOGY FOR ESTIMATING INCIDENCE OF LEUKEMIA
AND MAXIMUM LIFETIME RISK FROM EXPOSURE TO
FUGITIVE EMISSION SOURCES OF BENZENE
C.I Introduction
C.2 Atmospheric Dispersion Modeling and Plant
Emission Rates
C.3 Population Around Plants Containing Fugitive
Emission Sources of Benzene
C.4 Population Exposure Methodology
C.5 Leukemia Incidence and Maximum Lifetime Risk .
C.6 Uncertainties
C.7 Comparison of the HEM Dispersion Model
Subprogram and ISC-LT Dispersion Model ....
C.8 References
APPENDIX D STANDARDS-SETTING APPROACH
D.I Introduction
D.2 Quantitative Risk Estimation in the Regulatory
Process
D.3 Selection of BAT as the Minimum Level of Con-
trol
Page
B-l
B-2
B-2
B-9
C-l
C-2
C-2
C-4
C-ll
C-12
C-20
C-30
C-34
D-l
D-2
D-2
D-3
vii
-------
LIST OF TABLES
Title Page
1-1 Irretrievable Losses That Would Occur if Standard
Implementation were Delayed 1-14
2-1 List of Commenters on the Proposed National Emission
Standard for Benzene Fugitive Emissions 2-2
2-2 Control Costs per Megagram of Benzene Reduced 2-34
2-3 Controls Costs per Megagram of Total Emissions Reduced. . . . 2-36
2-4 Flare Emission Studies Completed As of October 1982 2-47
2-5 Annualized Model Unit Control Costs and Savings
of the Benzene Fugitive Emissions Standard 2-94
2-6 1985 Nationwide Costs of the Benzene Fugitive
Emissions Standard 2-95
2-7 Comparison of Leak Frequencies for Valves and Pumps 2-116
2-8 Comments and EPA Responses Referenced to Other EPA
Rulings 2-129
A-l Annualized Control Costs for Valves in Gas/Vapor Service. . . A-3
A-2 Annualized Control Costs for Valves in Liquid
Service A-6
A-3 Annualized Control Costs for Pumps - New Units. , A-8
A-4 Annualized Control Costs for Pumps - Existing Units A-12
A-5 Annualized Control Costs for Compressors - New and
Existing Units A-15
A-6 Annualized Control Costs for Pressure Relief Devices -
New Units A-17
A-7 Annualized Control Costs for Pressure Relief Devices -
Existing Units A-19
A-8 Annualized Control Costs for Open-Ended Lines - New and
Existing Units A-21
A-9 Annualized Control Costs for Sampling Connections - New and
Existing Units A-23
A-10 Annualized Control Costs for Product Accumulator Vessels. . . A-25
A-ll Control Costs Per Megagram of Total Emissions Reduced .... A-27
A-12 Control Costs Per Megagram of Benzene Reduced A-29
vm
-------
LIST OF TABLES (Concluded)
Title
B-l Valve Emission Factors and Mass Emission Reductions .
B-2 Fraction of Valves Screened and Operated on
B-3 Pump Emission Factors and Mass Emission Reductions .
B-4 Fraction of Pumps Screened and Operated on
C-l Plants and Locations for Fugitive Emission Sources of
Benzene
C-2 Baseline Dispersion Modeling and Exposure Data . . .
C-3 Estimated Annual Leukemia Incidence
C-4 Comparison of Maximum Annual Average Benzene
Concentration of Exposed Persons Using HEM
Dispersion Model and HEM with ISC-LT Input
C-5 Comparison of Total Exposure to Benzene Using HEM
Dispersion and HEM with ISC-LT Input
Page
B-5
B-6
B-7
B-8
C-9
C-14
C-21
C-32
C-32
IX
-------
LIST OF FIGURES
Title Page
2-1 Emission Factors for Light Liquid Service Pump Seals for
Benzene, Coke Oven By-product, Cumene, and Petroleum
Refinery Units 2-118
2-2 Emission Factors for Light Liquid Service Valves for
Benzene, Coke Oven By-product, Cumene, and Petroleum
Refinery Units 2-119
2-3 Emission Factors for Gas Service Valves for Benzene, Cumene,
and Petroleum Refinery Units 2-120
B-l Schematic Diagram of Modeled Leak Detection and Repair
Program B-3
-------
1.0 SUMMARY
1.1 INTRODUCTION
On January 5, 1981, the Environmental Protection Agency (EPA)
proposed the national emission standard for benzene fugitive emissions
(46 FR 1165) under authority of Section 112 of the Clean Air Act.
Public comments were requested on the proposal in the FEDERAL REGISTER.
EPA received 30 comment letters on the proposed standard. Industry
representatives submitted most of the comments. Also commenting were
representatives of State and local air pollution agencies and a repre-
sentative of an environmental group. The comments that were submitted,
along with responses to these comments, are summarized in Chapter 2 of
this document. The summary of comments and responses serves as the
basis for the revisions made to the standard between proposal and
promulgation.
1.2 SUMMARY OF CHANGES SINCE PROPOSAL
The proposed regulation was extensively revised for promulgation.
Significant changes were made in the following areas:
- Standards for pumps and compressors
- Requirements for difficult-to-monitor and unsafe-to-monitor
valves
- Alternative standards for valves
- Delay of repair provisions
- Requirements for control devices
.- Calibration gas requirements for Reference Method 21
- Alternative requirements for vendors and manufacturers
- Reporting requirements
- Benzene usage cutoff
1-1
-------
1.2.1 Standards for Pumps and Compressors
Since proposal, EPA has analyzed the annualized cost of controlling
benzene emissions and the resultant emission reduction for each alternative
control technique for each fugitive emission source. Based on comparison
of incremental costs and emission reductions, EPA reconsidered the
basis for selection of the best available technology (BAT). After
selection of BAT, EPA considered the next more restrictive level of
control before selecting the final standard. (See Section 2.3 for a
full discussion on the selection of the final standard.) As a consequence
of this reconsideration, the standard has changed since proposal for
two fugitive emission sources: pumps and compressors.
The proposed standard for pumps included the use of dual mechanical
seals. Since proposal, the standard has changed to a leak detection
and repair program, for pumps. The leak detection and repair program
requires monthly leak detection of benzene-handling pumps in light liquid
service. Leak detection is to be performed with a portable organic
compound analyzer according to Reference Method 21. If a reading of
10,000 ppm or greater of organic compound is obtained, a leak is detected.
Initial repair of the leak must be attempted within 5 days, and the
repair must be completed within 15 days. Delay of repair will be
allowed for pumps that can not be repaired without a process unit
shutdown. Delay of repair is also allowed when the plant owner or
operator determines that repair of the pump requires using a dual
mechanical seal system. Delay of repair is not expected^to be needed
for most situations, however, because pumps are commonly spared. The
bases for selection of the leak detection and repair requirements for
pumps are discussed in Section 2.4.2 of this document as responses to
comments pertaining to these requirements.
The equipment standard for pumps also has been incorporated into
the final standard. If an owner or operator prefers, he or she may
comply with the equipment standard. The details or provisions of the
equipment standard have not been changed substantially since proposal.
For new and existing compressors, the control costs incurred for
each megagram of emission reductions were determined for equipment
controls. A leak detection and repair program was not considered for
the final standard because installation of equipment is the only
1-2
-------
viable control technique for compressors. A leak detection and repair
program for compressors is not effective because the repair of a leak
on a compressor requires the installation of equipment. Because
installation of the equipment requires a process unit shutdown and
because compressors generally are not spared, repair would be delayed
until the next turnaround. Such an approach would result in little,
if any, emission reduction. Because of this and the considerations
presented in Section 2.3, use of control equipment was selected as the
standard for compressors. EPA expects very few compressors in benzene
service.
1.2.2 Requirements for Dlfflcult-to-Monitor and Unsafe-to-Monitor
Valves :•.-•-.
Some valves are difficult to monitor because access to the valve
is restricted. Difficult-to-monitor valves can be eliminated in new
process units but can not be eliminated in existing process units.
For new units, all valves will be subject to the proposed leak detection
and repair program. However, for existing process units, the standard
has been changed since proposal to allow an annual leak detection and
repair program for valves that are difficult to monitor. Valves that
are difficult to monitor are defined as valves that would require
elevating the monitoring personnel more than 2 meters above any permanent
available support surface. This means that stepladders would be used
to elevate monitoring personnel under safe conditions. However,
valves that cannot be safely monitored by the use of step ladders
could be classified as difficult to monitor.
In addition, some valves- are unsafe to monitor. Valves that are
unsafe to monitor cannot be eliminated in new or existing units. The
final standard has been changed to allow an owner or operator to
submit a plan that defines a leak detection, and repair program conforming
with the routine monitoring requirements of the standard as much as
possible given that monitoring should not .occur under unsafe conditions.
Valves that are unsafe to monitor-are defined as those valves that
could, as demonstrated by the owner or operator, expose monitoring
personnel to imminent hazards from temperature, pressure, or explosive
process conditions. EPA expects few valves in benzene service that are
unsafe to monitor.
1-3
-------
1.2.3 Alternative Standards for Valves
At proposal, two alternative standards were presented for valves.
Both of these alternatives called for 1 year of monthly monitoring to
obtain data on which to base the alternative standard. One alternative
standard was based on an allowable percentage of valves leaking.
Because an industry-wide allowable, leak percentage is not possible due
to variability of leak frequency among process units, an allowable
percentage of valves leaking would have been determined for each process
unit based on data collected in that unit. A minimum of one performance
test was required annually. The other alternative standard for valves
allowed the development of work practices that would achieve the same
result as the proposed leak detection and repair program. This alter-
native allowed an owner or operator to change the monitoring interval
if the resultant emission reductions were the same as those associated
with the standard.
Based on comments received on the proposed alternative standards
and on analysis of the results from recent screening and maintenance
studies, the alternative standards for valves were reexamined. As a
result, the alternative standards were changed. In making these changes,
EPA considered information gathered on benzene-producing and consuming
process units.
The first alternative standard specifies a 2 percent limitation as
the maximum percent of valves leaking within a process unit, determined
by a minimum of one performance test annually. EPA agrees with the
commenters that this will provide an incentive to maintain a good
performance level and promote low-leak unit design. This standard
provides the flexibility of a performance, level that could be met by
implementing any type of leak detection and repair program and
engineering controls chosen at the discretion of the owner or operator.
Even though an industry-wide allowable leak percentage was not possible
for valves, this alternative standard would allow a process unit to
comply with an allowable percentage of valves leaking without having to
determine a specific performance level by a year-long monthly monitoring
program. If the results of a performance test show a percentage of
valves leaking higher than 2 percent, however, the process unit would
1-4
-------
not be in compliance with the standard. Finally, if an owner or operator
determines that he or she no longer wishes to comply with this alternative
standard, he or she can submit a notification in writing to the Admini-
strator stating that he or she will comply with the routine (monthly/
quarterly) leak detection and repair program specified in §61.242-7(a)-(e).
The second alternative standard specifies two statistically-based
skip-period leak detection and repair programs. Under skip-period Teak
detection, an owner or operator can skip from routine monitoring
(monthly/quarterly) to less frequent monitoring after completing a
number of consecutive monitoring intervals with performance levels less
than 2 percent. This approach provides that the performance level is
achieved for each skipped period for better than.a 90 percent certainty.
Two sets of consecutive periods and fraction of periods skipped were
determined for benzene process units. First, after two consecutive
quarterly periods with less than 2 percent of. valves leaking, an owner
or operator may skip to semiannual monitoring. Second, after five
consecutive quarterly periods with less than 2 percent of valves leaking,
an owner or operator may skip to annual monitoring. This alternative
standard also requires that, if a process unit exceeds the 2 percent of
valves leaking, the owner or operator must revert to the monthly/quarterly
leak detection and repair program that is specified in §61.242-7(a)-(e).
Compliance with this alternative standard would be determined by inspection
and review of records. ..••••
1.2.4 Delay of Repair Provisions
EPA recognized at proposal that, in a few cases, repair of leaking
sources may be delayed for specific reasons while the owner or operator
is making a good faith effort to comply with repair time limits.
Commenters presented additional reasons for delay of repair. EPA
reviewed these reasons and considered some of them reasonable. Therefore,
the delay of repair provisions have been expanded in the final standard.
In the proposed, standard, delay of repair was allowed where
repair is technically or physically infeasible without a process unit
shutdown. An example of such a situation would be a leaking valve
•K
that could not be isolated from the process stream but which requires
1-5
-------
complete replacement or replacement of multiple internal parts. In
this case, the process unit would have to be shut down to effect repairs
on the valve, since the valve could not be physically isolated from
the process stream. The final standard includes the same provision as
the proposed standard in that it does not require that the process unit
be shut down for repairs. In addition to the provision already in the
proposed standard, several new provisions have been added to the final
standard.
One new provision allows for delay of repair beyond a process
unit shutdown for valves when unforeseeable circumstances deplete
valves used in repair. Another new provision has been added to avoid
causing extended delays in returning a process unit to production if
the unit shuts down briefly due to unforeseen circumstances. Delay of
repair beyond an unforeseen process unit shutdown will be allowed if
the shutdown is less than 24 hours in duration. Repair of leaking
fugitive emission sources for which repair has been delayed would be
required at the next process unit shutdown.
As part of the repair requirements, EPA is clarifying its intent
for spare fugitive emission sources that do not remain in benzene
service. Delay of repair of benzene fugitive emission sources for
which leaks-have been detected will be allowed for sources that no
longer contain benzene in concentrations greater than 10 percent.
These fugitive emission sources must be purged to a system that complies
with the requirements for closed-vent systems and control devices.
Delay of repair will not be allowed, however, for spare equipment that
contains process fluids and is prepared to be placed on-line; such equipment
is considered to be in benzene service.
In addition, a provision has been added to allow delay of repair
for certain leaking pumps. Sometimes, leaking pumps cannot be repaired
under the leak detection and repair program unless the owner or operator
installs dual seals with barrier fluid systems. For these leaking
pumps, a delay of repair for a period of 6 months will be allowed to
install the required equipment. Another new provision is added to
allow delay of repair for valves if the owner or operator shows that
leakage of purged material after immediate repair is greater than the
fugitive emissions that are likely to result from delay of repair.
1-6
-------
1.2.5 Requirements for Control Devices " • • ;
1.2,. 5.1 Combustion Devices and Vapor Recovery Systems. The
tempe rat are and residence time .specified for combustion devices in the
proposed standard (0.5 seconds at minimum 76,0 °C) were based on data
analyzed in -an EPA memo ("Thermal Incinerators and Flares") dated
August 22, 1980 (Docket Wo. A-79-27-II-B-35), The data base contained
in this memo included Union Carbide laboratory studies, EPA and industry
field tests, and 147 tests on existing Incinerators in Los Angeles
County. These data indicate that greater than 98 percent efficiency
is attainable by incinerators operating at 816 °C (1500 °F) and 0.75 seconds
residence time. The memo concludes that 98 percent efficiency, or
less than 20 ppmv (total organics minus methane and ethane) in the
exhaust stream, is achievable in many situations at less than 871 °C
(1600 °F) and 0.75 seconds residence time. The temperature and residence
time required in the proposed standard were set at a level representative
of 95 percent efficiency, the degree of emission reduction required of
control devices in the proposed standard.
While thermal incinerators are proven control devices for destruction
of benzene emissions, they are not the only enclosed combustion devices
that could be used. In fact, boilers and process heaters already
existing on-site are expected to be used for eliminating the small
benzene streams covered by the standard. In order to ensure that
these combustion systems achieve the requisite degree of control while
reducing the burden associated with demonstrating compliance with this
standard, the temperature-residence time requirements for enclosed
combustion devices (i.e., 0.5 seconds at a minimum of 760 °C) have been
retained in the final standard. However, any temperature and residence
time combination is allowed by the standard provided the owner or
operator demonstrates that the control device achieves a 95 percent
destruction efficiency.
Other combustion systems, such 7as catalytic incinerators, are
also applicable to the control of small benzene streams. Systems
which employ catalysts, however, typically operate at lower temperatures
and would not be able to meet these operating requirements. Therefore,
the temperature-residence time requirements will not apply to combustion
systems which employ catalysts. This change will permit the use of
1-7
-------
catalytic combustion units for control of benzene fugitive emissions'
without an equivalency determination. However, the owner or operator
must demonstrate that the control device achieves a 95 percent destruction
efficiency.
The requirements for vapor recovery systems have not changed since
proposal. They require a 95 percent reduction of benzene emissions to
the atmosphere.
1.2.5.2 Flares. At proposal, flares were not specifically
listed as an acceptable control option for the reduction of fugitive
emissions of benzene. The results of available flare efficiency
studies were not considered relevant. The gas streams tested were not
considered representative of the streams to be controlled. In some
cases the flare design was not representative of flares in the industry.
In others the analytical method was questionable. No .method for measuring
flare efficiency (evaluating flare performance) was available. Theoretical
calculations indicated that flare efficiency could be as low as 60 percent
for destruction of benzene in low-flow intermittent streams sent to a
large flare. This efficiency was cited in several background documents
(Ethylbenzene/Styrene, Benzene Fugitive, SOCMI Fugitive VOC) and
served as a primary consideration in not specifically allowing the use
of flares.
Since proposal, the use of flares was reconsidered for the benzene
fugitive emissions standard. Further actual flare measurement results
have become available, most notably from the CMA-EPA study (IV-A-28)*,
since the 60 percent theoretical estimate was made. In the CMA-EPA
study, steam-assisted flares, air-assisted flares, and flares operated
without assist were investigated over a wide range of exit velocity,
gas composition, and flare gas heat content conditions. After review
of available flare efficiency data (see Section 2.4.1.1), EPA has
concluded that smokeless flares operated with a flame present and exit
velocities less than 18 m/sec (60 ft/sec) with flare gas heat contents
greater than 11.2 MJ/scm (300 Btu/scf) for steam-assisted flares or
exit velocities' less than 18 m/sec (60 ft/sec) and flare gas heat
*References to Docket Entry Numbers for Docket No. A-79-27 are presented
in this manner throughout this document.
1-8
-------
contents greater than 7.45 MJ/scm (200 Btu/scf) for flares operated
without assist are acceptable alternatives to enclosed combustion
devices (incinerators, boilers, process heaters) and vapor recovery
systems, such as carbon adsorbers and condensation units. Air-assisted
flares operated smokelessly with a flame present are also permitted
if the heat content of the flared gas is above 11.2 MJ/scm (300 Btu/scf)
and the exit velocity meets maximum velocity criterion, which is dependent
upon the heat content of the gas. They may be applied to control of
emissions from pump seals (or degassing reservoirs), compressor seals
(or degassing reservoirs), pressure relief devices, and other fugitive
emission sources. The requirement for the presence of a flame can be
ensured by monitoring the flare's pilot light with an appropriate heat
sensor, such as a thermocouple. To ensure somkeless operation, visible
emissions from a flare would be limited to less than 5 minutes in any
2-hour period. EPA plans to update these requirements as new information
is obtained.
1.2.6 Calibration Gas Requirements for Reference Method 21
EPA considered two calibration gases for the use of Reference
Method 21 (See 40 CFR Part 60 Appendix A) in screening fugitive emission
sources of benzene—hexane and methane. Prior to proposal, hexane was
specified in the standard as the calibrant for the draft ,of Reference
Method 21. At proposal, the calibration gas was changed to methane
because methane is more readily available in the required concentration
range. The change was made in response to comments on the draft regu-
latory package. However, during the public comment period for the
proposed standard, other commenters objected to the change from hexane,
saying the change eliminated the use of photoionization monitors and
would mean that more leaks were detected. EPA has considered the
differences in the results that would be obtained with the two calibrants '
and,finds the differences insignificant. The differences are in the
same range as the variability seen in repeated emission measurements
from the same source. Data collected using hexane and methane can be
used interchangeably within +30 percent at the action level. Therefore,
because the differences are insignificant and because the allowance of
hexane as a calibration gas will provide for the use of photoionization
1-9
-------
monitors, EPA has changed the leak detection requirements to allow
hexane as an alternate calibration material.
1.2.7 Alternative Requirements for Vendors and Manufacturers
The final standard has been changed to allow a vendor or manufac-
turer to apply for permission to use alternative control systems or
equipment. This change was made to increase efficiency in the permis-
sion process and to provide plant owners and operators the incentive to
purchase improved control systems and equipment as they are developed.
Even though the provision allowing vendors and manufacturers to
apply for permission has been added, it should be remembered that the
ultimate responsibility for properly using an alternative method is
with the owner or operator, not the vendor or manufacturer.
1.2.8 Reporting Requirements
The promulgated standard includes reporting provisions requiring
periodic reports of leak detection and repair efforts'within a process
unit. The amount of reporting and the associated burden with the
reporting have been reduced from those in the proposed standard. In
particular, quarterly reporting as required in the proposed standard
has been reduced to semiannual reporting. The specific reporting
provisions of the final standard have been changed in response to
industry's comments on the proposed reporting provisions. EPA believes
that these provisions will provide an efficient and effective means of
enforcing the standard.
1.2.9 Benzene Usage Cutoff
An analysis by EPA indicated that the cost effectiveness of the
standard for equipment at plant sites processing less than 1,000 Mg/yr is
unreasonably high. Because this cost is unreasonably high in comparison
to the minimal emissions reduction achievable, EPA provided an exemption
for plants designed for low benzene usage. Plants with usage rates of
1,000 Mg of benzene/yr or less are exempt from the standard. It is
expected that this cutoff will exempt most research and development
facilities and other small-scale operations.
1.3 SUMMARY OF IMPACTS OF PROMULGATED ACTION
1.3.1 Alternatives to Promulgated Action
Regulatory alternatives are discussed in Chapter 6 of the background
information document (BID) for the proposed standard, EPA-450/3-80-032a.
1-10
-------
These regulatory alternatives reflect the.different levels of emission
control from which one was selected as the proposed best available
technology, considering costs, nonair quality health, and environmental
and economic impacts for benzene fugitive emission sources. Since
proposal, these alternatives have changed slightly. At proposal,
Regulatory Alternatives III and IV were selected as the basis for the
standard for existing and new sources, respectively. Since proposal,
EPA selected the final standard from alternative control techniques for
each fugitive emission source rather than selecting the standard from
various combinations of control techniques and sources. The only
difference between the proposed and final standard with regard to
selection of control techniques is the requirements for existing
compressors and new pumps. In the proposed standard EPA required
monthly monitoring of existing compressors. In the final', standard, EPA
is requiring control equipment for existing compressors as discussed in
Section 1.2.2. The final standard for new pumps now requires monthly
monitoring instead of the installation of dual seal systems. Therefore,
the requirements for new-and existing pumps are the same in the final
standard (monthly monitoring). Requirements for the other sources
remain unchanged. ,
1.3.2 Environmental Impacts of Promulgated Action . , .
Environmental impacts of the standard are described in,Chapter 7 .
of the BID for the proposed standard. Since proposal, EPA has: reanalyzed
the environmental impacts of the standard and has found the overall
effect on these impacts has not substantially changed. EPA has revised
the baseline emission estimate (discussed in Section 2.8.2) and the
estimate of benzene fugitive emissions to atmosphere that will be
reduced by the standard. The revisions were made as a consequence of
revising methods of analysis (i.e., baseline emission estimate) and
incorporating the leak detection and repair model for valves and
pumps, as described in Appendix B of this document.
The standard will reduce benzene fugitive emissions to atmosphere
from 7,920 Mg/yr to 2,470 Mg/yr for existing units. This reduction
represents a 69 percent decrease in emissions for existing units from
the current industry baseline level of emissions. Assuming that standards
of performance for fugitive emission sources of VOC (those covering
1-11
-------
synthetic organic chemical manufacturing and petroleum refinieries)
would not cover benzene fugitive emission sources, the standard would
reduce benzene fugitive emissions from 2400 Mg/yr~to 680 Mg/yr for new
units constructed between 1980 and 1985. Because, however, these
standards of performance would reduce fugitive emissions of benzene,
the amount reduced over what would be reduced without the benzene
fugitive emissions standard is less than indicated above. Because the
emission reductions are about the same as the proposed estimates and
because the water quality and solid waste impacts are generally propor-
tional to the emission reductions, the water quality and solid waste
impacts are similar to those presented in the BID for the proposed
standard.
With the changes noted above, the analysis of the environmental
impact in the BID for the proposed standard now becomes the final
Environmental Impact Statement for the promulgated standard.
1.3.3 Energy and Economic Impacts of Promulgated Action
Section 7.5 of the BID for the proposed standard describes the
energy impacts of the standard. The changes made in the standard have
decreased slightly the energy impacts of the standard because the
energy value and crude oil equivalent of benzene emission reductions
are lower.
Chapters 8 and 9 of the BID for the proposed standard describe
the cost and economic impacts of the proposed .standard. Since proposal,
the cost analysis of the standard has been revised to incorporate new
emission reduction estimates from the leak detection and repair model
for valves and pumps. As discussed in Chapter 2 (Sections 2.3 and 2.6),
the economic impact of the final standard is reasonable.
1.3.4 Health Impact of Promulgated Action
Since the standard was proposed, EPA has revised the baseline
benzene fugitive emissions estimate, the estimated emission reductions
resulting from the standard, and the risk assessment for leukemia risk
from benzene fugitive emissions (see Section 2.8.2 of this document
for a discussion of changes to the baseline and emission reduction
estimates, and see Appendix C for complete descriptions of the revisions
to the risk assessment). Benzene fugitive emissions are estimated to
be 7,920 Mg/yr in 1980 in the absence of a benzene fugitive emissions
1-12
-------
standard. The final standard is expected to reduce these emissions by
69 percent. In its revised estimate, EPA has calculated that, in the
absence of a benzene fugitive'emissions standard, the leukemia incidence
would be about 0.45 cases per year within this population due to exposure
to benzene fugitive emissions from existing sources and about 0.12
cases per year from exposure to new sources of benzene fugitive emissions
(assuming relevant standards of performance might not affect benzene
fugitive emissions). Further, EPA has estimated that for the maximally
exposed persons, the chance of contracting leukemia as a result of
exposure to fugitive benzene emissions would be about 1.5 x 10~3 per
lifetime for existing sources. Due to the assumptions which were made in
calculating these maximum lifetime risk and leukemia incidence numbers,
there is uncertainty associated with the numbers presented here. The
use of the risk numbers is explained in the section entitled "Need for
the Standard." However, with the standard in place, the standard will
reduce these health impacts by 69 percent.
1.3.5 Other Considerations
1-3.5.1 Irreversible and Irretrievable Commitment. Section 7.6.1
of the BID for the proposed standard concludes that the standard will
not result in any irreversible or irretrievable commitment of resources.
It was also concluded that the standard should help to save resources
due to the energy savings associated with the reduction in emissions.
These conclusions remain unchanged since proposal.
1-3.5.2 Environmental and Energy Impacts of Delayed Standards.
Tables 1-1 and 7-13 in the BID for the proposed standard summarize the
environmental and energy impacts associated with delaying promulgation
of the standard. The revised air and energy impacts are shown in
Table 1-1 of this document. The emission reductions and associated
energy savings shown would be irretrievably lost at the rates shown
for each of the 5 years.
1-13
-------
o
1
1
to
Lu
H- 1
f^f.
— ^
O
o
o
o
1
Z3 Q
O LU
^f
1— l
*-*-!
tu si
»— « UJ
1— O
iti 2£
fy^; H-I
i ±<
(
i
1
,—4
0)
ja
tooo
r— «i— O
3 E r- <
CTLU
UJ «
CU CO
r- CU O
O Isl •!—
e 4->
CD CD O
T3 CQ 3
3 T3
C_ 4— CU
o o ai
0
C 1 —
O)
IS) «
c: co
cu c
CQ O
*l J I *>
o o
3
CD "O
3 CU
i — o;
re
>• c
o
>>"-
CD CO
£_ tO
CU ••—
e E
LU LU
re
CD
CO
c «
O ~O
•i- S-
4-3 re
O T3
3 C
™^3 CO
ce: co
c cu
0 J=
to
CO O
•1— 4-3
LU CU
c re
CU 4-3
tsj 3
C l/*^
CU T-
CQ C-
4-}
Sc
f
re
cu
O sO O ID t-l |OO
oo co cr> en o LO
i-H |^"
^-1 LO i— 1 tO O |CM
«* to LO tr> 10 r^
CM
P-. i-H «* CO CM ICM
LO to to 10 r~- CM
'oo
r—
fO
1^
0
I—
re
i— 1 C\J OO •* LO CU
co co co co co >>
r— < r- < t— 1 <-H I— 1 LO
OO
o-
to =a:
re •=£ T3
z. &-• cu
cu c ^ ~o • o
4-> -i— CD tO t_C_ Q.CU
re c LU re o o c
i — cO-r-Z TD4-> C-CU
3 4J r— CO Q. Is)
o M- T3 co re re c
i — CCCU 4->4— COCU
re 3 re > co jc -Q
0 . J= T- C 4->
CO CO CU -r- CU •!— t— O
•r- CU C CD CO tO O
CU O Is] 4- Q. CO 3
3 £- C O > Q r—
r— Q. CU O> £-C 1— ire
re JQ c: Q. o CQ >
> 01 cu o
C t- N CU CU >,
tO ••- CU C JC >> ^1 CD
•r-C.GCU 4->-Q 4->t-
I — 4- O t- CO O C
eu cu o c 4-> cu
t- r— -E 4- O
. r— 4-> -i- £- CO
-a E re Q 4-> cu ^: '
£-3 t- I— 1 O 4- 4->
• "O
CO CO C OO C i — T-
Q. O C O O -r- >
O) O O 4-> T- O T-
.e -r- co "O
4-> tO CD tO i- to CO
Ct r ft* flj .-^ -pi ^v
r™ uj y^ *i^ \— i *^
OOO-i- 4-E 3-O
4-> T- E CO CO £-
CO O CU Qi O T3
CO CO O CO CO
i — i- > CU C ' 4- C
_a E c • cu o -i-
re co >> cu *— » M re
4-> . £_ IS) , JQ C r— 4->
3COQJC r— CO J2-Q
JD C C CO ^ — . -O JQ O
•r- CO -r- -Q 3 --~
t_ Is) 4— 4-> CDi 1~D to
4->ca>cu coc I — co
4-> cu i- ~a T- 3
re JQ •!- to >, oo •—
>~, S co r— i re
c cu -CJ i cr> Q. o >
O C >, •> T- rH
•i- .^. -a t- i-- 4-> 4->
4Ji — CU4-> '«-< i — XC
O CU 4-> to 3 CO •
3tOfl33 £ CM •— £_
"O re r— *o CD T— i re o
CU-Q3C CD>> «>4->
t- CD-I— ^«, X3 IO -i- O
cu cu "3 3 re
C "O £- T3 1 — T3 CO O"4-
O •!— C CU T- CO
•i- 3 cu re <* c c
CO 1 t- LO -i— CO c — O
co >s re ' — - o re c -i— T-
•i—C- TD lQ4-> CUOtO
E4->cuco • -O M t-
cotoci— coo ccucu
3 O r— CU T3 >
ocos- T-CU £-0
•i- -i- 4-> 3 «- O O
to t- c a> i— o
tocoo ere 4- >,
>r« QJ q— o CO > • -Q
E CO C M 4-> OO
O>3t03 C>, C i— 1 tO
4_j re cu CD cu i e
cucucocu J2t_ i — r--o
c JQ £- c- cu re -r-
a> rere 4-c >a>4->
N CO O LU -i- i — O
c o +-> to 3 .a 3
cucc4-> cu crre-a
jQCOCUT- 3* COl — CU
C_ E C r— OO £_
-acuc3 re >— < i — •>
I .r- "D C
4-> 4— re co r~~ o£-o
re -t— 4-> co >> re T—
E "O 4-3 CO CD CJ CO "O tO
•r- reo t-i — -occo
4~> cu e o cu -Q 3 re •!•*
cojrot- ere £-4->E
LU4-3CQ. LUI— CJtOOl
re j3 o
1-14
-------
2.0 SUMMARY OF PUBLIC COMMENTS ;
Commenters1 affiliation, comment EPA docket number, and date of
comment for every comment received are listed' in Table 2-1. Thirty
letters and documents on the proposed standard and its background
information document (BIO) were received. Significant comments-have
been divided into the following eight sections:, .
2.1. Legal Considerations and Need for the Standard •
2.2 Health Effects and Risk Assessment of Benzene Exposure
2.3. Selection of the Final Standard
2.4. Control Technology
2.5. Recordkeeping and Reporting
2.6. Costs
2.7. Test Methods and Procedures '
2.8. General Issues ,
Comments, issues, and responses are discussed in the following
sections of this chapter. Changes to the regulation are summarized in
Section 1.2 of Chapter 1.
In several instances, commenters have requested that EPA incorporate
all of their comments on other EPA rulemakings. For these comments,
where EPA's response is identical to the response given in the other
rulemaking, EPA is simply referring to the identical response given in
the other rulemaking. In Section 2.8 (General Issues), EPA has included
a table that lists the significant comments of this sort and refers to
the location of the responses.
2.1 LEGAL CONSIDERATIONS AND NEED FOR THE STANDARD
2.1.1 Benzene Listing
Several 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 fugitive NESHAP should
2-1
-------
Table 2-1. LIST OF COMMENTERS ON THE- PROPOSED NATIONAL
EMISSION STANDARD FOR BENZENE FUGITIVE EMISSIONS
Docket Reference Number and Date9
1. IV-F-1; July 14, 1981
2. IV-F-2; July 14, 1981b
3. IV-D-1; January 9, 1981
4. IV-D-2; February 4, 1981
5. IV-D-3; February 19, 1981C
6. IV-D-4; March 11, 1980d
7. IV-D-5; September 11, 1980d
8. IV-D-6; December 22, 1980d
9. IV-D-7; March 16, 1981C
10. IV-D-8; April 1, 1981
11. IV-D-9; November 7, 1980d
12. IV-D-10; April 20, 1981
June 15, 1981
June 5, 1981C
July 13, 1981
August 7, 1981
August 7, 1981s
September 4, 1981
September 10, 1981
September 9, 1981
September 11, 1981
September 8, 1981
September 14, 1981
September 14, 1981
September 11, 1981e
September 11, 1981
September 10, 1981
•September 11, 1981
September 14, 1981
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
IV-D-11;
IV-D-12;
IV-D-13;
IV-D-14;
IV-D-15;
IV-D-16;
IV-D-17;
IV-D-18;
IV-D-19;
IV-D-20;
IV-D-21;
IV-D-22;
IV-D-23;
IV-D-24;
IV-D-25;
IV-D-26;
IV-D-27;
Comment Affiliation
Hearing transcript
Chemical Manufacturers Association
Kirk land & Ellis for API
Louisiana Dept. of Natural
Resources
Kirkland & Ellis for API
Phillips Petroleum Company
Union Carbide Corporation
Sun Petroleum Products Company
Kirkland & Ellis for API
Air Products and Chemicals, Inc.
Dow Chemical U.S.A.
Erie County, New York, Dept. of
Environment and Planning
Cities Service Company
Kirkland & Ellis for API
American Petroleum Institute
Chemical Manufacturers Association
Monsanto
Merck Chemical Manufacturing Div.
Chevron USA ,
Shell Oil Company
Pennzoil Company
Gulf Oil Company, U.S. •
Chemical Manufacturers Association
Universal Oil Products
Kerr-McGee Corporation
Monsanto Company
/
Dow Chemical Company
Phillips Petroleum Company
American Petroleum Institute
2-2
-------
Table 2-1. LIST OF COMMENTERS ON THE PROPOSED NATIONAL EMISSION
STANDARD FOR BENZENE FUGITIVE EMISSIONS (Concluded)
Docket Reference Number and Date9
30. IV-D-28; September 11, 1981
31. IV-D-29; September 14, 1981
32. IV-D-30; September 8, 1981
33. IV-D-31; September 14, 1981
34. IV-K-1.; September 11, 1981
Commenter Affililiation
Texaco, Inc.
American Cyanamid Company
Standard Oil Company
Natural Resources Defense
Council, Inc.
Standard Oil Company (Indiana)
The docket number for benzene fugitives standard development is
A-79-27. The docket is located at EPA Headquarters in Washington D.C.
b • . ' ' . '
This reference is a transcript submitted by the commenter at the
public hearing and is essentially identical to the oral testimony.
"No response is given because the letter requested an extension to the
public comment period; therefore, the comment letter is not summarized
in this document.
1
These letters comment on test reports that were finalized prior to
proposal and were inadvertently left out of the pre-proposal docket.
2
This letter was inadverrtently placed in the benzene fugitives dockets;
therefore, no comment summary or response is given.
2-3
-------
be withdrawn. Various reasons were commonly cited to support this
contention. These comments are summarized below.
Comment: Some commenters felt that benzene does not constitute
the kind of risk deemed hazardous under Section 112 of the Clean Air
Act (IV-D-28; IV-D-24; IV-F-1; IV-D-27; IV-K-1).
Response: Response to this comment can be found in "Response to
Public Comments on EPA's Listing of Benzene Under Section 112"
(EPA-450/5-82-003), which was prepared to address the listing of
benzene under Section 112 of the Clean Air Act.
Comment: Several commenters stated that there exists little or
no evidence to substantiate risk from emissions of benzene. One
commenter stated that there is no evidence (IV-D-28), another stated
that there is insufficient evidence (IV-D-17), and a third stated that
the listing was based on an inadequate record (IV-D-19).
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: Three commenters felt that the listing and rulemaking
proceedings for benzene are premature since they are based on a draft
EPA policy regarding airborne carcinogens (IV-D-8; IV-D-25; IV-D-26).
One of the commenters 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 Administration Procedure Act (IV-D-8). One
commenter felt that EPA exceeded its legal authority and offended good
scientific practice in utilizing the airborne carcinogen policy to
list benzene (IV-D-25).
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.
2.1.2 Need for the Standard
Several commenters contended that the proposed benzene fugitive
emissions standard is not needed and, therefore, should be withdrawn.
These comments address the following: (1) significance and relative
proportion of risk associated with benzene fugitive emissions;
2-4
-------
(2) duplication 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.
Comment: Three commenters stated that EPA has not demonstrated
that benzene fugitive emissions pose a significant risk that merits the
adoption of a benzene fugitive emissions standard (IV-D-27, IV-F-1;
IV-K-1). One of these commenters stated that Section 112 requires that a
NESHAP be established at the level which in the Administrator's judgment
provides "an ample margin of safety to protect the public health from
such hazardous air pollutant." According to the commenter, last year the
Supreme Court 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 EPA has calculated are not "significant"
as that term has been used by the Court (IV-K-1).
Two commenters felt that the risk from exposure to fugitive
benzene is insignificant compared to other commonly accepted societal
risks (IV-D-24; IV-D-27). One commenter noted that the risk from
fugitive benzene emissions is insignificant in comparison to background
leukemia risk (IV-D-28, IV-D-27). The commenter further compared the
risk from fugitive benzene emissions to other yovernment determinations
of risk acceptability and noted that, under these determinations, the
risk from exposure to benzene from fugitive emission sources would
have been considered not worthy of regulation (IV-D-27).
Response: On June 8, 1977, EPA listed benzene as a hazardous air
pollutant under Section 112 of the Clean Air Act. After careful
consideration, EPA has not decided to delist benzene as a hazardous
air pollutant, as discussed in EPA-450/5-82-003.
the commenters are judging the significance of benzene fugitive
emissions based on quantitative risk estimates. In general, quantitative
risk estimates at ambient concentrations involve an analysis of the
effects of a substance in high-dose epidemiclogical or animal studies,
2-5
-------
and extrapolation of these high-dose results to relevant human exposure
routes at low doses. In the case of benzene, the effects observed
were the result of high-dose epidemiological studies. The mathematical
models used for such extrapolations are based on observed dose-response
relationships for carcinogens and assumptions about such relationships
as the dose approaches very low levels or zero. Quantitative risks to
public health from emissions of an airborne carcinogen may be estimated
by combining the dose-response relationship obtained from this
carcinogenicity strength determination with an analysis of the
extent of population exposure to a substance through ambient air.
Most exposure analyses are based on air quality models, available
estimates of emissions from sources of a substance, and approximations
of population distributions near these sources. EPA considers this
the best practicable approach. Even though ambient monitoring data
might be used to estimate quantitative risks to public health, these
data are available only for a few locations near these sources.
Thus, use of ambient monitoring data is not practicable. However,
EPA has data to confirm that the public is exposed to benzene. For
example, concentrations up to 51 micrograms per cubic meter (on a
24-hour average) were found around a petrochemical plant in Philadelphia,
Pennsylvania. .
The air quality models used in exposure analyses generally estimate
exposures of up to 20 kilometers. During exposure analyses, population
and growth statistics are examined in conjunction with ambient concen-
trations. Using these factors and existing carcinogenicity strength
determinations, estimates are then made of the degree of risk to
individuals and the range of increased cancer incidence expected from
ambient air exposures associated with a substance at various possible
emission levels.
The assumptions and procedures discussed above for extrapolation
and for exposure estimates for benzene emissions are subject to
uncertainty. The ranges in estimates at proposal (presented below)
of maximum lifetime risk and annual leukemia incidence from exposure
to fugitive emission sources of benzene represent the uncertainty
of estimates made concerning the benzene concentrations to which
2-6
-------
workers were exposed in the occupational studies of Infante, Aksoy,
and Ott that served as the basis for developing the benzene unit
risk factor (discussed 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 and
in Appendix E of Benzene Fugitive Emissions—Background Information
for Proposed Standard. EPA-450/3-80-032a). The ranges represent 95
percent confidence limits on two sources of uncertainty in the
benzene risk estimates. One source derives from the variations in
dose/response among the three occupational studies upon which the
benzene unit risk factor is based. A second source involves the
uncertainties in the estimates of ambient exposure. In the former
case, the confidence limits are based on the assumption that the
slopes of the dose/response relationships are unbiased estimates of
the time slope and that the estimates are log normally distributed.
In the latter case, the limits are based on the assumption that
actual exposure levels may vary by a factor of two from the estimates
obtained by dispersion modeling.
There are several other uncertainties associated with the
estimated leukemia incidence that are not quantified in these
ranges. Leukemia incidence was calculated based on a no-threshold
linear extrapolation of leukemia risk associated with a presumably
healthy white male cohort of workers to the general population,
which includes men, women, children, nonwhites, the aged, and the
unhealthy. These widely diverse segments of the population may or
may not have differing susceptability to leukemia than do workers
in the studies. Assumptions must be made in order to estimate
exposed populations in census tracts. 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).
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 as a permanent resident or as a transient.
Assumptions that must be made to estimate ambient concentrations by
dispersion modeling and exposed populations by census tract also
2-7
-------
introduce further uncertainties into the risk estimates in addition
to the factor of two contained in the stated ranges. 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; (4) meteorology, which is seldom available for a specific
plant. The particular dispersion modeling 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. For example, an analysis shows that
using the more complex ISC model rather than the less complex' model
used to estimate ambient benzene concentrations for fugitive
emission sources (equipment leaks) would increase the estimated
leukemia incidences for these sources by about 100 to 200
percent (Docket Item IV-B-18). 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. Furthermore,
leukemia is the only health effect of benzene considered. Other
health effects, such as aplastic anemia and chromosomal aberrations
are not quantified. Additionally, the benefits to the general
population of controlling other hydrocarbon emissions from these
fugitive emission sources are not quantified. Finally, these
estimates do not include the cumulative or synergistic effects of
concurrent exposure to benzene and other substances. Although
the current health risk numbers are no longer presented as a
range, they still contain the same uncertainties just described.
The decision to employ estimates of carcinogenic risks despite.
their lack of precision rests on the belief that, although they
are subject to uncertainties, current analytical models and techniques
can, with due consideration of the uncertainties, provide useful
2-8
-------
estimates of relative carcinogenic strength and of the probable
general ranges of excess cancer incidence and individual risks. This
view has been supported by the National! Academy of Sciences,a the
National Cancer Advisory Board,b and others.0 In the case of benzene,
EPA believes that in combination with the nature of the health evidence,
the volume of emissions, and the size and distribution of the
exposed population, the risk estimates are useful in the characterization
and relative comparison of the extent of the benzene health hazard.
At the time the benzene fugitive emissions standard was proposed,
EPA estimated benzene fugitive emissions to be 8,250 Mg/year in 1980
with no further regulation. The proposed standard was expected to
reduce these emissions by approximately 75 percent. At proposal, EPA
estimated that 65 million people live within 20 kilometers of existing
units that produce or use benzene in the petroleum refining and chemical
manufacturing industries. EPA calculated that, in the absence of a
benzene fugitives standard, the annual leukemia incidence would be in
the range of 0.15 to 1.14 cases per year within this population due to
exposure to benzene fugitive emissions. Further, EPA estimated that
for the maximally exposed individual, the chance of contracting leukemia
as a result of exposure to fugitive benzene emissions would be within
a range of 1.7 x 10-4 to 1.2 x 1Q-3 per lifetime.
Based on (1) the magnitude of benzene exposures from fugitive
emission sources of benzene, (2) the resulting estimated maximum
individual risks and estimated incidence of fatal cancers in the
Drinking Water and Health, Part 1, Chapers 1-5, Draft, National
Research Council, National Academy of Sciences, Washington, D.C.,
1977 (Docket No. OAQPS-79-14-II-A-24).
b
"General Criteria for Assessing the Evidence for Carcinogencity of
Chemical Substances," Report of the Subcommittee on Environmental
Carcinogenesis, National Cancer Advisory Board, Journal of the National
Cancer Institute. 58:2, February, 1977 (Docket No. OAQPS-79-14-II-A-27),
c
Hoel, David G., et alI. "Estimation of Risks of Irreversible, Delayed
Toxicity," Journal of Toxicology and Environmental Health 1:133-151,
1975 (Docket No. OAQP$-79-14-II-A-28). ~
2-9
-------
exposed population for the life of existing sources in the category,
(3) the projected increase in benzene emissions as a result of new
sources, (4) the potential to. reduce significantly the risks and
incidence by control requirements, and (5) consideration of the
uncertainties associated with quantitative risk estimates (including
the effects of concurrent exposures to other substances and to
other benzene emissions), EPA found that fugitive emission sources
of benzene pose a significant risk of cancer and warrant the
establishment of a national emission standard under Section 112.
Since the standard was proposed, EPA has revised the baseline
benzene fugitive emissions estimate, the estimated emission reductions
resulting from the standard, and the risk assessment for leukemia risk
from benzene fugitive emissions. (See Section 2.8.2 of this document
for a discussion of changes to the baseline and emission reduction
estimates, and see Appendix C for complete descriptions of the revisions
to the risk assessement). Benzene fugitive emissions are estimated to
be 7,920 Mg/yr in 1980 in the absence of a benzene fugitive emissions
standard. The promulgated standard is expected to reduce these emissions
by approximately 69 percent. An estimated 20 to 30 million people live
within 20 kilometers of existing process units that contain fugitive
emission sources of benzene. This estimate changed since proposal
because the revised risk assessment provides a better mechanism to
determine the number of people that live within 20 kilometers of
existing process units that contain fugitive emission sources of
benzene. EPA has also calculated that, in the absence of a benzene
fugitive emissions standard, the annual leukemia incidence woul.d be
about 0.45 cases per year within this population due to exposure to
benzene fugitive emissions from existing sources. Further, EPA has
estimated that for the maximally exposed group of people, the
chance of contracting leukemia as a result of exposure to fugitive
benzene emissions would be 1.5 x 10~3 per lifetime for existing
sources.
These revised risk estimates are consistent with those that
supported EPA's original determination at proposal. Thus, EPA continues
to consider fugitive emission sources of benzene as a significant
source category of benzene emissions for which risks and incidence can
be reduced significantly.
2-10
-------
As a means of evaluating the need for the standard, the commenters
compare benzene fugitive emissions risk numbers with other types of
regulated or commonly encountered risks. However, these comparisons
must be used only with caution for several reasons. First, as
described above, the benzene fugitive risk estimate and most other
risk estimates, especially those dealing with carcinogens, low
dosages, and varied atmospheric dispersion, contain considerable
uncertainty. Second, the commenters ignore several important
considerations by comparing the benzene fugitive emissions risk to
background risk, commonly accepted risk, or risk addressed by other
government regulations. Making value judgments about risk and
comparing various risks should be performed only in the context of
other factors that weigh in the decision making. These factors may
include (1) whether the risk is undertaken voluntarily, (2) whether
the effect will be immediate, (3) whether other alternatives are
available, (4) whether the exposure is essential or a luxury, (5)
whether the consequences are reversible, and (6) whether the risk
is associated with a common or a "dread" hazard, such as nuclear
holocaust or carcinogenesis. These factors seriously complicate
any comparison of the absolute values of risk attached to different
policies, activities, or technologies.
Congress writes legislation such as the Clean Air Act and implicitly
sets priorities for government regulation of risks. In setting these
priorities, Congress has found that it is not feasible to regulate
every risky activity, technology, or event, but that it is desirable
to regulate those significant risks that can be reasonably reduced.
The congressional process directs priorities towards those risks
that are most repugnant to the public. Ultimately,.government
agencies, through the Congressional process, are given regulatory
authority; but these agencies are each likely to receive and,
therefore, interpret their authority differently, thus addressing
risk differently. Accordingly, EPA is required by Section 112 of
the Clean Air Act to identify and regulate hazardous air pollutants.
As described previously in this response, EPA has determined, based
on -risk and other factors, that benzene fugitive emission sources
warrant federal regulatory action under Section 112.
2-11
-------
The other commenter stated that the "risk levels that EPA has
calculated are not 'significant1 as that term has been used by the
Court." EPA assumes that the commenter refers to the court interpretation
in Industrial Union Department, AFL-CIO v. American Petroleum Institute,
65 L. Ed. 2d 1010, 100 S. Ct. 2844 (1980). This interpretation of the
significance of risk was made in the context of The Occupational
Safety and Health Act of 1970, not the Clean Air Act. It is not
necessarily appropriate to transfer interpretations from one to the
other. In any case, the Court in fact never indicated what actually
constitutes a "significant" risk except to give obvious examples of
what constitutes plainly acceptable and plainly unacceptable risks.
The Court stated: "If, for example, the odds are one in a billion that
a person will die from cancer by taking a drink of chlorinated water,
the risk clearly could not be considered significant. On the other
hand, if the odds are one in a thousand that regular inhalation of
gasoline vapors that are two percent benzene will be fatal, a reasonable
person might well consider the risk significant and take appropriate
steps to decrease or eliminate it" (48 LW 5034). The Court then
stated that it was the duty of the OSHA Administrator to determine,
using rational judgment, the relative significance of the risks associated
with exposure to a particular carcinogen. EPA has determined that the
risk from benzene fugitive emissions represents a significant risk.
In consideration of this, and other factors as described above, EPA
considers fugitive emission sources of benzene to be a significant
source category of benzene emissions for which the risks and incidence
can be reduced significantly.
Comment: Several commenters stated that regulation of benzene
fugitive emissions duplicates existing federal and State regulations
and guidelines. Several of these commenters asserted that fugitive
emissions of benzene will be reduced under other air pollution programs
required by the Clean Air Act. Two commenters stated that some SIP's
involving NAAQS for ozone parallel EPA's proposed provisions for
monitoring, leak repairs, recordkeeping, and reporting (IV-D-20;
IV-D-30). One commenter noted that the Control Techniques Guideline
provisions are expected to reduce existing fugitive emissions by about
60 percent, noting that 93 percent of fugitive emissions affected by
2-12
-------
the proposed standard emanate from sources in ozone nonattainment
areas (IV-D-27). This same commenter added that new sources will be
subject to the proposed VOC new source performance standard (NSPS) for
synthetic organic chemical plants and refineries. Another commenter
stated that 96 percent of the population exposure to benzene emissions
from stationary sources is attributable to emission sources in ozone
nonattainment areas (IV-D-18). This commenter noted that most, if not
all, of these areas will be subject to VOC fugitive emission control
requirements imposed by the 1979-1980 SIP revisions which become
effective prior to 1982.
Several commenters asserted that benzene fugitive emissions are
already adequately controlled under existing OSHA regulations. One
commenter felt that the implementation of the OSHA standard in conjunction
with SIP's should be sufficient for effective control of benzene
fugitive emissions (IV-D-20). One commenter contended that emissions
covered by the standard will impact the workplace before any ex-plant
environmental exposures are encountered, and the emissions must,
therefore, be controlled to meet OSHA rules (IV-D-24). The same
commenter noted that the Supreme Court decision on OSHA's proposed
benzene workplace exposure limit change highlighted the validity of
the present workplace limit (IV-D-24). Another commenter referred EPA
to comments on the SOCMI VOC fugitives standard, dated March 26, 1981
(EPA docket item number A-79-32-IV-D-6). The commenter stated that
the engineering controls considered by OSHA are identical to those
described by EPA in the SOCMI VOC fugitives background information
document.
Response: EPA has decided to regulate emissions of benzene under
Section 112 of the Clean Air Act (42 FR 29332). Given this decision,
EPA considered this comment in the context of whether EPA should
select the alternative of taking no action to regulate benzene fugitive
emissions and relying instead upon the OSHA standard for benzene,
State implementation plans (SIP's) that control VOC emissions, and the
proposed standards of performance for fugitive emission sources of VOC
within the synthetic organic chemical manufacturing industry and
within petroleum refineries. However, after considering this alternative,
2-13
-------
EPA concluded that reliance on this alternative would not be a reasonable
approach to reducing emissions of benzene, a hazardous pollutant
within the meaning of the Clean Air Act.
EPA believes that reliance on the OSHA standard is an unreasonable
approach to reducing public risks associated with fugitive emissions
of benzene. The OSHA benzene standard limits worker exposure to
benzene. Many of the techniques used to attain the OSHA benzene
standard include indirect emission controls such as process modifications,
worker rotations, process or worker isolation, ventilation controls,
modification of work practices, and emissions dilution. These techniques
allow benzene emissions away from the workplace to remain uncontrolled,
or to be removed only from the workplace without requiring that they
be removed from the ambient air. Therefore, EPA concluded that this
approach would protect public health inadequately.
EPA decided against relying upon SIP's to control YOC emissions
from existing plants because not all existing plants are located in
NAAQS nonattainment areas. Furthermore, those that are located in
nonattainment areas might only be required to control emissions using
reasonably available control technology (RACT) rather than the promulgated
standard which provides:additional emission reductions at reasonable
costs. In its revised baseline as discussed in Section 2.8.2, EPA
included emissions from petroleum refineries in NAAQS nonattainment
areas, which are reduced by implementing controls specified in the
petroleum refinery VOC control techniques guideline (CTG) document.
EPA determined that the standards of performance under Section 111 for
SOCMI and refinery fugitive VOC emissions might also be inadequate for
controlling existing sources of benzene fugitive emissions. These
standards will not apply to existing units and, therefore, would not
result in emission reductions to the same extent as the promulgated
standard. Therefore, EPA concluded that this approach would protect
public health inadequately. (Also see Section 3.5.1 of EPA 450/
5-82-003.)
Next, EPA reviewed the problems that might result from any redundancy
or duplication of effort required by these different approaches. As a
result of this review, EPA determined that the only problem might be
associated with redundant and duplicative recordkeeping and reporting.
2-14
-------
This problem could only occur for sources covered by the Clean Air Act.
EPA has eliminated this problem within Section 61.110(b) of the proposed
benzene fugitive emissions standard by eliminating any dual recordkeeping
and reporting between the benzene fugitive emission standard and the
standards for SOCMI and petroleum refinery emission sources of VOC.
Similarly, any records and reports that are required by the States
would be accepted by EPA as long as they contain at least the information
required in the promulgated standard. The OSHA and benzene fugitive
emission standards require no duplication of effort. Duplication
between the benzene fugitive emission standard and the SIP regulations
is similarly nonexistent. Where two regulations cover the same fugitive
emission source, the owner or operator will simply comply with the
more stringent of the two regulations and thereby comply with both.
The owner or operator would not be required to install two sets of
control devices or to perform two separate leak detection and repair
programs for the same units.
Comment: One commenter asserted that the fugitive benzene proposal
seems premature because it addresses less than 4 percent of the health
problem perceived by EPA, while the most significant sources of health
risk identified by the EPA, for example, "urban exposures" and automobile
emissions, have not yet been addressed directly by rule, regulation,
or standard (IV-D-18). •
One commenter stated that the incidental exposure to benzene from
fueling gasoline-powered engines (0.5 to 2 percent benzene) would
probably exceed the concentration of benzene from a refinery (IV-D-10).
One commenter stated that regulation of benzene fugitive 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 National Ambient Air Quality Standards
(IV-F-1).
Response: The main issue in these comments concerns whether to
regulate mobile source-related benzene emissions. 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.
2-15
-------
In any case, EPA's determination that benzene fugitive emissions
are a significant source category is an issue completely separate
from the issue of whether or not mobile sources or other sources
are regulated. As described in the response to the previous comment,
EPA has determined that fugitive emission sources of benzene are a
significant source category of a hazardous air pollutant and thus
require the establishment of a national emission standard under
Section 112 of the Clean Air Act.
Comment: One commenter cited the 24-unit SOCMI fugitive emissions
study* to state that using EPA's own emission estimating technique
along with the most recent data specific to the chemical industry,
benzene fugitive emissions are already at a level for Alternative III
in the proposed regulation for existing sources, which the Administrator
has determined will provide the ample margin of safety required in
Section 112 (IV-F-1, IV-D-21). In addition, one commenter stated that
there are very little reliable data on actual uncontrolled benzene
fugitive emissions, thus casting further doubt on the Agency's having
met its statutory requirements of demonstrating significant risk from
benzene exposure as a basis for the proposed NESHAP's (IV-K-1).
Response: In selecting Regulatory Alternative III, EPA did not
determine that the residual risk was deemed acceptable in absolute
terms. Rather, EPA proposed Regulatory Alternative III for existing
sources in light of the exorbitant additional incremental costs that
would be required to further reduce the risk. EPA first identified
the best available technology (BAT), considering environmental,
economic, and energy impacts, to control fugitive emission sources of
benzene from the various regulatory alternatives considered. EPA then
examined the residual risks due to benzene emissions remaining after
application of BAT to determine whether or not they are unreasonable
in view of the risk reduction that would be gained by requiring a
level of control beyond BAT and the associated increase in costs.
*U.S. EPA, Office of Research and Development (ORD). "Frequency of
Leak Occurrence for Fittings in Synthetic Organic Chemical Plant
Process Units." EPA 600/2-81-003. September 1980. Docket Number
A-79-27-II-A-34.
2-16
-------
Therefore, the determination of whether or not residual risks are
unreasonable was contingent upon the incremental costs associated with
additional reductions of emissions.
In reanalyzing the impacts of the proposed standard, EPA continues
to use petroleum refinery data to estimate uncontrolled benzene emissions.
As discussed in Section 2.8.4, new data developed in the SOCMI fugitive
emissions study and associated with equipment in benzene service do
not estimate emission factors for all types of benzene-service
equipment and, thus, are insufficient for direct estimation of
benzene emissions. Additionally, these data were not gathered for
the purpose of generating emission factors. The petroleum refinery
data base used at proposal of the standard, however, was gathered
to determine VOC emission factors for all leak sources. EPA still
considers, therefore, the controls proposed under Regulatory
Alternative III to be effective in reducing emissions from fugitive
emission sources of benzene.
The third commenter (IV-K-1), in a separate comment on the data
base, explained in greater detail the basis for the comment on data
specific to fugitive benzene emissions. The summary of this detailed
version and EPA's response are found in Section 2.8.4 of this document.
Comment: One commenter ran through a series of calculations
concluding that the actual average individual lifetime leukemia risk
is equal to, or less than, 2.05 x 10"9, or two chances in one billion,
that an individual will contract leukemia as a result of benzene
fugitive emissions. The commenter arrived at this lower estimate by
using altered assumptions that changed two factors in the calculation:
First, the commenter lowered the EPA Carcinogen Assessment Group's
unit risk factor by one order of magnitude. Second, the commenter
lowered EPA's estimate of baseline emissions by 60 percent to account
for CTG regulatory impacts (IV-D-27).
Response: EPA uses maximum lifetime risk as one of the risk
criterion for selecting source categories and levels of emission
control. However, the commenter computed the average lifetime risk,
which is not comparable to maximum lifetime risk.
EPA questions the validity of the commenter's assumptions relating
to the unit risk factor. EPA has discussed the commenter's assumptions
2-17
-------
about the baseline emissions estimate in Section 2.8.2 of this
document and has responded to the comment about the unit risk
factor in EPA-450/5-82-003, which was prepared to address the
listing of benzene under Section 112 of the Clean Air Act.
2.1.3 Standard is Overdue
Comment: One commenter 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 (IV-D-31).
Response: Standards development can be a long and complex task.
EPA attempted to set this standard in a quality and timely fashion.
Accordingly, several years passed since June 1977 before the standard
was proposed and now the standard is being promulgated.
2.1.4 Clean Air Act Authority for EPA to Set Control Requirements
Comment; One commenter felt that EPA lacks the authority to
water down the control requirements of Section 112. The commenter
agreed with 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 (IV-D-31).
Response: Response to this comment can be found in Appendix D,
General Policy.
2.1.5 Authority for Risk-Benefit Analysis
Comment: One commenter asserted that EPA lacks the authority to
engage in risk-benefit analyses as it has in the benzene fugitive
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" (IV-D-31).
2-18
-------
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.
2.1.6 Ability to Quantify Margin of Safety
Comment: According to the commenter, Congress recognized that
under the current state of scientific knowledge, an ample margin of
safety cannot be quantified. The commenter stated that Congress did
not require EPA to delay regulation until this precise quantification
was possible, nor did Congress force EPA to undertake "sham" quantitative
risk analyses (IV-D-31).
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, in discussing the use of quantitative risk
assessment.
2.1.7 Suspension of Benzene Proceedings
Comment: Several commenters maintained that EPA should temporarily
suspend or postpone proceedings on the benzene source-specific standards
until EPA makes final decisions on the listing of benzene (IV-F-1;
IV-D-19; IV-D-25; IV-D-28; IV-D-3). One commenter noted that a brief
suspension of the proceedings will not result in undue delay (IV-D-3).
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.
2.2 HEALTH EFFECTS AND RISK ASSESSMENT OF BENZENE EXPOSURE
Several commenters criticized the methods EPA used to estimate
risk from fugitive benzene emissions. Some commenters felt that EPA
had overestimated risk. One felt that EPA had underestimated it.
These and other related comments are summarized below.
2.2.1 Overestimation of Risk
Several commenters felt that EPA had overestimated the risk from
fugitive benzene emissions. Two common reasons were given for EPA's
Overestimation of risk: (1) that the EPA Carcinogen Assessment Group
(CAG) overestimated the benzene dose-response relationship; and (2) that
2-19
-------
EPA's exposure assessments are flawed by inadequate data and unnecessary
or inaccurate assumptions. These comments are summarized below.
2.2.1.1 Dose-Response. Comment: Two commenters stated that
EPA's assumption that leukemia risk can be extrapolated from high
doses to very low doses is not justified by available direct evidence
(IV-D-27; IV-D-29). One of the commenters 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. The commenter contended that available
empirical evidence suggests the absence of health effects below 10 parts
per million. According to the commenter, 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, the commenter adds there is
no direct evidence that benzene is carcinogenic or leukemogenic at
levels below 100 parts per million. The commenter noted that 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. The commenter feels
that no leukemia risk can be substantiated at these low levels of
exposure (IV-D-27).
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: One commenter noted that, in its ruling on OSHA's
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 19 ppm
level had ever in fact caused leukemia." According to the commenter,
EPA has based its evaluation of public exposure to benzene storage
vessels on an estimated maximum out-of-plant concentration of 18.7
parts per billion, significantly below the levels addressed in the
OSHA proceedings.
2-20
-------
The commenter acknowledged that there is some epidemiologic data
which appear to support an association between leukemia and benzene at
high concentrations. However, the commenter added that the leukemogenic
action of benzene at these high levels is preceded by blood changes,
such as cytopenia and pancytopenia, and that these pre-1eukemic changes
do not occur at levels below about 35 ppm. According to the commenter,
one study of benzene-exposed pliofilm rubber workers which 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, the commenter
noted that no excess leukemia was found. As the commenter stated, the
exposures in the Tatter two groups, although not precisely quantifiable,
were clearly much greater than the nonoccupational exposures in the
.community (IV-D-24). Another commenter made a similar general statement
(IV-F-1).
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 in the background document for benzene
emissions from benzene storage tanks (EPA-450/3-80-034b).
Comment: One commenter maintained that the EPA has overestimated
the benzene risk factor by an order of magnitude. The commenter noted
that a valid risk factor for benzene exposure, which was developed by
Dr. Steven Lamm and confirmed by Dr. Richard Wilson, is 0.031 x 10~6
cases per year per person-ppb-year, whereas the EPA has utilized a
factor of 0.34 x 10~6 cases per year per person-ppb-year. According
to the commenter, taking the valid factor into account, a maximum of
only two cases would be anticipated (nationwide) due to 20 years of
exposure to existing benzene fugitive emissions. Utilizing EPA's
assumption of an overall efficiency of emission control of 77 percent,
the commenter concluded that this regulation would save 77 percent of
from 0.3 to 2.1 people or 1 person (IV-D-24).
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.
2-21
-------
Comment: One commenter 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 incidence, and those, leukemias that did occur were 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
(IV-D-27).
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.
2.2.1.2 Exposure Assessment. Comment: Two commenters 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 (IV-D-27; IV-F-1).
One commenter maintained that EPA's benzene fugitive emissions
exposure analysis relied upon incomplete and inaccurate meteorologic
data. Rather than use site-specific climatological data as required
by EPA guidelines, the commenter remarked that the analysis relies
entirely on conditions in Houston to apply to refineries and chemical
plants in 23 States and territories, from New York to California.
According to the commenter, 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." The commenter concluded
that since climatological data for approximately 300 sites throughout
the U.S. are available in EPA archives, EPA's total reliance on
Houston meteorology was not justified (IV-D-27). The commenter also
stated that EPA arbitrarily oriented the fugitive emission sources of
2-22
-------
a hypothetical model refinery in order to maximize the ambient concen-
trations at the plant boundary. According to the commenter, this was
done despite the fact that the actual location of emission sources at
the Sauget, Illinois, plant could easily be ascertained (IV-D-27).
One commenter asserted that EPA failed to validate the results of
its air quality modeling as a check on its accuracy, as required by
EPA guidelines. According to the commenter, in this exposure
analysis, 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
resulted in an unacceptably high degree of uncertainty in EPA's
exposure estimates; in some instances, the exposure estimate may be'
off by a factor of 100 or more (IV-D-27).
Response: The commenter is correct in noting that the benzene
fugitives risk assessment did not make use of plant-specific data
relating to emissions, meteorology, or plant configurations. However,
as explained below, the plant-specific approach probably would not
improve the precision or accuracy of the results enough to justify
/ f
the level of effort to use more specific data. EPA has concluded
that a plant-specific approach would be too costly and not necessary
for benzene fugitive emission sources. In response to this comment,
EPA has revised its original risk assessment for benzene fugitive
emissions. Even though the revised risk assessment methodology has
changed since proposal, it is similar to the one used at proposal.
The revised risk assessment methodology is described in Appendix C.
The revised methodology uses an improved calculation technique and
input values. For example, the unit risk factor has been recalculated;
there is a new baseline emission estimate (see Section 2.8.2);
different plants are included in the data base; there are new
emission reduction estimates (see Section 2.8.2); and better population
and meteorology data have been used.
EPA considered the option of using plant-specific data for all
parameters in order to run an exposure model for each plant. EPA
considered the uncertainty that would result using the plant-specific
data approach with the uncertainty that would result using the model
2-23
-------
plant and extrapolation approach. EPA also considered 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. Of course, the use of more precise and
accurate data allows better results to be calculated. However, 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 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 (ISC-LT) and the Human Exposure
Model, even with perfect input data, are subject to substantial
uncertainty. (The ISC-LT model, even with state-of-the-art input
data, is estimated to have a 95 percent confidence interval of plus
or minus a factor of two.)
EPA has not exaggerated the precision of the results of the model
plant extrapolation method, nor has 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 (see Section 2.1.2 of this document). The results
are presented in highly aggregate, nonspecific terms, in a fashion
that exhibits much less uncertainty than if EPA tried to squeeze more
detailed, refined results from the extrapolation. Using the model
plant extrapolation method, less accurate deviations in the results for
specific plants tend to average out when the total national incidence
is computed. Attempting to validate the results of the air quality
modeling would require an extremely detailed, burdensome, and costly
plant-specific approach. 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
2-24
-------
approach, EPA has elected to use the model plant extrapolation
approach.
Comment: One commenter noted that the risk estimate used by EPA
assumed no government regulation of benzene fugitive emissions, while
in fact the vast majority of such emission sources will be required to
reduce emissions by State implementation plans for ozone. According
to the commenter, approximately 93 percent of the risk from fugitive
benzene emissions comes from sources in ozone nonattainment areas.
These sources will be subject to emission control requirements at
least as stringent as those contained in EPA's control techniques
guidelines (CTG's) for refineries and synthetic organic chemical
plants. The commenter notes that EPA concedes that these requirements
will reduce benzene emissions by 60 percent (IV-D-27).
Response: In its revised risk assessment (see Appendix C of this
document), EPA has used a new baseline for benzene fugitive emissions
that accounts for emission reductions that will be achieved in 1982 as
a result of refinery CTG control requirements (see Section 2.8.2
of this document for discussion of the new baseline). CTG control
requirements are incorporated in the State implementation plans.
Petroleum refining units in nonattainment areas in 1982 are assumed
to be implementing CTG controls. All other units are assumed to
have uncontrolled emissions. EPA expects that CTG controls will
reduce total national fugitive benzene emissions in 1985 from the
uncontrolled level by about 8 percent. The CTG control requirements
are included in the baseline estimate of risk.
Comment: One commenter added that deficiencies exist in the
population concentration estimates contained in the exposure analysis.
According to the commenter, EPA assumed that population is distri-
buted uniformly in all directions at each site, which introduces an
uncertainty factor of 10 to 100 into the overall exposure estimate
(IV-D-27).
Response: EPA's revised risk estimate (see Appendix C) was based
upon a more sophisticated population exposure model, which utilized a
population data base characterized as having a high level of resolution.
The Human Exposure Model (HEM) was used to estimate the population
that resides in the vicinity of each receptor coordinate surrounding
2-25
-------
each plant. The HEM does not assume population is distributed evenly
around each plant. The population "at risk" to benzene exposure was
considered to be persons residing within 20 km of the plants. The
population around each plant was determined by specifying the geo-
graphical coordinates of that plant.
A slightly modified version of the "Master Enumeration District
List—Extended (MED-X)" data base, a Census Bureau 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 used to produce a randomly
accessible computer file of only the data necessary for the exposure
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 plant's geographical' coordinates and the concentration patterns
computed by the model plant extrapolation method were used as input to
the HEM. For each receptor coordinate, the concentration of benzene
and the population estimated by the HEM to be exposed to that parti-
cular concentration are identified. The-HEM multiplies these two
numbers to produce population exposure 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 exposure. 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 generally much smaller
than ED/BG1s; 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 exposures are calculated differently for
the ED/BG's located within each region.
2-26
-------
For ED/B6 centroids located between 0.1 km and 2.8 km from the
emission source, populations are divided between neighboring concen-
tration 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 the population centroids between 2.8 km and 20 km from the
source, a concentration grid cell, the area approximating a rectan-
gular 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 from the source,
the entire population of the ED/BG is assumed to be exposed to the
concentration that is geometrically interpolated radially and
arithmetically interpolated 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 were 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 and
linear interpolation.
Comment: One commenter stated the analysis failed to account for
population activity patterns and population mobility, thereby over-
estimating exposure levels for persons residing in the affected area
surrounding these plants (IV-D-27).
Response: As discussed in Section 2.1.2, there are several
uncertainties associated with exposure assessments. One of these
uncertainties concerns population activity patterns and mobility.
2-27
-------
While EPA's exposure assessment does not account for people moving
into or out of this area, there is no reasonable way to know if the
maximum exposed persons actually leave the area. In light of all
of these factors, EPA has concluded that its exposure assessment
is still a useful tool.
Comment: One commenter criticized EPA's estimate of "maximum
individual' lifetime risk" by noting that EPA has no evidence that any
individual ever lives an entire lifetime 0.1 kilometers from the plant
at a point of maximum benzene concentration (IV-D-27).
Response: Response to this comment can be generally found in
EPA-450/5-82-003, which was prepared to address the listing of
benzene under Section 112 of the Clean Air Act.
The maximum individual lifetime risk, as the commenter understood,
is the risk associated with exposure to the maximum concentrations.
Maximum concentrations are only modeled estimates and may overestimate
or underestimate the actual concentrations. As discussed in docket
item IV-B-18, the maximum concentrations and, consequently, the
maximum individual lifetime risks (which were estimated and used to
make, to the limited extent they were used, decisions) appear to be
underestimates. Provided the air at 0.1 kilometer from plant is
located in a neighborhood, the opportunity for exposure exists.
Using the HEM, exposures to maximum concentrations are generally
limited to distances greater than 0.2 kilometer and to locations
where people reside. In the absence of perfect information
regarding the magnitude and deviation of exposure, it is prudent to
assume that, as a "maximum," an individual could face continuous
exposure to a maximum concentration.
2.2.2 Underestimation of Risk
Comment: One commenter felt that EPA had understated the risk of
exposure to benzene fugitive 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 multipli-
cative effect than that such chemicals will cancel each other out.
This commenter cited many sources of uncertainty in the risk assessment
2-28
-------
and concluded that EPA may have drastically understated the real
leukemia risk associated with benzene. According to the commenter,
the estimates given by EPA at proposal may well underestimate the
health benefits of the increment between the proposed requirements
and use of vapor recovery or thermal destruction of emissions. The
commenter added that it is unacceptable that the noncarcinogenic
effects of benzene exposure have virtually dropped out of 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 Regulatory Alternatives IV or V, in favor of Alternative
III (IV-D-31).
Response: EPA considered this comment in the light of the decision
to promulgate a standard for benzene fugitive emissions (equipment
leaks of benzene). In the context of this standard, EPA estimated
exposure to benzene fugitive emissions and concluded that these
emissions warrant federal regulation under Section 112. In addition,
EPA estimated exposure to these emissions after application of BAT
and after application of more restrictive control and concluded
that BAT is the appropriate level of control for this standard.
These are the only uses of exposure assessment in this standard,
and neither of these uses is affected by the commenter's contention
that the exposure assessment underestimates leukemia risk.
While the commenter may be correct that interactions and synergisms
(resulting from exposures to multiple chemicals) may be additive or
multiplicative (or antagonistic) and therefore result in truly
greater (or smaller) risks to persons exposed to benzene, EPA is unable
to estimate these effects and, therefore, has not considered them.
It should be noted that many of the factors used in making the
exposure assessment have uncertainties associated with them and
that these uncertainties can result in underestimation as well as
overestimation. These uncertainties have been considered as much
as practicable by EPA in setting this standard. EPA, aware of
these uncertainties, used the exposure assessment in the two ways
described above in a reasonable manner to establish this standard.
2-29
-------
Comment: One commenter 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).
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 shorter than 20 years. Roughly, a plant will have a lifetime
of 20 years. Since there are little data available that estimate
plant lifetimes, EPA considers 20 years to be a reasonable estimate.
The length of the lifetime of plants using or producing benzene,
however, was not a factor considered in selecting the final standard.
2.2.3 Consistency in Methodologies
Comment: Two commenters stated that there should be some consistency
in risk assessment methodologies between the four current benzene
proposals (IV-D-8; IV-D-24). 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 the
commenter's review of the four current benzene proposals, a great deal
of duplication has occurred with little or no health benefit to the
public (IV-D-24).
Response: EPA has tailored the risk assessment methodologies
among the four benzene standards to be as consistent as practical.
However, an integrated analysis is not feasible, and thus some duplication
is unavoidable. Primarily due to the large numbers of plants, it is
not feasible for the fugitive benzene emissions or benzene storage
2-30
-------
risk assessments to achieve the same degree of detail and precision as
was achieved in the maleic anhydride or ethylbenzene/styrene (EB/S)
risk assessments. The maleic anhydride and EB/S assessments calculated
risk separately for each of 8 and 15 plants, respectively, using
plant-specific data supplied voluntarily by the plants as part of
their public comments. The benzene fugitives and benzene storage data
bases contain about the same number of plants. To supply plant-specific
data on emissions, meteorological data, and plant configuration would
require that these plants implement a leak detection "and repair program,
monitor meteorological conditions, and inventory detailed specifics on
plant configuration. As described in a previous comment in Section 2.2.1.2,
EPA believes that such an effort is unreasonably costly and unnecessary.
Comment: One commenter noted that the exposed population cited
in this regulation was 65 million persons, while the exposed population
cited for benzene emissions from benzene storage vessels was 85 million
persons. Since the population sizes are based on fixed size areas
surrounding benzene producers and users facilities, they should be
identical. According to the commenter, the concentration outside the
plant boundary may differ for the two emission sources, but the populated
area surrounding the plant should not (IV-D-24).
Response: The commenter is correct in observing that the exposed
population in the two risk assessments should be the same. As a
result of revisions to the risk assessments made since proposal, EPA
has adjusted this figure to read that 20 to 30 million persons will be
exposed to benzene emissions from either fugitive emission sources or
benzene storage vessels.
2.3 SELECTION OF THE FINAL STANDARD
Comment: Many people commented on the basis for selection of the
proposed standard. Several commenters questioned the cost effectiveness
and impacts of Regulatory Alternatives III and IV for existing and new
sources, respectively. Some of the commenters recommended the selection
of less stringent regulatory alternatives, and some recommended the
selection of a more stringent regulatory alternative. Other commenters
stated that selection should be based on the cost and emission reduction
impact of each control technique rather than regulatory alternatives.
2-31
-------
Response: After considering these comments, EPA selected the
final standards for new and existing equipment in benzene service.
Selection of the basis of the final standard was a two-step process
and was similar to the approach used when the standard was proposed.
The first step was the selection of the best available technology
(BAT). Best available technology for new and existing equipment in
benzene service is technology which, in the judgment of the Administrator,
is the most effective level of control considering economic, energy,
and environmental impacts and any technological problems associated
with the retrofitting of existing equipment. After consideration
of these impacts for each alternative control technique, one set of
control techniques was selected as BAT for new and existing equipment
in benzene service.
After selecting certain control techniques as BAT, EPA evaluated
the estimated health risks remaining after application of BAT to
determine if they are unreasonable in view of health risk reductions
and cost (economic) impacts that would result if the next more
stringent level of control were applied. This provides a comparison
of the costs and economic impacts of control with the benefits of
further risk reduction. The benefits of risk reduction are expressed
in terms of the estimated leukemia incidence within 20 kilometers
of the equipment covered by the standard and the estimated maximum
lifetime risk at the point of maximum exposure. The results of
this comparison determine whether, in the judgment of the Administrator,
the residual risks remaining after application of BAT are unreasonable.
If the risks remaining after application of BAT are judged to be
unreasonable, further controls would be required.
The costs of the control techniques for benzene equipment
leaks are very small relative to the capital and operating'
costs of affected process units. As a consequence, none of these
control techniques impacts the ability of an owner or operator to
raise capital or measurably impact product prices or energy require-
ments. Therefore, EPA selected BAT primarily based on a comparison
of costs and emission reductions associated with each alternative
control technique. In making this decision, EPA is accepting the
2-32
-------
suggestions of commenters to consider further cost per unit emission
reduction estimates and to consider these estimates for each type
of equipment covered by the standard in the selection of BAT. In
selecting BAT, EPA initially selected control techniques that
achieve the greatest emission reduction with reasonable control
costs per rnegagram of emission reduction. The emission reductions
and the average and incremental costs per megagram of benzene and
total emissions (including benzene and other volatile organic
compounds (VOC)) are summarized in Tables 2-2 and 2-3, respectively,
for each type of equipment covered by the standard. After initially
selecting one set of control techniques as BAT for each type of
equipment covered by the standard, EPA analyzed economic and other
impacts of this set of control techniques. To the extent that
these impacts were reasonable, the control techniques were selected
as BAT and then'were used in estimating the risks remaining after
application of BAT.
For each type of equipment, the average cost effectiveness of
each control technique was calculated based on the net annualized
cost and the annual emission reduction from the uncontrolled level.
Starting with the most stringent control technique, which achieves
the greatest emission reduction at the greatest annualized cost,
EPA examined the incremental cost effectiveness between the most
stringent control technique and the next less restrictive control
technique. The incremental cost effectiveness between any two
alternative control techniques was based on the difference in net
annualized costs divided by the difference in the annual emission
reductions of the alternate control techniques. If the incremental
cost in comparison to the incremental emission reduction is judged
unreasonable, then the next increment is examined until a control
technique with a reasonable incremental cost in comparison to the
incremental emission reduction is available.
Costs per megagram of emission reduction (average and incremental)
were calculated in terms of total emissions (benzene and other VOC)
as well as benzene alone. Control of benzene equipment leaks
2-33
-------
Table 2-2. CONTROL COSTS PER MEGAGRAM OF BENZENE REDUCEQa
Type of
Equipment
Valves
Pumps
Compressors
Pressure
Relief
Devices
Open-ended
Lines
Sampling
Connection
Systems
Product
Accumulator
Vessels
Control Technique
Annual leak detection and
repair
Quarterly leak detection
and repair
Monthly leak detection
and repair"1
Sealed bellows valves
Annual leak detection and
repair
Quarterly leak detection
and repair
Monthly leak detection
and repair^
bual mechanical seal
systems
Degassing reservoir
vents i
Quarterly leak detection
and repair'
Monthly leak detection
and repair1
Equipment control ft1
Caps on open endsf
Closed-purge sampling^
Closed-vent systemf
Benzene Emission
Reductionb
• (Mg/yr)
New
163
639
736
998
77
266
307
372
3.5
53
58
82
54
87
27
Existing
799
'2,750
3,160
4,280
290
959
1,140
1,360
—h
190
207
295
187
318"
106
Average
S/Mg Benzene0
New
e
e
— e
8,500
870
—e
— e
2,100
e
e
e
96
430
880
94
Existing
e
_.e
— e
11,000
870
e
e
2,400
—h
e
—e
180
470
900
' 97
Incremental
S/Mg Benzene*!
New Existing
--e ..e
— e ..e
210 120
33,000 44,000
870 870
— e ..e
--e ..e
13,000 15,000
— e ..h
— e ..e
300 290
1,400 1,700.
430 470
880 900
94 97
2-34
-------
Table 2-2. CONTROL COSTS PER MEGAGRAM OF BENZENE REDUCED*
(concluded)
aCosts and emission reductions are presented on a nationwide basis and
are derived from Docket Item IV-B-14.
^Benzene emission reductions are presented on a nationwide basis
as explained in Docket Item IV-B-14.
cAverage dollars per megagram (cost effectiveness) = net annualized cost
* annual benzene emission reduction. These cost-effectiveness numbers
can be calculated on a component basis, on a model unit basis, or on a
nationwide basis. In any case, the resulting cost effectiveness will be
essentially the same. The numbers in this table have been calculated
on a nationwide basis by multiplying the net annual cost per component
(BID Tables A-l through A-ll) by the total number of components nationwide
(BID Tables 2-6 and 2-7) and then dividing the resulting nationwide net
cost by the nationwide emission reduction, as explained in Docket
Item IV-B-14.
^Incremental dollars per megagram = (net annualized cost of the control
technique - net annualizecTcost of the next less restrictive control
technique) *• (annual benzene emission reduction of control technique -
annual benzene emission reduction of the next less restrictive control
technique).
eDashes denote savings.
^Control technique selected as the basis for the final standard.
SEmission reduction associated with one new compressor.
"Existing compressors in benzene service are not known to exist;
however if one does, the emission reduction and control costs per
megagram of benzene would be the same as for a new compressor.
^Controls for pressure relief devices are based on the following:
75 percent of relief devices are already controlled. For the
remaining uncontrolled sources, "equipment controls" means 75
percent of relief devices will be vented to a flare, 12.5 percent
will be controlled by rupture disk/block valve systems, and
12.5 percent will be controlled by rupture disk/3-way valve systems.
2-35
-------
Table 2-3. CONTROL COSTS PER MEGAGRAM OF TOTAL EMISSIONS REDUCED3
Total Emission
Reduct1onb
Type of
Equipment
Valves
•
Pumps
Compressors
Pressure
Relief
Devices
Open-ended
Lines
Sampling
Connection
Systems
Product
Accumulator
Vessels
Control Technique
Annual leak detection and
repair
Quarterly leak detection
and repair
Monthly leak detection
and repair'
Sealed bellows valves
Annual leak detection and
repair
Quarterly leak detection
and repair
Monthly leak detection
and repair"1
Dual mechanical seal
systems
Degassing reservoir
vents '
Quarterly leak detection
and repair'
Monthly leak detection
and repair
Equipment controlf»'
Caps on open ends^
Closed-purge sampling^
Closed- vent system11
(Mg/yr)
New
313
1,005
1,150
1,540
124
413
484
584
5.
83
90
128
83
136
42
Existing
1,306
4,440
5,090
6,960
468
1,560
1,830
2,210
59 --h
308
336
475
313
510
171
Average
$/Mg
New Existing
e ..e
— e ..e
—e _.e
4,900 6,900
540 540
— e e
— e ..e
1,400 1,500
— e ..h
— e ..e
— e e
61 110
280 280
560 560
60 60
Incremental
$/Mg
New
e
e
140
20,000
540
e
e
8,200
e
e
190
940
280
560
60
Existing
e
e
74
26,000
540
__e
e
8,700
— h
e
180
1,100
280
560
60
2-36
-------
Table 2-3. CONTROL COSTS PER MEGAGRAM OF TOTAL EMISSIONS REDUCED3
(concluded)
aCosts and emission reductions are presented on a nationwide basis and
are derived from Docket Item IV-B-14.
''Total emission reductions are estimated for benzene and other VOC
and are presented on a nationwide basis as explained in Docket Item
IV-B-14.
cAverage dollars per megagram (cost effectiveness) = net annualized cost
* annual emission reduction. See Table 2-2, footnote c.
dlncremental dollars per megagram - (net annualized cost of the
control technique - net annualized cost of the next less restrictive
control technique) *• (annual emission reduction of the control
technique - annual emission reduction of the next less restrictive
control technique).
eDashes denote savings.
^Control technique selected as the basis for the final standard.
SEmission reduction associated with one new compressor.
"Existing compressors in benzene service are not know to exist;
however, if one does, the emission reduction and control costs per
megagram of total emissions would be the same as for a new compressor.
"Controls for pressure relief devices are based on the following:
75 percent of relief devices are already controlled. For the remaining
uncontrolled sources, ''equipment controls" means 75 percent of
relief devices will be vented to a flare, 12.5 percent will be
controlled by rupture disk/block valve systems, and 12.5 percent
will be controlled by rupture disk/3-way valve systems.
2-37
-------
results in the destruction of other organic compounds (mainly
VOC's) as well as benzene; therefore, control of VOC is an added
benefit of controlling benzene. In making decisions about the
acceptability of the cost of emission reductions achieved by a
control technique, it is appropriate to consider the VOC as well as
the benzene emission reductions. However, VOC emission reductions
were considered only in the sense that VOC emission reductions can
add weight to selecting a control technique as BAT.
The basis for selecting BAT for each type of equipment in
benzene service is discussed below. It should be noted that the
control costs for each type of equipment do not represent the
actual amounts of money spent at any particular plant site. The
cost of emission reduction systems will vary according to the
chemical product being produced, production equipment, plant layout,
geographic location, and company preferences and policies. However,
these costs and emission reductions are considered typical of
control techniques for benzene equipment leaks and can be used in
selecting the level of control to be required by the standard.
Valves. EPA first considered the use of sealed bellows valves.
However, a requirement for sealed bellows valves is considered
unreasonable. The incremental cost for the use of sealed bellows
valves compared to a monthly leak detection and repair program for
valves is about $49 million/yr for existing valves and results in
an emission reduction of about 1,100 Mg/yr of benzene; this represents
an incremental cost effectiveness of about $44,000/Mg. Furthermore,
the use of sealed bellows valves results in an incremental emission
reduction of 1,870 Mg/yr of total emissions (including benzene and
other VOC). Because the incremental cost effectiveness ($44,000/Mg
of benzene) of this control technique is relatively high and because
the additional emission reduction of VOC does not add enough weight
to convince EPA that the costs are reasonable, EPA decided not to
require the use of sealed bellows valves.
Next, EPA considered several leak detection and repair programs
for valves. The leak detection and repair programs differed in the
monitoring frequency that could be implemented. As Tables 2-2 and
2-3 show, all of the monitoring programs would result in net credits
2-38
-------
because the value of recovered product resulting from Implementation
of each program is greater than the cost of each program. The
largest emission reduction is associated with the monthly program.
The incremental cost for a monthly program compared to a quarterly
program is about $48,000/yr for existing valves and results in an
incremental benzene emission reduction of about 400 Mg/yr; this
represents an incremental cost effectiveness of about $120/Mg of
benzene. Because EPA considers the incremental cost effectiveness
associated with the monthly program to be reasonable, EPA selected
a monthly leak detection and repair program as BAT for valves.
Pumps. The costs and emission reductions associated with the
control of pumps were determined for three leak detection and
repair programs and the use of dual mechanical seal systems. The
incremental cost associated with the use of dual mechanical seal
systems (compared to the monthly program) on existing pumps is
about $3.3 million/yr and results in an incremental emission reduction
of about 220 Mg/yr of benzene; this represents an incremental cost
effectiveness of $15,000/Mg. Furthermore, the use of dual mechanical
seal systems reduces an incremental 380 Mg/yr of total emissions
(including benzene and other VOC). Because the incremental cost
effectiveness ($15,000/Mg of benzene) is relatively high and because
the additional emission reduction of VOC does not add enough weight
to convince EPA that the costs are reasonable, EPA decided not to
require the use of dual mechanical seal systems.
Next, EPA considered leak detection and repair programs for
pumps. Monthly and quarterly leak detection and repair programs
both result in net annual savings. The annual program results in a
cost of about $250,000/yr for a benzene emission reduction of about
290 Mg/yr; this represents a cost effectiveness of about $870/Mg of
benzene. EPA decided not to select an annual program as BAT because
it results in higher costs and lower emission reductions in comparison
to either a monthly or a quarterly program. The monthly program
achieves a higher degree of control than the quarterly program at a
slightly lower credit. The monthly program (compared to the quarterly
program) results in a savings and an incremental emission reduction
of about 180 Mg/yr of benzene; this represents an incremental cost
2-39
-------
effectiveness of essentially zero for existing pumps in benzene
service. Because EPA considers the incremental cost effectiveness
associated with monthly monitoring of pumps reasonable, monthly
leak detection and repair was selected as BAT for pumps.
Compressors. One control technique is considered viable for
compressors in benzene service—the installation of equipment such
as control of the barrier fluid system. The installation of control
equipment results in an emission reduction of about 3.5 Mg/yr per
compressor at a savings in cost for each compressor. The control
technique results in a savings because the value of product retained
by controlling the barrier fluid system exceeds the cost of the
control equipment. Since this cost is reasonable, control equipment
was selected as BAT for compressors.
For existing compressors, EPA proposed a monthly leak detection
and repair program. At proposal, EPA selected, as the basis of the
standard for existing pumps and compressors, a monthly leak detection
and repair program because some existing pumps could not be reasonably
retrofitted with dual mechanical seal systems and because the cost
of retrofitting compressors and replacing pumps (in comparison to the
cost of a leak detection and repair program) was considered exorbitant
in light of the resulting incremental emission reductions. Since
proposal, EPA has concluded that the effectiveness of a leak detection
and repair program for compressors in benzene service would be zero.
Also, as requested by commenters, EPA has analyzed compressors
separately from pumps. This analysis yields the cost and emission
reduction estimates above. EPA has thus concluded that control equip-
ment is reasonable for existing compressors and, therefore, selected
it as BAT for existing compressors as well as new compressors.
Pressure relief devices. The annualized costs and emission
reductions associated with monthly and quarterly leak detection and
repair programs and with the use of control equipment (rupture
disks and flares) were determined for pressure relief devices in
gas service. As Tables 2-2 and 2-3 show, both the quarterly and
monthly leak detection and repair programs are less expensive than
2-40
-------
installation of .equipment controls, but they result in lower emission
reductions. These programs result in an incremental cost effective-
ness of about $300/Mg of benzene for the monthly program (compared
to the quarterly program) and a credit for the quarterly program.
Equipment controls would result (compared to a monthly program) in
incremental emission reductions of about 30 percent and incremental
costs of $l,500/Mg of benzene and $l,700/Mg of benzene ($940/Mg of
total emissions and $l,100/Mg of total emissions) for new and
existing equipment, respectively, compared to a monthly leak detection
and repair program. However, for new pressure relief devices, the
incremental cost effectiveness of comparing equipment controls with
monthly monitoring could range from about $l,200/Mg of benzene or
$760/Mg of total emissions (for rupture disk/block valve systems
and venting to a flare) to about $3,000/Mg of benzene or $l,900/Mg
of total emissions (for rupture disk/3-way valve systems). And for
existing pressure relief devices, the incremental cost effectiveness
could range from about $l,300/Mg of benzene or $820/Mg of total
emissions for flare systems to about $3,800/Mg of benzene of $2,400/Mg
of total emissions for rupture disk/3-way valve systems. Because
the overall incremental cost effectiveness of equipment controls
($l,500/Mg of benzene or $940/Mg of total emissions for new equipment
and $l,700/Mg of benzene or $l,100/Mg of total emissions for existing
equipment) and average costs are considered reasonable and because
equipment controls provide a significant incremental emission
reduction, they were selected as BAT for pressure relief devices.
Open-ended lines, sampling connection systems, and product
accumulator vessels. EPA considered caps or closures as the control
technique for the standard for open-ended lines. Costs of $430/Mg
and $470/Mg of benzene are reasonable for controlling equipment
leaks of benzene from new and existing open-ended lines, respectively.
EPA.selected caps or closures as BAT for open-ended lines.
EPA considered closed-purge sampling as the control technique
for the standard for sampling systems. Costs of $880/Mg and $900/Mg
of benzene are reasonable for controlling equipment leaks of benzene
from new and existing sampling systems, respectively. EPA selected
closed-purge sampling as BAT for sampling systems.
2-41
-------
EPA considered closed-vent systems connected to a control
device as the control technique for the final standard for product
accumulator vessels. For existing units in benzene service, the
installation of closed-vent systems connected to a control device
will result in a nationwide net annual cost of $10,300 and an annual
emission reduction of about 100 Mg of benzene; this represents a
cost effectiveness of about $100/Mg. Since the cost associated
with this control technique is reasonable, EPA selected closed-vent
systems as BAT for product accumulator vessels.
Economic impact considerations of BAT. As mentioned above,
once BAT was identified for each type of equipment covered by the
standard, EPA analyzed the economic impact of the initial set of
BAT control techniques. As a result and as explained in the next
section of this preamble, EPA concluded that the control techniques
initially selected as BAT have reasonable economic impacts. In
addition, EPA has also concluded that other impacts, environmental
and energy, associated with these control techniques are reasonable.
Thus, they were selected as BAT for equipment in benzene service.
Selection of the final standards. After selecting certain
control techniques as BAT (those identified above), EPA evaluated
the estimated health risks remaining after application of BAT to
determine if they are unreasonable in view of health risk reductions
and cost (economic) impacts that would result if a more stringent
level of control were applied. Because the most stringent, viable
control technique for each type of equipment covered by the standard
is already selected for all types of equipment except for valves
and pumps, EPA identified a more stringent level of control by
reviewing the control techniques for valves and pumps. The more
stringent level of control used for this analysis includes the use
of dual mechanical seal systems on pumps in addition to the require-
ments selected as BAT. This control technique was selected for
analysis because it adds the next most cost-beneficial control
technique. Thus, if EPA decided not to require this control
technique in addition to those control techniques selected as BAT,
then EPA would not require less cost-beneficial control techniques,
such as sealed bellows valves.
2-42
-------
Health and cost impacts were first examined for existing
equipment covered by the standard to determine whether a more
stringent level of control should be required. Requiring a more
stringent level of control instead of BAT would reduce estimated
leukemia incidence within 20 kilometers of the equipment covered by
the standard from about 0.14 cases per year to about 0.13 cases per
year for existing equipment. It would reduce the estimated maximum
lifetime risk at the point of maximum exposure from about 4.5 x
10~4 to about 4.2 x 10~4. Due to the assumptions used in calculating
the health numbers, there is uncertainty associated with the numbers
presented here. The uncertainties associated with these numbers
are explained in Section 2.1.2 entitled "Need for the Standard."
Requiring the more stringent level of control rather than BAT would
increase capital cost from $5.5 million to $19.5 million and would
increase net annualized costs from $400 thousand to a cost of $3.7
million for existing equipment. Because of the relatively small
health benefits to be gained with the additional costs of requiring
the more stringent level of control instead of BAT for existing
equipment, EPA considers the risks remaining after application of
BAT to existing equipment not to be unreasonable. For this reason,
EPA judged the level of control selected ais BAT to provide an ample
margin of safety and decided not to require a more stringent level
of control than BAT for existing equipment.
Health and cost impacts were next examined for new equipment
covered by the standard to determine whether a more stringent
control level should be required. As with existing equipment, EPA
considered the use of dual mechanical seal systems on pumps as the
more stringent control level that is next most cost beneficial.
Thus, if EPA decides not to require the use of these seals, then
EPA would not require less cost beneficial control technologies,
such as sealed bellows valves. Requiring the more stringent level
of control--the use of dual mechanical seals on pumps in addition
to BAT—could reduce estimated leukemia incidence within 20
kilometers of the equipment covered by the standard from about 0.038
cases per year to about 0.035 cases per year for new equipment.
2-43
-------
It would reduce the estimated maximum lifetime risk at the point of
maximum exposure from about 4.5 x 10~4 to about 4.2 x 10"4. The
uncertainties associated with the numbers are explained in Section 2.1.2.
Requiring the more stringent level of control rather than BAT
would increase capital costs from $1.4 million to $5.1 million and
net annualized costs of $100 thousand to $900 thousand for new
equipment. Because of the relatively small health benefits to be
gained with the additional costs of requiring the more stringent
level of control instead of BAT for new equipment, EPA considers
the risks remaining after application of BAT to new equipment not,
to be unreasonable. For this reason, EPA judged the level of
control selected as BAT to provide an ample margin of safety and
decided not to require a more stringent level of control than BAT
for new equipment.
2.4 CONTROL TECHNOLOGY
2.4.1 Equipment Specifications
2.4.1.1 Flare Systems. Comment: Many commenters objected to
EPA's exclusion of flare systems as a control device (IV-D-13; IV-D-18;
IV-D-20; IV-D-21; IV-D-22; IV-D-24; IV-D-27; IV-F-1). The comments
addressed the following issues: (1) flare efficiencies, (2) on-going
EPA flare study, (3) safety aspects of flares, and (4) cost effectiveness
of flares.
Response: At proposal, flares were not considered an acceptable
control technique for elimination of fugitive emissions of benzene.
The results of studies that were available were considered inapplicable
to the streams to be controlled. In some studies the flare design was
not representative of flares in the industry. In others the analytical
method was questionable. At that time no approved method for measuring
flare efficiency (evaluating flare performance) was available.
Theoretical calculations indicated that flare efficiency could be
as low as 60 percent for destruction of benzene in low-flow, intermittent
streams sent to a large flare. This efficiency was cited in several
background documents (Ethylbenzene/Styrene, Benzene Fugitive, SOCMI
Fugitive VOC) and served as a primary consideration in not allowing
the general use of flares. However, this theoretical computation was
2-44
-------
based on assumptions that may not be applicable to the design situation
under study.
The use of flares, therefore, was reconsidered for the benzene
fugitive emission standard. Commenters pointed out potential operational
difficulties associated with the use of incinerators that could be
avoided with the use of flares. A major difficulty seen was in designing
systems for the low-volume and intermittent flow to the control device.
In addition, consideration was given to the extensive use of flares by
industry to handle emergency releases. Since flares are currently in
widespread use in the chemical and petroleum refining industries, they
represent a large investment in control by the industries.
The following presents a review of flares and operating conditions
used .in five studies of flare combustion efficiency. Each study can be
found in complete form in the docket.
Palmer (IV-J-3) experimented with a 1/2-inch ID flare head, the
tip of which was located 4 feet, from the ground. Ethylene was flared
at 50 to 250 ft/sec at the exit, (0.4 x 106 to 2.1 x 106 Btu/hr).
Helium was added to the ethylene as a tracer at 1 to 3 volume percent
and the effect of steam injection was investigated in some experiments.
Destruction efficiency (the percent ethylene converted to some other
compound) was 97.8 percent.
Siegel (II-D-92) made the first comprehensive study of a commercial
flare system. He studied burning of refinery gas on a commercial
flare head manufactured by Flaregas Company. The flare gases used
consisted primarily of hydrogen (45.4 to 69.3 percent by volume) and
light paraffins (methane to butane). Traces of H2$ were also present
in some runs. The flare was operated from 0.03 to 2.9 megagrams of
fuel/hr (287 to 6,393 Ib/hr), and the maximum heat release rate was
approximately 235 x 106 Btu/hr. Combustion efficiencies (the percent
VOC converted to C02) averaged over 99 percent.
Lee and Whipple (IV-J-5) studied a bench-scale propane flare.
The flare head was 2 inches in diameter with one 13/16-inch center
hole surrounded by two rings of 16 1/8-inch holes, and two rings of
16 3/16-inch holes. This configuration had an open area of
57.1 percent. The velocity through the head was approximately
2-45
-------
3 ft/sec and the heating rate was 0.3 M Btu/hr. The effects of
steam and crosswind were not investigated in this study. Destruction
efficiencies were 99.9 percent or greater.
Howes, et al. (IV-A-20) studied two commercial flare heads at
John link's flare test facility. The primary purpose of this test
(which was sponsored by the EPA) was to develop a flare testing
procedure. The commercial flare heads were an LH air-assisted head
and an LRGO (Linear Relief Gas Oxidizer) head manufactured by John
Zink Company. The LH flare burned 2,300 Ib/hr of commercial propane.
The exit gas velocity based on the pipe diameter was 27 ft/sec and
the firing rate was 44 x 106 Btu/hr. The LRGO flare consisted of
three burner heads 3 feet apart. The three burners combined fired
4,200 Ibs/hr of natural gas. This corresponds to a firing rate of
83.7 x 106 Btu/hr. Steam was not used for either flare, but the LH
flare head was in some trials assisted by a force draft fan.
Combustion efficiencies for both flares during normal operation were
greater than 99 percent.
An excellent detailed review of all four studies was done by
Joseph, et al. (IV-A-21), and a summary of the studies is given in
Table 2-4. A fifth study by McDaniel, et al. (IV-A-28) determined
the influence on flare performance of mixing, Btu content, and gas
flow velocity. A steam-assisted flare was tested at the John Zink
facility using the procedures developed by Howes. The test was
sponsored by the Chemical Manufacturers Association (CMA) with the
cooperation and support of the EPA. All of the tests were with an
80 percent propylene, 20 percent propane mixture diluted as required
with nitrogen to give different Btu/scf values. This was the first
work that determined flare efficiencies at a variety of nonideal
conditions where lower efficiencies had been predicted. All previous
tests were of flares burning gases that were very easily combustible
and did not tend to soot. This was also the first test that used
the sampling and chemical analysis methods developed for the EPA by
Howes.
The steam-assisted flare was tested with exit flow velocities rang-
ing from 0.02 to 60 ft/sec, with Btu contents from 200 to 2,183 Btu/scf
2-46
-------
CM
CO
en
1
Q;
LU
CQ
O
o
o
u_
0
oo
Q
UJ
1
UJ
0-
o
o
oo
UJ
o
oo
2:
o
I — 1
oo ,
oo
i— i
UJ
UJ
o:
*^j
u-
i
CM
CU
r~™
-Q
(d
**
u
^- cu
IB »F-
t— U
U. -F-
"*-
UJ
*J !.
Q-***N
•ss
3 03
So
J= O
1— FH
in
IB
1
IB
C
o>
•F-
in
&
a.
0)
U
O
o
s-
o
in
g.
I/O
o
IB
O)
in
c
en
en
y\
i
CO
en
F-l
CM
I
^
,
O
cu
c
cu
!>
UJ
IB
•o
s
L.1
O
1
1
•"-1
c
°-
3
•o
•
•— *
UJ
. .
CM
en
cu
E
15
o.
en
A
CO
o
cu
c
IB
O.
o
O-
c
in
cu
"o cl
a: IB
u
cu
CU IB
o •&
in _
a CM
i
i
*— i
cu
.a
IB
u
c
o
c
.—1
CO
CTi
^i
cu
C2.
0.
3
OS
CU
cu
en
^
co
i
en
*
(/] |
= ej
i— S-
Q.T3
CMC in
O "§1 !.
ir> -r- a
I >—
cu in
0 .
F- IB
IB •!-
•F- "O
u
s- =
CU IO
O CM
CM
en
o
•— <
cu
JT
3
in
•F- IB
IB
S. O
CU
in >,
in *J
O in
CU
Q >
" =
^^^
1
l-H
'eu
ai
O PO
a>
en
en
i
CO
r--.
IF)
1
O
o
cu
cu
1
o
s.
a.
c
in
cu
O
IB
u
s.
0
CO
CM
1
»— 1
a.
LU
i
<
0
&
en
F-4
~ -
,
IB
CU
0)
C
18
a
Su
2-47
-------
and with steam to gas (weight) ratios varying from 0 (no steam) to
6.8611. Steam-assisted and air-assisted flares were tested with fuel
gas heat contents as low as 300 Btu/scf. Flares without'assist were
tested down to 200 Btu/scf. This efficiency was also found to be
achievable for air-assisted flares combusting gases with heat contents
over 300 Btu/scf and with exit gas velocities below a maximum value
(depending upon the heat content of the gas stream). All of these
tests, except for those with very high steam to gas ratios, showed
combustion efficiencies of over 98 percent. Flares with high steam to
gas ratios (about 10 times more steam than that required for smokeless
operation) had lower efficiencies (69 to 82 percent) when combusting
2,183 Btu/scf gas. .
After consideration of the results of these five tests, EPA has
concluded that 98 percent combustion efficiency can be achieved by
steam-assisted flares with exit flow velocities less than 60 ft/sec
combusting gases with heat contents over 300 Btu/scf and by flares
operated without assist with exit flow velocities less than 60 ft/sec
gases with heat contents over 200 Btu/scf. Flares are not normally
operated at the very high steam to gas ratios that resulted in low
efficiency in some tests because steam is expensive and operators
make every effort to keep steam consumption low. Flares with high
steam rates are also noisy and may be a neighborhood nuisance.
EPA has a program under way to determine more exactly the efficiencies
of flares used in the petroleum refining and chemical industries, and
a flare test facility has been constructed. The combustion efficiency
of four flares (1 1/2 inches to 12 inches ID) will be determined, and
the effect on efficiency of flare operating parameters, weather factors,
and fuel composition will be established. The efficiency of larger
flares will be estimated by scaling.
According to the current knowledge of flare design, the best
available flare design (i.e., the state-of-the-art flare design) is
the smokeless flare. The smokeless flare introduces air into the
flame by injection of steam or air. This injection of steam or air
increases the mixing of the flared compounds with the flame zone,
thereby increasing the destruction of the compounds. Smoking flares
are environmentally less desirable because they emit particulate. It
2-48
-------
is difficult, however, to maintain smokeless operation unless the
off-gas flow to the flare is constant. When the off-gas flow rate
increases, there is a short period of time before the smoke sensor
responds and additional steam {or air) reaches the flare tip, .During
this period, the flare smokes. Smoking may also occur during large
emergency discharges because insufficient steam lor air} is available
in the plant to make these infrequent discharges nonsmoking, A number
of engineering practices currently used in industry help to achieve
continuous smokeless operation. These include staged elevated- flares,
dual flare tips (small tip for low-flow, large tip for emergency
releases), and continuous flare gas recovery systems. These systems
are further discussed later in this section.
Taking all of these factors into consideration, EPA decided to
allow use of smokeless flares operated with a flame present to control
fugitive emissions of, benzene. In order to ensure that the smokeless
flare operates with a flame present, the flare's pilot light is to be
monitored with an appropriate heat sensor., such as a thermocouple. To
ensure smokeless operation, visible emissions from a flare would be
limited to less than 5 minutes in any 2-hour period. In addition,
steam-assisted flares would have to be operated with exit velocities
less than 60 ft/sec combusting gases with heat contents greater than
300 Btu/scf. Flares operated without assist would have to be operated
with exit velocities less than 60 ft/sec combustion gases with heat
contents greater than 200 Btu/scf. Air-assisted flares would have to
be operated with exit velocities below a maximum value, depending upon
the gas heat content which must be greater than 300 Btu/scf. Flares
operated within these requirements are considered as acceptable alternatives
to'enclosed combustion devices (incinerators, boilers, process heaters)
and vapor recovery systems, such as carbon adsorbers and condensation
units. They may be applied to control of emissions from pump seals
(or degassing reservoirs), compressor seals (or degassing reservoirs),
and pressure relief devices.
As mentioned above, EPA has a program under way to determine the
effectiveness of flares not studied to date. As this data and information
are collected and evaluated, EPA plans to update the requirements for
flares. It is not expected that the requirements would become more
2-49
-------
restrictive. Until the requirements are updated, plant owners and
operators are allowed to determine whether other flare systems are
equivalent to the systems required in the standard.
Comment: One commenter expressed concern that excluding flares
as an enclosed combustion device would cause potentially unsafe conditions
when venting emissions from product accumulator vessels. The commenter
detailed safety problems associated with venting receivers to fired
heaters, vapor recovery systems, and incinerators. The commenter
recommended that benzene-containing receiver vessels continue to be
vented to properly designed flare systems unless there is evidence
that benzene destruction is insufficient (IV-D-22).
Response: As previously indicated, EPA has determined that
smokeless flares designed for the expected flowrates will be accepted
as a control device for benzene fugitive emissions, including emissions
from product accumulator vessels.
Comment: Three commenters stated that nonflare systems are not
as cost effective as flare systems (IV-D-18; IV-D-22; IV-F-1), and
another commenter noted that flare systems can achieve or approach the
same degree of emission reduction as enclosed combustion devices or
vapor recovery systems at significantly lower costs (IV-F-1). One of
the commenters stated that using an incinerator to control intermittent
vents would cost about $3 million (including piping) (IV-D-18).
Response: As previously indicated, EPA has determined that
smokeless flares designed for the expected flowrates will be accepted
as a control device for benzene fugitive emissions.
2.4.1.2 Pump and Compressor Seals. Several commenters questioned
the requirements on pump and compressor seals. Several commenters
noted the inflexibility of the dual seal requirements. Other comments
were received about the barrier fluid systems, costs of dual seals,
and sealless pumps and compressors.
Comment: Two commenters felt that the requirement to use dual
mechanical seals on new pumps is inflexible or overly restrictive
(IV-D-24; IV-D-13). One commenter, who submitted two sets of comments,
felt similarly about new compressors (IV-D-27; IV-D-13). The commenter
felt that the requirement for specific types of mechanical seals on
compressors is inflexible and cannot be applied in all cases.
2-50
-------
One commenter noted that in one of'its refineries dual seals leak
more than single seal pumps. The commenter recommended that the dual
mechanical seal requirement be deleted and that a properly installed
benzene sensor and alarm system be used to detect pump seal leaks.
The commenter has found that area-type benzene monitoring systems are
very reliable in actual plant use. The commenter further noted that
the combination of single mechanical seal pumps and a benzene sensor-alarm
system would cost less than dual seal systems and would be more effective
in reducing emissions (IV-D-20).
Response: Since proposal, EPA decided to allow the less stringent
monthly leak detection and repair program as the standard for new and
existing pumps (See Section 2.3). Therefore, the final regulation for
new and existing pumps requires a monthly leak detection and repair
program using a 10,000 ppm leak definition, as determined by the leak
detection method specified in §61.245. Leaking pumps must be repaired
as soon as practicable but within 15 days.
The standard also provides alternatives to the work practice
standard. The two choices are as follows: (1) installation of a
properly designed mechanical seal system with an associated barrier
fluid system as specified in §61.242-2(d), and (2) installation of
a closed-vent system as described in §61.242-2(f). The standard
for pumps gives the owner or operator the flexibility to choose the
best means of controlling emissions from pumps at his or her process
unit. In any given process unit, however, an owner or operator may
select to use equipment for some pumps and apply the work practice
standard to the remaining pumps.
One of the commenters appears to be questioning the effectiveness
of dual versus single seal systems on pumps. Contrary to the commenter's
experience at one refinery, EPA studies have found that dual seals
do not leak more often than single seals. Also, EPA has found that
dual mechanical seals may not be any more effective than single
seals unless the barrier fluid used to control VOC emissions is a
heavy liquid or, in the control of benzene emissions, unless the barrier
fluid is not benzene. For example, if a VOC or benzene-containing
barrier fluid is used on a dual seal, the outer seal would function
similar to a single seal. However, pumps equipped with dual seals
2-51
-------
would have less emissions than pumps with single seals if the
appropriate barrier fluid is used. As stated previously in this
response, the dual seal requirement on pumps has been deleted but
has been retained as an alternative standard. If the owner or
operator chooses to install dual mechanical seals instead of imple-
menting a leak detection and repair program, a barrier fluid
system using a nonbenzene and non VOC barrier fluid and equipped
with a sensor must be used.
The use of single seals with a sensor-alarm system may be less
expensive than a dual seal system, but the effectiveness of this
approach depends on the specific sensor-alarm system. Therefore,
it is not practicable to define general standards to cover this
approach. However, an owner or operator can request equivalency
for such an approach. *
As discussed in Section 2.3, the effectiveness of a leak
detection and repair program for compressors essentially would be
zero because compressors generally are not spared and, therefore,
repair would be delayed until the next turnaround. Because EPA
judges that equipment controls for compressors are available and
the cost is reasonable (see Section 2.3), equipment was selected as
BAT for existing and new equipment.
Comment: One commenter noted that it was unclear if tandem seals
(dual seals where the barrier fluid is not at a greater pressure than
the stuffing box) are allowed. The commenter recommended that the
definition of tandem seals be clarified and that tandem seals should
be allowed as a control device. The commenter believed that tandem
seals along with the weekly visual checks and 5-day/15-day repair
requirements would eliminate the need to degas and incinerate since
sources will not leak for any extended period of time (IV-D-30).
Response: As stated in the preamble of the proposed benzene
fugitive emissions standard (46 FR 1178, January 5, 1981, Column 3),
the term "dual mechanical seal systems" includes both double and
tandem seal designs. A printing error excluded the complete definition
of a tandem mechanical seal system in the preamble. As defined in a
memorandum of printing errors in the FEDERAL REGISTER publication
2-52
-------
(IV-B-3), tandem mechanical seal systems have two seals in a front-
to-back arrangement. The preamble describes the remainder of the
term as having a barrier fluid with pressure maintained lower than
the pump stuffing box pressure. Under the proposed standard,
either the tandem or double (back-to-back) seal arrangement would have
been allowed for new pumps in benzene service.
As stated in the preceding response, however, the standard has
been revised since proposal. The promulgated standard allows the
plant owner or operator to select the use of a work practice standard
for new and existing pumps in benzene service (monthly leak detection
and repair program with a leak definition of 10,000 ppmv) instead
of the equipment standard. If the owner or operator selects the
equipment (dual seal) standard for new pumps, then he would not
need to include his pumps in a monthly leak detection and repair
program. When dual seal systems are used, EPA believes that equivalent
emissions reduction can be achieved if leakage of process fluid
into the barrier fluid is controlled by either (1) degassing and
venting emissions to a control device or (2) continuously replacing
the fluid with fresh barrier fluid and disposing of the contaminated
barrier fluid.
Comment: One commenter stated that requiring a seal sensor to
monitor catastrophic seal failure and visual inspections for liquid
dripping from the seal is redundant and recommended that the operator
be permitted to do one or the other (IV-D-24).
Response: As a requirement of Section 112(e) of the Clean Air
Act, provisions must be made to ensure the proper operation and main-
tenance of control systems required by equipment standards. If an
owner or operator chooses to install equipment rather than comply
with a work practice standard, outer seal failure will be noted
through periodic visual inspections (i.e., weekly visual inspections).
However, some form of indicator is needed to indicate catastrophic
failure of an inner seal of a dual seal arrangement. These practices
(seal sensor and weekly visual inspections) are not redundant,
since they are both necessary for ensuring proper maintenance of
the dual seal system.
2-53
-------
Comment: Two commenters questioned the barrier fluid system
requirement. One commenter referred to comments made on the SOCMI
VOC fugitives NSPS (IV-D-21). This comment and EPA's response 'are
presented in the SOCMi VOC Fugitives NSPS BID for the promulgated
standard (EPA-450/3-80-033b). Another commenter noted that there
is no definition of barrier fluid and recommended that the vapor
pressure of the barrier fluid be less than one atmosphere at seal
operating temperature. The commenter believed that if a barrier
fluid is simply any benzene"free" fluid, the system may not meet
the proposed standard. The commenter stated that a specification
of "benzene-free" as less-than-or-equal-to 10 percent benzene by
weight provides no clue as to an acceptable fluid vapor pressure.
The commenter also stated that since the barrier fluid, whether
process fluid or contained fluid, will be emitted by the seal as it
cools the seal faces, the fluid's heat capacity, enthalpy, and
vapor pressure at seal operating temperatures are key issues (IV-D-24).
The commenter further believed that barrier fluid systems operating
at a pressure above the stuffing box pressure may lead to unsafe
conditions and that leakage into some process fluids might result
in fires or explosions (IV-D-24).
Response: As described in previous responses in this section,
the standard for new pumps in benzene service has been revised since
proposal. The final standard, in addition to allowing the equipment
alternatives previously specified, requires the use of a work practice
standard for both new and existing pumps. The use of the equipment
alternatives results in better than equivalent emission reduction;
therefore, EPA is allowing equipment controls as an alternative standard
for pumps. In any given process unit, an owner or operator may choose
to use equipment for some pumps and apply the work practice standard
to the remaining pumps. If he chooses to use the equipment on a pump,
he will not need to include that pump in a monthly leak detection and
repair program. As discussed in a previous response in this section,
EPA selected equipment controls as BAT for both new and existing
compressors.
A nonbenzene barrier fluid must be used with mechanical seal
systems for compressors and with dual seal systems for control of
2-54
-------
benzene emissions from pumps if an owner or operator chooses to install
equipment in lieu of leak detection and repair of pumps. If a benzene-
containing barrier fluid were used instead, the seal would not be
effective in terms of sealing benzene from atmosphere. Leakage through
and failure of the outer seal would result in benzene emissions.
For pumps and compressors covered by standards of performance
(Section 111), EPA decided also to specify a vapor pressure for VOC
barrier fluids used with seal systems for pumps or compressors.
Because these systems could leak VOC if not required to use a heavy
liquid barrier fluid and because these systems would otherwise be
covered by petroleum refinery and synthetic organic chemical manu-
facturing industry standards, EPA decided to limit barrier fluids
to make the dual seal system effective. This only applies where
the barrier fluids would be limited by standards of performance
under Section 111. In summary, an owner or operator can select any
barrier fluid that does not contain greater than 10 percent benzene
and is a heavy liquid VOC or a nonVOC.
One of the commenters expressed concern about the safety of
barrier fluid pressures above the pump stuffing box pressure. This
pressure allowance is common practice, but it is not recommended or
required in all cases. The barrier fluid requirement is flexible
and allows the owner or operator to consider safety, design, and
operating conditions of specific processes.
Comment: Two commenters remarked about sealless pumps and
compressors. One of the commenters questioned the method for
specifying the acceptability of enclosed pumps and asked how to
measure the concentration of an enclosed pump if there is no shaft
=f
that penetrates the pump housing (IV-D-24).
The other commenter noted that sealless pumps, which are extremely
effective, are not required, according to the proposed standard,
because they can be used at only a limited number of emission points.
The commenter added that sealless compressors are not required because
they are said not to be widely available. Therefore, the commenter
recommended that sealless pumps and compressors be required where they
can be used. The commenter added that production and availability of
2-55
-------
seal!ess compressors would increase if they were required for future
plants (IV-D-31).
Response: Since leak!ess equipment has limited applications in
the chemical and petroleum refining industries, leakless equipment is
not required by the regulation. EPA cannot specify where leakless
equipment can and cannot be used at this time. However, leakless
equipment is allowed as an alternative to the standards for pumps and
compressors without a determination of alternative means of emission
limitation. They are exempt from the routine monitoring requirements
by setting a performance standard of no detectable emissions.
The provisions for leakless technology are presented in §61.242-2(e)
for pumps. The provisions deal only with pumps whose shafts do not
penetrate the pump housing, i.e., sealless and canned pumps. These
pumps are excluded from leak detection and repair only if they are
operated with no detectable emissions above background as measured by
methods set forth in the regulation. Similar provisions are given in
§61.242-3(i) for compressors and in §61.242-7(f) for valves. Leakless
seal technology is not precluded by these provisions. Other types of
leakless seal technology that may be developed could become an accepted
control alternative if adequately demonstrated through the alternative
means provisions of §61.244.
The no detectable emissions limit of 500 ppmv above background
concentration that is required for enclosed pumps and compressors is
based on the measurement method of Reference Method 21. The portable
VOC analyzer is used to determine the local ambient VOC concentration
in the vicinity of the source to be evaluated, and then a measurement
is made at the surface of the potential leak interface. The potential
leak interfaces include bolted flanges and other assemblages. If a
concentration change of less than 5 percent of the leak definition is
observed, then a "no detectable emissions" condition exists. The
definition of 5 percent of the leak definition was selected based on
the sensitivity of the instrument, which reads +_5 percent of the
meter scale. "No detectable emissions" would exist when the observed
concentration change between local ambient and leak interface
surface measurements is less than 500 ppmv.
2-56
-------
Comment: Two commenters remarked about the costs associated with
dual seal systems. One commenter reported that one of the refineries
contacted in the area estimated a minimum cost of $200,000 for
installation of secondary seals with an estimated $35,000 annual
maintenance and monitoring cost (IV-D-10). One commenter evaluated,
the use of dual mechanical seals with a barrier fluid and controlled
degassing vents. The commenter concluded that safe installation of
this type of system is very expensive and would only control minor
amounts of benzene (IV-D-29).
Response: Since proposal, a monthly leak detection and repair
program has been selected for control of benzene fugitive emissions
from new and existing pumps (see Section 2.3). For new and existing
compressors, EPA has selected equipment control. EPA has estimated
costs of installing safe systems and has concluded that they are
reasonable.
2.4.1.3 Sampling Systems. Four commenters submitted the following
comments on the sampling system requirements:
Comment: One commenter recommended that the last five words of
paragraph '(2), Sec. 62.112(c) of the proposed standard, "without
benzene emissions to atmosphere," should be changed to, "with minimal
emissions to atmosphere." The commenter noted that designing sampling
systems with zero benzene emissions would require using a benzene-free
purge. The commenter recommended that only specially designed sampling
valves should be required (IV-D-2Q). One commenter recommended that
other sampling systems besides a purge system be allowed, for example,
venting to a control device with 95 percent recapture (IV-D-21).
Another commenter stated that sampling systems are not state-of-the-art
for worker protection. The commenter also recommended that a definition
of in-situ sampling systems be added to the standard (IV-D-24).
Response: In the BID for the proposed standard, it was estimated
that closed-loop sampling systems are almost 100 percent effective
in eliminating sampling purge emissions. As noted in the preamble
of the proposed standard (46 FR 1180, January 5, 1981), however, no
available data indicate that application of any control technique
would be able to comply with a "no detectable emissions" standard
(500 ppmv or less VOC concentration above background). Some benzene
2-57
-------
could be emitted during sample transfer to a closed collection
device.
The intent of the standard is to eliminate sample line purging
to atmosphere, ground, or sewer drain. A "zero" (no detectable)
emissions limit is not practicable as noted by the establishment of
an equipment standard instead of an emissions limit because sampling
systems can have detectable emissions. Reference Method 21 would
not be practicable for determining emissions from these systems.
EPA recognizes that closed-loop sampling systems have limitations
with respect to low-pressure processes or tankage and, in some instances,
safety requirements. The regulation, therefore., does not specify a
"closed-loop sampling system," but it does require a "closed-purge
system." This will allow any system that collects all benzene purged
during sampling and recycles or destroys the collected benzene.
Closed-loop sampling systems are used in the BID for the proposed
standard to evaluate the cost of controlling fugitive emissions from
sampling systems (closed-loop sampling systems are generally the most
costly system to implement a closed purged system).
The intent for sampling systems has been clarified in the final
standard with the changes presented in response to other comments.
By elaborating on the requirements for "closed-p.urge systems" and
"in-situ" sampling systems, the proper equipment design criteria
for sampling systems have been better described:
"Closed-purge system" means the equipment comprising a closed-loop
sampling system or any system that collects benzene purged in the
sampling process and either recycles or disposes of benzene.
"In-situ sampling systems" are non-extractive or in-line sampling
systems that involve measurement or sampling of process stream
conditions without extraction of the sample from the process stream,
resulting in no purged emissions of benzene.
In addition to closed-purge sampling, EPA has considered since
proposal allowing a closed-vent vacuum system connected to a control
device. Closed-vent vacuum systems connected to a control device
collect the sample purge and then transport the sample purge to a
control device. If these systems are not open to atmosphere, then
their reduction in emissions of benzene would be equivalent to
2-58
-------
collecting the purge in a collection system that is not open to
atmosphere.. Based on these considerations, EPA has decided to
allow closed-vent vacuum systems connected to control devices.
Comment: One commenter stated that there were the following
problems in controlling emissions of benzene in accordance with the
proposed regulations:
(1) Attempts to perform closed-loop sampling of nitration batches
at one plant were unsuccessful.
(2) One plant anticipated problems with closed-loop sampling of
raw material deliveries and reactor content measurements (IV-D-29).
Response: As stated in the previous response, EPA specifies in
the proposed and promulgated regulations "closed-purge systems" as the
control equipment for sampling systems, not "closed-loop sampling
systems." Requiring "closed-purge systems" allows any system that
collects all benzene purged during sampling and recycles or destroys
the collected benzene. An example system is that of the Fetterolf-"Ram-
seal" presented; in Docket A-79-27-ir-D-56.
2.4.1.4 Continuous Benzene Emissions Control. Comment: One
commenter expressed concern for the requirement that all control
systems be operated 100 percent of the time when benzene emissions
may occur. The commenter was concerned that the provision does not
allow for expected or unexpected maintenance or repair of the
control system. The commenter recommended that a bypass of emissions
be allowed where the control systems must be taken out of service
for maintenance or an emergency only where the next emissions that
will be bypassed would not exceed the excess emissions that would
result from a shutdown and start-up of the process unit (IV-D-21).
Response: The standard requires that control devices be operated
when emissions of benzene are vented to them. This means that closed-
vent systems transporting the benzene emissions to control devices
can not be vented uncontrolled to atmosphere. For larger emissions,
current industry practices do not allow uncontrolled venting of
emissions from closed-vent systems. Thus, alternative control
devices will be available to control these emissions. For smaller
emissions, such as those from dual mechanical seal systems, the
2-59
-------
closed-vent system can be isolated from the control device long
enough to repair or maintain the control device.
2.4.1.5 Open-ended Lines. Comment: One commenter stated that
EPA did not recognize block-and-bleed operations in reference to
open-ended lines (IV-D-21).
' Response; Provisions were made in the proposed standard for
block-and-bleed operations. Sections 61.112(d)(6) and (7) of the
proposed standard required:
"(6) Each open-ended valve shall be equipped with a cap,
blind, plug, or a closed second valve that is attached to seal
the open end at all times except during operations requiring flow
through the open-ended valve.
(7) Each open-ended valve equipped with a second valve, as
required in Section 61.112(d)(6) shall be operated such that the
open-ended valve is completely closed before the second valve is
closed."
Where a block-and-bleed system is being used, the bleed valve (second
valve) must "seal the open end at all times except during operations
requiring process fluid flow through the open-ended line." The bleed
valve, therefore, can remain open when venting the space between the
two block valves. Bleed valves are exempt when they are operational,
but require a cap when not in use.
2.4.1.6 Enclosed Combustion Devices. Comment: Two commenters
questioned the required design conditions for enclosed combustion
devices. One commenter noted that a temperature requirement with a
residence time is not specific enough to guarantee 95 percent benzene
destruction. The commenter recommended that an offgas oxygen level
should be specified if pyrolysis operating conditions are inadequate.
The commenter recommended that the level of destruction should be
specified and that the designer/operator should specify the operating
conditions. The commenter proposed the criterion of, "equipped with a
barrier fluid degassing reservoir that is connected by a closed-vent
system to a combustion device or to a vapor recovery system designed
for a minimum of 95 percent destruction or capture of benzene input to
the combustion device or vapor recovery system, respectively." The
2-60 .
-------
commenter added that this criterion could also be applied to
Section 61.112(a)(7) of the proposed standard (IV-D-24).
The commenter also expressed concern that closed-vent systems and
enclosed combustion devices introduce fire and explosion hazards into
many workplaces (IV-D-24). The other commenter noted that, in reference
to product accumulator vessel vents being connected to a combustion
device, a minimum temperature of 760 °C (1,400 °F) is probably excessive
in achieving 95 percent destruction of benzene vapors (IV-D-20).
Response: Temperature and residence time are critical parameters
influencing benzene destruction efficiency. They remain as one criterion
that can be used to demonstrate compliance with the standard; however,
EPA has made it clear in the final regulation that other control
system design criteria can be used to demonstrate compliance with the
standard. In addition to thermal incinerators, other control devices
with high efficiencies are available and operate under less severe
conditions. For example, catalytic incinerators operate with high
control efficiency under much reduced temperature-residence time
requirements. Also, boilers or process heaters combust benzene emissions
from low flow sources, such as accumulator vessel vents and seal oil
degassing system vents. For some sources, flares will be selected as
the control devices. In yet other circumstances, vapor recovery
devices may be preferable to enclosed combustion. Well-designed,
existing enclosed combustion devices and vapor recovery systems can
achieve the emission reduction efficiencies of better than 95 percent.
Thus, EPA selected a benzene reduction efficiency of 95 percent for
enclosed combustion devices.
2.4.1.7 Pressure Relief Devices. Comment: Two commenters
/
questioned the requirement for controlling benzene emissions from
pressure relief devices. One commenter questioned the practical use
of rupture disks with block valves as an effective means for reducing
emissions to less than 200 ppm. The commenter added that existing
safety/relief valves converted to special discs having o-ring seals
(in order to meet emission limits) could not be recommended for services
above 400 °F. The commenter noted that a negligible degree of additional
control would result especially since there are few relief valves
compared to the number of control valves (IV-D-20). The other commenter
2-61
-------
stated that since the no detectable emissions level is above the OSHA
STEL criterion of 25 ppm (i.e., 15-minute ceiling level based on the
ability to monitor), a warning should be included in the standard
noting that monitoring of emissions should be done in accordance with
OSHA rules (IV-D-24).
Response: During normal operations rupture disks provide effective
control of fugitive emissions from safety/relief valves. In order to
determine whether the disk is properly sealing the system against
leaks, a pressure gauge can be installed in the pocket between the
disk and the safety/relief valve. The proposed standard, however,
does not preclude the use of control techniques other than rupture
disks. The proposed standard for no detectable emissions may be
achieved by transporting emissions via a closed-vent system to a
control device (i.e., flare system). In regard to the comment on the
use of relief valves with o-ring seals, EPA does not have information
indicating that these would result in less than 500 ppmv (no detectable
emissions). If they do, then the owner or operator is responsible for
understanding their limitations.
Although pressure relief devices represent a small fraction of
the emission sources as indicated by one of the commenters, they
contribute a relatively large proportion of the total volume of
fugitive emissions. Pressure relief devices represent over
11 percent of the total uncontrolled fugitive emissions. The
uncontrolled emission factor for pressure relief devices
(3.9 kg/day) is 15 times greater than the emission factor for light
liquid valves and 6 times greater than the uncontrolled factor for
gas valves.
EPA is relying on the OSHA standards to protect occupational
safety and health. The standard does not require plant operators
to perform activities not otherwise performed by current industry
practices. EPA expects owners or operators to adhere to occupational
safety and health regulations in implementing the standard. However,
it is not appropriate to note this in the standard and, therefore,
it is not included.
2-62
-------
2.4.1.8 Valves. Comment: One commenter suggested that diaphragm
or bellows valves be required where they can be used despite the fact
that their use is limited (VI-D-31).
Resjponse: Diaphragm and bellows valves have limited applicability
in the chemical and petroleum refining industries. Temperature or
pressure constraints on the diaphragm elastomer or the susceptibility
of bellows to corrosion restrict the use of these leakless valves. As
explained in Section 2.3, the costs are considered unreasonably high.
2.4.1.9 Product Accumulator Vessels. Comment: One commenter
stated that one plant indicated that controlling emissions from product
accumulator vessels to "no detectable emissions" would require very
great expense for very little control (IV-D-29).
Response: As discussed in the preamble to the proposed standard,
the control technique of connecting existing and new product accumulator
vessels to a control device with a closed vent system was evaluated
during selection of the proposed standard. EPA selected this
control technique as the basis for the proposed standard. However,
as the commenter points out, the proposed standard required that
accumulator vessels be operated at a state of emissions having a
concentration less than 200 ppm above background using Method 21
(See proposed §61.115(b)), that is, having "no detectable
emissions." This requirement did not mean that emissions must be
eliminated completely, but rather the emissions must be vented to a
control device as stated in proposed §61.112(e)(3). Method 21 was
used to determine "no detectable emissions" to ensure the closed
vent system does not leak.
Because this commenter did not clearly understand EPA's intent
, as expressed in the proposed standard, EPA has redrafted this
requirement to make it clear that accumulator vessel vents must be
vented to a control device. In the section on control devices and
closed vent systems, Method 21 is still used to ensure the closed
vent system does not leak. EPA also deleted the proposed §61.112(e)(2)
because double coverage of ethylbenzene or styrene plants is no longer
a concern.
2-63
-------
Because the commenter questioned the costs of this requirement,
EPA reviewed the costs of controlling accumulator vessel vents. In
Appendix F to the BID for the proposed standard, EPA estimated the
nationwide benzene emissions total would include 130 Mg/yr from
accumulator vessels and the annualized cost would be about $11,000.
The commenter did not question these estimates or provide a basis
to evaluate these estimates further. In Appendix A, Table A-10 of
the BIO for the promulgated standard, EPA reevaluated the cost and
effectiveness of controlling product accumulator vessels and estimated
the cost effectiveness to be about $100/Mg. EPA considers such a
cost effectiveness reasonable for BAT. Therefore, EPA did not
change the substantive requirements for product accumulator vessels.
2.4.2 Leak Detection and Repair Requirements
Several commenters recommended changes in the proposed leak
detection and repair requirements. Most of the comments referred to
adding provisions for inaccessible or unsafe-to-monitor valves, extensions
for repair beyond shutdown, and reconsidering the monitoring frequency
or leak definition. Other comments were concerned with test methods
and procedures, back-up systems, inspection and maintenance data,
repair schedule, safety of monitoring relief valves, and alternative
standards for valves.
2.4.2.1 Inaccessible Valves. Comment: Six commenters recommended
that a separate monitoring requirement be provided for inaccessible
(difficult-to-monitor) valves (IV-D-18; IV-D-21; IV-D-29; IV-D-30;
IV-F-1; IV-D-13; IV-D-27; IV-K-1).
One commenter stated that some valves are inaccessible for safety
reasons or because of elevation and/or configuration. As the commenter
noted, many of these valves can be eliminated in an entirely new
plant, but compliance with the proposed regulation may be difficult in
an older plant. This commenter proposed that an alternative requirement
be provided for inaccessible valves (IV-F-1).
One commenter estimated that platforms needed to provide safe
access for routine monitoring purposes cost between $10,000 and $20,000
each. This cost, the commenter noted, is difficult to justify considering
the small incremental control potentially achievable. The commenter
2-64
-------
recommended that inaccessible or difficult-to-access components be
treated specially or exempted altogether (IV-D-18).
Another commenter noted that valves that are inaccessible without
erecting scaffolding, climbing out on piping, or hanging from a ladder
pose a safety hazard to the monitoring personnel. The commenter
suggested that "accessible" be defined as "able to safely reach from
the ground or an existing permanent platform," which would include
90 percent of the valves in a petrochemical facility (IV-D-30).
Another commenter suggested that an annual leak detection and
repair program be specified for difficult-to-monitor valves (i.e.,
restricted access to valve bonnet or valves located on elevated pipe
racks) as is specified in the proposed standard of performance for
refinery VOC fugitive emissions (IV-D-13; IV-D-27).
Response: EPA acknowledges that some valves are difficult to
monitor because access to the valve bonnet is restricted or the valves
are located in elevated pipe racks. Oifficult-to-monitor valves can
be eliminated in new process units, but they may not be eliminated
in existing process units. The proposed standard, therefore, has been
amended to provide for difficult-to-monitor valves. The final standard
defines difficult-to-monitor valves as valves that require elevating
the monitoring personnel more than 2 meters above any permanent
available support surface. The commenter's suggested definition
expresses EPA's intention in requiring safe access to difficult-to-monitor
valves. EPA's definition includes specifically the safe use of step
ladders, which may be required to elevate monitoring personnel under
safe conditions, but scaffolds will not be required. For existing
fugitive emission sources of benzene, EPA is requiring an annual leak
detection and repair program for valves that are difficult to monitor.
An owner or operator, therefore, will designate valves that cannot be
safely monitored by the use of step ladders and will comply with an
annual leak detection and repair program.
2.4.2.2 Unsafe-to-Monitor Valves. Comment: One commenter
recommended that valves unsafe to monitor because of extreme temperature,
pressure, or explosion hazard be inspected by a plan developed by each
refinery (IV-D-13; IV-D-27).
2-65
-------
Response; EPA recognizes that some valves are unsafe to monitor
based on the judgment of the owner, or operator. Unsafe-to-monitor
valves are defined as those that expose monitoring personnel to imminent
hazards from temperature, pressure, or explosive process conditions.
EPA does not believe valves in benzene service will be unsafe to
monitor. Although unsafe-to-monitor valves in benzene service may
not exist, they could not be eliminated if they do exist; thus EPA
is giving special consideration for unsafe-to-monitor valves. The
final standard will require an owner or operator to prepare a plan
that defines a leak detection and repair program conforming with
the routine monitoring requirements of the standard as much as
possible, except monitoring should not occur under unsafe conditions.
2.4.2.3 Extensions for Repair Beyond Shutdown. Comment: Two
commenters recommended that the proposed leak repair requirements
include extensions for repair beyond a process unit shutdown in limited
situations. The commenters listed some of, these situations, which
include obtaining replacement or spare parts not normally in stock,
replacement of the entire valve assembly, items requiring long delivery
times, or manufacturer delays in supplying parts. The commeriters
recommended that industry be allowed extensions of repair beyond a
process unit shutdown in these situations (IV-F-1; IV-D-13; IV-D-21;
IV-D-27). One of the commenters further recommended that the provision
for delayed repairs should apply to valves as well as all other components
(IV-D-21).
Response: In the proposed standard, delay of repair was allowed
for leaks that could not be repaired without shutting down a process
unit. In general, these leaks would be repaired at the next scheduled
unit shutdown. EPA recognizes that in certain situations the lack of
spare parts might prevent repair of all leaking valves during a unit
shutdown. Spare parts for valves can usually be stocked so that all
leaks that can not be repaired without shutting down the unit can be
repaired during the shutdown. Spare parts include packing gland bolts
and valve packing material. In a few instances, the entire valve
assembly would need to be replaced. In the final regulation, EPA is
adding provisions to allow delay of repair beyond a shutdown for
valves that require replacement of the entire valve assembly, provided
2-66
-------
that the owner or operator can demonstrate that sufficient stock of
spare valve assemblies has been maintained and that the supplies have
been depleted.
Another new provision is added to the regulation to allow delay
of repair for valves if the owner or operator shows that leakage of
purged material due to immediate repair is greater than the fugitive
emissions that are likely to result from delay of repair. EPA does
not expect this situation to occur often. Based on an engineering
analysis, EPA has estimated the emissions from draining a line to
allow immediate isolation of a valve for off-line repair for comparison
with estimated emissions that are likely to result from allowing a
valve to continue leaking (IV-B-20). This analysis is based on a
leaking control valve that has inlet and outlet block valves and a
bypass loop to allow continued operation of the process unit.
Assuming that all of the benzene trapped between the inlet and
outlet block valves is released to atmosphere when the control
valve is isolated from the process (a worst-case situation), EPA
estimated that for an average size process line 0.077 meters (3
inches) in diameter and 2 meters (6 feet) in length, 8.35 kg of
benzene would be emitted. This 8.35 kg represents about 4 days of
benzene emissions if the valve were allowed to continue to leaking.
EPA would not consider allowing delay of repair for this valve to
be reasonable. EPA believes this delay of repair provision for
valves would be reasonable for only a fairly large and long process
line or process vessel and for a fairly long time until the next
process unit shutdown. For example, EPA roughly estimated the
maximum volume of benzene leaking over a 180-day period and compared
it to the corresponding maximum line length and diameter of pipe.
The emissions resulting from a valve leaking 180 days (367 kg of
benzene) corresponds with a pipe length of 22 meters (73 feet). In
general, times until shutdowns, pipeline length, and pipeline
diameter are smaller than these estimates; thus, use of this provision
will be infrequent. Additionally, it is unlikely that all of the
benzene purge would be emitted to atmosphere. Some plants purge to
a closed-drain system or other systems to reduce personnel exposure
to short-term, high concentrations of benzene like those resulting
2-67
-------
from draining a process line. For this reason, emissions estimated
above for immediate valve repair are likely overstated. Consequently,
delay of repair for valves is reasonable if the valves cannot be
repaired without a shutdown. This situtation is expected to occur
very infrequently. The final standard allows a plant to determine
whether off-line repair emissions from draining a line are greater
than emissions resulting from repair delay ("continued leakage").
If this demonstrates that emissions from draining a line are greater
than emissions from allowing a valve to continue leaking until the
next process unit shutdown,.then delay of repair will be allowed.
A definition of "process unit shutdown" has also been added to
the regulation to clarify EPA's intent of avoiding extended.delays
in returning a process unit to production if the unit shuts down
briefly due to unforeseen circumstances. Delay of repair beyond an
unforeseen process unit shutdown will be allowed if this shutdown
is less than 24 hours in duration (Docket No. A-79-32-IV-J-3).
Repair of leaking equipment for which repair has been delayed would
be required at the next scheduled process unit shutdown.
As part of the repair requirements, EPA is clarifying its
intent for spare equipment that does not remain in benzene service.
Delay of repair of equipment for which leaks have been detected
will be allowed for equipment that is isolated from the process and
that no longer contains benzene in concentrations greater than
10 percent. This equipment must be purged to a system th'at complies
with the requirements for closed-vent systems and control devices.
Emissions from equipment can be collected and transported through a
properly designed and installed closed-vent system as verified by
Reference Method 21. Control devices allowed by EPA that destroy
benzene equipment leaks are flares, enclosed combustion devices,
and vapor recovery systems that will achieve at least a 95 percent
reduction efficiency of benzene emissions. As discussed in Section
2.4.1.1, smokeless flares operated with a flame present may be
used. Enclosed combustion devices are required to be designed to
provide a minimum residence time of 0.50 seconds at a minimum
temperature of 760°C, and vapor recovery systems are required to
operate with an efficiency of 95 percent or greater.
2-68
-------
2.4.2.4 Monitoring Frequency. Comment: Four commenters
recommended that the monitoring frequency be reconsidered. Various
reasons for the recommended changes were cited (IV-D-13; IV-D-20;
IV-D-21; IV-D-27; V-F-1; IV-K-1). Another commenter supported
EPA's choice of a monthly inspection frequency of equipment that
may leak and stated that longer intervals would allow large leaks
to go undetected for 6 months or a year (IV-D-31).
One commenter stated that the monthly monitoring requirement is
excessive and partially based on assumed factors that are not supported
by any data. The commenter recommended that data from on-going studies
be considered, that the basis for the monitoring frequency be re-
evaluated, and that the monitoring frequency be established on a
cost-effective basis (IV-D-13, IV-D-27).
One commenter stated that quarterly leak detection monitoring is
too frequent when considered in light of the small potential losses.
The commenter recommended that the Agency require no greater frequency
than annual leak detection monitoring (IV-K-1).
Another commenter recommended that a true quarterly monitoring
program rather than a hybrid (monthly/quarterly) program be implemented
by the proposed regulation for the following reasons (IV-F-1). The
commenter believed that leak frequency data developed for the chemical
industry is significantly different from that developed for the petroleum
refining industry. In the absence of representative data, the commenter
further believed that it is not logical to assume anything other than
a linear leak occurrence/recurrence rate with time and a recurrence
rate that varies in proportion to occurrence. The commenter felt
that, without any field data, EPA had to use engineering judgment to
conclude that more frequent monitoring will result in finding four
times as many leaks in valves monitored monthly rather than annually
and twice as many leaks if monitored monthly rather than quarterly.
The commenter felt that the proposed monthly/quarterly monitoring
requirements are not cost effective. A directed maintenance program
during a shutdown will yield optimum results at minimum cost. The
commenter recommended that no further monitoring be required on a
piece of equipment that can not be repaired on-line until just before
shutdown, at which time the leak can be fixed. The commenter also
2-69
-------
believed that monitoring and sampling requirements are excessive for
pumps, suggesting that quarterly monitoring be required (IV-D-21).
Response: Since proposal, EPA has re-evaluated leak detection
and repair cost data for valves and pumps using the model developed
specifically for leak detection and repair programs (Appendix B). As
discussed in Section 2.3, EPA has concluded that monthly leak detection
and repair is reasonable for valves and pumps because the results from
the model indicate that the costs of monthly programs for valves and
pumps are reasonable. The costs for various leak detection and repair
intervals are presented in Appendix A.
Monthly/quarterly leak detection and repair programs were examined
because EPA considered this program to provide the most reasonable
approach in implementing monthly monitoring. Based on the results of
the LDAR model (see Appendix B), however, the monthly/quarterly
program does not achieve as much emissions reduction as the monthly
program. EPA is allowing the use of the monthly/quarterly leak detection
and repair program, however, because the monthly/quarterly program
as designed by EPA would provide better control than the control
estimated by the LDAR model. The intent is to require monthly
monitoring for valves that leak often or occasionally but quarterly
monitoring for valves that do not leak very often.
The comment regarding quarterly versus annual leak detection and
repair for pumps (IV-K-1) applies to the level of control suggested by
the petroleum refinery control technique guidelines, or Regulatory
Alternative II, in the BID for the proposed standard. As discussed
above, after considering the comment, EPA did not select this control
level for the final standard.
2.4.2.5 Leak Definition. Comment: Three commenters disagreed
with the 10,000 ppmv level of leak detection set by EPA (IV-D-21;
IV-D-24; IV-D-31). One of the commenters felt that the criterion of
10,000 ppm VOC is unduly restrictive in the monitoring requirements
for pipeline valves, open-ended valves, and existing pumps and compressors.
This value corresponds to only 2,900 ppm of benzene according to the
commenter. Since the concentration is related to a rate and benzene
emissions were not measured in the EPA studies, the commenter noted
2-70
-------
this rate may be more restrictive than that allowed under the proposed
VOC fugitive emission regulation (IV-D-24).
Another commenter felt that the 10,000 ppm cut-off should be
lowered and that EPA is not justified in exempting all leaks below
10,000 ppm from the repair requirement. The commenter expressed
concern that EPA apparently is not certain that making repairs in the
1,000 to 10,000 ppm range will result in net increases in emissions.
The commenter contended that the BID for the proposed standard, pages C-16
through C-17, shows that most repaired leaks result in lower emission
levels. The commenter asserted that EPA should have examined whether
an intermediate cut-off (for example, 2,500, 5,000, or 7,500 ppm)
would better maximize net reductions (IV-D-31).
Response: The first commenter states that the 10,000 ppmv leak
definition is too low. The leak definition is not based on acceptable
or unacceptable emission rates. Instead, EPA selected a value based
on seal failure as a result of testing fugitive emission sources. In
addition, most organic concentrations are either very low (much
less than 10,000 ppmv) or very high (much greater than 10,000 ppmv),
so small changes in concentration are not meaningful. The 10,000 ppmv
leak definition is the highest level that a typical organics monitoring
instrument can detect directly. In order to measure higher concen-
trations with the instruments most commonly used, additional care and
calibration for devices such as dilution probes are required to obtain
accurate results.
The Maintenance Study (IV-A-6) examined the effects of simple
on-line maintenance (tightening bolts) for valves using an action
level of 10,000 ppmv. A repair efficiency of only 29 percent was
found as a result of this maintenance. However, this efficiency
corresponded to a 71 weight percent reduction in emissions. At a
higher leak definition, the repair efficiency may be increased, but
a lower weight percent reduction in emissions would result. Therefore,
the 10,000 ppmv leak definition was selected instead of a higher
level.
The second commenter correctly pointed out that there would be
only a potential for a net increase in emissions if an action level
between 1,000 and 10,000 ppm were selected. A net increase in mass
2-71
-------
emissions might result if higher concentration levels result from
attempted repair of a valve with a screening value between 1,000 and
10,000 ppm. Many leaks can be repaired at concentrations less than
1,000 ppm. However, the key criterion in selecting a leak definition
is the mass emissions reduction achievable. Any leak definition
chosen would only be an indicator of whether a source was emitting
benzene in quantities large enough to warrant repair. In this regard,
a leak definition of 10,000 ppm accomplishes this goal. Based on the
findings of the Maintenance Study, the 10,000 ppm leak definition
results in an overall 71 weight percent reduction of VOC emissions
using simple on-line maintenance. EPA believes that there is only a
slight potential for emission reduction for sources having benzene
concentrations between 1,000 and 10,000 ppmv; therefore, using a lower
leak definition would not increase reductions in emissions signifi-
cantly. This potential is offset by a chance that increases in emissions
may occur by attempting maintenance at levels lower than 10,000 ppmv,
thereby negating the benefit of using a leak definition around 1,000 ppm.
Therefore, a leak definition of 10,000 ppmv was selected instead of a
lower level.
2.4.2.6 Spare Equipment. Comment: One commenter recommended
that greater back-up requirements are needed to enable quick repairs
at significant emission points that otherwise cannot be readily repaired
without shutting a plant down. According to the commenter, back-up
systems would be practical and enable quick leak repair for control
valves, block valves, safety/relief valves, pumps, and compressors.
The commenter suggested as an example that piping could often be built
in series for two pumps, and a spare pump could be exchanged among
leaking pump sites (IV-D-31).
Response: In allowing delays of repair, EPA recognizes that
there may be instances where equipment cannot be repaired on-line.
Delay of repairs are allowed mainly if a process unit must be shutdown.
Back-up systems can help reduce these shutdowns, but having a complete
set of spare equipment is not practicable. Pumps that require a
process unit shutdown for repair already have a spare. (Ever^ dump is
assumed to be spared in the cost impact analysis for the model units.)
Thus, requiring spares for all pumps would not be needed and would
2-72
-------
increase emissions of benzene with little, if any, benefit. For
valves, spare valves might provide repairs for sources that otherwise
would continue to leak. However, because EPA believes the number of
these continuing leaks to be small, the increase in emissions due to
the extra valves is not offset by the decrease in emissions due to
repairing those that continue to leak. Since equipment specifications
(not leak detection and repair) are required for pressure relief
devices and compressors, spares for these components are not appropriate,
2.4.2.7 Data from Control Techniques Guideline. Comment: One
commenter believed that EPA underestimated numbers in its control
techniques guideline for fugitive emission inspection and maintenance
data (e.g., number of valves that can be tagged per day, number of
valves that can be inspected per day, cost of repairing leaks, number
of field people required for monitoring, the number of clerical people
needed to transmit data and prepare reports, amount of supervision to
conduct the program, amount of emissions abated) (IV-D-20).
Response: The commenter is referring to the control techniques
guideline (CTG) document entitled Control of Volatile Organic Compound
Leaks from Petroleum Refinery Equipment (II-A-10). The estimates
of leak detection and repair data presented in the CTG are based on
actual data supplied by industry. The commenter appears to be
questioning the costing procedures used in the proposed standard
because they were similar to the procedures, used in the CTG. For
example, the estimated number of valves that can be inspected, manhours
needed for leak detection and repair, and subsequent costs are based
on actual studies of monitoring and maintenance program manpower
requirements performed by Exxon (II-D-28), general survey practices
followed by Phillips Petroleum Company, and the State of California
Air Resources Board (II-I-35).
The basis for the leak detection and repair costs for the final
standard is presented in the AID (IV-A-24), except an hourly labor
rate of $15.50 was used (based on 1979 estimates). The labor rate
estimate included wages plus 40 percent of wages for labor-related
administrative and overhead costs (Refinery CTG estimates). Admi-
nistrative and support costs to implement the regulation were
estimated to be an additional 40 percent of monitoring and maintenance
2-73
-------
labor (Refinery CTG estimates). Two monitoring instruments per model
unit were assumed to be required for the leak detection and repair
program at a cost of $4,300 per instrument. In addition, an annual
maintenance and calibration cost of $2,800 was assumed for the
monitoring instruments (Docket No. A-79-27-II-A-10 and A-80-44-II-E-7).
EPA considers these estimates for the leak detection and repair
program to be reasonable.
2.4.2.8 Repair Schedule. Two commenters submitted remarks
concerning the schedule for repairing leaks once detected.
Comment: One commenter felt that the timing requirements for
seal repair are too restrictive, stating that a fixed time schedule
will reduce the productivity of operating facilities for no good
reason. The commenter noted that in most plants, a catastrophic seal
failure will be repaired quickly as standard operating practice. In
addition, the commenter suggested that if the benzene leak rate of
liquid dripping from a seal is not excessive and does not require
repair, the operator should not take action until a normal shutdown
for routine maintenance or until the benzene leak rate becomes excessive
(IV-D-24). , . '
The other commenter, however, disagreed and argued that a 15-day
repair time is too long, especially for large leaks. The commenter
believed that in most cases firms should have readily at hand both the
personnel and the equipment to effect repairs. The commenter agreed
with a first attempt at repair within 5 days and added that it should
be the maximum allowed except in carefully specified situations (IV-D-31).
Response: The 15-day repair time for valves, pumps, compressors,
and other fugitive emission sources (for example, pressure relief
devices in liquid service, flanges and other connectors) is considered
adequate for repair of all of these sources but critical valves that
cannot be bypassed. Provisions have been made for delaying these
valves until the unit is shutdown. The 15-day interval was selected
because it provides time for better determination of methods for
isolating pieces of leaking equipment when equipment isolation is
needed for repair beyond simple field repairs. The 15-day repair
interval allows more efficient handling of repair tasks while maintaining
an effective reduction in fugitive emissions. Longer repair intervals
2-74
-------
provide less effective reduction in emissions and do not substantially
improve the efficiency in handling repair tasks. Shorter repair
intervals recommended by the second commenter could cause scheduling
problems in repairing valves and pumps that need to be removed from
the process for repair. The 15-day interval provides the owner or
operator sufficient time to determine what spare parts are needed for
these isolated valves and pumps and time for flexibility in scheduling
repair for these valves and pumps. The first commenter is referring
to the weekly visual inspection requirement for pump seals. The
commenter does not define what constitutes an "excessive" leak rate of
benzene dripping from a seal. If any owner or operator does not take
action to repair a pump that shows signs of dripping liquids, then a
potentially large leak and loss of valuable product would continue
unnecessarily. The weekly visual inspection requirement for pump
seals is a measure towards preventive maintenance and a supplement to
monitoring, and it is common sense. As stated in Chapter 4, page 4-6,
of the BID for the proposed standard, observation of liquid leaking
along the pump shaft indicates a seal failure and signals the need for
repair.
2.4.2.9 Valve Maintenance. Two commenters questioned specific
maintenance (repair) requirements for valves.
Comment; One commenter clarified a public hearing comment by
recommending that the concept of follow-up maintenance on valves
should be explained in more detail. In general, however, the commenter
agreed that the valve maintenance procedure is acceptable, that is, an
attempted repair on a valve that does not bring the concentration
below 10,000 ppm would be followed by tagging the valve for repair at
the next turnaround. The commenter recommended, however, that EPA
explain this concept in more detail (IV-D-21).
Response: As discussed in the preamble to the proposed standard,
an initial attempt to repair a leaking valve should be accomplished as
soon as practicable, but no more than 5 days after detection of the
leak. Attempting to repair the leak within 5 days will help maintenance
personnel identify the leaks that can not be repaired with simple
field repair or without shutdown of the process unit. The initial
repair attempt may consist of simple repair, which can be performed as
2-75
-------
leaks are detected (undirected repair), or may be scheduled for attention
at a later time, not to exceed the 5-day limit. The initial attempt
at simple valve repair shall include, but not be limited to, tightening
of the valve stem packing gland, tightening or replacement of valve
bonnet bolts, and in some cases, application of leak sealant.
Valves that continue to leak after simple field repair attempts
must be repaired within 15 days following initial leak detection. The
15-day interval provides time for properly isolating leaking valves
that require more than simple field repair. The 15-day interval also
provides the owner or operator with sufficient time for determining
precisely which spare parts are needed and flexibility in scheduling
repair for leaking valves.
Delay of repair beyond 15 days will be allowed for leaks that can
not be repaired without shutting down a process unit. In general,
these leaks will be repaired at the next unit shutdown. Leaking
valves will not be allowed to operate beyond process unit shutdown,
except in unique cases where replacement of entire valve assemblies is
required. In these unique cases, the valve will be allowed to leak
beyond the scheduled process unit shutdown, provided the owner or
operator can demonstrate that a sufficient stock of spare valve assemblies
had been maintained. In addition, delay of repair beyond 15 days will
be allowed for leaks if the unit shutdown is less than 24 hours in
duration.
Comment: The other commenter suggested that more emphasis should
be placed on preventive maintenance procedures and cited as an example
the possibility of replacing the packing in valves at regular intervals
before it becomes brittle and subject to leakage (IV-D-31).
Response: Certainly the ideal way to eliminate benzene fugitive
emissions is to prevent them from occurring altogether. Some of the
equipment and performance requirements in the final standard provide
for this where possible. One means of reducing leaks from valves is
through scheduled preventive maintenance. . This type of program would
not, however, eliminate leaks due to the numerous variables affecting
the equipment leak occurrences. General maintenance is already performed
in process units covered by the standard but, while requiring routine
2-76
-------
maintenance would result in some extra emission reductions, the extra
effort is not justified.
For example, a new valve stem packing may be loose, or the seal
of a valve bonnet flange may be imperfect, thereby allowing leaks.
These types of leaks would not be corrected by simple replacement of
valve stem packings at specified intervals. Monitoring would be
necessary to detect and repair leaks not resulting from valve stem
packing deterioration. Monitoring at specified intervals will also
detect leaks due to premature valve stem packing failure. Further,
the alternative performance standard for valves allows less than
1.0 percent of the valves to leak at any time. This performance
standard is designed to provide incentive to plant owners or operators
to perform preventive maintenance in order to avoid some of the costs
of recurrent monitoring. The selection of the alternative performance
standard for valves is discussed in Section 2.4.11.
2.4.2.10 Safety of Monitoring Relief Valves. Comment: One
commenter questioned the safety of requiring all safety/relief valves
to be monitored, stating that if a process upset occurs while monitoring
one of these valves, the valve would relieve, and serious injury or
death could result. The commenter added that the only potential leak
area on a safety valve is the open end of a valve designed to vent to
atmosphere. The commenter recommended, therefore, that EPA eliminate
this monitoring requirement on safety/relief valves (IV-D-30).
Response: The standard for new and existing pressure relief
devices in gas service in the petroleum refining and chemical manufac-
turing industries requires an annual performance test using Reference
Method 21 to verify that the device is maintained at no detectable
emissions. The annual testing is similar to testing done by EPA and EPA
contractors in collecting data for the standard and similar to testing
required by States under implementation plans. These tests could be
scheduled during periodic inspections of pressure relief devices, which
are typical of many industry safety practices. Monitoring should be
done by personnel who understand the precautions needed when monitoring
pressure relief devices. If a pressure relief device is likely to
relieve when monitoring occurs, then special precautions, such as
monitoring of process conditions (temperature and pressures), should be
2-77
-------
taken by the process unit operator. Based on EPA's experience in
collecting data for pressure relief devices, monitoring of these devices
can be done safely.
In addition to the annual performance test, pressure relief
devices must be checked within 5 days after each relief discharge to
demonstrate the attainment of the no detectable emissions limit. When
a relief valve/rupture disk system is used, EPA anticipates the rupture
disk would need to be replaced after each relief discharge to ensure
the standard is achieved. The relief valve would be monitored when the
system is put back in service. For a pressure/relief valve that is
equipped with soft-seat o-rings that can achieve the standard, the
relief valve would be required to be monitored after each relief dis-
charge (within 5 days) to demonstrate attainment of the no detectable
emissions limit. However, the pressure relief device does not need to
be monitored (annually or within the 5-day limit) if it is piped to a
closed vent system connected to a control device (e.g., flare header to
flare).
2.4.2.11 Alternative Standards for Valves. Comment: One commenter
recommended that the alternative standards for valves in the refinery
VOC fugitives NSPS be incorporated into the proposed benzene fugitives
NESHAP. For example, the commenter suggested that the 2 percent
allowable number of leaks and the skip-period leak detection/repair
program could be incorporated into the proposed benzene fugitives
NESHAP (IV-D-13; IV-D-27).
Response: EPA realizes that the emission reduction and annualized
cost of the leak detection and repair program depend in part on the
number of leaking valves that are detected during monitoring. If very
few valve leaks are detected in a process unit, then the amount of
benzene that could be reduced by the program for valves is much smaller
than the amount that could be reduced in a unit having more leaks.
Additionally,, the annualized cost of the leak detection and repair
program would be larger for a unit with fewer leaks than in a unit
with more leaks, because the annualized cost includes a recovery
credit based on the amount of benzene reduced by the program. Thus,
the annualized cost per megagram of benzene emission reduction for the
2-78
-------
leak detection and repair program varies with the number of valves
which leak within a unit.
For example, a monthly leak detection and repair program for
valves in benzene service typically results in an annualized credit of
about $20 per valve and achieves an annual benzene emission reduction
of 0.09 Mg per valve for an average-size process unit. In contrast,
for an average-size process unit with 1.0 percent of the valves leaking
on the average, a monthly leak detection and repair program results in
an annualized cost of about $8 per valve and achieves an annual emission
reduction of 0.004 Mg. For an average-size process unit with 0.5 percent
of the valves leaking on the average, a monthly leak detection and
repair program results in an annualized cost of about $9 per valve and
achieves an annual emission reduction of 0.0011 Mg (IV-B-5).
EPA judges that the emission reduction and annualized cost
relationship is unreasonably high for process units having on the
average fewer than about 1 percent of valves leaking (IV-B-5).
Based on this judgment, an allowable percent of valves leaking was
determined that reflects the average of about 1 percent of valves
leaking, as discussed below.
An allowable percent of valves leaking was selected that included
the variability inherent in leak detection of valves. The variability
in leak detection of valves can be characterized as a binominal distri-
bution around the average percent of valves leaking. An allowable
percent of valves leaking of 2 percent, to be achieved at any point in
time, would provide an owner or operator a risk of about 5 percent that
greater than 2 percent of valves would be found leaking when the average
of about 1 percent was actually being achieved. EPA also considered
the range of percent leaking valves within process units and found that
at about 2 percent a break between low-leak plants and higher leak
plants occurred. Based on these considerations, EPA selected an allowable
percent of valves leaking of 2 percent.
EPA is providing two alternative standards that would exempt
valves in benzene service within process units from the required
monthly leak detection and repair program. Plant owners or operators
may identify and elect to achieve either of the alternative standards.
The alternative standards will allow owners or operators to tailor
2-79
-------
benzene fugitive emissions control programs to their own operations.
An owner or operator will report which alternative standard he or she
had identified and elected to achieve.
The first alternative standard will limit the maximum percent of
valves leaking within a process unit to 2 percent. This type of
standard will provide the flexibility of a performance standard. An
industry-wide performance standard which could be achieved by all
process units was not possible for'valves. This was due to the varia-
bility in valve leak frequency and variability in the ability of ;a
leak detection and repair program to reduce these leaks among all .
process units within the industry. However, this alternative standard
will allow any process unit the option of complying with an allowable
percent of valves leaking for a particular unit.. Choosing this alter-
native standard will allow for the possiblity of different leak detection
and repair programs and substitution of engineering controls at the
discretion of the owner or operator. This alternative standard will
also eliminate a large part of the recordkeeping and reporting associated
with the monthly leak detection and repair program for valves.
Performance tests as specified in 40 FR Part 61 require three
runs. Three runs for performance tests to determine the percent of
valves leaking are unnecessary. Thus, performance tests are exempt
from Part 61 in the final standard. However, this alternative standard
will require a minimum of one performance test per year. Additional
performance tests could be requested by EPA. If the results of a '
performance test showed that greater than 2 percent of the valves
leak, the owner or operator could be cited for violation of the standards.
Inaccessible valves, which would not be monitored on a routine basis
under §61.242-7(h), would be included in the annual test since an
annual test of these valves is not considered unreasonable.
In certain circumstances, an owner or operator may want to request
a waiver of future tests as provided in the General Provisions of
40 FR Part 61. This will provide flexibility for plant owners and
operators where, for whatever reason, routine leak detection and
repair is not needed to effectively control emissions. This will
include plants that use superior equipment, plants that have effective
occupational safety and health programs, or plants that simply do not
2-80
-------
leak for unknown reasons. Based on performance tests that demonstrate
the achievability of the 2 percent standard and information that
indicates that this standard would be achieved on a continuing basis,
EPA could waive the annual performance tests.
The second alternative standard will allow the use of skip-period
leak detection for valves in benzene service. Under skip-period leak
detection, an owner or operator could skip from routine leak detection
for valves to less frequent leak detection. This skip-period leak
detection program will require that a performance level of 2 percent
be achieved on a continuous basis with more than 90 percent certainty.
An owner or operator will choose one of two skip-period leak detection
programs for valves and then implement that program. The first skip-period
leak detection program could be used when fewer than 2 percent of the
valves had been leaking for two consecutive quarterly leak detection
periods. The first skip-period leak detection program will allow an
owner or operator to skip every other quarterly leak detection period;
that is, leak detection can be performed semiannually. Under the
second skip-period leak detection program, if fewer than 2 percent of
the valves had been leaking for five consecutive quarterly leak detection
periods, the owner or operator may skip three quarterly leak detection
periods; that is, leak detection can be performed annually. When more
than 2 percent of valves are found to leak, monthly leak detection
will be required to be resumed.
2.4.2.12 Need for Leak Detection and Repair Requirements.
Comment: One commenter noted that, in general, the leak monitoring
t
and repair program requirements are excessive, duplicate previous
regulations for hydrocarbon fugitive emissions, and only create unnecessary
paperwork for industry and the regulatory agency. The commenter noted
that at its refineries, the benzene handling units are manned 24 hours
a day, 7 days a week by qualified operators whose duties include the
constant checking and observing of all equipment in their assigned
area for fugitive emissions. In addition, the commenter stated that
it has installed a gas chromatograph that continually records the
level of benzene in the ambient air at the complex. The commenter
feels that these combined measurements are more than adequate to
prevent benzene fugitive leaks from going undetected (IV-D-20).
2-81
-------
Response: EPA recognizes that some refineries and chemical
plants already have programs that include checking and recording
ambient benzene levels. These programs, however, are not uniformly
found in all refineries and chemical plants. In addition, EPA has
found that gas chromatograph systems in general are highly subjective
and rely on operator diligence to a very high degree. The leak detection
and repair programs that are included in the final standard are intended
to identify benzene leaks that may not be detected by visual, audible,
or olfactory means, which tend to be very subjective. To the extent
that programs already in place provide effective emission reduction,
EPA is allowing plant owners or operators to select one of two alternative
standards for leak detection and repair of valves. EPA has given the
owners or operators the flexibility to tailor fugitive emissions control
programs to their own operations. (See Section 2.4.11 for a discussion
of the alternative standards for valves.)
2.5 RECORDKEEPING AND REPORTING
Several commenters raised issues pertaining to both reporting and
recordkeeping requirements. Others addressed reporting and recordkeeping
as separate topics.
2.5.1 Recordkeeping
Comment: One commenter felt that indicating a leak with a log
entry is preferable to tagging every leak and leaving the tag in place
for three months, since many leaks will be repaired as they are found
(IV-D-13).
Response: A tag is not specifically required by the standard;
only some sort of identification is required, and a log entry is
acceptable as long as the equipment component can be identified from
the log entry. Some identification would be needed for all fugitive
emission sources of benzene in the leak detection and repair program.
This distinguishes these sources from sources complying with other
requirements. Further, identification of a leak is necessary to
enable follow-up inspections to be made when using the allowance for
quarterly monitoring of certain valves. Valves must be identified as
leakers for two consecutive months after the leak is found before they
can be monitored quarterly. Thus, even though a tag is not required,
some sort of identification is required.
2-82
-------
Comment: One commenter suggested that, in order to eliminate
duplicate records for the leak detection and repair program, EPA
should allow the use of existing automatic data systems to maintain
the required records along with the other plant operating and maintenance
records. The commenter added that maintenance job orders to repair
fugitive emission leaks could be coded to allow them to be retrieved
from the computer data bank (IV-D-16).
Response: The owner or operator is allowed to maintain records
in any desired form. The only stipulation is that all records must be
true, accurate, and readily available for EPA inspection. The use of
automatic data systems to maintain records is allowed.
2.5.2 Reporting
Comment: One commenter felt that the reporting requirements are
purely for the ease of enforcement purposes, for data collection
purposes, or require the submittal of duplicate information. The
commenter suggested that EPA either delete the requirements or justify
the need for the routine reporting requirements in determining compliance
with the standard (IV-F-1).
One commenter suggested that after submitting the initial report,
plants should report only changes in the number of valves or leaks
detected and repaired (IV-D-11). Two commenters recommended that only
leaks not repaired should be reported quarterly (IV-D-13; IV-D-18).
One commenter stated that reporting the number of valves in each
process unit is unnecessary since the number rarely changes (IV-D-13).
One commenter added that records of details of unsuccessful repair
attempts, while possibly of interest to the owner or operator, should
not be made a reporting requirement (IV-D-18).
Response: Effective enforcement of the standard is necessary
especially in light of the hazardous nature of benzene. Reporting
requirements are very helpful for efficient and effective enforcement
of the standard. Contrary to what the commenter suggests, the requirements
are not made purely for the ease of enforcement or for the purpose of
data collection; nor are the reporting requirements duplicative.
Reports will be used in conjunction with records and inspections to
enforce the standard.
2-83
-------
In response to the comments on reporting requirements, EPA has
reduced the amount of information that must be reported by the plant
owner or operator. The information required in reports is the same
information that a plant manager would likely want to evaluate for his
or her program. The report will include the number of leaks that
occurred within the process unit during the reporting period, the
number of leaks that could not be repaired within 15 days, and the
general reasons for unsuccessful or delay of repair past the 15-day
period.
The requirement to report reasons for unsuccessful or delay of
repair is necessary to allow EPA to assess whether the owner or operator
is making reasonable attempts at repair and understands the workings
of the standard. EPA expects that delays will occur only because
repair would result in process unit shutdown. Such delays can be
readily explained by the owner or operator. Since EPA does not expect
many of these delays to occur, EPA considers reporting the reasons for
them to be reasonable.
The requirement to report the number of leaks found will assist
EPA in determining whether or not the number of leaks not repaired
within 15 days indicates reasonable attempt at repair. EPA will gauge
the significance of the number of leaks not repaired within 15 days in
relation to the number of leaks found.
EPA has decided to reduce the reporting requirements from quarterly
to semiannual reporting. This reduction is made in order to reduce
reporting costs and orient the focus of the reports towards revealing
longer-term trends in leak repair programs. The objective of the
enforcement program is to single out plants that, over the long-term,
repair leaks with less efficiency than most other plants. Such long-term
trends are best identifiable over 6-month periods. Quarterly reporting
would be more frequent and more costly than necessary. Annual reporting
would not be frequent enough because EPA enforcement personnel need to
inspect plants before the plant records become outdated and less
useful in helping to determine what has been occurring within the
plant. Furthermore, EPA enforcement programs are planned 1 year in
advance, and, since plant records may be destroyed after 2 years, it
2-84
-------
is likely that, under an annual reporting program, a plant inspection
could not be planned until after plant records have already been
destroyed.
In the proposed standard, EPA included a requirement to report
leak location and I.D. number. This would have allowed EPA to determine
whether certain fugitive emission sources of benzene in a plant were ,
causing repeated problems. However, in order to reduce reporting
requirements for industry and to reduce Agency review requirements,
EPA has decided to eliminate leak location and I.D. number from the
reporting requirements.
Since no reporting format is stipulated in the standard, reports
required by other regulations may simply be photocopied and submitted
in compliance with the benzene fugitive emissions standard as long as
the report satisfies the informational requirements of §61.117.
The selected program involves minimum recordkeeping and a combination
of inspections and reporting. This program is effective, efficient,
reasonable, and should fit well with management of the standard by
plant personnel. During the first 2 years of the program, the average
annual burden of reporting and recordkeeping to industry would:be
about 20 person-years and would cost industry about $520,000 (IV-B-13).
The burden is distributed among about 240 process units and, on an
annual basis, represents about 1 person-month per process unit. At
these costs, the program provides a reasonable level of compliance
monitoring.
The selected recordkeeping requirements are the minimum that
could be achieved without precluding the possibility of enforcing the
standard. In addition, they represent the minimum level of documentation
that plant personnel would require to evaluate implementation of the
standard. Without retrospective data, inspections would be useless
and reporting would be impossible. Reporting, being less expensive
than inspections to both industry and EPA, is an effective mechanism
for reducing the cost of inspections. Reports reduce the amount of
time required to conduct inspections and make it possible to reduce
the number of plants that need to be inspected. Plant inspections
enable EPA to determine whether or not an owner or operator understands
the standard and is in compliance with the standard: Inspections will
2-85
-------
be conducted when reported information is ambiguous or clearly indicates
the need for an inspection. The reporting requirements will enable
EPA to reduce the inspection rate and still maintain an effective
compliance monitoring program.
A range of five alternatives was evaluated by EPA in selecting a
reporting and recordkeeping program for the standard (IV-B-9).
The alternatives, which represent different combinations of recordkeeping.,
reporting, and inspection levels of effort, were formulated to correspond
with realistic programs suggested by industry and EPA personnel. The
alternatives were evaluated in terms of their costs to industry and EPA
and in terms of their effectiveness in maintaining a reasonable level
of compliance monitoring.
Comment: One commenter noted that, although the proposed rules
state that compliance with the reporting requirements of proposed
§61.117 relieves the obligation to file a pre-construction approval
application under 40 FR §61.07, nothing is said about the applicability
of 40 FR §61.05. According to the commenter, 40 FR §61.05 provides
that certain constructed or modified sources must obtain written
approval of the Administrator, unless exempted by the President under
Section 112(c)(2) of the Clean Air Act. The commenter suggested that
if benzene fugitive emission rules are to be adopted, then the proposed
40 FR §61.117 should be revised to address the general prohibitions
of 40 FR §61.05 as well as the specific application requirements of
40 FR §61.07 (IV-D-19).
Response; EPA agrees with this commenter. Accordingly, the
provisions in §61.117 were expanded to include §61.05(a). The sole
purpose of this provision is to reduce the administrative burden associated
with minor changes to existing process units. Nothing in this provision
should be interpreted to mean that compliance with the standard is
not required. EPA is retaining review and approval of the construction
of equipment in benzene service if it is associated with a new process
unit. EPA's ro.le in assuring effective control of hazardous air
pollutants and the lack of an unreasonable administrative burden
associated with the approval process convinced EPA to retain its review
and approval process for new construction. Comment: Two commenters
felt it is unreasonable to require plant managers to certify plant
2-86
-------
compliance with the standard when supervisory personnel are more
familiar with compliance status (IV-D-16; IV-D-18). One of the com-
menters (IV-D-16) recommended that the manager sign a "knowledge and
belief" certification that would best describe his function in reporting
leak detection and repair requirements. The commenter explained that
this certification would require the signer to personally examine the
information to be submitted and to satisfy himself by inquiry to those
immediately responsible for obtaining the information (i.e., supervisory
personnel) that the information is true, accurate, and complete.
Response: According to the General Provisions for National Emission
Standards for Hazardous Air Pollutants, Section 61.02, "owner or operator"
is defined as any person who owns, leases, operates, controls, or
supervises a stationary source. Under this definition, any supervisory
personnel may be designated to certify plant compliance.
2.5.3 Miscellaneous Comments Addressing Both Recordkeeping and
Reporting
Several commenters addressed reporting and recordkeeping together
as a single topic.
Comment: Several commenters characterized the proposed reporting
and recordkeeping requirements as excessive, unnecessary, cumbersome,
or onerous for both industry and EPA (IV-F-1; IV-D-19; IV-D-20; IV-D-13;
IV-D-21; IV-D-27; IV-D-2; IV-D-26). These commenters suggested reducing,
streamlining, or justifying the proposed requirements.
Three commenters stated that the proposed reporting and recordkeeping
requirements would duplicate already existing VOC control reporting
and recordkeeping requirements (IV-D-19; IV-D-20).
One commenter made the following suggestions to reduce reporting
and recordkeeping requirements for both industry and EPA: (1) that
the reporting and recordkeeping requirements be abridged to include
only pertinent data, i.e., (e) equipment identification, (b) ppm
benzene leak detected, (c) date and individual detecting the leak,
(d) date of repair, and (e) ppm benzene detected, if any, after repair;
and (2) that each owner or operator should structure his/her own
program, as required, to assure that needed repair is made within
15 days unless a shutdown is needed (IV-D-18). One commenter felt that
the reporting and recordkeeping requirements were not developed within
2-87
-------
the guidelines of the Federal Reports Act, 44 USC §3501, which states
that "information needed by Federal agencies shall be obtained with a
minimum burden upon business enterprises .... Unnecessary duplication
of efforts in obtaining information through the use of reports, question-
naires, and other methods shall be eliminated as rapidly as practicable."
This commenter recommended that the proposed reporting and recordkeeping
requirements be significantly reduced as required by the Federal Reports
Act (IV-D-19).
Response: The reporting and recordkeeping requirements are designed
to require absolutely the minimum level of industrial effort necessary
to ensure effective implementation of the standard. Responding to the
comments, EPA has reduced the reporting requirements as described in
Section 2.5.2 of this document. Further reduction in reporting or
recordkeeping requirements would eliminate their usefulness to EPA as a
tool for determining compliance with the standard.
The rationale for recordkeeping under the standard is to document
information relating to the use of specific equipment and the results
of the leak detection and repair program. A log will be maintained
for information pertaining to leaking sources. The log will contain
the instrument identification number, the leaking source identification
number, the date of detection of the leaking source, the date of the
first attempt to repair the source, and the date of final repair.
These requirements are less strenuous than those suggested by the
commenter above in IV-D-18. For the purpose of enforcing equipment
standards, records will be maintained of the dates of installation,
startup, control equipment repair, control equipment modifications,
and dates and descriptions of any control equipment failures. These
records will be needed to provide information necessary to allow
enforcement personnel to assess the effectiveness of implementation
and maintenance of equipment standards.
The recordkeeping requirements in no way force the owner or
operator to keep duplicate records. Records that are kept to comply
with other standards may be used to comply with the benzene fugitive
emissions standard as long as they satisfy the.recordkeeping require-
ments in §61.116. Since the benzene fugitive emission standard does
not require records to be kept in a particular form, any form will be
2-88
-------
acceptable as long as it contains the required information. Similarly,
since no reporting format is stipulated in the standard, reports
required by other regulations may simply be photocopied and submitted
in compliance with the benzene fugitive emissions standard as long as
the report satisfies the informational requirements in §61.117.
The comment suggesting that the owner or operator be allowed to
structure his or her own program is unclear. Neither recordkeeping
and reporting requirements nor leak detection and repair requirements
dictate how the owner or operator should structure a repair program.
The leak detection and repair program requires certain equipment to be
used in certain cases, but the owner or operator may structure any
program to achieve compliance with equipment requirements within
15 days of finding a leak.
2.6 COSTS
The cost of the proposed standard is discussed in Chapter 8 of
the BID for the proposed standard, and its economic impacts are in
Chapter 9. Since publication of the BID for the proposed standard,
EPA estimates that the annualized cost of the standard is smaller than
estimated at proposal. The main reason for this change is that the final
standard is less stringent and accordingly less costly than the proposed
standard. The BID for the proposed standard concludes that any potential
price increases resulting from imposition of the proposed standard
would be well under 1 percent, and that the profits and market positions
of individual manufacturers would not be changed. In view of the lower
cost of the promulgated standard explained below, these conclusions are
underscored.
Comments and responses on Chapters 8 and 9 are assigned in this
section to three categories: impacts on small facilities, cost effective-
ness, and benefit-cost considerations. This categorization is not
rigid because some comments are quite broad.
2.6.1 Impacts on Small Plants
This subsection addresses two principal concerns: the effect of
the standard on small businesses, and the application of the standard
to small-volume plants and to plants that use minor amounts of benzene
or that use benzene intermittently.
2-89
-------
The Regulatory Flexibility Act (Public Law 96-354, September 19,
1980) directs Federal agencies to pay close attention to minimizing any
potentially adverse impacts of a standard on small businesses, small
governments, and small organizations. Accordingly, EPA has reviewed
the final standard in accordance with the Regulatory Flexibility Act.
This standard will have no known effects on small governments and small
organizations. It may affect some small businesses, but the impacts
will be few and minor. Essentially all firms that will be required to
comply with the standard either are not small businesses, or are
subsidiaries of large firms. Because the overall net annualized (1985)
cost of the standard is expected to be small ($100 thousand per year
for new units and $400 thousand per year for existing units), there
should be no adverse impacts on firms regardless of whether they are a
small business.
In the analysis above, a small business is one that employs fewer
than 750 persons. This level was set by the Small Business Admini-
stration (SBA) as a criterion for extending SBA loans and related
assistance (13 FR Part 121, Schedule A). The definition applies to
firms that manufacture cyclic crudes and cyclic intermediates, pharma-
ceutical s, and many other chemicals. The BID for the proposed standard
lists 77 existing companies that may be affected by the standard.
Most of these companies manufacture cyclic crudes and many other
chemicals. With the possible exception of two companies, all of these
firms either employ more than 750 persons, or are subsidiaries of
large firms.
Comment: One commenter felt that the leak detection and repair
requirements would impose substantial costs on small-volume users of
benzene with no appreciable benefit to public health. According to
the commenter, small-volume pipeline systems at pharmaceutical plants
may contain seve'ral hundred valves that would need to be monitored
monthly when in benzene service; the sporadic need for additional
manpower to perform monitoring and clerical functions arising from
recordkeeping requirements would have a major impact on operating
expenses. The commenter added that the economic and administrative
burden of complying with the standard would be heavy for small-volume
2-90
-------
users, as compared to large benzene production units, in proportion to
the level of fugitive emissions from such facilities (IV-D-16).
Response: EPA is exempting most intermittent users or producers
of benzene, such as pharmaceutical plants. As discussed in Section 2.8.1,
EPA is exempting from the standard equipment at plant sites with a
design usage rate of 1,000 Mg/yr or less. This is true regardless of
whether benzene is produced or used. The 1,000 Mg/year exemption would
exclude most research facilities, pilot plants, and intermittent users
of benzene from the standard.
The possibility that pharmaceutical operations could be adversely
affected by the final standard is very small. This is true for several
reasons. First, most pharmaceutical plants use very little benzene.
According to estimates contained in Market Input/Output
Studies - Benzene Consumption as a Solvent (EPA-560/6-77-034, October
1978, p. 41), 1978 benzene consumption by pharmaceutical manufacturers
was about 0.72 Gg. No companies consumed more than 1,000 Mg/year in
1978. The commenter states that they consumed about 325 Mg/year
during 1981. Thus, it is unlikely that pharmaceutical operations
would be affected by the standard. Second, benzene consumption by the
pharmaceutical industry is declining rapidly. The market input/output
study just noted estimates that consumption declined from 2.14 Gg in
1976 to 0.72 Gg in 1978, a decline of about 66 percent over the 2-year
period. Third, the number of companies using benzene has also declined
and is expected to continue to fall. For the 2^year period 1976 to
1978, the study estimates that the number of pharmaceutical companies
using benzene declined from 10 to 5. And finally, even though pharma-
ceutical operations with a benzene throughput in excess of 1,000 Mg/yr
are subject to the standard, they have substantial equipment inventories
in benzene service and, therefore, emit benzene in enough quantity to
warrant coverage by the final standard. EPA has reviewed the compliance
costs for these operations and has concluded that these costs are
reasonable (IV-B-17).
2.6.2 Cost Effectiveness
Some commenters claimed that the cost effectiveness was not
calculated properly, or with the correct data. Others said that the
cost effectiveness did not justify the standard. In these comments,
2-91
-------
"effectiveness" was measured in terms of emission reduction in some
cases, and lives saved in others.
Comment: Two commenters felt that EPA had not selected the most
cost-effective alternative as the basis for the proposed standard
(IV-F-1; IV-D-8). One of the commenters (IV-F-1) said that industry's
experience in air pollution abatement control programs has led it to
conclude that capital costs in excess of $3,000 per magagram are not
cost effective and should be rejected unless the other alternatives do
not substantially achieve the necessary degree of control. The commenter
concluded that the new SOCMI data indicate that the cost effectiveness
is higher than the industry guideline of $3,000 per megagram.
Response: The basis for the standard is discussed in Section 2.3.
Selection of BAT is based on an examination of the incremental cost
effectiveness among various control techniques for each fugitive
emission source. Whether to require more restrictive control than BAT
is based on judging the risks remaining after BAT is applied and the
cost of reducing these risks. Since proposal, EPA has selected a less
restrictive standard than the one proposed in January 1981; consequently,
the cost associated with the standard has decreased. The overall cost
of the standard is expected to be less than $100/Mg of benzene and total
emissions (benzene and other VOC) reduced. Even though some processes
could experience an overall cost increase as a result of the standard,
EPA considers the cost impact reasonable.
In response to the commenter and as discussed in Sections 2.3 and
2.8.4, EPA based emission estimates on refinery emission factors rather
than on SOCMI emission factors because recent benzene-specific emission
data from refineries and chemical plants are more similar to refining
units than to SOCMI units. Therefore, the commenter's conclusion that
the new SOCMI data (discussed in the AID, EPA-450/3-82-010) indicating
that the cost effectiveness is higher than the industry guideline of
$3,000 per megagram is not based on the same emission estimates used by
EPA. However, using capital cost as the basis for cost effectiveness
as the commenter did, (which is conceptually difficult to interpret in
that there is no time period in the numerator while the denominator is
on an annual basis) costs are considerably lower ($930/Mg of benzene
for new sources and $l,100/Mg of benzene for existing sources) than the
2-92
-------
commenter's guideline of $3,000 per megagram. It is unclear how the
commenter estimated "operating costs." Using net annualized costs,
however, the overall cost effectiveness for the standard is less than
$100/Mg of benzene and total emissions (benzene and other VOC) reduced.
Table 2-5 presents the revised annualized model unit control costs
and savings for the standard, based on the control techniques discussed
in Section 2.3. Table 2-6 presents the revised 1985 nationwide costs
of the standard for both existing and new units.
Comment: One commenter said that EPA should investigate the
incremental cost effectiveness of each discrete control requirement,
and not the cost effectiveness of arbitrarily combined groups of control
alternatives (IV-D-27).
Response: Selection of the final standard is based, as requested
by the commenter, on an incremental cost-effectiveness analysis of each
discrete control requirement.
Comment: One commenter felt that cost effectiveness dictates that
Alternative IV at a minimum, or preferably Alternative V, should have
been selected instead of Alternative III in order to fulfill the mandate
of Section 112. This judgment is based on the commenter's observation
that the net price increase would be less than one-fourth of 1 percent
in benzene prices for Alternative IV and less than 4 percent for
Alternative V. The commenter considered this a "trivial price to
pay for saving additional lives," noting that, in rulemaking on the
vinyl chloride standard in 1975, EPA decided that a price impact as
high as 10 percent would have been acceptable. The commenter added
that cost estimates are usually exaggerated, and firms often develop
innovative, cheaper compliance techniques (IV-D-31).
Response: EPA has selected the final standard after considering
whether the incremental risk reduction that would be achieved by applying
additional control beyond BAT warrants the incremental cost. As explained
in Section 2.3, EPA made the decision not to require additional control.
EPA's rationale for selecting standards under Section 112 is discussed
in EPA-450/5-82-003.
Comment: One commenter stated that the benzene standards involve
minimal emissions and minimal population risk concerns at very great
costs for implementation (IV-D-24).
2-93
-------
Table 2-5. ANNUALIZED MODEL UNIT CONTROL COSTS AND,SAVINGSa
OF THE BENZENE FUGITIVE EMISSIONS. STANDARD0
(Thousand May 1979 Dollars)
Item
Installed Capital Cost
Total Annual f zed Cost
Recovery Credit
Net Annualized Cost or Savings
VOC/Yr Emission Reduction
Benzene/Yr Emission Reduction
Cost (Savings) per Hg
Total Emissions Reduced
Cost (Savings) per Hg Benzene Reduced
New
16
10.1
6.9
3.2
19 Hg
12 Hg
0.17
0.27
A
Existing
17
, 10.2
6.9
3.3
19 Hg
12 Hg
0.17
0.28
Model Unit0
B C
New Existing New Existing
31 32 48 49
20.1 20.2 30.1 30.4
20.5 20.5 33.8 33.8
(0.40)d (0.30)d (3.7)d (3.4)d
56 Hg 56 Hg 93 Hg 93 Hg
31 Hg 31 Mg 71 Mg 71 Mg
(0.007)d (0.005)d (0.040)d (0.036)d
(0.013)d (0.010)d (0.052)d (0.048)d
Costs are for new and existing units and include monitoring instruments but do not include
cost for compressors because compressors in benzene service are not known to exist.
Recovery credits are based on total emission reductions (benzene and other VOC) and $370/Hg.
The standard requires monthly leak detection and repair programs for valves and pumps, and
equipment specifications for pressure relief devices, open-ended lines, sampling connections,
and product accumulator vessels.
cHodel units have the following numbers of components:
A
5
34
87
3
35
9
1
Hodel Unit
B
15
100
264
9
105
26
2
C
25
167
439
16
175
44
2
Pumps
Valves
Gas
Light Liquid
Pressure Relief Devices (gas)
Open-ended Lines
Sampling Connections
Accumulator Vessels
Several assumptions are made to compute model unit costs. For pressure relief devices 75
percent are assumed already controlled in the absence of the standard. For the 25 percent
of pressure relief devices that are uncontrolled, it is assumed that 75 percent will be
controlled with a closed vent system to flare and 25 percent will be controlled with a
rupture disk system. For relief valves using rupture disks, one-half will be controlled
with block valves and one-half will be controlled with 3-way valves. For accumulator
vessels, 95 percent are assumed already controlled in the absence of the-standard.
Numbers in parenthesis denote savings.
eTotal emissions include benzene and other VOC.
2-94
-------
Table 2-6. 1985 NATIONWIDE COSTS OF THE BENZENE FUGITIVE
EMISSION STANDARD
(Thousands May 1979 Dollars)
I teni
Installed Capital Cost
Total Annual i zed Cost
Recovery Credit0
Net Annual ized Cost
Total Emission Reduction Per Yr
Benzene Emission Reduction Per Yr
Cost Per Mg Total Emissions Reduced
Cost Per Mg Benzene Reduced
New Units3
1,400
900
800
100
2,200 Mg
•l.SOOMg.
0.045
0.067
Existing Units
5,500
3,400
3,000
400
8,400 fig
5,200 Mg
0.048
0.077
New unit construction through 1985 is assumed to be:. A, 42; B, 7; and C, 11.
Existing units assumed for 1985 are: A, 131; B, 75; and C, 18.
cBased on total (benzene and other VOC) emission reduction and $370/Mg.
2-95
-------
Response: The "benzene standards" to which the commenter refers
are the regulations for benzene storage vessels, ethyl benzene/styrene
plants, maleic anhydride plants, and benzene fugitive emissions. This
document only covers benzene fugitive emissions; other documents cover
the other source categories.
As discussed in Section 2.1.2, EPA has determined that benzene
fugitive emission sources pose significant risk to the public residing
in the vicinity of benzene-producing and using plants. The final
standard will reduce total benzene fugitive emissions from existing units
by about 69 percent at a small industry-wide net annualized cost of
$77/Mg of benzene emission reduction. For new units in 1985, the
standard will also reduce emissions by about 72 percent at a net cost
of $67/Mg of benzene emission reduction. (This is based on not counting
the likely effects of standards of performance for VOC fugitive emissions
in the synthetic organic chemical manufacturing and petroleum refining
industries. To the extent that these standards impact benzene sources,
the emission reduction and costs will be proportionately smaller;
however, the cost effectiveness will remain the same.) EPA does not
consider these cost-effectiveness numbers to be unreasonable.
Comment: One commenter noted that, utilizing the costs presented
in the BID for the proposed standard, the cost of saving one life in
20 years will be $116 million. The commenter felt that this is far
greater than is considered reasonable in conventional risk/benefit
terms. The commenter noted that the Congress, in its actions to
minimize auto-related deaths, has established a cost/benefit relation-
ship of $0.5 million to $1.25 million per life saved as being reasonable
to the benefit derived (IV-D-24).
Response: The estimated cost of saving one life in 20 years
presented by the commenter is based on costs for the standard at
proposal. In addition, EPA believes that the commenter's estimate is
high. The commenter did not present his method for deriving his estimate.
Since proposal, the annualized cost of the standard is estimated to be
small. (See Tables 2-5 and 2-6, above.) Imposition of this standard
will reduce the incidence of cancer and the exposure of people to
benzene with essentially no economic impact on the plants covered by
2-96
-------
the standard. Therefore, EPA considers the benzene fugitive emissions
standard to be reasonable.
Selection of the proposed and promulgated standards Was based on
emission reductions and costs and consideration of health impacts, not
on the dollars spent per life saved. Even though the commenter's
"cost-effectiveness" numbers may be calculated correctly, they are not
a valid indicator of the benefits derived from the standard. The
health impacts considered during selection of the standard are used,
not in absolute terms but in relative terms, only in comparing BAT
with a more stringent alternative to determine whether or not the
residual risks after BAT are unreasonable. A more meaningful and
accurate measure of "cost effectiveness" is the cost per unit emission
reduction, which in this case is less than $100/Mg of benzene and VOC.
2.6.3 Benefit-Cost Considerations
Most comments on benefit-cost considerations cited Executive
Order 12291; these comments implied that the benefits of the standard
do not justify the costs. "Benefits" is a much broader term than
"effectiveness" and relates to the broad spectrum of improvements that
flows from a reduction of emissions.
Comment: Five commenters stated that the Agency has failed to
perform an adequate cost-benefit analysis as required by Executive
Order 12291 (IV-D-8; IV-D-27; IV-D-19; IV-K-1; IV-F-1). One commenter
stated that the proposed standard represents overkill, since the
improved, air quality achieved by regulating the sources is not fully
justified by any benefits shown in the proposed standard. The commenter
recommended that EPA reconsider the cost versus benefits of the proposed
regulation (IV-D-10).
Response: Executive Order 12291 specifies that a regulatory
action, to the extent permitted by law, must not be undertaken unless
the potential benefits to society from the regulation outweigh the
potential costs to society. An exhaustive benefit-cost analysis is
not required here because the benzene fugitive emissions standard does
not constitute a major rule within the meaning of the Executive Order
and is not appropriate here because the costs and time required to
conduct such an analysis are prohibitive.
2-97
-------
As discussed in Section 2.6.2, costs per megagram of emission
reduction were used in selecting the promulgated standard. This
regulation was submitted to the Office of Management and Budget (OMB)
for review as required by Executive Order 12291. Any comments from
OMB to EPA and any EPA responses to those comments are available for
public inspection in Docket No. A-79-27, Central Docket Section, West
Tower Lobby, Gallery 1, Waterside Mall, 401 M Street, S.W., Washington,
D.C. 20460.
According to the directives of Executive Order 12291, "major
rules" are those that are projected to have any of the following
impacts:
• An annual effect on the economy of $100 million or more,
• A major increase in costs or prices for consumers, individual
industries, Federal, State, or local government agencies, or
geographic regions, or
• Significant adverse effects on competition, employment,
investment, productivity, innovation, or on the ability of
United States-based enterprises to compete with foreign-based
enterprises in domestic or export markets.
As noted above in Tables 2-5 and 2-6, the standard will result in a small
net annualized cost. None of the other effects referred to in the
Executive Order is anticipated. Thus, the standard does not constitute
a major rule because the costs of the standard and its overall impact
on the economy will be negligible.
Although benefits have not been rigorously defined and weighed
against costs, the standard will create significant benefits. Such
benefits are related to the reduction of risks of contracting leukemia
due to exposure to benzene.
In addition to the avoidance of the health effects noted above,
regulatory action will reduce the rate of emission of VOC's to the
atmosphere. Because VOC's are precursors of photochemical oxidants,
the ambient concentrations of such oxidants will be reduced. Among
the benefits of reduced exposure to ozone are reductions in the following:
2-98
-------
• Human health effects - ozone exposure has been shown to
cause increased rates of respiratory symptoms, such as
coughing, wheezing, sneezing, and shortness of breath;
increased rates of headache, eye irritation, and throat
irritation; and physiological damage to red blood cells.
One experiment links ozone exposure to human cell damages
known as chromosomal aberrations.
• Vegetation effects - reduced crop yields as a result of
damages to leaves and/or plants have been shown for several
crops including citrus, grapes, and cotton. The reduction
in crop yields was shown to be linked to the level and
duration of ozone exposure.
• Materials effects - ozone exposure has been shown to accelerate
the deterioration of organic materials, such as plastics and
rubber (elastomers), textile dyes, fibers, and certain
paints and coatings.
• Ecosystem effects - continued ozone exposure has been shown
to be linked to structural changes of forests, such as the
disappearance of certain tree species (Ponderosa and Jeffrey
pines) and death of predominant vegetation. Hence, ozone
causes a stress to the ecosystem.
2.7 TEST METHODS AND PROCEDURES
Comment: One of the commenters stated that the test method for
determining compliance with the 200 ppmv and 10,000 ppmv leak levels
is a general hydrocarbon analysis, and, therefore, nonbenzene hydrocarbons
are not distinguished from benzene in the analysis. The commenter
believed that this test method would require a facility with a stream
containing 10.5 percent benzene by weight to control leaks to the same
degree as a stream that handles 100 percent benzene and whose fugitive
benzene emissions are about 10 times as concentrated (IV-D-20).
Response: The commenter is referring to Reference Method 21.
As noted below, Reference Method 21 is used to classify fugitive emission
sources according to whether they leak or not. Thus, the method is
used to determine which sources of benzene are leaking benzene to the
atmosphere, not to determine how much benzene is being leaked. The
commenter is correct by stating that different streams containing
2-99
-------
different percentages of benzene would be required to control leaks to
the same degree. However, EPA did not estimate emission reductions and
cost effectiveness based on streams containing 100 percent benzene in
the line. The estimates were based on-estimates of streams in benzene-
using and producing plants. These percentages ranged from small amounts
(about 25 percent) to large amounts (100 percent). The selection of
10 percent as a cutoff for "in benzene service" addresses the situation
of controlling very small amounts of benzene and was discussed in the
preamble to the proposed standard. Also, to the extent that a stream
with less than an average percent benzene in it is controlled by the
standard, the VOC's reduced would be an added benefit (see Section 2.3).
Comment: One commenter remarked that Reference Method 21 cannot
be used to calibrate all the instruments allowed by the procedure
(IV-D-24). The commenter referred to its March 18, 1981, comments on
the SOCMI VOC fugitives CTG. The commenter said that the method allows
the use of photoionization devices which will not respond to the
calibration gases specified in the method. The commenter recommended
that, if an instrument cannot be calibrated, it should not be used to
measure emissions.
Response: The reference method has been written to be applied to
fugitive emission source screening in general, with specific application
requirements being established in each regulation. The method stipulates
that the monitoring instrumentation selected by the owner or operator
must be responsive to the chemical in the process line, in this case,
benzene. The method states that photoionization instruments might
meet the requirements, but it does not state categorically that photo-
ionization instruments may be used. However, since this type of
analyzer may be useful in certain process units, and it does not
respond to methane, an alternative calibration material has been added
in the regulation. Based on the comparison of results in the response
to the next comment, hexane is specified as the alternate material.
At an ionization potential of 11.7-11.8eV, photoionization analyzers
will respond to hexane as a calibration material. The use of the
alternative material is not limited to a single type of analyzer.
Comment: One commenter noted that EPA's monitoring procedure
allows the use of alternate instrumentation that will yield different
2-100
-------
results for the same leak. This inconsistency, the commenter noted,
will lead to enforcement problems. The commenter recommended that the
leak rate be defined in terms that will provide consistent monitoring
results regardless of the application or analytical instrument operator
(IV-D-24).
Response: Reference Method 21 gives specifications for the instrument
to be used in monitoring fugitive emission sources of organic compounds.
The technique is intended to identify leaks only, not to provide a
rigorous analytical concentration of organic compounds. A specific
statement has, therefore, been added to Reference Method 21 to clarify
the intention to identify leaks only.
The variation in response factor, due to compound or instrument,
is not expected to affect significantly the number of leaks determined
through screening because screening values are usually much greater
than 10,000 ppm for leaks and much less than 10,000 ppm for nonleaks.
As demonstrated by the Maintenance Study (IV-A-6, Appendix E), results
on valves, the largest percentage of valves found leaking were screened
at or above 50,000 ppm. This trend was also noted in the petroleum
refining studies (II-A-30). Thus, when a source is found to be leaking,
it is likely to have a high screening value.
Laboratory experiments using two organic compound analyzers indicated
a wide variation in response factors for a number of organic chemicals
from 0 to 571. However, 90 percent of the chemicals tested had responses
between 0.1 and 10 (Response Factors of VOC Analyzers, Report No. 4,
IV-A-15; Response of Portable VOC Analyzers, Report No. 5, IV-A-19).
Differences were also seen between the two types of analyzers tested.
When considered in analyzing leak frequencies (Analysis Report, IV-A-18),
the response factor variation, however, did not produce significant
changes in the percent leaking estimates resulting from the SOCMI
24-unit study (II-A-34). Although a small reduction in the estimated
leak frequencies is indicated for gas valves in high leak process units,
the estimates in all other cases were almost indistinguishable from the
unadjusted estimate. Furthermore, the differences when present were
in the same range as the variation in reproducibility described in
Docket No. A-79-27-IV-A-6.
2-101
-------
Comment: One commenter believed that the portable detectors
required in the leak detection program may be impractical due to their
inaccuracy, calibration difficulties, and frequent maintenance and
replacement (IV-D-16).
Response: The data presented in the EPA report, "Frequency of
Leak Occurrence for Fittings in Synthetic Organic Chemical Process
Plant Units" (II-A-34), indicated that from 1 to 1-1/2 hours per day
were required for calibration and maintenance of the monitoring instrument.
Based on the number of sources that were screened in each unit, this
calibration time amounted to 16 to 25 percent of the total time for •
screening.
The calibration time during the EPA studies was expected to be
longer than for screening alone since concentration measurements were
being recorded that would be used in further analysis. This necessitated
calibration with more than one standard concentration on a more frequent
basis (2 to 3 times daily). Also, because concentrations up to 100,000 ppm
were being measured, a dilution probe had to be calibrated several
times daily. However, routine screening would require calibration
with only one standard concentration. Also, a dilution probe would
not be required. Calibration time for routine screening is estimated
to require about 10 to 25 percent of that required during EPA tests,
or about 2 to 6 percent of the total time for screening.
The SOCMI 24-unit study report also listed a number of problems,
equipment-related and procedure-related, that were encountered during
the study. Procedure-related problems, as well as equipment problems,
to a large extent were due to the contractor's being remotely located
from his laboratories and repair facilities. Problems with instruments
in the field take longer to fix because of shipping delays. Furthermore,
many inefficiencies were encountered because the personnel performing
the studies were research staff who spent only short periods of time
on projects of this type, thereby lacking the experience necessary for
troubleshooting.
Time lost due to equipment failure is expected to be minimized by
maintaining the critical spare instrument parts (including readout
meter, battery pack, regulator repair kit, pressure gauges, hydrogen
flow valves, and filters) identified during the 24-unit study. Additionally,
2-102
-------
personnel familiar with the troubleshooting procedures should
facilitate instrument operation and repair. The proximity of
instrument .shops and labs within plants affected by the standard
should also improve routine screening efficiency.
The cost estimates in the BID for the proposed standard include
costs for two instruments, one of which was considered a spare.
Having a spare should decrease instrument downtime. Moreover, an
ample allowance of time was made for calibration and maintenance of
instruments. The 40 percent administrative and overhead charge allotted
includes time for calibration, and an additional $2,700 per year was
allotted for instrument materials and maintenance. In view of these
differences between research field studies and routine screening
activities, the calibration and maintenance costs allotted in the BID
for the proposed standard are reasonable, and no adjustments have been
made.
Comment: One commenter suggested that EPA determine whether a
sufficient supply of detectors is available (IV-D-16).
Response: EPA has no reason to believe that there is not a
sufficient supply of detectors available. Detectors such as those
allowed by the standard have been in use since promulgation of State
and OSHA regulations.
2.8 GENERAL ISSUES
2.8.1 Applicability
Several commenters questioned the applicability of specific
requirements on certain sources and facilities.
2.8.1.1 Exemption Requests. Five commenters requested exemption
from the standard for various facilities or sources.
Comment: One of these commenters recommended that certain
small-volume or intermittent benzene uses, such as use in pharmaceutical
production, be exempted from the standard (IV-D-16). The commenter
stated that in a pharmaceutical plant much of the process equipment is
multipurpose and is used for batch operations, rather than continuous
processes characteristic of large-scale organic chemical plants and
petroleum refineries. The variability in process operations and in
raw material usage would make it difficult to implement the leak
detection program as required by the standard. The commenter noted
2-103
-------
that pharmaceutical plants contain facilities incorporating extensive
small-volume pipeline systems, where a single processing train could
contain several hundred potential fugitive emission sources subject to
the standard. The commenter also stated that, in general, the volume
of benzene handled by pharmaceutical process equipment in continuous
operation is small compared to large volumes of benzene handled in a
benzene production unit. The commenter believed that since benzene
emissions from their batch processes are small, the risk of benzene
exposure to public health is small. The commenter stated that small
volume users and users other than organic chemical plants were not
considered in EPA's evaluation of the economic and environmental
impact of the standard.
Response: The intent of the standard is to reduce benzene emissions
from fugitive emission sources that contact benzene. Although some
processes are batch or intermittent, they can emit benzene in significant
quantities because these processes may require many pieces of equipment
with potential to emit benzene. However, EPA believes it is reasonable
to exempt plants from the standard when the cost of the standard is
unreasonably high in comparison to the achieved emission reductions.
Therefore, EPA decided to determine a cutoff for exempting these plants
based on a cost and emission reduction analysis.
Two approaches for setting the cutoff were considered: (1) number
of sources at a plant site and (2) benzene usage per plant site.
A cutoff based on the number of sources at a plant site would be
consistent with the model units used as a basis for the impacts of the
standard. However, this cutoff could not be applied readily to small
or intermittent users of benzene, such as pharmaceutical manufacturers
and research and development facilities. These facilities often
require frequent repiping; therefore, it is difficult to keep track of
how many sources are located at a plant site.
A cutoff based on benzene design usage is reasonable because it
is easily understood and can be applied readily to a plant site.
However, it is inconsistent with the model unit development where
throughput and emissions are generally not related, and its determination
is less straightforward than basing the cutoff on number of sources.
However, because the design capacity of a plant using or producing
2-104
-------
benzene can be applied easily, it was chosen as the basis for the
cutoff.
For plants in which the benzene emission reduction achieved by
implementing the standard is about 4 Mg of benzene per year, EPA has
determined that the cost of achieving this emission reduction is
unreasonable. In order to exclude plants on this basis, EPA selected a
minimi urn cutoff of 1,000 Mg/year per plant site based on a benzene design
usage rate or throughput. Throughput is determined by a mass balance
during the design stages of process operations, accumulating all benzene
processed in 1 year. The derivation of this cutoff is presented in
Docket Number IV-B-17. For plants with a benzene design usage
rate greater than 1,000 Mg/year, the cost of the standard is reasonable.
Comment: Another commenter suggested that the applicability of
the proposed rule be defined specifically to exclude research and
development activities and pilot plant studies which are already
designed to limit worker exposure to benzene (IV-D-26). One commenter
suggested an 18-month exemption for small temporary sources that would
allow laboratory, developmental pilot plant work, and full scale
experimental runs or commercial tests without the control and monitoring
requirements (IV-D-29). .
Response: The commenter suggests an exemption for small temporary
sources, such as laboratories and developmental pilot plants. Even
though worker exposure to benzene in these operations already may be
limited, the OSHA standards are not applicable in regulating benzene
fugitive emissions to atmosphere as discussed in the response to the
first comment in Section 2.1.2. As indicated in the previous response,
a plant design usage or throughput rate of benzene equal to or.
less than 1,000 Mg/yr per plant has been selected as a cutoff. Based on
the minimal amount of benzene processed at laboratory facilities and
pilot plants, the 1,000 Mg/yr cutoff is likely to exempt most of these
sources from the standard. Large-scale pilot plants or commercial tests
might process more than 1,000 Mg/yr. If this happens, these plants would
not have difficulty in achieving the standard since there are no major
equipment requirements and the work practice requirements are reasonable,
considering costs and emission reductions achievable. An 18-month
exemption for small temporary sources with a plant design usage
2-105
-------
rate of benzene greater than 1,000 Mg/yr is not needed because these
sources would not experience unreasonable costs and would achieve the
same emission reduction as production units.
Comment: One commenter suggested that flanges should be removed
from the definition of fugitive emission source, since testing by
Radian and others shows that flanges are a negligible source of emissions.
The commenter also recommended an exemption for pipeline systems
transporting benzene, especially from a plant to a remote terminal or
benzene-consuming facility. Due to the remote nature of some stations,
the commenter noted that the requirement of weekly visual inspections
of pumps and compressors and daily visual inspections required of new
pumps and compressors is unduly burdensome to the owner or operator of
the pipeline. The commenter added that requiring daily visual inspections
for replacement pumps may delay the replacement of existing pumps and
may result in increased emissions from the older pumps (IV-D-30).
Response: EPA agrees with the commenter that flanges are a small
source of emissions. Flanges in refineries contribute 2.2 percent of
all benzene fugitive emissions. However, flanges comprise 61 percent
of the total number of benzene fugitive emission sources. Thus, a
great deal of effort would be required to control these emissions.
For this reason, EPA excluded flanges in benzene service from routine
leak detection and equipment requirements in the proposed standard.
However, flanges do leak occasionally; therefore, the proposed standard
required that, if leaks are detected from flanges, they must be repaired
within the same allowable time as other sources (15 days). EPA has
not changed this requirement since proposal.
The commenter expressed concern that the pump and compressor
requirements are unduly burdensome for remotely located pipeline
systems. As discussed in Section 2.3.2, the equipment requirement
(dual mechanical seal system) for new pumps has been deleted and
replaced with a monthly leak detection and repair program, which is
also required for existing pumps. This revision eliminates daily
inspection of the sensor (or as an alternate, installation of an
audible alarm) to detect failure of the seal system. Weekly visual
inspections for liquids dripping from new and existing pump seals are
still required except for pumps that are remotely located.
2-106
-------
Remotely located emission sources are ones that are not located
within the boundary of the plant site. This equipment must be visually
inspected as often as practicable and at least monthly.
Compressors are required to be controlled with degassing vent
systems. Therefore, daily sensor checks (or installation of an audible
alarm) are still required unless the compressor is remotely located.
Weekly visual inspections of compressors have never been required.
For remotely located pumps and compressors, therefore, daily
sensor checks (or installation of audible alarm) and weekly visual
inspections of leaking equipment are not required, although they
should be inspected as often as practicable. However, the monthly
leak detection and repair requirements do apply to these remotely
located components except compressors. Compressors must use seal
systems; therefore, leak detection requirements are not needed.
Comment: One commenter expressed concern that EPA did not consider
equipment size or accessibility in its monitoring requirements for
pumps, compressors, valves, flanges, and pressure relief devices.
According to the commenter, the standard would apply to all equipment
in benzene service whether it is 1/8-inch tubing or a 48-inch pipe.
The commenter suggested, therefore, that the standard apply only to
meaningful and accessible sources, i.e., process lines 6-inch pipe
size and above (IV-D-29).
Response: EPA did not use equipment size as a criterion of
applicability in developing the benzene fugitive emissions standard.
Benzene fugitive emissions are not related to size of pipe. Instead,
data from an EPA study on fugitive hydrocarbon emissions in petroleum
refineries (II-A-30, pages 131-189) indicate that there is generally
little or no correlation between line size and leak rate. Furthermore,
the data indicate that the majority of valves in gas and light-liquid
service and flanges have line sizes less than 6 inches. The commenter's
suggestion that the standard apply to process lines 6 inches or greater
would probably exempt most of the equipment from the standard. An
exemption for equipment based on size, therefore, is not justified.
The commenter also suggests that accessibility of equipment
should have been considered in developing the standard. Accessibility
of valves is discussed in Section 2.4.2.1.
2-107
-------
Comment: One commenter suggested that provisions be added to the
definition of modification for exempting de mini mis levels and for
reconstruction that yields no net increase in emissions (IV-D-21).
The commenter referred to comments on the SOCMI VOC fugitives NSPS
(A-79-32-IV-D-17).
Response: The commenter's suggestion that de minimis levels of
benzene be exempted under the modifications provisions does not consider
what is actually being covered by the standard. The commenter appears
to think that each process unit is the basis for coverage under the
standard. Actually, the designated source for the standard is each
fugitive emission source (e.g., pump or valve) in benzene service.
Therefore, there would be no modifications because each source, whether
it is new or existing, would have the same emission rate. No provision
to exempt de minimis levels of benzene is necessary. The reconstruction
provisions of the regulation have been deleted since proposal. Thus,
the comment on reconstruction needs no specific response. However,
more importantly, the standard does differ for new and existing sources.
2.8.1.2 Ten Percent Benzene Cutoff. Several commenters questioned
the applicability of sources based on the 10 percent cutoff concentration
of benzene.
Comment: Two commenters stated the 10 percent cutoff would exclude
benzene emissions from gasoline stations (IV-D-29) and motor fuel
evaporation, containing about 4 percent benzene (IV-D-2). Another
commenter noted the existence in Buffalo, New York, of a refinery which
has no process streams containing greater than 10 percent benzene
(IV-D-10). One commenter recommended that EPA make provision for the
possibility of plant designers' diluting certain streams to avoid the
need for controls. The commenter added that where streams fluctuate
above and below 10 percent benzene, EPA should make clear that the
whole stream and its associated valves, seals, and pumps are subject
to the standard (IV-D-31).
Response: The selection of the 10 percent cutoff concentration
of benzene was based on estimates indicating that 90 percent or more of
the total benzene fugitive emissions originate from components that
process materials containing 10 or more percent by weight benzene. The
10 percent cutoff was proposed to be used to distinguish between process
2-108
-------
streams where benzene can be expected and those streams where benzene
might not be expected, where the quantity of benzene is relatively minor,
or where benzene is incidental. EPA proposed to exclude from coverage
those sources that handle streams below 10 percent by weight benzene
based on the relatively small amount of benzene contained by these
sources and the large number of sources with streams below the cutoff.
If a refinery or chemical plant, such as the one mentioned by the
commenter (IV-D-10), does not have any process streams containing greater
than 10 percent benzene, then fugitive emission sources in that plant
are not covered by the standard. As mentioned in the preamble to the •
proposed standard, the 10 percent cutoff could be revised if data
become available that indicate emissions are higher than are currently
estimated for streams below the cutoff. Commenters presented no data on
this decision and, therefore, no change was made to the 10 percent cutoff.
In surveying all possible sources of benzene fugitive emissions, a
cutoff was deemed necessary to avoid covering a large number or relatively
minor sources of benzene. A zero percent cutoff would cover sources
containing any amount of benzene,. A 1 percent cutoff would exclude both
crude oil and refinery streams containing trace amounts of benzene.
Benzene concentrations between 1 and 10 percent are found in gasoline
and gasoline blending components. These streams were exempted because
of the large number of sources within this range and the relatively
small amount of benzene emissions. The 10 percent cutoff also excludes
chemical streams where benzene is a contaminant*
One commenter (IV-D-31) suggested the possibility that plant
designers might dilute certain streams to avoid meeting the 10 percent
cutoff and, thus, avoid the need for controls. The likelihood of a
plant using this tactic to avoid coverage under the standard is minimal.
Diluting a process stream would not be a cost-effective practice for
industry, since in the end it would take more energy or benzene to produce
an intermediate or final product. Because the costs of this standard
are so small, dilution would likely cost the operator more than the
cost of compliance with the standard. Also, the commenter noted that
streams sometimes fluctuate. The standard covers stream that are
intended to be greater than 10 percent. If a stream can reasonably be
expected to fluctuate above 10 percent, it is already covered by the
2-109
-------
standard. The need for a special provision in the standard to avoid
these situations, therefore, is unwarranted.
2.8.2 Selection of Baseline
Comment: One commenter questioned EPA's baseline for calculating
incremental benefits of control alternatives. The commenter stated
that Regulatory Alternative II should be used as a baseline in non-
attainment areas instead of Regulatory Alternative I, since Alternative II
is based on CTG controls already being implemented in refineries through
SIP's. The commenter further recommended that emission reduction and
cost estimates be recalculated on the basis of CTG controls prescribed
by Regulatory Alternative II (IV-D-13; IV-D-27).
Response: At proposal, EPA estimated that uncontrolled benzene
fugitive emissions were about 8,300 megagrams per year for existing
units and about 2,500 megagrams per year for new units by 1985. This
estimate did not include the effects of the CTG for petroleum refineries.
Since proposal, these effects have been included in the estimated
impacts of the standard. Baseline emissions are defined on two levels:
on an industry-wide basis and on a process-unit basis. An industry-wide
baseline, which depicts expected emissions given current regulations,
is used in evaluating whether or not to regulate emissions from the
subject source category. The industry-wide baseline for fugitive
emissions of benzene represents the sum of uncontrolled emissions from
chemical and pharmaceutical plants, uncontrolled emissions from petroleum
refineries in National Ambient Air Quality Standard (NAAQS) attainment
areas for ozone, and SIP controlled emissions from petroleum refineries
in NAAQS nonattainment areas for ozone. Uncontrolled fugitive benzene
emission estimates for petroleum refineries in NAAQS nonattainment
areas are reduced, as requested by the commenter, by implementing controls
specified in the petroleum refinery VOC control techniques guideline
(CTG) document (II-A-10). Thus, industry-wide baseline emissions of
benzene, which have been re-estimated to be about 7,920 megagrams per
year for existing units and about 2,470 megagrams per year for new
units by 1985 correctly account for the contribution of fugitive emissions
of benzene from uncontrolled chemical plants and refineries and from
refineries operating using CTG controls (IV-B-11).
2-110
-------
Baseline emissions also are defined on a process unit basis in
order to compare the overall effectiveness of each alternative control
technique in reducing emissions from the least restrictive control
level used on any process unit. Most process units are not required
to use CTG controls because many process units are located in NAAQS
attainment areas for ozone. Thus, process unit baseline emissions
represent uncontrolled emissions (the least restrictive control level
used on any process unit) and correctly do not include emissions that
occur while using CTG controls. However, as requested by the commenter,
emission reductions using CTG controls was one of the control options
considered for the final standard. More importantly, because EPA has
used incremental costs and emission reductions in selecting the basis
of the final standard (which has changed since proposal), the baseline
does not affect selection of the level of the standard.
2.8.3 Format of the Standard
Comment: Several commenters criticized the basic format of the
standard, recommending that the regulation should be based on performance
criteria or an acceptable ambient exposure level rather than a specification
standard or work procedure standard. One commenter noted that a
performance-based standard would allow more flexibility and would be
compatible with existing OSHA rules for benzene. The commenter suggested
that EPA should be willing to accept compliance to a limit of detection
OSHA criterion (e.g., 1 ppmv) as compliance with the benzene fugitive
emissions standard, since OSHA is still considering possible revision
to the benzene workplace standard (IV-D-24). Two of the commenters
stated that industry should select its own control method as long as
it meets a performance limit. The commenters recommended that an
emission level (acceptable leak rate) be prescribed for all pumps and
compressors (new and existing) instead of equipment specifications
(IV-D-13; IV-D-24; IV-D-27). Another commenter stated that a work
procedure standard is easiest to enforce, but is rigid and requires
special paper work that would be excessively burdensome to industry.
According to the commenter, a performance standard would allow each
individual source to adopt the most cost-effective methods for benzene
control (IV-D-29). This commenter also recommended that EPA establish
an "acceptable ambient exposure level" for benzene and then allow each
2-111
-------
individual source to develop a program to limit exposures of individuals
living around the plant site (IV-D-29).
Response:. As discussed in the preamble to the proposed standard,
Section 112(e) of the Clean Air Act requires that an emission (performance)
standard be established for control of a hazardous air pollutant
unless it is not feasible to prescribe or enforce such a standard. A
performance standard is not feasible in the following circumstances:
(1) if the pollutant can not be emitted through a conveyance designed
and constructed to emit or capture the pollutant or (2) if the application
of measurement methodology is not practicable due to technological or
economic limitations. If either or both of these circumstances exist,
then the Clean Air Act gives EPA the authority to require a design,
equipment, work practice, or operational standard, or a combination.
EPA agrees with the commenters that a performance standard does
allow more flexibility in complying with the standard than the other
regulatory formats. However, EPA has found that a performance standard
is only feasible for leak.less equipment and pressure relief devices.
A control technique is available that eliminates fugitive emissions
from pressure relief devices, and Reference Method 21 can be used to
detect leaks of fugitive emissions from pressure relief devices. In
general, a performance standard is not feasibile for other fugitive
emission sources because the application of available measurement
methods is not practicable due to technological or economic limitations.
For example, valves, pumps, and sampling systems generally are not
designed to release fugitive emissions into a conveyance. Furthermore,
measurement of an emission level for the number of valves, pumps, and
compressors by the bagging method would be expensive, time-consuming,
and impractical. However, for valves, pumps, and compressors that can
achieve a no detectable emission limit, a performance standard is
provided because it is equivalent to BAT for these sources.
EPA is allowing two alternative standards for valves in a process
unit. As discussed in Section 2.4.2.11, these alternative standards
will limit the maximum percent of valves leaking within a process unit
to 2 percent. This will give plant owners and operators more flexibility
in complying with the standard for valves because it allows for the
possibility of different leak detection and repair programs and substitution
2-112
-------
of engineering controls. This flexibility is not available for other
fugitive emission sources, however, because there are not enough of
these other sources to base the alternative standards on statistical
procedures.
Section 112 of the Clean Air Act requires an emission standard,
not an ambient air quality standard. Furthermore, it is not possible
to set an "acceptable, ambient exposure level," as recommended by the
commenter. A work practice standard was selected for pumps and valves
with a leak definition of 10,000 ppm by volume. This leak definition
is a technology-based requirement and not a health-based "exposure
level."
Comment: One commenter recommended an approach that evaluates
specific areas with known problems and that includes control regulations
directed to those areas. For example, SIP's could be used to identify
control strategies where problem areas have been documented (IV-D-10).
Response: EPA has examined a variety of existing regulatory
programs (including relying solely on State programs) and has concluded
that despite existing controls, benzene emissions from fugitive emission
sources pose a significant risk to public health. Having reached this
decision, EPA lacks authority to promulgate any standards other than a
national standard for this source category. Section 112 clearly contem-
plates that NESHAP's be uniform; Section 112 speaks of promulgating a
"standard," which is meant to protect public health. Moreover, Section
112(e)(2)(B) speaks of a "particular class of sources," which suggests
that Congress intended standards to be applicable to such a class as a
whole. Finally, nothing in Section 112 or its legislative history
offers any indication that EPA could promulgate anything other than a
uniform national standard.
It should be noted that defining source categories according to
the population density around a source would ignore the constant
changes in population. Such an approach would be very difficult to
implement; it would entail determining a standard for each source.
Furthermore, this approach would not consider the risk to the most
exposed group of people.
2-113
-------
2.8.4 Emissions Data Base
Several commenters questioned the data used to estimate emission
rates of benzene from fugitive emission sources.
Comment: Two commenters stated that existing data do not accurately
reflect actual benzene emissions from fugitive emission sources. The
commenters cited the EPA screening study on 23 SOCMI units as the basis
for their concern. One of the commenters felt that until better data
are produced, EPA should not continue developing the proposed standard
(IV-F-1). The same commenter further stated (in another set of comments)
that the BID for the proposed standard overestimates benzene fugitive
emissions by 25 to 33 percent (IV-D-21). The other commenter stated
that EPA did not base its emissions source data upon fugitive emissions
from benzene units only. The commenter indicated that recent studies
suggest benzene emissions may ,be two to three times less than that
relied upon by EPA in the proposed benzene regulation (IV-K-1).
Response; The commenters question the basis for the emission
factors used by EPA in estimating benzene fugitive emissions. During
preparation of the BID for the proposed standard, EPA examined data
associated with equipment in benzene units and found the data insufficient
for direct estimation of benzene emissions, as discussed below. In
the BID for the proposed standard, refinery VOC fugitive emission
factors were used to estimate VOC emissions from benzene-service
equipment. These VOC emissions were then "corrected" to benzene
emissions by multiplying by the average weight percent benzene in the
process stream. This procedure is based on the assumption that the
ratio of benzene emissions to VOC emissions is the same as the ratio
of the amount of benzene to VOC in the process stream.
Since preparation of the BID for the proposed standard, new data
have become available, which EPA has considered. Some of the new data
are cumene data from the EPA screening studies to which the commenters
refer. Other data concern equipment in benzene service within coke
oven by-product plants. In addition to these new data, EPA has developed
and considered VOC emission factors for benzene-service equipment
(IV-B-22) based on an "adjusted" approach as explained in "Fugitive
Emission Sources of Organic Compounds—Additional Information on Emissions,
Emission Reductions, and Costs" (AID) (IV-A-24).
2-114
-------
Available leak frequency data are summarized in Table 2-7, and
VOC emission factors are shown in Figures 2-1 through 2-3. As with the
data available at proposal, the new data were found to be insufficient
for direct estimation of benzene emissions. After analyzing all available
data, EPA concluded that the refinery VOC emission factors and the
concentration of benzene within the process stream should be used as
the basis for emission estimates for fugitive emission sources of
benzene. Thus, EPA has not changed the approach used at proposal to
estimate benzene emissions from fugitive emission sources, as detailed
below. The estimates for benzene emissions from controlled sources
have been revised using the LDAR model, which is described in Appendix B.
In the BID for the proposed standard, EPA considered several data
sets to estimate benzene emissions. None of the data sets was limited
to benzene-service equipment, thus no benzene-specific emission estimates
could be developed. However, EPA concluded that VOC emission estimates
could be adjusted to estimate benzene emissions from fugitive emission
sources. Much of the leak frequency data shown in Table 2-7 was
available at proposal; however, the only VOC emission factors available
were for refinery equipment. The leak frequency data from the cumene
units, Unit E, and Unit F do not appear to be substantially different
from the refinery data, especially considering the confidence intervals
associated with the data.
Limited testing of the benzene content of process streams and
fugitive emissions from these process streams performed during the
testing at Units E and F supports the approach of adjusting VOC emission
estimates to obtain estimates of benzene emissions (IV-A-2). These
data indicate a similarity between the weight fraction benzene in the
process stream and the weight fraction benzene in a leak. Because the
refinery leak frequency data were not substantially different from the
other available leak frequency data that included benzene-service
equipment, and because the refinery emission factors were the only
comprehensive set of emission factors available that included benzene-
service equipment, the refinery emission factors were used to estimate
VOC emissions from benzene-service equipment. These estimates were
then used to quantify benzene emissions by correcting the VOC emission
estimates based on the percent benzene in the process stream.
2-115
-------
CO
CL.
Cu
a
CO
LU
_J
^*
f^f
o
U_
CO
LU
O
z:
LU
o- -
LU
o;
u_
2
fQ
1—
0.
5
CO
U4
|
UJ
or
LU
to
0
o*
1
J
UJ
o
>
s
LU
a
O1
3
i—
i
UJ
0
UJ
c/>
to
s
_o
in o
Ol
(O
4J Ol
C C
J- ID
-
fc 10
Q* flj
a _j
0) "S
E en
zi~
•o
to
V— 1
i— I M- VO 4-
ewi o *-« r-* i •— < LO .^H
cri * iii • i i i i
CO 01 00 0
• Ol *"^ • •
CM CMLnfi-li-lQ- S
Ol
o *o CM otncomPoo
^ "*• *~l CMCMPO roro
•
U5 00
01 • • 01 in
• • ^* o c\j • •
«*• i i i i -H . *r T* T
•-"•-J .vo com.cMi-iroin
epio 10 >-o«a-
^
o in 1-4 01 in CM o mi— ' o
?o 01 10 m o^ ^^ ^}> in r^*
SJ5 S SS^^SS
CO
f»4
es co
CM <4- r~!
•-< 0 rH
1 CM 1
T eo i 1-1 in «a-
«— < • CM O -CO^OOO
IO i— ( 1 i-4 1 1 I |
-------
I
o
(O
u_
c
o
•r*
CO
to
CU
o
E
cu
-a
o
o
CU r—
O tO
s- >
Q) S-
Q- CU
4->
to a
CO »-H
O O
(O O
t- E
= 3
O C-i
•r- CU
to S-
cu
cu ta (d o
cu cu re
•Or- i— M-
cu
(0 O S- O
r— CU'r-
3T3 E to
O CU -I- to
r— to <4- -r-
S- CU
CU
N
1 1 1 1
-a
cu
o
cu
"cu
• t/v
Ol n
1 <->
• •
Ol ' f
K_ . ' ,
v*/ ^\J 1 \*S
in
^
C3
*
1 1 1 1
••
^s
3 >j
CU £-
r— CU
O E
S- T-
4-> <4—
CU CU
a. a:
cu
cu
E
0
i~ O
CU 3
> -o
C O
s-
CU Q-
.-•' 1
0 >>
C J CQ
CD
CU
N
E
CU
CO
—
4->
S
E
C.
3
CT
LU
"—
5
1
Q.
3
cr
UJ
cu
o
S-
cu
(J~>
cu
E
cu
N
c:
cu
ca
o o m o Lo o
VJ CM i-H i— 1 O
3 O O CD C3
S- -r-
O £=
fO CU
CU E
tO -I—
M-
a. cu
E S-
Q. E
CU CL)
O i—
•r- O
> S-
i- 4-5
CU CU
I/) Q.
jo -a
=5 03
cr
• fM> •>
i— CU
E
4-> CU
-E £
C7) 3
•i- U
i- -U
O U
f- 3
-a
to o
s- s_
O Q.
+-> I
u >>
fO -Q
M-
E
E CU
O >
•i- O
to
(f> a>
I
CM
CU
3
Ol
aounos/ui|/6>|
NOISSIW3 OOA
2-118
-------
S-
3
o
E
O
CO
CO
O
O
UJ
O
O
O)
E
o
-p
E
O)
CD S-
Q_ O)
-M
to E
O—O
O ^J
+-> , (!)
ns O r-
4- C
E -i- E
10 O M- O
r- OJ •!-
3T3 1- to
O «*-
O) O)
0. CH
,
CJ> CQ
O)
OJ
N
0)
CQ
•M
E
3
cr
cr
UJ
o
o>
oo
E to
N ••-
E C
O) =1
O S-
ll
o o
•i- S-
> -p
S- 0)
o> a.
CO
i -o
-a c
cr «
•r- I
O >>
(Q J^
E O)
O >
•r- O
CO
CO CD
•r- ^4
UJ O
i
CVJ
o
OJ
o
o
I—I
o
CD-
aounos/UM/6>i
NOISSIW3 OOA
2-119
-------
£.
o
o
03
C
^o
to
in
•i—
CD
o
CD
-a
c
o
o
4J
OJ r-
O 03
S- >
OJ S-
0. O)
to c
en i—i
-a
s- E
o 03
o >>
03 O
M- C
OJ
O-CTi
•1- <1J
to s-
10 4-
O) 03
QJ
-a i—
OJ
•(-> C
03 O
'3 -a
O OJ
I— tO
03 03
o .a
CD
o to
E S-
^s
03 O
OJ 03
>> E
Co
aj T-
E to
•r- CO
4- •(-
CJ E
S. OJ
-a E
S- E O
O 03 »i-
4-> to
o >i in
03 O T-
CD O>
o cr^si
•i- OJ 03
CO S- CD
to <*- i—
•r-
E -s<£ O
OJ 03 E
CD I
"O r— -^i
OJ 03
•(-^ E OJ •
03 O r— tO
3 -a •—i o
O OJ S 4->
r— to o o
O3 03 O 03
o -a co <4-
I o — o * D
1 1 1 1 1
•a
OJ
o
OJ
p-»
CD
co
n 1 n
VJ | VJ
Ol /-\
" '
* a
,
i i i i i
^
3
O)
*0
4O
OJ
Q.
OJ
E
CD
E
0
—
CD
E
CD
ISI
E
CD
CO
—
O O O CD O O
ID «3- CO OJ i-H
CD CD O O O
• • • • •
0 0 0,,_. . _, ...... ,Q , 0
•
>^
s-
O)
E
*+-
OJ
_
-M
QJ
gz
{*)
•P«
3
cr
UJ
,»_>
I.K
E
S- OJ
O) E
CO Q.
OJ 3
E CT
O) UJ
N
E
CD
CQ
OJ
E
O)
N
E
OJ
J3
S-
o
4-
to
CD
r— tO
03 4->
> •!-
E
OJ 3
(J
> S-
S- OJ
OJ E
tO -r-
1 4-
to aj
03 i-
cn
S- 3
o
tO S-
&» 4^
O CD
4-> Q.
O
03 -a
t,imm r~
03
O "
•r- OJ
00 E
to O)
•r- E
E 3
UJ O
,
1
CM
a>
3
O)
aounos/aq/6>i
NOISSIW3 30A
2-120
-------
Since proposal, several additional data sets have been analyzed.
These include the leak frequency data shown in Table 2-7 for benzene-
service equipment, coke oven by-product plant equipment, and the BTX
and HDA unit equipment. In addition, new VOC emission factors have
become available for several sources. Figures 2-1 through 2-3 compare
recently available VOC emission factors for cumene units, coke oven
by-product plants, and benzene-service equipment to the refinery factors.
The benzene-service equipment leak frequency data are from the
"Analysis" report (IV-A-18). These data are for equipment in the SOCMI
24-unit study that had benzene as the primary or secondary chemical in
the process stream. Most of these data are from cumene units, while
the remainder is from ethylene units. The leak frequency for benzene-
service gas valves and single-seal light liquid pumps is higher than
the corresponding leak frequency for the refining data, and the light
liquid valve leak frequency is slightly lower. VOC emission factors
were developed for the benzene-service equipment using the technique
discussed in the AID as follows (IV-B-22): Emission factors for leaking
and nonleaking sources were developed based on the refinery and SOCMI
data. These emission factors and the benzene-service equipment leak
frequencies were then used to calculate overall VOC emission factors
for benzene-service equipment. These emission factors generally fall
within the confidence intervals of the other emission factors and
(1) are higher than the refinery VOC emission factors for light liquid
pumps, (2) are about the same for light liquid valves, and (3) are
higher or lower than the valve (gas service) emission factor depending
on whether the basis is the refinery or SOCMI estimation, respectively.
Testing was performed at three coke oven by-product plants to
develop estimates of fugitive emissions. The coke oven leak frequency
data shown in Table 2-7 are for equipment in benzene service. Very
few gas-service valves were tested, so these data were combined with
the light liquid-service equipment in the data presentation. Table 2-7
shows the light liquid valve leak frequency for coke oven by-product
plants is one-half that of the refinery data, while for light liquid
pumps the coke oven leak frequency is twice that of the refinery data.
VOC emission factors were developed specifically for the coke oven
by-product plant benzene-service equipment, and they are shown in
2-121
-------
Figures 2-1 through 2-3. Because of the limited amount of data collected
in these plants, the confidence intervals for the VOC emission factor
estimates are very large. These confidence intervals completely overlap
the refinery factor confidence intervals. In addition, the actual
estimates are within the confidence intervals of the refinery emission
factors -for both light liquid pumps and valves. Additional testing was
performed at coke oven by-product plants to compare the weight fraction
benzene in the process stream to the weight fraction benzene in a leak
from that process stream. These data show that the average difference
in the benzene fraction was -15 percent; that is, the weight fraction
benzene in the leak averaged only 15 percent greater than the weight
fraction benzene in the process line. (If the absolute values of
the differences between the process stream and leak fraction benzene
are used, the average difference is 19 percent). The largest percent
differences were found for the sources with the lowest leak rates.
These differences are well within the accuracy of the determination of
the percent benzene and the emission rate determinations, so they are
not significant (IV-B-23).
The BTX and HDA unit leak frequency data shown in Table 2-7 are a
subset of the refinery study data (II-A-30). For all sources the
confidence intervals for these data overlap, partially due to the
limited nature of the benzene testing in the refinery study.
Leak frequency data were available for the cumene units at proposal
of the standard. However, since proposal these data have been used
along with other data on emission rates from the SOCMI Maintenance
study (IV-A-6) to estimate VOC emission factors for cumene units as
shown in Figures 2-1 through 2-3. For all three sources, the confidence
intervals for the cumene emission factors overlap those of the refinery
data, and the estimated VOC emission factors are lower than those of
the refinery data.
In selecting the method to determine benzene emissions from
benzene-service equipment, all of the preceding data were considered.
As at proposal, the only method available to estimate emissions of
benzene from fugitive emission sources is the adjustment of VOC emission
estimates. Further, since VOC emissions would also be controlled by a
benzene standard with the resultant reduction in product loss and VOC
2-122
-------
emissions to atmosphere, it is desirable to use a technique that
quantifies both the benzene and VOC emissions and emission reductions.
Data that have become available since proposal in general are insufficient
in that they do not estimate VOC emission factors for all types of
benzene-service equipment and that limited data are used where factors
are estimated. In addition, these data were not gathered for the
purpose of generating emission factors. Data were gathered, however,
that confirm the assumption that the weight fraction benzene in the
process line is indicative of the percent benzene in an equipment leak.
In the refinery data base used at proposal, VOC emission factors were
determined for all leak sources. The refinery data base resulted from
an extensive study of fugitive emissions for the purpose of developing
emission factors. This study provided a representative estimate of
baseline fugitive emissions because the study occurred prior to the
recent increase in awareness of fugitive emissions. Finally, the VOC
emission factors developed in the refinery studies are not substantially
different from the benzene unit and benzene-service equipment emission
factors shown in Figures 2-1 through 2-3. Based on these considerations,
EPA selected the refinery data and the concentration of benzene within
the process stream as the basis for emission estimates for fugitive
emission sources of benzene.
Comment: One commenter observed that the data base contains
little benzene-specific information and added that the data that were
available were inconsistent. The commenter noted that EPA has not
completed any fugitive emission studies that deal specifically with
benzene or benzene mixtures. The commenters asserted that EPA's
discussions of control technology rely on transfer of technology from
one industry segment to another without use of all literature data.
In support of this position, the commenter cited two studies that
report benzene emissions from single mechanical seals that are from
1/4 to 1/10 of the values used by EPA to represent VOC fugitive losses.
These low values, according to the commenter, are consistent with the
chemical industry's ability to comply with the OSHA workplace limit of
10 ppmv. The commenter recommended that EPA resolve the reasons for
variations in fugitive loss data and suggested that several NIOSH
studies on quantifying fugitive emissions might provide the basis for
2-123
-------
resolving the differences (IV-D-24). According to another commenter,
the cumene data from the EPA screening study indicate that benzene
fugitive emissions are actually 0.25 to 0.33 percent of plant losses
as compared to 0.53 to 0.92 percent losses in the proposal BID .(IV-K-1).
Response: EPA tested a variety of units in benzene service
either as an individual source test or as part of larger screening and
maintenance studies for refining and chemical process units. Approximately
3,400 valves and pumps in benzene units were tested. Benzene test
data from petroleum refineries and chemical process units are described
in the BID for the proposed standard, Appendix C. As described in the
previous response, Table 2-7 presents new test data that have been
reported since the completion of the BID for the proposed standard.
In Table 2-7, 1,035 valves and pumps in benzene service are shown.
Thus, EPA has adequately studied benzene-specific information.
In responding to the second part of the comment, EPA did rely on
technology transfer between the refining and chemical industries in
setting the standard. Technology transfer applies to the selection of
control techniques, not the determination of emission estimates. There
can be great variation in emission rates for fugitive emissions because
of the nature of fugitive emissions. Most sources do not leak or have
very low leak rates, and a few sources have high leak rates. It is
very possible to calculate low emission rates for certain sources such
as pump seals based on limited sampling of selected sources. EPA
collected and evaluated data on fugitive emission sources and concluded
that using the petroleum refining data (emission factors) and the
concentration of benzene in the process stream is the best way to
estimate benzene fugitive emissions. Since proposal, new data have
confirmed this conclusion.
Addressing the final comment, estimation of benzene emissions
based on plant losses has not been performed by EPA. The source and
validity of data used to support these estimates is unknown.
2.8.5 Consistency with Other Regulations and Guidelines
Several commenters suggested that the regulation should be technically
consistent with other regulations and guidelines.
Comment: One commenter recommended specifically that the standard
be technically consistent with the new source performance standard
2-124
-------
(NSPS) and the control technique guideline for VOC fugitive emissions
in SOCMI facilities and compatible with OSHA standards (IV-D-24).
Another commenter stated that the SOCMI NSPS and the benzene fugitive
emissions standard are not equivalent in spite of what EPA maintains.
The commenter noted that the actual monitoring and control requirements
are identical, but the theoretical emission reduction is 87 percent
for the proposed SOCMI NSPS and 73 percent for new sources in the
proposed benzene standard, indicating that the former standard is
stricter (IV-D-21, IV-F-1).
Response: The benzene fugitive emissions standard, the SOCMI VOC
fugitive standard, and the SOCMI CTG's are technically consistent.
They are all based on the same or compatible equipment requirements
and the same work practice and monitoring requirements. For example, :
the leak detection and repair requirements for valves and pumps are
similar. Although the draft SOCMI CTG specifies quarterly inspections
and the NSPS and NESHAP specify monthly inspections of these sources,
the basic control technique is the same. All three use the same type
of instrument and require the same repair interval. In addition, all
three require caps on open-ended lines. The equipment and work practice
requirements are almost identical for the SOCMI fugitives NSPS and
benzene fugitives NESHAP.
The second commenter (IV-D-21) correctly pointed out that the
emission reduction estimated under the alternative selected as the
\
basis of the proposed SOCMI VOC NSPS is 87 percent for new sources.
However, the proposed benzene fugitive NESHAP was estimated to achieve
an 80 percent emission reduction for new sources in the fifth year of
the standard. The 73 percent emission reduction that the commenter
quotes is the estimate for existing sources in benzene service. There
is an obvious reason for the difference in the emission reduction
estimates for the two standards. The standards are based on different
model plants. The numbers of components (i.e., pumps, valves, etc.)
per model unit are different for the standards. The emissions and
emission reductions for each source type are different so that changing
the relative number of sources will change the overall emission reduction
even though the control techniques are identical.
2-125
-------
Comment: One commenter stated that new provisions of the recommended
refinery VOC fugitives NSPS should be incorporated into the proposed
•benzene fugitives NESHAP, since the two standards address similar
types of equipment and sources (II-D-13).
Response; As discussed in Sections 2.3 and 2.4, many of the new
provisions of the proposed refinery VOC fugitives standard have been
incorporated into the benzene fugitives standard. For example, the
equipment requirement of dual mechanical seals and barrier fluid
systems has been deleted from the proposed standard for new pumps.
Instead, the final standard requires a work practice standard (leak
detection and repair) for new as well as existing pumps as in the
proposed refinery VOC standard for new pumps. Another provision of
the proposed refinery VOC standard that has been incorporated into the
benzene fugitives standard is modifying the monitoring frequency for
difficult-to-monitor valves when the monitoring requires elevation of
personnel more than 2 meters above a permanent available support surface.
An annual leak detection and repair program is required for these
valves designated by the plant owner or operator to be difficult-to-monitor.
2.8.6 Incorporation of Other Comments or Documents
Comment: Many commenters requested that EPA incorporate as part
of the administrative record their comments on other proposed EPA
rulemakings and policies '(IV-D-13; IV-D-24). One commenter requested
that EPA include in the record the November 6, 1980, letter from R.R.
Russell, Council on Wage and Price Stability, to D. Costle, EPA (IV-D-8).
Another commenter .requested that EPA make part of the record the
June 1980 "Comparisons of Estimated and Actual Pollution Control
Capital Expenditures for Selected Industries," by Putnam, Hayes, and
Bartlett (IV-D-31).
Response: With regard to the request of the first group of
commenters, EPA has incorporated these referenced comments on the
proposed SOCMI VOC fugitives NSPS, SOCMI CTG, maleic anhydride,
ethylbenzene/styrene, and benzene storage NESHAP's into the record for
the benzene fugitives NESHAP. Table 2-8 lists citations for the
comments and the EPA responses on other rulemakings and policies for
which identical responses have been incorporated into the benzene
fugitives standard. As requested by the second commenter (IV-D-8),
2-126
-------
CO
CD
-> o
"JrS
3 SO
o o tn
i. t. «T
010. i
-^: <£
o s- Q-
10 ouj
ai
o
re
in i-
c o
Q 4J
— to
v»
(A 0)
fig
UJ N
C
aj a>
C CO
0)
si 3
c 5
I
cn<4- t/> O
^ co
U C "O I
18 O 0) f> '
CO -F- 4-> —,
4J 18 O
i re o>tr>
S F— «3-
M S» 3 I
•24 o e «*
C H- O O.
re c i- uj
h- i—i Q. ~-^
c -i- -a o i
o ••- s: c*-*
•r- 4J IO
•W 0) Ul 1- Wl
•t- JC i— >, *-> T>
en *J (O I. re L.
u. >,•!- in E -on
tSI g 3 O CO
cj Si -o t- re i
o c .c c c *-> o
> .^ o 1-1 •— co eo
Q-
LlJ
Of.
UJ
I
o
S-
u
c
aj
o
Ul
o
S
« «c
c s- re
a> o i/o
51 =
«£
Q.
2-12.7
-------
EPA has included the referenced letter into the docket as A-79-27-IV-H-1.
Finally, the report referenced by the third commenter (IV-D-31) has
been included in the record as Docket Number A-79-27-IV-A-13.
2.8.7 Use of Alternative Control Systems
Comment: Two commenters questioned the provisions dealing with
use of alternative control systems. One commenter felt that vendors
should be allowed to provide alternative means of emission limitation
(IV-D-21). The other commenter believed that the owner or operator
should not be required to collect extensive data to determine the
alternative means. According to the commenter, EPA should have suffi-
cient data based on past studies to determine the alternative means of
control methods (IV-D-24).
Response: The alternative means provisions of the proposed standard
as stated in §61.244 do not directly allow vendors to apply for permis-
sion to use an alternative means of emission reduction. Section 61.244(b)
lists guidelines for determination of alternatives to the equipment
requirements of the standard. A primary consideration in this process
is that "each owner or operator applying for permission shall be respon-
sible for collecting and verifying test data for an alternative means
of emission limitation " A vendor could provide the owner or operator
with the required test data; however, the ultimate responsibility for
the accuracy of the data is with the owner or operator instead of the
vendor. In order to increase efficiency in the process and to provide
plant owners and operators the incentive to purchase improved control
systems and equipment as they are developed, EPA decided to revise the
final standard to allow a vendor or manufacturer to apply for alterna- •
tives to control systems or equipment. However, a determination of the .
alternative means may depend on how a plant owner or operator,uses the
control systems and equipment. Therefore, the application for permis-
sion to use an alternative means must come from the plant owner or
operator unless a general determination of the alternative means with
specific criteria can be made, in which case a vendor may request such
a determination.
The logical way for a plant to document an alternative means
of emission limitation is for the owner or operator to collect
plant-specific data to show that the alternative achieves a reduction
2-128
-------
in emissions equivalent to requirements of the standard. Industry is
in the best position to collect such data. EPA does not have adequate
data to determine equivalence for every alternate control method, and,
therefore, must rely on industry for such data. Moreover, industry
has taken the position repeatedly that it should be allowed to select
its own means of emission limitation whenever possible. It follows
then that industry should collect its own test data in support of its
alternative control techniques just as EPA has done in selecting the
requirements of the standard.
2.8.8 Siting Criteria
Comment: One commenter criticized EPA for not considering the
siting criteria proposed by the cancer policy in minimizing construction
of new sources in populated areas or in subjecting new sources placed
in populated areas to higher control requirements. Instead, the com-
menter stated that EPA has applied other approaches on a case-by-case
basis in developing the standard. The commenter suggested that EPA
should incorporate the siting criteria into the standard (IV-D-31).
Response: As at proposal, EPA has not decided how to .implement
the siting criteria discussed by the commenter. The siting criteria
are not formulated and, accordingly, can not be implemented in this ,
standard. They may be implemented later as discussed at proposal.
2.8.9 Withdrawal of the Standard
Comment: One commenter felt that the proposed rules and standards
should be withdrawn in part because, individually, existing and new
components of the sources in benzene service (such as pumps, compressors,
and valves) have small or undetectable emission rates (IV-D-18).
Response: Emission rates examined on an individual source basis
(i.e., pumps, valves, etc.) may appear small. However, when the large
numbers of benzene-producing equipment in a refinery or chemical plant
are taken into consideration, the emission levels can be quite significant
on a process unit or plant-wide basis.
2-129
-------
-------
APPENDIX A
ANNUALIZED COSTS AND COST EFFECTIVENESS
FOR BENZENE FUGITIVE EMISSION SOURCES
A-l
-------
APPENDIX A
ANNUALIZED COSTS AND COST EFFECTIVENESS FOR BENZENE
FUGITIVE EMISSION SOURCES
Tables A-l through A-10 present annualized costs and cost effectiveness
for various control techniques by benzene fugitive emission sources.
Included are costs of implementing a leak detection and repair program
at different intervals versus costs of equipment specifications. Tables A-ll
and A-12 summarize for all equipment types the VOC and benzene emission
reduction, average, and incremental cost effectiveness associated with
each control technique. A discussion of the basis for the selection of
the control techniques for each fugitive emission source is presented in
Section 2.3 of this document.
A-2
-------
Table A-l. ANNUALIZED CONTROL COSTS FOR VALVES IN GAS/VAPOR SERVICE
(May 1979 Dollars)
CONTROL TECHNIQUE
COSTS
Installed Capital Cost
Annuallzed Capital
A. Control Equipment0
B. Initial Leak Repair6
Annual i zed Operating Costs
A. Maintenance
B. Miscellaneous
C. Labor
1. Monitoring"
2. Leak Repair^
3. Administrative
and Support J
Total Annual Cost Before
Credit
Recovery Credit''
Net Annualized Cost1
Total Emission reduction
(Mg/yr)
VOCro
Benzene Model An
Benzene Model Bn
Benzene Model C"
Cost Effectiveness ($Mg)°
VOCm
Benzene Model An
Benzene Model Bn
Benzene Model C"
Weighted Average New
($/Mg BenzeneJP Existing
Annual
Inspections
0
0
0.40
0
0
0.51
2.94
1.38
5.23
13
(7.77)
0.035
0.022
0.019
0.026
(220)
(350)
(410)
(290)
J2I5N
(350)
Quarterly
Inspections
0
- 0
0.40
0
0
2.04
3.27
2.12
7.83
52
(44.17)
0.14
0.088
0.077
0.105
(320)
(500)
(570)
(420)
(500)
(510)
Monthly
Inspections
0
0
0.40
0
0
6.09
3.34
3.77
13.60
. 59
(45.40)
0.16
0.1008
0.088
0.12
(280)
(450)
(520)
(380)
(450)
(460)
Sealed Bel lows
Valve
New Existing
. 2,500b
408<*
0
125f
. 1009
0
0
0
633
85
548
0.23
0.14
0.13
0.17
2,400
3,900
4,200
3,200
3,700
--
3,700b
600d
0
185f
1489
0
0
0
933
85
848
0.23
0.14
0.13
0.17
3,700
6,100
6,500
4,900
*«•
6,060
A-3
-------
Table A-l. ANNUALIZED CONTROL COSTS FOR VALVES
IN GAS/VAPOR SERVICE3 (continued)
(May 1979 Dollars)
All costs and emission reduction estimates are for one piece of equipment
in benzene service.
bTelecon. Hclnnis, J.R., PES, Inc., to V. Caton and C. Hetrick, Chempump
Division of Crane, Warrington, PA. August 23, 1979. Cost of sealed
bellows valve.
°Cost of monitoring instrument is not included in this analysis.
Based on 10-year equipment life and 10 percent interest (CRF - 0.163).
*Annualized charge for initial leak repairs for a valve inspection
program is obtained by: number of leaks x repair time x 15.50/hr x 1.4
(overhead) x 0.163 (CRF for initial repair, 10 years at 10 percent
Interest). Assume 10 percent of gas valves and 11 percent of light
liquid valves leak in initial survey. Initial leak repair is constant
for all monitoring intervals.
Gas:
Light liquid:
Labor Rate*
% leaking = t Leaks x Repair Time(hrs) x ($/hr) x
0.10 = 0.10 x 1.13 x 15.50 x
0.11 = 0.11 x 1.13 x 15.50 x
1.4 x 0.163
T~4 053
1.4 0.163
$/yr
0.40
0.44
•Includes wages plus 40 percent for labor-related administrative and
overhead costs.
0.05 x capital cost.
90.04 x capital cost.
Monitoring Labor Cost (from leak detection and repair (LDAR) model):
Annual:
Quarterly:
Quarterly/
Monthly:
Monthly:
I Gas Service
Valves
Fraction
Screened
4.24
11.79
Monitoring
# Valves
Screened
3^94
4.24
11.79
x
x
X
X
X
Monitoring
Time (hours)
2/l>5
2/60
- 2/60
2/60
X
X
X
X
X
Labor Rate
($/hr) -
15.50 =
15.50
15.50
15.50
Labor
Cost
w
2.04
2.19
6.09
i
Leak Repair Cost (from LDAR Model):
Annual:
Quarterly:
Quarterly/
Itonthly:
Monthly:
Fraction
1 Gas-Service of Sources
Valves
1
1
1
1
x Operated On
x
x
x
x
0.1685
0.1866
0. 1877
0.1909
= t Leaks
» 0.168
= 0.187
' 0.188
- 0.191
Repair
Time
x J
x
x
x
x
[Hours)
1.13
1.13
1.13
1.13
Labor Rate
x
x
x
x
x
($/hr)
15.50
15.50
15.50
15.50
Leak
Repair
Cost
•• ($/vr)
•• 2.94
•• 3.27
•• 3.29
•• 3.34
•^0.40 x (Monitoring cost + leak repair cost).
A-4
-------
Table A-l. ANNUALIZED CONTROL COSTS FOR VALVES
IN GAS/VAPOR SERVICE3 (concluded)
(May 1979 Dollars)
Recovery credit is based on total VOC emission reduction with a credit
f± H ?hene\ Tl?1s,cred1t is conservative since benzene is often a
feed, and other chemicals that leak are more valuable. Recovery credit
value is obtained by multiplying VOC emission reduction by $370/Md
benzene (2nd quarter, 1979 value). *»'u/ng
Total annual cost (before credit) - recovery credit. .......
Annual:
Quarterly:
Quarterly/
Monthly:
Monthly:
For
100
Emission
Factor Percent Conversion
(kq/day) X Reductinn * )-n Mn/w,.
0.64
0.64
0.64
0.64
X
X
X
X
0.149
0.597
0.611
0.703
X
X
X
X
T
0.365
0.365
0.365
0.365
Total VOC Emission
Reduction ( Mq/vr)
0.035
0.14
0.14
0.16
factor of 0.64
reduction and cost effectiveness estimates are presented
KI „» + U * asfi on the Percentage of benzene in each process:
63 percent benzene, B = 55 percent benzene, C « 75 percent benzene.
reductions"?
let annualized cost by total VOC and benzene
Effectiveness for New and Modified/Reconstructed
Net Annual Cost ($/yr)
Reduction x
(Mg/yr)
Model; Unit A 's
Net Annual Cost
42(0.63 + 7(0.
42 + 7
($/yr)
55) + 1]
t- 11
r New & M/R "
iel Unit B's
(0.751'
[ # New & M/R
Model Unit C
's
Net Annual Cost
vuc Emission
Reduction x 0.64
(Mg/yr)
Weighted Average Cost Effectiveness for Existing Model Units -
Net Annual Cost ($/yr)
(# of Existing
Model A's x
W...W.. *
(Ma/vr) '» nf frvic»-.
% Benzene\ /# of^ Existing
in A 1 T( Model B's
Model Unit A's + Model Unit B's +
Net Annual Cost ($/yr)
Reduction x
(Mg/yr)
= Net Annual Cost (S/vr^
VOC Emission
Reduction x 0.62
(Mg/yr)
oji •<• /DIU.SS + 18(0.75)
131 +75+18
% Benzene \ /f of Existing
in B )^ Model C's
# Existing ' —
Model Unit C's
y % Benzene \
x in C j
A-5
-------
Table A-2. ANNUALIZED CONTROL COSTS FOR VALVES IN LIQUID SERVICE0
(May 1979 Dollars)
CONTROL TECHNIQUE
COSTS .
Installed Capital Cost
Annuallzed Capital
A. Control Equipment0
B. Initial Leak Repair8
Annuallzed Operating Costs
A. Maintenance
B. Miscellaneous
C. Labor
1. Monitoring"
2. Leak Repair*
3. Administrative
and Support^
Total Annual Cost Before
Credit
Recovery Credit^
Net Annuallzed Cost1
Total Emission reduction
(Hg/yr)
VOCM
Benzene Model A"
Benzene Model Bn
Benzene Model Cn
Cost Effectiveness ($Mg)°
VOC"»
Benzene Model An
Benzene Model 8n
Benzene Model Cn
Weighted Average New
(S/Hq 8enzene)P Existing
Annual
Inspections
0
0
0.44
0
0
, 0.51
2.96
1.39
5.30
7.4.
(2.10)
0.020
0.013
0.011
0.015
(110)
(160)
(190)
(140)
(160)
(180)
Quarterly
Inspections
0
0
0.44
0
0
2.04
3.27
2.12
7.87
22
(14.13) '
0.06
0.038
0.033
0.045
(240)
(370)
(430)
(310)
(370) •
(380)
Monthly
Inspections
0
0
0.44
0
0
6.09
3.34
3.77 '
13.64
26
(12.36)
0.069
0.043
0.038
0.052
(180)
(290)
(330)
(240)
(280)
(290)
Sealed Bel
Valve
New
2,500t>
408d
0
125f
1009
0
0
0
633
33
600
0.09
0.050
0.050
0.068
6,700
10,000
12,000
8,800
10,300
--
lows
Existing
3,700b
60Qd
0
185*
1489
0
0
0 .
933
33
900
O.n<)
0.060
0.050
0.068
10,000
15,000
18,000
13,000
_•
16,000
A-6
-------
Table A-2. ANNUALIZED CONTROL COSTS FOR VALVES
. IN LIQUID SERVICE9 (concluded)
(May 1979 Dollars)
a -
9: See Table A-l, footnotes a through g.
Monitoring Labor Cost (from LDAR model):
Annual :
Quarterly:
Quarterly/
Monthly:
Monthly:
Light Liquid
# Valves
1
1
1
1
x
x
X
X
X
Fraction
Screened
3'.94
4.24
11,79
9 Valves
Screened
0.99
3.94
4.24
11.79
x
x
X
X
X
Monitoring
Time (hours)
2/60
2/60
2/60
2/60
x
X
X
X
X
Labor
Rate
($/hr)
15.50
15.50
15.50
15.50
Monitoring
Labor
Cost
• ($/yr)
0.51
2.04
2.19
6.09
\eak Repair Cost (from LDAR Model):
Annual:
Quarterly:
Quarterly/
Monthly:
Monthly:
#
Light Liquid
Valves x
1
1
1
1
x
X
X
X
Fraction
of Sources
Operated On
0.1686
0. 1867
0.1878
0.1909
= # Leaks
- 0.
- 0.
= 0.
= 0.
169
187
188
191
x
x
X
X
X
Repair
Time
(Hours)
1,13
1.13
1.13
1.13-
x
X
X
X
X
Labor
Rate
($/hr)
15.50
15.50
15.50
15.50
Leak
Repai r
Cost
($ yr)
- 2.96
=• 3.27
• 3.29
- 3.34
j - 1: See Table A-l, footnotes j through 1.
For Tight liquid valves, the LDAR model estimates percent reduction in
mass emissions. The emission factor for light liquid valves is derived
from refinery data (Docket No. A-79-27-IT-A-30, p. 266).
Total VOC Emission
Reduction (Hq/yr)
0.020
0.059
0.061
0.069
For sealed bellows valves, an emission factor of 0.26 kg/day and 100 percent
control efficiency are assumed.
"See Table A-l, footnote n.
°See Table A-l, footnote o.
pSee Table A-l, footnote p.
Inspection
Interval
Annual :
Quarterly:
Quarterly/
Monthly:
Monthly:
Emission
Factor
(kq/day)
0.26
0.26
0,26
0.26
Percent Conversion
x Reduction x to Mq/yr
x
x
x
x
0.212
0.627
0.640
0.725
x
x
x
X
0.365
0.365
0.365
0.365
A-7
-------
Table A-3. ANNUALIZED CONTROL COSTS FOR PUMPS - NEW UNITS'
(May 1979 Dollars)
COSTS
Installed Capital Cost
A. Seal
B. Barrier Fluid System
C. Degassing Vents
Annual 1 zed Capital
A. Control Equipment0
I. Dual Mechanical
Seals
t Seald
• Installation6
2. Barrlec Fluid
System r
3. Degassing Vents
4. Replacement Seal9
B. Initial Leak Repair"
C. Initial Seal Replacement1
Annuallzed Operating Costs
A. Maintenance^
1. Dual Mechanical
Seals
2. Barrier Fluid
System
3. Degassing Vents
B. Miscellaneous
1. Dual Mechanical
Seals
2. Barrier Fluid
System
3. Degassing Vents
Annual
Inspections
0
0
0
0
0
0
0
40
14
4.7
0
0
0
0
0
0
CONTROL
Quarterly
Inspections
0
0
0
0
0
0
0
46
14
4.7
0
0
0
0
0
0
TECHNIQUE
Monthly
Inspections
0
0
0 •
0
0
0
0
49
14
4.7
0
0
0
0
0
0
Dual Mechanical Seals with
Barrier Fluid System and Valve
Degassing Vents
590
1,530
4,090
199
40
249
667
0
0
0
30
77
205
24
61
164
A-8
-------
Table A-3. ANNUALIZEDaCONTROL COSTS FOR PUMPS -
NEW UNITS3 (continued)
(May 1979 Dollars)
CONTROL TECHNIQUE
COSTS
C. Laoor
1. Monitoring
2. Leak Repair"'
3. Administrative
and Support
Total Annual Cost Before
Credit
Recovery Credit0
Net Annual ized Costc'p
Total Emission Reduction
(Mg/yr)
vocq
Benzene Model Ar
Benzene Model Br
Benzene Model Cr
Cost Effectiveness ($/Hg)s
VOC
Benzene Model Ar
Benzene Model 8r
Benzene Model Cr
Weighted Average
(S/Mg Benzene)1
Annual
Inspections.
10
84
38
191
73
113 .
0.21
0.13
0.12
0.16
540
370
940
710
870
Quarterly
Inspections
17
97
46
225
260
' (35)
0.70
0.44
0.39
0.53
(50)
(80)
(90)
(66)
(78)
Monthly
Inspections
38
102 .
56
264
300
(36)
0..82
0.52
0.45
0.62
(44)
(69)
(80)
(58)
(69)
Dual Mechanical Seals with
Barrier Fluid System. and Valve
Degassing Vents
0
0
' 0
1,716
366
1,350
. 0.99
0.62
0.54
0.74
1,400
2.200
2,500
1,800
2,100
A-9
-------
Table A-3. ANNUALIZED CONTROL COSTS FOR PUMPS -
NEW UNITS3 (continued)
(May 1979 Dollars)
aSee Table A-l, footnote a.
bLetter from J.A. Pearson, The Goodyear Tire and Rubber Company, to J.R.
Farmer, U.S. EPA. November 16, 1979. (Docket No. A-79-27-II-D-77).
Pump costs from BID I, p. 8-5. New seal cost = $573-230 (single seal
credit) + $246 (installation) = $590.
cSee Table A-l, footnote c.
dAnnualized cost of dual seal = 573 x 0.58 CRF = 332
Single seal credit = (230) x 0.58 CRF = (133).
eS1xteen hours installation at $15.50 per hour = $248 x 0.163 CRF = $40.
fSee Table A-l, footnote d.
9Replacement seal costs 1/2 that of a now seal because the old seal has
salvage value.
Annual cost » 1/2(225) = 113 (last quarter 1978). Cost index =
1.05 x 113 = $119/seal replaced
(Reference: Fugitive Emission Sources of Organic Compounds—Additional
Information on Emissions, Emission Reductions, and Costs, U.S. EPA,
OAQPS, EPA-450/3-82-010. April 1982. p. 5-19.
For annual, quarterly, and monthly monitoring, number of leaks = 0.34,
0.39, and 0.41, respectively (see footnote m below).
''Annualized charge for initial leak repairs for inspection program is
obtained by: number of leaks per component x repair time x $15.50/hr. x
1.4 (overhead) x 0.163 (CRF for initial repair over 10 years at 10
percent interest). Assume 24 percent of pumps leak in initial survey.
Initial leak repair is constant for all monitoring intervals.
1 pump x 0.24 leaks/pump x 16 repair hrj/leak x $15.50/hr x 1.4 x 0.163 CRF
» $13.58.
^Initial seal replacement cost = percent of pumps initially leaking x
replacement seal cost x capital recovery factor (0.24 x $119 x
0.163 =• $4.70).
^ See Table A-l, footnotes f and g, respectively.
Monitoring labor cost (from LDAR Model): :
Annual:
Quarterly:
Itonthly:
Fraction # Pumps Monitoring*
x Screened = Screened x Time (hours) x
x 1 = 1 x IU/55 x
x 4 = 4 x 10/60 x
x 12 - - 1' x 10/60 x 15.50
'Assumes 2-man monitoring team per pump.
Weekly visual inspection cost » $7/yr for all inspection intervals (assumes
0.5 minutes/source, 52 times/yr, $15.50/hr).
Monitoring
Labor Cost
Labor Rate per Source
($/hr) = •' '
15.50 =
15.50 = $10.
• - $31.
roLeak repair cost (from LDAR model):
Annual:
Quarterly:
Monthly:
-------
Table A-3. ANNUALIZED CONTROL COSTS
PUMPS - NEW UNITS*
(May 1979 Dollars)
FOR (Concluded)
See Table A-l, footnote j.
°See Table A-l, footnote k.
pSee Table A-l, footnote 1.
^The LDAR model estimates control efficiency (percent reduction in mass
emissions) for pumps. The emission factor for pumps is derived from
refinery data (Docket No. A-79-27-II-A-30, p. 266).
Total VOC
Emission Reduction
(Mg/.yr)
0.21
0.70
0.82
*100 percent control efficiency is assumed for dual mechanical pump
seal systems.
rSee Table A-l, footnote n.
sSee Table A-l, footnote o.
Annual :
Quarterly:
Monthly:
Emission
Factor
(kg/day)
2.7
2.7
2.7
x
x
X
X
Percent
Reduction
0.218 .
0.709
0.833
Conversion
x to Mg/yr
x
x
x
0.365
0.365
0.365
Table A-l, footnote p.
( ) denotes savings.
A-ll
-------
Table A-4. ANNUALIZED CONTROL'COSTS FOR PUMPS - EXISTING UNITS'
(May 1979 Dollars)
Annual
COSTS Inspections
Installed Capital Cost
A. Seal
8. Barrier Fluid System
C. Degassing Vents
Annual) zed Capital
'A. Control Equipment*"
1. Dual Mechanical
Seals
. Seald
« Installation3
2. Barriee Fluid
System
3. Degassing Vents
4. Replacement Seal9
8. Initial Leak Repair11
C. Initial Seal Replacement
Annual ized Operating Costs
A. Maintenance^
1. Dual Mechanical
Seals
2. Barrier Fluid
System
3. Degassing Vents
B. Miscellaneous
1. Dual Mechanical
Seals
2. Barrier Fluid
System
3. Degassing Vents
0
0
0
0
0
0 ,
0
40
14
4.7
0
0
0
' 0
0
0
CONTROL
Quarterly
Inspections
0
0
0
0
0
0
0
46
14
4.7
0
0
0
0
0
0
TECHNIQUE
Monthly
Inspections
0
0
• o
, 0
0
0
0
49
14
4.7
0
0
0
0
,0
0
Dual Mechanical Seals with
Barrier Fluid System and
Degassing Vents
870
• , 1,530
4,090
332
48
249
667
0
o-
0
44
37
' 205
35 ..
29
164
A-12
-------
Table A-4. ANNUALIZED CONTROL COSTS FOR PUMPS -
EXISTING UNITS3 (Continued)
(May 1979 Dollars)
CONTROL TECHNIQUE
COSTS
C. Labor
1. Monitoring
2. Leak Repair1"
' 3. Administrative
and Support
Total Annual Cost Before
Credit
Recovery Credit0
Net Annual ized Costc>p
Total Emission Reduction
(Mg/yr)
vocq
Benzene Model Ar
Benzene Model Br
Benzene Model Cr
Cost Effectiveness ($/Mg)s
VOC
Benzene Model Ar
Benzene Model Br
Benzene Model Cr
Weighted Average
($/Mg Benzene)*
Annual
Inspections
10
84
38
191
78
113
0.21
0.13
0.12
0.16
540
870
940
710
870
Quarterly
Inspections
17
97
46
225
260
(35)
0.70
0.44
0.39
0.53
(50)
(80)
(90) .
(66)
(81)
Monthly
Inspections
38
102
56
264
300
(36)
. 0.82
0.52
0.45 ,
0.62
(44)
(69)
(80)
(58)
(71)
Dual Mechanical Seals with
Barrier Fluid System and
Degassing Vents
0
0
0
1,810
366
1,444
0.99
0.62
0.54
0.74
1,500
2,300
2,700
1,900
2,400
A-13
-------
Table A-4. ANNUALIZED CONTROL COSTS FOR (Concluded)
PUMPS - EXISTING UNITS3
(May 1979 Dollars)
aAll costs and emission reduction estimates are for one light liquid
pump in benzene service.
Letter from J.A. Pearson, The Goodyear Tire and Rubber Company, to J.R.
Farmer, U.S. EPA. November 16, 1979. (Docket No. A-79-27-II-D-77)
Pump costs from BID I, p. 8-5. Capital cost of retrofitting an existing
pump seal = $573 + $297 (installation) = $870.
cSee Table A-l, footnote c.
dAnnualized cost of dual seal = $573 x 0.58 CRF = $332.
Nineteen hours installation at $15.50 per hour = $295 x .163 CRF = $48.
fSee Table A-l, footnoted.
9See Table A-3, footnote g.
hSee Table A-3, footnote h.
1Initial seal repair cost: percent of pumps initially leaking x replacement
seal cost x capital recovery factor (0.24 x $119 x 0.163 = $4.70).
"'"See Table A-l, footnotes f and g, respectively.
See Table A-3, footnote 1.
'See Table A-3, footnote m.
See Table A-l, footnote j.
°See Table A-l, footnote k.
P$ee Table A-l, footnote 1.
qSee Table A-3, footnote q.
rSee Table A-l, footnote n.
sSee Table A-l, footnote Q.
See Table A-l, footnote p.
( ) denotes savings.
A-14 '
-------
Table A-5. ANNUALIZED CONTROL COSTS FOR COMPRESSORS - NEW AND EXISTING UNITS3
(May 1979 Dollars)
CONTROL TECHNIQUE
COSTS
Installed Capital Cost
Annuali zed Capital
A. Control Equipment0
B.- Initial Leak Repaird
Annual ized Operating Costs
A. Maintenance6
B. Miscellaneous1"
C. Labor9
1. Monitoring
2. Leak Repair
3. Administrative
and Support
Total Annual Cost Before
Credit
Recovery Credit
Net Annual ized Cost1
Total Emission Reduction
(Mg/yr)
VOCJ
Benzene Model A*
Benzene Model B
Benzene Model C*
Cost Effectiveness ($/Mg)1
vock
Benzene Model A*
Benzene Model Bk
Benzene Model C
Weighted Average
($/Mg Benzene)*
Degassing
Reservoir Vents
7.3006
1,190
0
365
292
0
0
0
1,847
2,040
(193)
5.5
3.5
3.0
4.1
(35)
(55)
(64)
(47)
New (55)
Existing (57)
A-15
-------
Table A-5. ANNUAL IZEO CONTROL COSTS FOR COMPRESSORS - NEW
AND EXISTING UNITS3 (Concluded)
(May 1979 Dollars)
aSee Table A-l, footnote a.
Cost per seal. From: Organic Chemical Manufacturing Volume 3: Storage,
Fugitive, and Secondary Sources," EPA-450/3-80-025, December 1980.
Cost updates from Chemical Engineering. Economic Indicators, 86(16):7.
July 30, 1979. Costs have same basis as pump seals with a 'single
compressor seal connected to a vent. The compressor seal area could be
vented directly to a control device at similar cost.
cSee Table A-l, footnote d.
No initial leak repair charges for equipment specification.
See Table A-l, footnotes, f and g, respectively.
9No monitoring or leak repair labor costs are incurred for equipment
specification.
hSee Table A-l, footnote K.
Table A-l, footnote 1.
emission reduction is based on an uncontrolled emission factor of
15 kg/day and assumes 100 percent control efficiency for compressor
degassing reservoir vents.
kSee Table A-l, footnote n.
See Table A-l, footnote o.
mSee footnote p. Table A-l.
( ) denotes savings.
A-16
-------
Table A-6. ANNUALIZED CONTROL COSTS FOR PRESSURE RELIEF DEVICES -
NEW UNITS3
(May 1979 Dollars)
CONTROL TECHNIQUE
COSTS
Installed Capital Cost
Annual ized Capital
A. Control Equipment
B. Initial Leak Repair
Annual ized Operating Costs
A. Maintenance
B. Miscellaneous1
C. Labor
1. Monitoring^
2. Leak Repair
3. Administrative
and Support
Total Annual Cost Before
Credit
Recovery Credit
Net Annual ized Costf>m
Total Emission Reduction
(Mg/yr)
vocn
Benzene Model A°
Benzene Model B°
Benzene Model C°
Cost Effectiveness ($/Mg)p
VOC
Benzene Model A°
Benzene Model B°
Benzene Model C°
Weighted Average
($/Mg Benzener
Quarterly
Inspections
0
__f
0
__f
__f
17
0
7
24
340
(316)
0.91
0.57
0.5005
0.68
(350)
(550)
(630)
(460)
(540)
Monthly
Inspections
0
__f
0 .
__f
__f
50
0
20
70
370
(300)
0.99
0.62
0.54
O.fi
(303J
(480)
(560)
(410)
(480)
Rupture Disk System
Block Valve
l,810b
3809
0
90
70
0
0
0
540
520
20
1.4
0.88
0.77
1.1
14
23
26
18
22
3-way Valve
3,740°
695g
0
190
150
0
0
. 0
1,035
520
515
1.4
0.88
0.77
1.1
370
590
670
470
590
0-Ring
220d
36r
0
11
9
0
0
0
56
520
(464)
1.4
0.88
0.77
1.1
(330)
(530)
(600)
(420)
(520)
Closed Vent System
to PI are
2,120e
350r
0
110
85
0
0
0
545
520
25
1.4
0.88
0.77
1.1
18
28
32
23
28
A-17
-------
Table A-6. ANNUALIZED CONTROL COSTS FOR PRESSURE RELIEF DEVICES (concluded)
NEW UNITS3
(May 1979 Dollars)
See Table A-l, footnote a.
Refinery CT6. June 1978, p. 4-6 and cost update from Economic Indicators.
Chemical Engineering. 8i5(16):7. July 30, 1979.
cMemo from Cole, D.G., PES, Inc., to K.C. Hustvedt, U.S. EPA. Estimated
Costs for Rupture Disk System with a 3-way Valve. July 29, 1981
(Docket No. A-80-44-II-B-35). Costs were adjusted to reflect May 1979
dollars.
Cost of 0-rings based on relief valve control costs in the EPA report,
Additional Information Document (AID), Report Number EPA/3-80-033b,
page 5-23. Costs were updated to reflect May 1979 dollars using cost
indices from Chemical Engineering, Economic Indicators, April 9, 1979,
and July 30, 1979.
eCost of a closed vent system to transport the discharge or leakage
of safety/relief valves to a flare based on costs in the EPA report,
Additional Information Document (AID), Report Number EPA/3-80-033b,
page 5-24. Costs were updated to reflect May 1979 dollars using cost
indices from Chemical Engineering, Economic Indicators, April 9, 1979,
and July 30, 1979.
See Table A-l, footnote c.
^Obtained by multiplying capital recovery factor (2 years, 10 percent
interest = 0.58) by capital cost for rupture disk and capital recovery
factor (10 years, 10 percent interest = 0.163) by capital cost for all
other equipment (rupture disk holder, piping, valves, pressure relief
valve).
See Table A-l, footnotes f and g, respectively.
^Monitoring labor hours (i.e., # workers x # components x time to monitor
x times monitored per year) x $15.50 per hour. Assumes 2-man monitoring
team per relief valve, 8 minutes monitoring time per valve, monitored
quarterly or monthly.
k
See Table A-l, footnote j.
m
Table A-l, footnote k.
See Table A-l, footnote 1.
nVOC emission reduction is based on an uncontrolled emission factor of
3.9 kg/day and assumes 64 percent efficiency for quarterly inspections,
68 percent efficiency for monthly inspections, and 100 percent efficiency
for equipment controls.
°See Table A-l, footnote n.
pSee Table A-l, footnote o.
'•'See Table A-l, footnote p.
rSee Table A-l, footnote d.
A- 18
-------
Table A-7. ANNUALIZED CONTROL COSTS FOR PRESSURE RELIEF DEVICES -
EXISTING UNITS3
(May 1979 Dollars)
CONTROL TECHNIQUE
COSTS
Installed Capital Cost
Annual ized Capital
A. Control Equipment
B. Initial Leak Repair
Annual 1zed Operating Costs
A. Maintenance
B. Miscellaneous1
C. Labor
1. MonitoringJ
2. Leak Repair
3. Administrative.
and Support
Total Annual Cost Before
Credit
Recovery Credit
Net Annual ized Costf'n
Total Emission Reduction
(Mg/yr)
voc"
Benzene Model A°
Benzene Model' B°
Benzene Model C°
Cost Effectiveness ($/Mg)p
VOC
Benzene Model A°
Benzene Model B°
Benzene Model C°
Weighted Average
(S/Mg Benzene r
Quarterly
Inspections
0
__f
0
__f
__f
17
0
7
24
340
(316)
0..91
0.57
0.5005
0.6ff
(350)
. (550)
(630)
(460)
(560)
Monthly
Inspections
0
__f
0
—f
__f
50
0
20
70
370
(300.)
0.99
0.62
0.54
0.74
(303)
(480)
(560)
(410)
(490)
Rupture- Disk System
Block Valve 3-way Valve
3,310b
6209
0
170
130
0
0
0'
920
520
400
1.4
0.88
0.77
1.1
290
450
520
360
460
4,380C
8009
0
220
180
0
0
0
1,200
520
680
1.4
0.88
0.77
1.1
490
770
880
620
780
Closed Vent System
0-Ring to Flare
220d
36r
0
11
9
0
0
0
56
520
(464)
1.4
0.88
0.77
1.1
(330)
(530)
(600)
(420)
(520)
2,120e
' 350r
0
110
85
0
0
0
545
520
25
1.4
0.88
0.77
1.1
18
28
32
23
28
A-19
-------
Table A-7. ANNUALIZED CONTROL COSTS FOR PRESSURE RELIEF DEVICES (concluded)
EXISTING UNITS3
(May 1979 Dollars) ,
aSee Table A-6, footnote a.
bSee Table A-6, footnote b.
cSee Table A-6, footnote c.
dSee Table A-6, footnote d.
eSee Table A-6, footnote e.
See Table A-l, footnote c.
^See Table A-6, footnote g.
See Table A-l, footnotes f and g, respectively.
JSee Table A-6, footnote j.
kSee Table A-6, footnote k.
See Table A-l, footnote k.
mSee Table A-l, footnote 1.
nSee Table A-6, footnote n.
°See Table A-l, footnote n.
^See Table A-l, footnote o.
^See Table A-l, footnote p.
rSee Table A-l, footnote d.
A-20
-------
Table A-8. ANNUALIZED CONTROL COSTS FOR OPEN-ENDED LINES
NEW AND EXISTING UNITS3
(May 1979 Dollars)
CONTROL TECHNIQUE
COSTS
Installed Capital Cost
Annual i zed Capital
A. Control Equi pmentc
B. Initial Leak Repair
Annual i zed Operating Costs
A. Maintenance
B. Miscellaneous6
C. Labor
1. Monitoring
2. Leak Repair
3. Administrative
and Support
Total Annual Cost Before
Credit
Recovery Credit
Net Annuali zed Cost^
Total Emission Reduction
(Mg/yr)
vocn
Benzene Model A1
Benzene Model B1
Benzene Model C1
Cost Effectiveness ($/Mg)J
VCC
Benzene Model A
Benzene Model B1
Benzene Model C1
Weighted Average
($/Mg Benzene )K
Caps on Open Ends
50
8
0
3
2
0
0
0
13
7.4
5.6
0.02008
0.013
0.011
0.015
280
430
510
370
New 430
Existing 470
A-21
-------
Table A-8. ANNUALIZED CONTROL COSTS FOR OPEN-ENDED LINES - (concluded)
NEW AND EXISTING UNITS3
(May 1979 Dollars)
aSee Table A-l, footnote a.
Organic Chemical Manufacturing Volume 3: Storage, Fugitive, and Secondary
Sources, EPA-450/3-80-025, December 1980. (Cost updates from Chemical
Engineering. Economic Indicators. April 23, 1979, and July 30, 1979).
°See Table A-l, footnote d.
d - e
See Table A-l, footnotes f and g, respectively.
fSee Table A-l, footnote k.
9See Table A-l, footnote 1.
VOC emission reduction is based on an uncontrolled emission factor of
0.055 kg/day and assumes 100 percent control efficiency.
^ee Table A-l, footnote n.
JSee Table A-l, footnote o.
k
See Table A-l, footnote p.
A-22
-------
Table A-9. ANNUALIZED CONTROL COSTS FOR SAMPLING CONNECTIONS
NEW AND EXISTING UNITS3
(May 197.9 Dollars)
CONTROL TECHNIQUE
COSTS
Installed Capital Cost
'Annual ized Capital
A. Control Equi pment0
B. Initial Leak Repair
Annual ized Operating Costs
A. Maintenance
B. Miscellaneous
,C. Labor
1. Monitoring
2. Leak Repair
3. Administrative
and Support
Total Annual Cost Before
Credit
Recovery Credit
Net Annual ized Cost9
Total Emission Reduction
(Mg/yr)
voch
Benzene Model A1
Benzene Model B1
Benzene Model C
Cost Effectiveness ($/Mg)J
VOC
Benzene Model A1
Benzene Model B1
Benzene Model C1
Weighted Average
($/Mg Benzene)
Closed-Purge Sampling Systems
480
78 -'.''..-
0
24
19
' 0
0
0
121
48
73
0.13
0.082
0.072
0.098
560
890
1,010
740
New 880
Existing 900
A-23
-------
Table A-9. ANNUALIZED CONTROL COSTS FOR SAMPLING CONNECTIONS - (concluded)
NEW AND EXISTING UNITS3
(May 1979 Dollars)
dSee Table A-l,' footnote a.
bOrganic Chemical Manufacturing Volume 3: Storage, Fugitive, and Secondary
Sources, EPA-450/3-80-025, December 1980. Costs were updated to reflect
May 1979 dollars using Economic Indicators, Chemical Engineering.
86(9):7, April 23, 1979, and 86(16):7, July 30, 1979).
cSee Table A-l, footnote d.
d " eSee Table A-l, footnotes f and g, respectively.
fSee Table A-l, footnote k.
9See Table A-l, footnote 1.
hVOC emission reduction is based on an uncontrolled emission factor of
0.36 kg/day and assumes 100 percent control efficiency for closed-sampling
systems.
nSee Table A-l, footnote n.
JSee Table A-l, footnote o.
kSee Table A-l, footnote p.
A-24
-------
Table A-10. ANNUAL IZED CONTROL COSTS
FOR PRODUCT ACCUMULATOR VESSELS9
(May 1979 Dollars)
CONTROL TECHNIQUE • ; ' "
COSTS
Installed Capital Costb
Annual ized Capital
A. Control Equipment0
B. Initial Leak Repair
Annual ized Operating Costs
A. Maintenance
B. Miscellaneous6
C. Labor
1. Monitoring
2. Leak Repair
3. Administrative
and Support
Total Annual Cost Before
Credit
Recovery Credit
Net Annual ized Cost9
Total Emission Reduction
(Mg/yr)
voch
Benzene Model A1
Benzene Model B1
Benzene Model C1
Cost Effectiveness ($/Mg)
VOCJ
Benzene Model A1
Benzene Model B1
Benzene Model C1
fc
Weighted Average
($/Mg Benzene)
Closed Vent System
2,700
400
0
140
110
0
0 .
0,
650
0
650
10.2
6.8
5.9
8.1
60
96
110
80
New 94
Existing 97
A-25
-------
Table A-10. ANNUALIZE!) CONTROL COSTS FOR PRODUCT ACCUMULATOR VESSELS
(concluded)
(May 1979 dollars)
aSee Table A-l, footnote a.
^Capital cost is based on 61 meters of 5.1 cm carbon steel pipe, $2,700
installed per vent.
cSea Table A-l, footnote d.
dSee Table A-l, footnote f.
eSee Table A-l, footnote g.
Table *\-l, footnote k. No recovered product credit is expected for
accumulator vessels.
9See Table A-l, footnote 1.
emission reduction is based on an uncontrolled emission factor of
29.5 kg/day and assumes a control efficiency of 100 percent for a
closed vent system.
''See Table A-l, footnote n.
J*See Table A-l, footnote o.
^See Table A-l, footnote p.
A-26
-------
Table A-ll. CONTROL COSTS PER MEGAGRAM OF TOTAL EMISSIONS REDUCED6
Fugitive
Emission
Source
Valves
in gas/
vapor
service
Valves
in light
liquid
service
Pumps
Compressors
Pressure
Relief
Devi ces
Open-ended
Lines
Sampling
Connection
Systems
Product
Accumulator
Vessels
Control Technique
Annual leak detection and
repair
Quarterly leak detection
and repair
Monthly leak detection
and repair
Sealed bellows valves
' Annual leak detection and
repair
Quarterly leak detection
and repair
Monthly leak detection
and repair
Sealed bellows valves
Annual leak detection and
repair
Quarterly leak detection
and repair
Monthly leak detection
and repair
Dual mechanical seal
systems
Degassing reservoir
vents
Quarterly leak detection
and repair
Monthly leak detection
and repair
Rupture disk with
block valves
Rupture disk with
3-way valve
0-Ring
Closed vent system
to flare
Caps on open ends
Closed purge sampling
Closed vent system
Total Emission ,
Reduction Per Source
(Mq/yr)
0.035
0.14
0.16
0.'23
0.020
0.06
0.069
0.09
0.21
0.70
•
0.82
0.99
0.55
0.91
0.99
1.4
1.4
1.4
1.4
0.02008
• 0.13
10.8
. Average
$/MqC
New Existing
(220)e
(320)e
(280)e
2,400
(110}e
(240)e
(180)e
6i700
540
(50)e
(44)e
1,400
(35)e
(350)e
(303)e
14
370
I330)e
18
280
560
60e
(220)e
(320)e
(280)e
3,700
(110)e
(240)e
(180)e
10,000
540
(50)e
(44)e
1,500
(35)e
(350)e
(303)e
290
490
(330)e
18
280
560
60e
Incremental
S/Mg*
New
(220)e
(350)e
(62-)e
8,500
(110)e
(300)e
200
29,000
540
(302)e
(8)e
8,200
(35)e
(350)e
200
780f
2,000f
(400)e'f
790f
280
560
60e
Existing
(220)e
(350)e
(62)e
13,000
( 1101s
(300)e
200
43,000
540
(302)e
(8)e
8,700
(35)e
(350)e
200
l,700f
2,400f
(400)e'f
790f
280
560
60e
A-27
-------
Table A-ll. CONTROL COSTS PER MEGA6RAI1 OF TOTAL EMISSIONS REDUCED
(Concluded)
aCosts and emission reductions are based on one fugitive emission source.
Total emissions include benzene and other VOC.
Total emission reduction (Mg/yr) from Tables A-l through A-10.
°Average dollars per megagram (cost effectiveness) = net annualized cost
per component * annual emission reduction per component.
dIncremental dollars per megagram = ( net annualized cost of the control
technique - net annualized cost of the next less restrictive control
technique) * (annual emission reduction control technique - annual .
emission reduction of the next less restrictive control technique).
eValues in parentheses represent savings.
^Incremental costs are based on comparing equipment costs with monthly
leak detection and repair costs.
^Monthly program is compared to straight quarterly program.
A-28
-------
Table A-12. CONTROL COSTS PER MEGAGRAM OF BENZENE REDUCED3
Benzene Emission .
Fugitive
Emission
Source
Va 1 ves
in gas/
vapor
serv ice
Valves
in light
liquid
service
Pumps
Compressors
Pressure
Relief
Devices
Open-ended
Lines
Sampl ing
Connection
Systems
Product
Accumulator
Vessels
Control Technique
Annual leak detection and
repair
Quarterly leak detection
and repair
Monthly leak detection
. and repair
Sealed bellows valves
Annual leak detection and
repair
Quarterly leak detection
and repair
Monthly leak detection
and repair
Sealed bellows valves
Annual leak detection and
repair
Quarterly leak detection
and repair
Monthly leak detection
and repair
Dual mechanical seal
systems
Degassing reservoir
vents
Quarterly leak detection
and repair
Monthly leak detection
and repair
Rupture disk with
block valves
Rupture disk with
3-way valve
0-Ring
Closed vent system
to flare
Caps on open ends
Closed purge sampling
Closed vent system
Reduction Per Source
(Mq/yr)
New
0.022
0.089
0.102
0.15
0.013
0.038
0.044
0.058
0.13
0.45
0.52
0.63
3.5
0.58
0.63
0.90
0.90
0.90
0.90
0.013
0.083
6.9
Existing
0.022
0.087
0.099
0.14
0.012
0.037
0.043
0.056
0.13
0.43
0.51
0.61
3.4
0.56
0.61
0.87
0.87
0.87
0.87
0.012
0.081
6.7
Average
$/Mq Benzene
New
(350)e
•
(500)e
(450)e
3,700
(160)e
(370)e
(2SO)e
10,300
870
(78)e
(69)e
2,100
(55)e
(540)e
(480)e
22
590
(520)e
28
430
880
94
Existing
(350)e
(510)e
(460)e
6,100
(180)e
(380)e
(290)e
16,000
870
(81)e
(71)e
2,400
(57)e
(560)e
(490)e
460
780
(600)e
29
470
900
97
Incremental ,
S/Mq Benzenpd
New
(350)e
(540)e
(95)e
12,000
(160)e
(480)e
300
44,000
870
(460)e
14
13,000
(55)e
(540}e
320
l,200f .
3,000f
(610)e'f
l,200f
430
880
94
txi sting
(350)e
(560)e
(103)e
22,000
(180)e
(480)e
300
70,000
870
(490)e
13
15,000
(57)e
(560)e
320
2,700f
3,800f
(630)'e>f
l,300f
470
900
97
A-29
-------
Table A-12.
CONTROL COSTS PER HEGAGRAH OF BENZENE REDUCED (Concluded)
aCosts and emission reductions are based on one fugitive emission source
in benzene service.
bThe benzene emission reduction represents a weighted average of emission
Jeduct?ons for three model plants with different percentages of benzene
E^sv^^
are mufti Jlied by the VOC emission reduction for each control technique.
cost
technique).
eValues in parentheses denote savings.
Incremental costs are based on comparing equipment costs with monthly
leak detection and repair costs.
A-30
-------
APPENDIX'S
MODEL FOR EVALUATING THE EFFECTS OF LEAK DETECTION
AND REPAIR ON BENZENE FUGITIVE EMISSIONS FROM CERTAIN VALVES AND PUMPS
B-l
-------
APPENDIX B - MODEL FOR EVALUATING THE EFFECTS OF LEAK DETECTION
AND REPAIR ON BENZENE FUGITIVE EMISSIONS FROM CERTAIN VALVES AND PUMPS :
B.I INTRODUCTION
The purpose of Appendix 8 is to present a mathematical model for
evaluating leak detection and repair (LDAR) programs. The model
described in this appendix takes an empirical approach by incorporating
recently available data on leak occurrence and recurrence and data on
the effectiveness of simple in-line repair.
B.2 DESCRIPTION OF MODEL
The modeled LDAR program is based on the premise that all sources
at any given time are in one of four categories:
1) Non-leaking sources (sources screening less than the leak definition
of 10,000 ppmv);
2) Leaking sources (sources screening equal to or greater than the
leak definition);
3) Leaking sources which cannot be repaired on-line (screening equal
to or greater than the leak definition) and are awaiting a shutdown,
or process unit turnaround, for repair;
4) Repaired sources with early leak recurrence.
There are also four basic components to the model:
1) Screening of all sources except those in Category 3, above;
2) Maintenance of screened sources in Categories 2 and 4 above, in
order to reduce emissions to less than 10,000 ppmv;
3) Rescreening of repaired sources;
4) Process unit turnaround during which maintenance is performed for
sources in Categories 2, 3, and 4, above. Figure B-l shows a schematic
diagram of the LDAR program.
Since there are only four categories of sources, only four "leak
rates" apply to all sources. The repaired sources experiencing early
leak recurrence are assumed to have the same leak rate as sources
which were unsuccessfully repaired. The LDAR model does not evaluate
gradual changes in leak rates over time but assumes that all sources
in a given category have the same average leak rate.
B-2
-------
r—
'—.
•j
1
i
I
1
a
•
j
;
I
u
J
•M^^n
I"
*[
1'
*
b
£
j
5 :
f ;
<
t
S'
i'
:•
U i
I
•s
f|j !
:|t I
MS S
• |M •
]•.
• ^
$ s
:1
U "•
1*
t
5 j!
:| |
if 1
|!
x-s
I*
j 3
-, j
*
J
1 ^
I I
f J
i 1
4
5 I
31]
^|2
?
"
*
• ;
i 1
• >.
2
I i
U A
M »
i
1
•
K:
5 1 i 1
ir.
- Sf
. - ^
4
•
h 3
ii 1
a k> e
•
5"
•
JJ
** o
L *-
Js
'
/ 1
5
|l 4
•s •: |
k
1 i
5 i
& I
• U
• u
W Q
i ••
M ' . • S
M
»
1
I
•fl
•
•
'
^•M
!
"
i
i
4
"i
j
• U
ij
T
i
•=C
CD
O
a:
^
o;
t— H
f^
LU
a:
•ZL
o
f—
LU
LU
S
LJLJ
LU
I ri
i O
Z C3
t s:
•
1 - o
•« 3 ' '
i. 1 i
5-S * CD
:- \ •*
" * I-H
:] s Q
? 2 1 °
a " I t-"
I! 1 1
» ? 3 Si
3 • LXJ
H| 5
•^ • o ^j
* M — *
•5 ^ ~ T— 1
5^3 1
is s m
25: i-
|J 5 ^
J Ha C
« *~*~
• k. V
.^! *
B-3
-------
The LDAR model enables investigation of several leak detection
and repair program scenarios. General inputs pertaining to the LDAR
program itself may vary (for example, frequency of inspection, repairs,
and process unit turnarounds). Further, input characteristics of the
emission sources may vary. Inputs required in the latter group include:
1) The fraction of sources initially leaking;
2) The fraction of sources which become leakers during a period;
3) The fraction of sources with attempted maintenance for which
repair was successful;
4) The emission reductions from successful and unsuccessful repair.
Other assumptions associated with the LDAR model are as follows:
1) All repairs occur at the end of the repair period; the effects
associated with the time interval during which repairs occur are
negligible;
2) Unsuccessfully repaired sources instantaneously fall into the
unrepaired category;
3) Leaks other than unsuccessful maintenance and early recurrences
occur at a linear rate with time during a given monthly period; the
monthly occurrence rate is assumed linear within an inspection period;
4) A process unit turnaround essentially occurs instantaneously at
the end of a quarter and before the beginning of the next monitoring
period;
5) The leak recurrence rate is equal to the leak occurrence rate;
sources that experience leak occurrence or leak recurrence immediately
leak at the rate of the "leaking sources" category.
The input parameters required for the LDAR model and the outputs
that the model calculates are presented in the Additional Information
Document (AID).2 The major data outputs for valves and pumps appear
in Tables B-l and B-2 for valves in gas and light liquid service and
in Tables B-3 and B-4 for pumps in light liquid service. Tables B-l
and B-3 contain information used to estimate total VOC and benzene
emission reductions at various inspection intervals for the cost
analysis presented in Tables A-l through A-4 of Appendix A. Tables 3-2
and B-4 contain information on the fraction of valves and pumps screened
and the fraction of sources operated on to determine the leak detection
and repair costs in Tables A-l through A-4 of Appendix A.
B-4
-------
UJ
Li
t-t
g
Lu
cc
a
lr
z . cs
c V*
H, _,
CJ t—
— ; -T*
a cs
u; o: M
OO Ct >- _l
^y 1— I
2 si
i *•£
o S^
S >- =
0 cj i-
UJ <: z
o: as o
u. x
^*
0 a '
£ ^2
oo ^ S
±_i a: o
^> T OS
UJ x, <:
e= z
00 IS OS
00 - =
-
Q £ *
5 u en
^t <; uj
U.S-
oo _;
o; z •=
O C3 =•
1 — "™
cj en as
5- cn c
77- H, u.
LL- x
_. U. 07
0 ai
HH Ui M
oo t- cn
oo -
Lu H- ui »
co i:
LU Ui en ua
tr o? en
— J o x cn
:- z cs
-V HH
«•
CJ
3
a
UJ
OS
OS
X
X
cs
1
z
c
cn
cn
X
UJ
z
•*
g
tu
cn
a
3
Z d
G K,
t- en ~"
O 1- t-
3 >-" X
a z cs
Ui 3 HH
OS _l
X
TO _i t-
M W
i—
u cn
•s u;
u. ->
^i
z - u
MX M
1- Ul >
cn os
ui cn ui
tn en
u. • Z C3
u£ *^
-i
OS Z -t
_^
t*) O -J
«r »r <£ >-
O O Z _
O O C. u.
. . 1-1 u;
O O K l-
CJ OS
-
C CO-
>—
u cn
<: ui
u. ->
z 3
M — O 2>
S3 -< w
w -o cn as
. . cn a
O O M U.
x
ui cn
z
a o
Ul "-i
t- cn
« cn uj
X M U
>-< X >->
h- UJ ->
cn os
Lu CO Ui
u cn
u. - z es
as w
4
X
3
n in cn
«* o
o o
0 0
-< CJ
1
1
1
1
1
1
z
c
t—
L;
3
a
u
OS
a:
z
X
OS
••^
1
z
a
cc
iC
I— 1
LU
Z
•c
Ui
z
a
i—
. CJ
a
UJ
*j*
OS
X
X
cs
T
z
a
cn
cn
H^
LU
Z
UJ
X
a
z
p
2;
l£
h-
1
0- IX
O- (M
UT -0
* •
O O
UJ
CJ
t-t
g
U.'
en
a
3
z a
o «
i—
CJ t-
3 X
a tr
UJ M
as en '
>—
-
a i
*•
z S
o rx o 5.
in in cn as
. . O7 O
O O >- U.
X
ui en
ai
i- cn
CC OS
LU en ui
en cn
u. «
ex cn
>- z cs
OS M
x
3
•o o- cn
« O
o o
0 0
« P-J
z
0
H-
^
3
a
u
OS
^
X
X
CC
^
i
z
o
cn
en
x
Ui
z
UJ
z
o
hH
I—
CJ
3
a
LU
OS
X
x.
1
z
a
Hi*
cn
cn
H4
X
Ul
z
^
UJ
a
z
3
a
as
z
as
3
«T CM
O -<
r-i r-i
• •
o o
05 rx
CO CO
o o
0 O
0 0
n o*
*•« »-*
, ^
o o
-. o
M f*3
CJ PJ
0 0
o o
« r-i
B-5
-------
Table B-2. FRACTION OF VALVES SCREENED AND OPERATED ON
SUMMARY OF TOTAL FRACTION OF SOURCES SCREENED AND OPERATED ON FOR VAI.VFR PY YEAR
MONTHLY LIMITS
GAS SERVICE
LIGHT i
SFRVTHE
t
YEAR
1
•»
3
4
5
YEAR
1
2
3
4
S
YEAR
1
2
3
4
S
YEAR
1
2
3
4
S
TOTAL FRACTION OF TOTAL FRACTION OF
SOURCES SCREENED SOURCES OPERATED ON
12.7728 0.2825
11.5686 0.2110
11.8971 0.1792
11. .6917 0.2026
11.8991 0,1776
SUMMARY OF TOTAL FRACTION OF SOURCES SCREENED AMD t
QUART/MONTH UNITS
GAS SERVICE
TOTAL FRACTION OF TOTAL FRACTION OF
SOURCES SCREENED SOURCES OPERATED ON
5.2994 0.2776
4.1169 0.2065
4.3150 0.177]
4.1595 0.1983
4.2971 0.1756
SUMMARY OF TOTftL FRACTION OF SOURCES SCREENED AMD
QUARTERLY UNITS
GAS SFRVICF
TOTAL FRACTION OF TOTAL FRACTION OF
SOURCES SCREENED SOURCES OPERATED OH
4.9324 0.2760
3.8648 0.2051
3.9726 . 0.176?
3.9044 0.1970
3.9730 0.1748
SUMMARY OF TOTAL FRACTION OF SOURCES SCREENED ANI.1
YEARLY UNITS
GAS SERVICE
TOTAL FRACTION OF TOTAL FRACTION OF
SOURCES SCREENED SOURCES OPERATED ON
1.9900 0.2512
0.9749 0.1798
1.0000 0.1634
0.9037 6.1735
1.0000 0.1627
B-6
TOTAL FRACTION OF TOTAI FRACTION OF
SOURCES SCREENED SOURCES OPERA TED ON
12.7574 0.2937
11.5553 0.2119
11.897? 0.1793
11.6915 0.2026
11.8991 0.1776
1PERATFD OH F OR VALVFS PY YEAR
LIGHT LIQUID SERVICE
TOTAL FRACTION OF TOTAL FRACTION OF
SOURCES SHRKENFD SOURCES OPERATED ON
5.3121 0.2888
4.1121 0.2074
4.3170 0.1773
4.1594 0.1983
4.2971 0.1756
1PFRATED ON FOR VALVES RY YEAR
LIGHT IRUID SKRMICE
TOTAL FRACTION OF TOTAL FRACTION OF
SOURCES SCREENED SOURCES OPERATED ON
4.9281 0.287?
3.8603 0.2060
3.9725 0.1764
3.9043 0.1970
3.9730 0.1.748
1PFRATED OK FOR VALVFS BY YEAR
LIGHT LIOUII.I SERVICE
TOTAL FRACTION OF TOTAL FRACTION OF
SOURCES SCRFENED SOURCES OPERATED ON
1.9890 0.2622
0.9738 0.1808
1.0000 0.1636
0.9836 0.1736
1.0000 0.1627
-------
z
o
z
o
u
a
UI
CS 03
a en
ui i-
o
t/o
z:
o
i—i
h-
O
Q
LU
C£
oo
oo
oo
to
Q
-
t- _i
U X
cc z
L^ Q
z
a
Z 1
a
•~ z
ce 3
x a
X GC
f^ ^
vx 2»
"~ 3
cn i—
-
i- a
u
cn ce
u en ui
en cn
LU «
ox u
o
>- z »
a: w
00
«* V4
o o
0 0
-t CM
•
a _i
M CS
I- Ui
U I-
-
h- a
u
.
* *
o o
« cn
UI
CJ
I- UI =>
cn x
ui cn ui
cn cn
u,«
a z u
a
>• z >
ae M
<£
en
t£.
*^
C9
^
I
cn
in
•x.
UI
-
CJ _'
LU UI
a
a
•*^ 2
fi£ 3
X O
X LK
CD •
h- a
CJ
- z :>
ce 1-1
«•
cn
cn
z
UI
z
-------
Table B-4. FRACTION OF PUMPS SCREENED AND OPERATED ON
SUMMARY OF TOTAL FRACTION OF SOURCES SCREENED AND OPERATED ON FOR PUMPS BY YEAR
MONTHLY UNITS
VOC SERVICE
YEAR
1
n
3
4
5
TOTAL FRACTION OF
SOURCES SCREENED
13.0000
12.0000
12.0000
12.0000
12.0000
TOTAL FRACTION
SOURCES OPERATED
0.6480
0.4030
0.4080
0.4080
0.4080
OF
ON
SUMMARY OF TOTAL FRACTION OF SOURCES SCREENED AND OPERATED ON FOR PUMPS BY YEAR
QUARTERLY UNITS
VOC SERVICE
TOTAL FRACTION OF
YEAR SOURCES SCREENED
1 S.OOOO
2 4.0000
3 4.0000
4 4.0000
5 4.0000
TOTAL FRACTION
SOURCES OPERATED
0.6343
0.3943
0.3943
0.3943
0.3943
OF
ON
SUMMARY OF TOTAL FRACTION OF SOURCES SCREENED AND OPERATED ON FOR PUMPS Br "EAR
YEARLY UNITS
VOC SERVICE
YEAR
1
2
3
4
5
TOTAL FRACTION OF
SOURCES SCREENED
2.0000
1.0000
1.0000
1.0000
1.0000
TOTAL FRACTION
SOURCES OPERATED
0.5797
0.3397
0.3397
0.3397
0.3397
OF
ON
B-8
-------
B.3 REFERENCES
1. Wetherold, R.G., G.J. Langley, et.al. Evaluation of Maintenance
for Fugitive VOC Emissions Control. U.S. EPA, Industrial Environmental
Research Laboratory. EPA-600/52-81-080. May 1981 Docket
Number A-79-27-IV-A-16.*
2.
Fugitive Emission Sources of Organic Compounds—Additional Information
on Emissions, Emission Reductions, and Costs. U.S. EPA, Office
of Air Quality Planning and Standards, Emission Standards and
Engineering Division. EPA-450/3-82-010. April 1982. Docket
Number A-79-27-IV-A-24.*
*References can be located in Docket Number A-79-27 at the U.S. Environmental
Protection Agency Library, Waterside Mall, Washington, D.C.
B-9
-------
-------
APPENDIX C
METHODOLOGY FOR ESTIMATING INCIDENCE OF LEUKEMIA AND
MAXIMUM LIFETIME RISK FROM EXPOSURE TO
FUGITIVE EMISSION SOURCES OF BENZENE
C-l
-------
APPENDIX C
METHODOLOGY FOR ESTIMATING INCIDENCE OF LEUKEMIA AND MAXIMUM
LIFETIME RISK FROM EXPOSURE TO FUGITIVE EMISSION
SOURCES OF BENZENE
C.I INTRODUCTION
The purpose of this appendix is to describe the methodology and
to provide the information used to estimate the incidence of leukemia
and maximum lifetime risk from population exposure to benzene emissions
from fugitive emission sources. 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
exposure by summing the products of the concentrations and associated
populations, and calculation of annual leukemia incidence and maximum
lifetime risk from the concentration and exposure 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 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, EPA-450/5-82-003.
C.2 ATMOSPHERIC DISPERSION MODELING AND PLANT EMISSION RATES
The Human Exposure Model (HEM) was used to estimate concentrations
of benzene around approximately 130 plants that contain fugitive emission
sources of benzene.1 The HEM estimates the annual average ground-level
C-2
-------
concentrations resulting from emissions from point and area sources.
For point sources, the dispersion model within the HEM is a Gaussian
model that uses the same basic dispersion algorithm as the climato-
logical form of EPA's Climatological Dispersion Model.2 Gaussian
concentration files are used in conjunction with multi-year STAR data
and annual 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. Industry provided confirmation where ambiguities
existed.
Plant emission rates used as inputs to the dispersion model were
calculated as follows: (1) three model process unit emission rates
were developed; (2) model units were assigned to each plant according
to the actual process unit composition of the plant; and (3) assigned
model unit emission rates were summed for each plant to obtain the
total plant benzene fugitive emission rate. Model unit emission rates
are based upon data that reflect benzene emissions from fugitive
emission sources, as described in Section 2.8.4 of this document. EPA
Docket Number A-79-27-IV-B-11 explains the derivation of emission rates
in greater detail.
Ordinarily the Industrial Source Complex - Long Term (ISC-LT)
dispersion model would have been used to estimate benzene concentrations;
however, the HEM model was used instead. The HEM tends generally to
C-3
-------
predict more conservative concentrations than the ISC-LT dispersion
model. The ISC-LT model has been validated and is considered superior
to the HEM. However, the ISC-LT model was not used in this analysis
because the effort required to use properly the ISC-LT was not warranted,
considering the uncertainties in the basic data available for the
model (see Section 2.2.1.2 of this document). Because there are so
many plants, the resources that would be required to model each one
individually were considered unreasonable given that the difference in
the more detailed results would not be great enough to alter the basic
decisions of the standard. In order to evaluate the effect of using
the HEM instead of the ISC-LT dispersion model, the ISC-LT model was
used for nine plants and the results were compared with the HEM results.
This comparison is described in Section C.7 and in more detail in
Docket Number A-79-27-IV-B-18. The HEM output for all plants modeled
is contained in the EPA Docket Number A-79-27-IV-B-15.
C.3 POPULATION AROUND PLANTS CONTAINING FUGITIVE EMISSION SOURCES
OF BENZENE
The HEM was used to estimate the population that resides in the
vicinity of each receptor coordinate surrounding each plant containing
fugitive emission sources of benzene. 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 the data necessary for the estimation. A separate file
of county-level growth factors, based on 1978 estimates of 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 fugitive emission sources of benzene. The population
around each plant was identified by specifying the geographical coordinates
of that plant. Table C-l presents the plants and locations for fugitive
emission sources of benzene.
C-4
-------
Table C-l. PLANTS AND LOCATIONS FOR FUGITIVE EMISSION
SOURCES OF BENZENE
Plant
Location
Coordinates
LongitudeLatltude
Region II
1. American Cyanamid
2. DuPont
3. Merck & Co.
4. Reichhold3
5. Standard Chlorine
6. Tenneco3
7. Texaco
8. Ashland Oil
9. ICC Industries
10. Commonwealth Oil
11. Phillips Puerto Rico
12. Puerto Rico Olefins
13. Union Carbide
14. Amerada Hess
Region III
15. Getty
16. Standard Chlorine
17. Sun-Olin
18. Continental Oil
19. Atlantic Richfield
20. Gordon Terminals
21. Gulf Oil
22. Koppers3
23. Koppers
24. Standard Oil
(Ohio)/BP Oil
Boundbrook, NJ
Gibbstown, NJ
Rahway, NJ
Elizabeth, NJ
Kearny, NJ
Fords, 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 Rocks, PA
Philadelphia, PA
Bridgeville, PA
Petrolia, PA
Marcus Hook, PA
74°06'04"
75°17'50"
74°16'00"
74°13'05"
74°06'39"
74°19'08"
75°08'42"
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"
80°21'20"
80°03'10"
75°12'31"
80°04'41"
79°42'30"
75°25'26"
40°33'25"
39°50'25"
40037'00"
40°39'15"
40°45'03"
40°30'42"
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°39'21"
40°28'22"
39°54'18"
40°21'32"
41°45'41"
39049'12"
C-5
-------
Table C-l. PLANTS AND LOCATIONS FOR FUGITIVE EMISSION
SOURCES OF (BENZENE (Continued)
Plant
Region III (concluded)
25. Sun Oil
26. U.S. Steel
27. Allied Chemical
28. American Cyanarnid
29. Ashland Oilb
30. Mobay Chemical
31. PPG
32. Union Carbide
33. Merck & Co.
Region IV
34. Oim Walter Resources
35. Reichhold Chemicals
36. Ashland Oil
37. B.F. Goodrich
30. GAP
39. 01 in Corporation
40. Chevron
41. First Chemical
Region V
42. Clark Oil
43. Core-Lube
44. Koppers
45. Monsanto
46. National Distillers
(U.S.I.)
47. Northern
Petrochemicals
43. Reichhold Chemical sa
49. Shell Oil
Location
Marcus Hook, PA
Neville Island, PA
Mounds vi lie, WV
Willow Island, WV
Neal, WV
New Martinsville, WV
Natrium, WV
Institute, WV
Elkton, VA
Birmingham, AL
Tuscaloosa, AL
Ashland, KY
Calvert City, KY
Calvert City, KY
Brandenburg, KY
Pascagoula, MS
Pascagoula, MS
Blue Island, IL
Danville, IL
Cicero, II.
Sauget, IL
Tuscola, IL
Morris, IL
Morris, IL
Wood River, IL
Coordi
Longitude
75°24'51"
80°05'00"
80°48'04"
81° 19 "08"
82°35'37"
80°49'50"
80°51'06"
81°47'05"
78°39'03"
86°47'30"
87°28'21"
82°36'32"
88°19'51"
88°24'48"
86°07'15"
88°28'37"
88°29'45"
87°42'07"
87°32'30"
87°38'45"
90°10'11"
88°21'00"
88°25'42"
88°17'56"
90°04'24"
nates
Latitude
39°48'45"
40°30'00"
39°55'00"
39°21'45"
38°22'07"
39°43'30"
39°44'45»
38°22'40"
38°23'05"
33°35'30"
33°15'06"
38°22'30"
37°03'19"
37°02'51"
38°00'30"
30°19'04"
30°20'57"
41039'19"
40°07'10"
41048'30"
38°36'06"
39°47'53"
41°21'28M
42023'20"
38°50'26"
C-6
-------
Table C-l. PLANTS AND LOCATIONS FOR FUGITIVE EMISSION
SOURCES OF BENZENE (Continued)
Dlant
Region V (concluded)
50. Union Oil
(California)
51. Dow Chemical
52. Dow Chemical
53. Sun Oil
Region VI
54. Vertac/Transvaal
55. Allied Chemical
56. American Hoechst
57. Cities Service
58. Continental Oil
59. Cos-Mar, Inc.
60. Dow Chemical
61. Exxon
62. Gulf Coast Olefins
63. Gulf Oil
64. Gulf Oil
65. Pennzoil United
(Atlas Processing)
66. Rubicon
67. Shell Oil
68. Tenneco
69. Union Carbide
70. Sun Oil
71. Amerada Hess
72. American Hoechst
73. American Petrofina
of Texas
Location
Lemont, IL
Bay City, MI
Midland, MI
Toledo, OH
Jacksonville, AR
Geisrnar, 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
Geisrnar, LA
Norco, LA
Chalmette, LA
Taft, LA
Tulsa, OK
Houston, TX
Bay port, TX
Port Arthur, TX
Coordi
Longitude
88000'10"
83°52'22"
84°12'18"
83°31'40"
92°04'56"
91°03'12"
91°12'40"
93°19'01"
93°16'35"
91°04'09"
91°14'30"
91°10'17"
90°26'23"
89°58'26"
90°55'19"
93°46'13"
91000'37"
90°27'35"
89058'19"
90027'15"
96°01'15"
95014'15"
95°01.I15"
93°53'20"
nates
Latitude
41°40'20"
43°37'21"
43035'42"
41°36'52"
34°55'36"
3P°12'55"
30033'03"
30°10'58"
30°14'30"
30°14'16"
30019'50"
30°29'14"
29°59'16"
29°41'00"
30°05'44"
32°28'12"
30011'06"
29°59'42"
29°55'56"
29°59'17"
36°08'25"
29°41'39"
29°36'10"
29°57'30"
C-7
-------
Table C-l. PLANTS AND LOCATIONS FOR FUGITIVE EMISSION
SOURCES OF BENZENE (Continued)
Coordinates
Plant
Reqion VI (continued)
74. American Petrofina
(Cosden Oil)
75. American Petrofina/
Union Oil of
California
7b. Atlantic Richfield
77. Atlantic Richfield
(ARCO/Polyners)
78. Atlantic Richfield
(ARCO/Polymers)
79. Celanese
80. Charter
International
81. Coastal States Gas
82. Corpus Christi
Petrochemicals
83. Cosden Oil
34. Crown Central
85. Denka (Petrotex)b
86. Dow Chemical (A)
87. Dow Chemical (B)
88. Dow Chemical
89. DuPont
90. DuPont
91. Eastman Kodak
92. El Paso Natural Gas
93. El Paso Products/
(Rexene Polyolefins)
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
Houston, TX
Freeport, TX
Freeport, TX
Orange, TX
Beaumont, TX
Orange, TX
Longview, TX
Odessa, TX
Odessa, TX
Longitude
101°24'55"
93058'45"
95°07'30"
95013'54"
93058'15"
100°57'47"
95015'09"
97°26'44"
97°31'21"
93052'58"
95°10'30"
283.5
95°19'55"
95°24'09"
93°45'14"
94°01'40"
93044.44"
94°41'24"
102°19'29"
102020'00"
Latitude
32016'H"
30°00'00M
29°50'00"
29°43'10"
29°51'24"
35°32'U7"
29040-17"
27°48'42"
27°50'02"
29°57'46"
29°44'40"
3289.6
28°57'23"
28059'17"
30°03'20"
30000'51"
30003'24"
32°26'17"
31049'27"
31°49'22"
C-8
-------
Table C-l, PLANTS AND LOCATIONS FOR FUGITIVE EMISSION
SOURCES OF BENZENE (Continued),
Coordinates
Plant
Region VI (continued)
94. Exxon
95. GATX Terminal Group
96. Georgia-Pacific Corp.
97. Goodyear Tire and
Rubber
98. Gulf Oil Chemicals
99. Gulf Oil Chemicals
100. Hercules
101. Howell
102. Independent Refining
Corp.
103. Kerr-McGee Corp.
(Southwestern)
104. Marathon Oil
105. Mobil Oil
106. Monsanto
107. Monsanto
108. Oxirane
109. Petrounited Terminal
Services
110. Phillips Petroleum
111. Phillips Petroleum
112. Phillips Petroleum
113. Quintana-Howell
114. Shell Chemical
115. Shell Oil
116. Shell Oil
Location
Baytown, TX
Houston, TX
Houston, TX
Bay port, TX
Cedar Bayou, TX
Port Arthur, TX
McGregor, TX
San Antonio, TX
Winnie, TX
Corpus Chris ti, 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
Longi tude
• 95°01'04"
95°13'29"
95°03'00"
95°02'44"
94°55'10"
93°58'30"
97°16'30"
98°27'36"
94°20'28"
97°25'24"
94°54'47"
94°03'30"
95°12'44"
94°53'40"
95°06'29"
95°01'23"
101°22'05"
95°10'53"
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"
29°50'04"
27°48'16"
29°22'21"
30°04'00"
29°15I09"
29°22'44"
29050'00"
29°33'51"
35042'05"
29°43'59"
29°04'24"
27048'30"
29°38'15"
29°42'55"
31°49'05"
C-9
-------
Table C-l. PLANTS AND LOCATIONS FOR FUGITIVE EMISSION
SOURCES OF BENZENE (Concluded)
Coordinates
Plant
Region VI (concluded)
117. Standard Oil
(Indiana)
118. Standard Oil
(Indiana)/Amoco
119. Sun Oil
120. Texaco
121. Texaco/ Jeff arson
Chemical
122. Union Carbide
123. Union Carbide
124. USS Chemicals
Reqion VII
125. Chemplex
126. Getty Oil
127. Monsanto
Region IX
128. Atlantic Richfield
129. Chevron
130. Specialty Organ ics
131. Standard Oil of
California (Chevron
Chemical)
132. Union Carbide
133. Witco Chemical
134. Montrose Chemical
135. Stauffer Chemical
aPlant no longer produces or
in the analysis.
Plant does not use benzene
included in the analysis.
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
uses maleic anhydride;
in production of maleic
C- 10
Longitude
95°11'55"
94055'45"
97°31'38"
,93054'43"
93°56'00"
96°45'59"
94°56'33"
95°15'06"
90°17'29"
96°52'00"
90012'00"
118°14'30
122°23'36
117°55'56
118°24'41
118°20'50
118°14'13
115°00'40
115°00'40
therefore,
anhydride;
Latitude
29°13'06"
29°21'58"
27°49'57"
29°52'00"
29°57'50
28°30'38"
29°22'27"
29°42'18"
41°48'24"
37°47'10"
38035'00"
33043,49,,
11 37°56'12"
11 34°06118"
33°54'39"
33°51'11"
11 33°49'18"
11 36°02'28"
" 36°02'28"
it is no longer included
therefore, it is no longe
-------
C.4 POPULATION EXPOSURE METHODOLOGY
C.4.1 Exposure Methodology
The HEM uses benzene atmospheric concentration patterns (see
Section C.2) together with population information (see Section C.3) to
calculate population exposure. 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 exposure 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 exposure. 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 generally
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 exposure 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 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
C-ll
-------
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 exposures,
see Reference 1.)
C.4.2 Total Exposure
Total exposure (persons- |Lig/m3) is the sum of the products of
concentration and population, computed as illustrated by the following
equation:
N
Total exposure = f ^ (PiC-j)
where
P-J = 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 2.0 km.
The computed total exposure 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. (Note: "Exposure" as used in this appendix is the same as
"dosage" in the computer printout, Docket Number IV-B-15.)
C.5 LEUKEMIA INCIDENCE AND MAXIMUM LIFETIME RISK
C.5.1 Unit Risk Factor
The unit risk factor (URF) for benzene is 9.9 x 10-8 (cases per
year)/( Aig/m3-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 risk to ambient benzene exposures and updated in Response
to Public Comments on EPA's Listing of Benzene Under Section 112,
EPA-450/5-82-003.
C-12
-------
C.5.2 Annual Leukemia Incidence
Annual leukemia incidence (the number of cases per year) associated
with a given plant under a given regulatory alternative is the product
of the total exposure around that plant (in /ag/m3-persons) and the
unit risk factor, 9.9 x 10-8. Thus,
Leukemia cases per year = (total exposure) x (unit risk factor), (2)
where total exposure 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 fugitive
emission sources of benzene have various risk levels of leukemia
incidence 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:
Maximum lifetime risk = C-j jmax x (URF) x 70 years,
(3)
where
ci,max = tne maximum concentration among all plants at any receptor
location where exposed persons reside,
URF = the unit risk factor, 9.9 x 10-8, and
70 years = an individual's average 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-2 presents the model unit number, emission
rate, maximum annual average benzene concentration, and total exposure
for each plant under the baseline level.
C.5.4.1 Annual Leukemia Incidence. As an example for calculating
annual leukemia incidence, the Chevron plant in El Segundo, California
(Number 131), is used. Under the current (baseline) level of emission
C-13
-------
Table C-2. BASELINE DISPERSION MODELING AND EXPOSURE DATA
Plant Model Unit
Number Number
Region II
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Region III
15
16
17
18
19
20
21
22
23
10
10
60
9
2
9
30
21
2
45
46
4
25
24
23
.2
4
8
12
59
52
9
11
Emission
Rate
(g/sec)
4.50
4.50
0.72
0.93
2.70
0.93
2.02
0.73
2.70
3.45
2.81
0.55
4.02
2.10
0.99
2.70
0.55
1.73
0.28
0.23
2.69
0.93
0.90
Maximum
Annual
Average
Benzene
Concentration
0/g/m3)
1.69 x 102
5.00 x 101
2.71 x 101
b
5.00 x 10°
b
4.80 x 101
3.24 x 101
1.20 x 102
1.11 x 10"1
1.12 x 10"1
1.11 x 10"1
1.08 x 10"1
9C
7.17 x 10"^
2.35 x 101
5.00 x 101
5.37 x 10"2
2.50 x 101
5.00 x 10°
n
6.72 x 10U
6.39 x 101
b
1.04 x 10"1
Total
Exposure
(persons-/^g/m3)
2.91 x 105
1.51 x 105
6.25 x 104
b
4.06 x 105
b
1.26 x 105
1.70 x 104
3.75 x 104
4.25 x 104
4.29 x 104
4C
4.25 x 10^
4.14 x 104
4C
1.62 x 10^
6.08 x 103
9.86 x 103
4.74 x 102
3.51 x 103
1.96 x 103
4
1.48 x 10^
2.32 x 105
b
4.60 x 102
C-14
-------
Table C-2. BASELINE DISPERSION MODELING AND EXPOSURE DATA (Continued)
Plant Model Unit
Number Number
Emission
Rate
(g/sec)
Maximum
Annual
Average
Benzene
Concentration
(M9/m3)
Total
Exposure
(persons~jug/m3)
Region III (concluded)
24
25
26
27
28
29
30
31
32
33
Region IV
34
35
36
37
38
39
40
41
Region V
42
43
44
45
46
47
48
48
56
9
10
10
9
10
2
19
61
1
1
49
4
9
4
21
10
3
1
9
14
4
4
9
1.51
1.77
0.93
4.50
4.50
0.93
4.50
2.70
2.65
1.32
0.82
0.82
4.14
0.55
0.93
0.55
1,10
4.50
0,92
0.82
0.93
7.20
0.55
0.55
0.93
2.50 x 101
4.21 x 101
2.68 x 101
9.63 x 10"1
1.00 x 102
d
4.07 x 10"1
5.00 x 101
8.91 x 101
2.50 x 101
1.92 x 101
1.00 x 101
1.00 x 102
7.10 x 10~2
1.97 x 10"1
8.73 x 10"2
5.87 x 10"2
2.11 x 102
2.33 x 101
1.00 x 101
c
1.67 x 102
5.31 x 10~2
1.39 x 101
b
5.08 x 104
4.19 x 104
3.78 x 104
2.35 x 104
8.00 x 103
d
2.99 x 103
2.62 x 103
4.42 x 104
2.79 x 103
1.64 x 104
4.30 x 103
2.60 x 104
3.62 x 102
1.07 x 103
6.82 x 102
1.98 x 103
3.01 x 104
5.73 x 104
2.56 x 103
c
3.15 x 105
3.89 x 102
6.10 x 103
b
-------
Table C-2. BASELINE DISPERSION MODELING AND EXPOSURE DATA (Continued)
Plant Model Unit
Number Number
Emission
Rate
(g/sec)
Region V (concluded)
49
50
51
52
53
Region VI
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
21
23
26
13
56
2
4
20
42
4
20
5
40
4
48
2U
23
10
4
33
26
54
59
20
22
53
0.73
0.99
3.10
3.62
1.77
2.70
0.55
0.92
2.21
0.55
0.92
1.66
2.46
0.55
1.88
0.92
0.99
4.50
0.55
1.37
3.10
2.70
0.23
0.92
0.88
5.59
Maximum
Annual
Average
Benzene
Concentration
(^g/m3)
1.69 x 101
2.50 x 101
7.83 x 101
9.14 x 101
5.83 x 101
2.50 x 101
1.00 x 101
1.00 x 101
1.00 x 101
1.00 x 101
1.00 x 101
5.12 x 101
5.00 x 101
1.89 x 101
1.27 x 10"1
1.00 x 101
4.13 x 101
1.00 x 102
1.00 x 101
2.50 x 101
5.00 x 101
1.08 x 102
7.50 x 10°
4.49 x 101
1.00 x 101
1.00 x 102
Total
Exposure
(persons-^g/m3)
9.28 x 103
1.23 x 104
3.32 x 104
2.81 x 104
6.95 x 104
4.97 x 103
8.06 x 102
6.28 x 103
6.33 x 103
3.13 x 103
1.49 x 103
6.00 x 103
7.05 x 104
3.13 x 103
1.16 x 103
2.22 x 103
4.70 x 104
6.32 x 103
1.94 x 103
7.18 x 104
1.04 x 104
5.20 x 104
1.50 x 104
9.93 x 103
7.70 x 103
1.06 x 104
C-16
-------
Table C-2. BASELINE DISPERSION MODELING AND EXPOSURE DATA (Continued)
Plant
Number
Region VI
75
76
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
Model Unit
Number
(continued)
39
27
47
6
59
43
44
57
4
56
9
20
58
4
10
4
4
18
17
32
59
3
7
5
36
2
23
Emission
Rate
(g/sec)
1.92
4.21
2.98
0.64
0.23
1.63
2.43
3.10
0.55
1.77
0.93
0.92
5.24
0.55
4.50
• 0.55
0.55
1.47
0.83
2.21
0.23
0.92
5.14
1.66
4.50
2.70
0.99
Maximum
Annual
Average
Benzene
Concentration
(^g/m3)
2.50 x 101
2.06 x 102
5.00 x 101
5.00 x 10°
4.48 x 10°
5.31 x 101
1.03 x 102
1.32 x 102
1.00 x 10°
5.00 x 10°
d
2.18 x 101
5.00 x 101
2.50 x 10°
5.00 x 101
1.00 x 10°
5.00 x 10°
1.00 x 101
2.41 x 101
1.00 x 101
7.50 x 10°
4.49 x 101
1.00 x 102
2.50 x 101
5.00 x 101
2.50 x 101
7.50 x 101
Total
Exposure
(persons-fzg/m3)
1.59 x 104
5.10 x 104
1.26 x 105
3.27 x 103
2.50 x 103
9.88 x 104
3.78 x 104
1.42 x 104
3.04 x 103
5.80 x 104
d
3.59 x 103
2.08 x 104
3.02 x 103
2.39 x 104
3.02 x 103
2.22 x 103
1.04 x 104
7.44 x 103
2.52 x 104
9.44 x 103
1.04 x 104
5.69 x 104
4.64 x 103
2.12 x 104
3.22 x 103
4.71 x 104
C-17
-------
Table C-2. BASELINE DISPERSION MODELING AND EXPOSURE DATA (Continued)
Plant Model Unit
Number Number
Emission
Rate .
(g/sec)
Reqion VI (concluded)
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
Reqion VII
125
126
23
21
35
55
15
28
6
59
16
4
38
48
4
31
56
5
34
51
37
4
18
4
4
4
35
1.53
0.73
1.91
2.32
3.84
4.02
1.28
0.23
1.57
0.55
3.24
1.51
0.55
2.20
2.31
1.66
3.20
4.16
2.47
0.55
1.47
0.55
0.55
0.55
2.45
Maximum
Annual
Average
Benzene
Concentration
(j/g/m3 )
1.22 x 10"1
3.10 x 101
4.54 x 101
2.50 x 101
5.00 x 101
9.55 x 101
2.50 x 101
1.12 x 101
2.50 x 101
2.69 x 101
5.00 x 101
6.41 x 101
2.69 x 101
5.00 x 101
6.70 x 101
1.07 x 10"1
7.60 x 101
1.77 x 102
2.50 x 101
1.00 x 10°
5.00 x 101
1.00 x 101
1.79 x 101
5.00 x 10°
1.00 x 101
Total
Exposure
(persons-jig/m3)
6.51 x 102
2.32 x-104
2.08 x 104
2.01 x 104
2.12 x 103
5.59 x 104
1.06 x 104
2.86 x 103
2.28 x 103
2.01 x 104
3.00 x 103
1.42 x 104
6.99 x 103
4.74 x 104
1.52 x 104
2.57 x 102
2.90 x 104
2.43 x 104
2.43 x 104
5.77 x 103
6.68 x 102
5.47 x 103
2.96 x 104
1.25 x 103
6.81 x 103
C-18
-------
Table C-2. BASELINE DISPERSION MODELING AND EXPOSURE DATA (Concluded)
Plant
Number
Region IX
127
128
129
130
131
132
133
134
135
Model Unit
Number
9
41
23
2
35
4
8
2
'1
Emission
Rate
(g/sec)
0.93
1.54
0.99
2.70
1.91
0.55
1.73
2.70
0.82
Maximum
Annual
Average
Benzene
Concentration
(M9/N3)
1.00 x 101
8.07 x 101
5.20 x 101
1.41 x 102
8.71 x 101
2.51 x 101
9.07 x 101
5.00 x 101
1.00 x 101
Total
Exposure
(persons-/zg/m3)
4.14 x 104
,1.35 x 105
3.71 x 104
2.07 x 105
1.36 x 105
5.11 x 104
1.72 x 105
1.36 x 104
4.13 x 103
Model unit numbers (1-56) correspond to the ones listed in
Docket Number A-79-27-1I-A-29, Table A-2. Model unit number 57
contains 1 sulfolane unit, 1 UDEX unit, one ethylene unit, one
cyclohexane unit, and one reforming unit. Model unit number 58
contains one pyrolysis gas unit and 5 ethylene units. Model unit
number 59 contains a benzene storage unit, and model units 60 and
61 contain pharmaceutical process units.
•"Plant no longer produces or uses maleic anhydride; therefore* it is
no longer included in the analysis.
'Since population estimate is not included in the HEM for this plant,
concentration and dosage are based on estimates given at proposal.
These estimates were revised to reflect an increase in emission
rate for this plant.
Plant does not use benzene in production of maleic anhydride; therefore,
it is no longer included in the analysis.
C-19
-------
control, the leukemia cases per year are computed according to Equation 2
as follows:
Annual leukemia incidence =1.36 x 105 (from Table C-2) x 9.9 x 10-8
(cases per year)
Leukemia cases per year = 0.013
C.5.4.2 Maximum Lifetime Risk. As shown in Table C-2, Plant Number 41
(First Chemical) has the highest maximum annual average benzene concentration
of 2.11 x 102 ^g/m3. Using this maximum concentration and Equation 3,
maximum lifetime risk under the current (baseline) level of control is
calculated as follows:
Maximum lifetime risk = 2.11 x 102 x 9.9 x 10-8 x 70
Maximum lifetime risk = 1.46 x 10~3
C.5.5 Summary of Impacts
The methodology for calculating annual leukemia incidence (described
in Sections C.5.2 and C.5.4.1) was applied to each plant for three
levels of emission control. Annual leukemia incidence for each plant
under the three levels [baseline (current level), best available
technology (BAT), and a more stringent level of control (beyond BAT)]
is shown in Table C-3. The total estimated nationwide incidence of
leukemia under the assumed baseline level of control is 0.45 cases per
year. The estimated maximum lifetime risk, which was calculated in
Section C.5.4.2 under the assumed baseline level of control, is
1.46 x 10-3.
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-20
-------
Table C-3. ESTIMATED ANNUAL LEUKEMIA INCIDENCE
Plant
Number
Region II
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Region III
15
16
17
18
19
Baseline
2.88 x 10"2
1.49 x 10~2
6.19 x 10"3
a
4.02 x 10"2
a
1.25 x 10"2
1.68 x 10"3
3.71 x 10"3
3b
4.21 x 10 J
-3b
4.25 x 10 J
-3b
4.21 x 10 J
_3b
4.10 x 10 *
-3b
1.60 x 10 J
6.02 x 10"4
9.76 x 10"4
4.69 x 10"5
3.47 x 10"4
1.94 x 10"4
BAT
8.93 x 10"3
4.62 x 10"3
1.92 x 10"3
a
1.25 x 10~2
a
3.88 x 10"3
5.21 x 10"4
1.15 x 10"3
-3b
1.30 x 10 J
-3b
1.32 x 10 J
-3b
1.31 x 10 J
-3b
1.27 x 10 J
4.96 x 10"4
1.87 x 10"4
3.02 x 10"4
1.45 x 10"5
1.08 x 10~4
6.01 x 10"5
Beyond BAT
8.06 x 10"3
4.17 x 10"3
1.73 x 10"3
a
1.13 x 10"2
a
3.50 x 10"3
4.70 x 10"4
1.04 x 10~3
-3b
1.18 x 10 J
-3b
1.19 x 10 J
-3b
1.18 x 10 J
-3b
1.15 x 10 _
4.48 x 10"4
1.68 x 10"4
2.73 x 10"4
1.31 x 10"5
9.72 x 10"5
5.43 x 10"5
C-21
-------
Table C-3. ESTIMATED ANNUAL LEUKEMIA INCIDENCE (Continued)
Plant
Number
Region III
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Region IV
34
35
36
37
38
39
40
41
Baseline
(concluded)
1.46 x 10"3
2.30 x 10"2
a
4.55 x 10"5
5.03 x 10"3
4.15 x 10"3
3.74 x 10"3
2.33 x 10"3
7.92 x 10"4
c
2.96 x 10"4
2.59 x 10"4
4.38 x 10"3
2.76 x 10"4
1.62 x 10"3
4.26 x 10"4
2.57 x 10~3
3.58 x 10"5
1.06 x 10"4
6.75 x 10"5
1.96 x 10"4
2.98 x 10~3
BAT
4.53 x 10"4
7.13 x 10"3
a
1.41 x 10"5
1.56 x 10"3
1.29 x 10"3
1.16 x 10"3
7.22 x 10"4
2.46 x 10"4
c
9.18 x 10"5
8.03 x 10"5
1.36 x 10"3
8.56 x 10"5
5.02 x 10"4
1.32 x 10"4
7.97 x 10"4
1.11 x 10"5
3.29 x 10"5
2.09 x 10"5
6.08 x 10"5
9.24 x 10"4
Beyond BAT
4.09 x 10"4
6.44 x 10"3
a
1.27 x 10"5
1.41 x 10"3
1.16 x 10"3
1.05 x 10"3
6.52 x 10"4
2.22 x 10"4
c
8.29 x 10"5
7.25 x 10"5
1.23 x 10"3
7.73 x 10"5
4.54 x 10"4
1.19 x 10"4
7.20 x 10"4
1.00 x 10'5
2.97 x 10"5
1.89 x 10"5
5.49 x 10"5
8.34 x 10"4
C-22
-------
Table C-3. ESTIMATED ANNUAL LEUKEMIA INCIDENCE (Continued)
Plant
Number
Region V
42
43
44
45
46
47
48
49
50
51
52
53
Region VI
54
55
56
57
58
59
60
61
62
Baseline
5.67 x 10"3
2.53 x 10"4
c
3.12 x 10"2
3.85 x 10"5
6.04 x 10"4
a
9.19 x 10~4
1.22 x 10"3
3.29 x 10"3
2.78 x 10"3
6.88 x 10"3
4.92 x 10"4
7.98 x 10"5
6.22 x 10"4
6.27 x 10"4
3.10 x 10"4
1.48 x 10"4
5.94 x 10"4
6.98 x 10"3
3.10 x 10"4
BAT
1.76 x 10"3
7.84 x 10~5
c
9.67 x 10"3
1.19 x 10"5
1.87 x 10"4
a
2.85 x 10"4
3.78 x 10"4
1.02 x 10"3
8.62 x 10"4
2.13 x 10"3
1.52 x 10~4
.2.47 x 10"5
1.93 x 10"4
1.94 x 10"4
9.61 x 10"5
4.59 x 10"5
1.84 x 10"4
2.16 x 10"3
9.61 x 10"5
Beyond BAT
1.59 x 10"3
7.08 x 10"5
c
8.74 x 10~3
1.08 x 10"5
1.69 x 10"4
a
2.57 x 10"4
3.42 x 10"4
9.21 x 10"4
7.78 x 10"4
2.15 x 10"3
1.38 x 10"4
2.23 x 10"5
1.74 x 10"4
1.76 x 10"4
8.68 x 10~5
4.14 x 10"5
1.66 x 10"4
1.95 x 10"3
8.68 x 10"5
C-23
-------
Table C-3. ESTIMATED ANNUAL LEUKEMIA INCIDENCE (Continued)
Plant
Number
Region VI
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
Baseline
(continued)
1.15 x 10"4
2.20 x 10"4
4.65 x 10"3
6.26 x 10"4
1.92 x 10"4
7.11 x 10"3
1.03 x 10"3
5.15 x 10"3
1.48 x 10"3
9.83 x 10"3
7.62 x 10"4
1.05 x 10"3
1.57 x 10"3
5.05 x 10"3
1.25 x 10"2
3.24 x 10"4
2.48 x 10"4
9.78 x 10"3
3.74 x 10"3
1.40 x 10"3
3.01 x 10"4
5.74 x 10"3
c
3.55 x 10"4
BAT
3.56 x 10"5
6.82 x 10"5
1.44 x 10"3
1.94 x 10"4
5.95 x 10"5
2.20 x 10"3
3.19 x 10"4
1.60 x 10"3
4.59 x 10"4
3.05 x 10"4
2.36 x 10"4
3.26 x 10"4
4.87 x 10"4
1.56 x 10"3
3.88 x 10"3
1.00 x 10"4
7.69 x 10"5
3.03 x 10"3
1.16 x 10"3
4.34 x 10"4
9.33 x 10"5
1.78 x 10"3
c
1.10 x 10"4
Beyond BAT
3.22 x 10"5
6.16 x 10"5
1.30 x 10"3
1.75 x 10"4
5.38 x 10"5
1.99 x 10"3
2.88 x 10"4
1.44 x 10"3
4.14 x 10"4
2.75 x 10"4
2.13 x 10"4
2.94 x 10"4
4.40 x 10"4
1.41 x 10"3
3.50 x 10"3
9.07 x 10"5
6.94 x 10"5
2.74 x 10"3
1.05 x 10"3
3.92 x 10"4
8.43 x 10"5
1.61 x 10"3
c
9.94 x 10"5
C-24
-------
Table C-3. ESTIMATED ANNUAL LEUKEMIA INCIDENCE (Continued)
Plant
Number
Region VI
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
Baseline
(continued)
2.06 x 10"3
2.99 x 10"4
2.37 x 10"3
2.99 x 10"4
2.20 x 10"4
1.03 x 10"3
7.36 x 1U"4
2.49 x 10"3
9.34 x 10"4
1.03 x 10"3
5.63 x 10"3
4,59 x 10~4
2.10 x 10"3
3.19 x 10"4
4.66 x Id"3
6.44 x 10"5
2.30 x 10"3
2.06 x 10"3
1.99 x 10'3
2.10 x 10"4
5.53 x 10"3
1.05 x 10"3
2.83 x 10"4
BAT
6.39 x 10"4
9.27 x 10"5
7.35 x 10"4
9.27 x 10"5
6.82 x 10"5
3.19 x 10"4
2.28 x 10"4
, 7.72 x 10"4
2.90 x 10"4
3.19 x 10"4
1.74 x 10"3
1.42 x 10"4
6.51 x 10"4
9.89 x 10"5
1.44 x 10"3
2.00 x 10"5
7.13 x 10"4
6.39 x 10"4
6.17 x 10"4
6.51 x 10~5
1.71 x 10"3
3.26 x 10"4
8.77 x 10"5
Beyond BAT
5.77 x 10"4
8.37 x 10"5
. 6.64 x 10"4
8.37 x 10"5
6.16 x 10"5
2.88 x 10"4
2.06 x 10"4
6.97 x 10"4
2.62 x 10"4
2.88 x 10"4
1.58 x 10"3
1.28 x 10"4
5.88 x 10"4
8.93 x 10"5
1.30 x 10"3
1.80 x 10"5
6.44 x 10"4
5.77 x 10"4
5.57 x 10"4
5.88 x 10"5
1.55 x 10"3
2.94 x 10"4
7.92 x 10"5
C-25
-------
Table C-3. ESTIMATED ANNUAL LEUKEMIA INCIDENCE (Continued)
Plant
Number
Region
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
Region
125
126
127
Regi on
128
129
Baseline
VI (concluded)
2.26 x 10~4
1.99 x 10"3
2.97 x 10~4
1.40 x 10"3
6.92 x 10"4
4.69 x 10"3
1.50 x 10"3
2.54 x 10"5
2.87 x 10"3
2.40 x 10"3
2.40 x 10"3
5.71 x 10~4
6.61 x 10"5
5.41 x 10~4
2.93 x 10"3
VII
1.24 x 10"4
6.74 x 10"4
4.10 x 10"3
IX
1.34 x 10"2
3.67 x 10"3
BAT
7.01 x 10"5
6.17 x 10"4
9.21 x 10"5
'. 4.34 x 10"4
2.14 x 10"4
1.45 x 10"3
4.65 x 10"4
7.87 x 10"6
8.89 x 10~4
7.44 x 10"4
7.44 x 10"4
1.77 x 10"4
2.05 x 10"5
1.68 x 10"4
9.08 x 10"4
3.84 x 10"5
2.09 x 10"4
1.27 x 10"3
4.15 x 10"3
1.14 x 10"3
Beyond BAT
6.33 x 10~5
5.57 x 10"4
8.32 x 10~5
3.92 x 10"4
1.94 x 10"4
1.31 x 10"3
4.20 x 10~4
7.11 x 10"6
8.04 x 10"4
6.72 x 10"4
6.72 x 10"4
1.60 x 10"4
1.85 x 10"5
1.51 x 10"4
8.20 x 10"4
3.47 x 10"5
1.89 x 10"4
1.15 x 10"3
3.75 x 10"3
1.03 x 10"3
C-26
-------
Table C-3. ESTIMATED ANNUAL LEUKEMIA INCIDENCE (Concluded)
Plant
Number
Region IX
13U
131
132
133
134
135
TOTAL
Baseline
(concluded)
2.05 x 10"2
1.35 x 10"2
5.06 x 10"3
1.70 x 10"2
1.35 x 10"3
4.09 x 10"4
4.50 x 10"1
BAT
6.36 x 10~3
4.18 x 10"3
1.57 x 10"3
5.27 x 10"3
4.18 x 10"4
1.27 x 10"4
1.40 x 10"1
Beyond BAT
5.74 x 10"3
3.78 x 10"3
1.42 x 10"3
4.76 x 10"3
3.78 x 10"4
1.14 x 10"4
1.26 x 10"1
Plant no longer produces or uses maleic anhydride; therefore, it is
no longer included in the analysis.
See footnote c, Table C-2.
Plant does not use benzene in production of maleic anhydride; therefore,
it is no longer included in the analysis.
C-27
-------
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 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. The best model to use
(ISC-LT) 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. For example, an analysis
shows that using the more complex model used to estimate ambient benzene
concentrations for fugitive emission sources (equipment leaks) would
increase the estimated leukemia incidences for these sources by about
100 to 200 percent (Docket Item IV-B-18). 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. Assuming the inputs
to the dispersion model are accurate, the predicted benzene concentrations
are considered to be accurate to within a factor of 2.3
C.6.2 Exposed Populations
Several simplifying assumptions were made with respect to the
assumed exposed population. The number of people was assumed to
remain constant over time. In addition, those that are exposed are
assumed to remain at the same location 24 hours per day, 365 days per
year, for a lifetime (70 years) and are assumed to be exposed to a
constant benzene concentration over time. This assumption is more
likely to be valid for the children and the aged, but is not charac-
teristic of the general population. The assumption that exposed
populations remain at the same location is counterbalanced to some extent
(at least in the calculation of incidence) by the assumption that no one
C-28
-------
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 concen-
tration 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 uncertainties associated with the
occupational studies of Infante, Aksoy, and Ott and the variations
in the dose/response relationships among the studies.3 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.
These widely diverse population segments may have susceptibilities to
leukemia that differ from those of workers in the studies. 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).
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 as a permanent resident or as a transient.
Furthermore, while incidence of 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
C-29
-------
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 whether they
occur due to ambient benzene exposure levels. Additionally, benefits
to the general population as the result of indirect control of other
organic emissions in the process of controlling benzene emissions
are not quantified. Possible benzene exposures from other sources also
are not included in the estimate. For example, an individual living
near a fugitive emission source of benzene 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.7 COMPARISON OF THE HEM DISPERSION MODEL SUBPROGRAM AND ISC-LT
DISPERSION MODEL
An analysis was performed using the Industrial Source Complex - >
Long Term Dispersion Model (ISC-LT) to evaluate the effect of using the
dispersion model subprogram of the HEM.4 The ISC-LT is considered to be
a more complex and broadly applicable dispersion model than the dispersion
model subprogram of the HEM, and thus serves as a good basis for comparison.
In general, the main difference between the two models is that in the
ISC-LT model, the benzene sources were modeled as area sources with an
above-grade release height; the HEM dispersion model characterizes the
sources as point sources with a ground level release height.
C.7.1 Methodology
The modeling analysis was performed in three separate steps.
First, nine representative plants were chosen out of the approximately
130 existing plants that contain fugitive emission sources of benzene.
Second, maximum concentrations and total exposure were calculated
using the HEM with its own dispersion model subprogram. Finally, the
ISC-LT was used to recalculate fugitive benzene concentrations surrounding
each of the nine plants. These concentrations were then used as input
into the HEM population exposure subprogram, bypassing the dispersion
model subprogram. Maximum concentrations to which persons were exposed
and total exposure were then calculated by the HEM using the ISC-LT
concentrations as input.
C-30
-------
C.7.2 Results
Table C-4 shows the maximum annual average concentration of
benzene to which persons were exposed, as calculated by the two
dispersion models, for each of the nine plants modeled. The comparison
indicates that, in general, there are differences between the concen-
trations calculated by the HEM dispersion model subprogram and the
concentrations predicted by the HEM with the ISC-LT inputs. In
comparison to the ISC-LT results, the HEM maximum concentrations were
smaller at six of the nine plants. Table C-5 shows the total exposure
that was calculated for each of the nine plants. In all cases (i.e.,
nine out of nine), the results of the HEM using the HEM dispersion
model subprogram predict lower total exposure than the results of the
HEM using the ISC-LT input.
C.7.3 Discussion
The differences in results between the two models can be attributed
partly to the differences in the handling of point versus area source
dispersion of benzene fugitive emissions. The ISC-LT dispersion model
places individual process units (characterized as area sources) discrimi-
nately throughout each plant. In contrast, the HEM dispersion model
subprogram places all the process units at one point source, usually
near the center of the plant. By modeling the process units individually,
sources may be located closer to the .plant boundary, and, therefore,
people may be exposed to higher concentrations. The initial dispersion
in the horizontal direction associated with the ISC-LT area sources
could also have combined with source location to result in slightly
higher exposure to the predicted benzene concentrations.
C.7.4 Conclusions
The results of the analysis indicate that total exposures for all
nine plants are higher using the more sophisticated dispersion model
(ISC-LT) to predict pollutant concentrations for input into the HEM
than by using the HEM with its own dispersion model subprogram.
Neither the ISC-LT nor the HEM dispersion model subprogram predicts
higher maximum concentrations consistently due to plant-specific
criteria (i.e., source locations as compared to population densities).
The differences between the two model predictions are significant.
However, because of the degree of uncertainty in dispersion analysis
C-31
-------
Table C-4. COMPARISON OF MAXIMUM ANNUAL AVERAGE BENZENE
CONCENTRATION OF EXPOSED PERSONS USING HEMJHSPERSION
MODEL AND HEM WITH ISC-LT INPUT^
Plant
Number
2
21
26
30
45
60
83
119
131
Maximum Concentration
HEM Only
5.00 x 10}
6.39 x 10
2.68 x 101
4.07 x 10"1
1.67 x 102
5.12 x 101
1.00 x 10°
1.77 x 102
8.71 x 101
(wg/m3)
HEfl with
ISC-LT Input
1.00 x 102a
5.10 x 101
4.10 x 101
7.12 x 10"1
2.82 x 102
8.03 x 101
5.00 x 10°
0.99 x 102a
2.00 x 102
aThis plant was run with the ISC-LT in the urban node, which tends
to predict lower concentrations than the rural mode option.
Table C-5. COMPARISON OF TOTAL EXPOSURE TO BENZENE USING
HEM DISPERSION AND HEM WITH ISC-LT INPUT0
Total Exposure ( Persons -jug/m3)
Plant
Number
2
21
26
30
45
60
83
119
131
HEM Only
1.51 x 105
2.32 x 105
0.38 x 105
2.99 x 103
3.15 x 105
0.60 x 104
3.04 x 103
2.43 x 104
1.36 x 105
HEM with
ISC-LT Input
3.63 x 105
3.00 x 105
1.05 x 105
5.78 x 103
7.68 x 105
1.89 x 104
8.39 x 103
3.68 x 104
1.46 x 105
aThis plant was run with the ISC-LT in the urban mode, which tends
to predict lower concentrations than the rural mode option.
C-32
-------
and the degree of effort required to model all of the plants specifically,
the use of a more accurate dispersion model (ISC-LT) rather than the
one present in the HEM is not warranted.
C-33
-------
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
(Docket Number IV—A-ll), and Volume II, Publication Number
EPA-2/250-2 (Docket Number IV-A-12), May 1980.
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. Docket Number IV-A-8.
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 II-A-28. January 1979.
Memorandum from Eldridge, K., Pacific Environmental Services, Inc. to
Dimmick, W.F., U.S. Environmental Protection Agency. May 20, 1983.
Methodology and Results of the Comparison of the HEM Dispersion Model
Subprogram and the ISC-LT Dispersion Model. Docket Nubmer IV-B-18,
Table 5.
5. Reference 4, Table 6.
n
3.
4.
C-34
-------
APPENDIX D
STANDARDS-SETTING APPROACH
D-l
-------
APPENDIX D
STANDARDS-SETTING APPROACH
D.I INTRODUCTION
Several individuals and organizations commenting on the proposed
benzene fugitive emissions standard suggested changes in the standard-
setting approach used by EPA at that time. These comments and EPA's
responses are discussed below.
The comments are divided into three major subjects: (1) the
role of quantitative risk estimation in the regulatory process; (2) the
requirement of best available technology (BAT) as the minimum level of
control for significant source categories selected for regulation, and
(3) the determination of the appropriate level of control for nonthreshold
pollutants under Section 112.
D.2 QUANTITATIVE RISK ESTIMATION IN THE REGULATORY PROCESS
Commenters were significantly divided on the general issues of the
accuracy and reliability of quantitative risk assessment (QRA) for
carcinogens, as well as on the utility of such estimates in the regulatory
decision making process. QRA proponents, primarily industry and trade
association commenters, held that such assessments, when based on
reliable data, should play an important role at all decision making
stages on potential airborne carcinogens. Commenters largely from
public interest groups, State air pollution control agencies, and
private individuals expressed concern that the underlying uncertainties
in attempts to quantify cancer risks greatly reduce the reliability of
such estimates and argue for limiting or avoiding their use in the
regulatory process.
EPA has considered the comments on the role of quantitative risk
assessment in the regulatory decision making process for the benzene
fugitive emissions standard and concludes that QRA can provide meaningful
information at certain points in the decision making process. However,
the importance ascribed to such information must be in proportion to its
reliability. The utility of a particular estimate in the decision
D-2
-------
making process will depend on the magnitude of the estimated uncertainty.
In cases for which data are inadequate to permit estimation of uncertainty,
EPA will endeavor to assess qualitatively the reliability of the
calculated risks and consider them accordingly in the decision making
process. Where risk estimation is feasible and some measure of confidence
is obtainable, EPA will perform quantitative assessments for use in the
appropriate stages of the regulatory process. In the case of benzene
fugitive emissions, EPA used QRA in determining that fugitive emission
sources pose significant health risks to the exposed public, and in judging
that the health risks after application of the level of control represented
by BAT are not unreasonable in light of the cost of a more stringent control
level.
D.3 SELECTION OF BAT AS THE MINIMUM LEVEL OF CONTROL
Although most commenters endorsed EPA's proposed procedure for
consideration of economic and technological feasibility in the development
of emissions standards under Section 112, the minimum requirement of
BAT often was criticized as unnecessarily rigid and not reflective of
Congressional intent. A number of commenters argued that a control
level less stringent than BAT could be appropriate where the health
risks are low. Several commenters argued that the imposition of BAT
could result in excessive control, with no statutory support and no
evidence that the avoidable risks were unreasonable.
Some commenters argued that Section 112 contained the presumption
of a zero-emission standard, and contended that an evaluation of "residual
risk" does not provide the mandated "ample margin of safety." The
commenters stated further that EPA is not authorized to consider
technical and economic factors in setting emission standards.
Several commenters suggested that health risks should be considered
in establishing BAT or that cost benefit analysis should be performed.
Two commenters suggested as alternatives to BAT "reasonably available
control technology" and "best available retrofit technology." Finally,
one commenter recommended that EPA determine acceptable risk levels in
place of a BAT minimum control level.
Most carcinogens (such as benzene) seem to present finite risks at
any level of exposure, risks that increase as the level of exposure
increases. Standards for these pollutants could not eliminate all risk
D-3
-------
unless the standard prevents any exposure, which would in turn require
preventing any emissions. It does not appear that Congress intended
Section 112 standards to cause widespread shutdown of industries emitting
benzene or other nonthreshold pollutants. Therefore, as an alternative
to widespread shutdown of industries, EPA must establish emission
standards for benzene-emitting sources that pose significant risk at
levels that may still present some human health risk.
Another issue that has been encountered in using health risk
estimates is whether protection should focus on the risk to the most
exposed individuals or to the exposed population as a whole (cancer
incidence). Even when the many uncertainties in health risk estimates
are considered, results to date indicate that the total cancer incidence
associated with exposure to benzene, even on a nationwide basis, is
likely to be small compared to the incidence associated with factors
such as smoking and diet. However, individual risks for a limited
number of people living close to uncontrolled or partially controlled
emission sources may be relatively high.
Neither the language nor the legislative history of Section 112
reveals any specific Congressional intent on how to deal with these
issues and how to apply the phrase "provides an ample margin of safety
to protect the public health" to benzene or other nonthreshold pollutants
that present cancer risks at any level of exposure.
In view of this, EPA has applied the following interpretation of
Section 112 to the regulation of benzene fugitive emission sources.
Each source should be controlled at least to the level that reflects
best available technology (BAT), and to a more stringent level if, in
the judgment of the Administrator, it is necessary to prevent unreasonable
health risks. In determining BAT for benzene fugitive emission sources,
the Agency first identified alternative control levels that have been
demonstrated. These control levels may have actually been achieved by
representative plants containing fugitive emission sources of benzene
or can be based on technology transfer from other source categories.
The Agency determined the cost and associated impacts of the various
alternatives. The Administrator selected as BAT that alternative that
D-4
-------
achieved the most emissions reduction (and therefore risk reduction)
without having what he judges to be unreasonable impacts. Retrofit
costs were considered since BAT was determined for existing fugitive
emission sources. A particular regulatory alternative may be rejected
as BAT for a variety of reasons, among them that (1) it is judged to
result in a price increase that affects consumers adversely to an
unreasonable extent; (2) it will result in plant closures or unreasonably
discourage construction of new plants due to reduced return on investment
or capital inavailability; or (3) it has unreasonably high costs for
the amount of emissions reduction achieved. The level of control that
represents BAT may be different for new and existing sources of benzene
because of higher costs associated with retrofitting controls on existing
sources, or differences in control technology for new versus existing
sources. Whether the estimated health risks remaining after application
of BAT are unreasonable is decided in light of a judgmental evaluation
of the estimated maximum lifetime risk and cancer incidence remaining
after application of BAT, the impacts, including economic impacts, of
further reducing those risks, the readily available benefits of the
substance or activity producing the risk, and the availability of
substitutes and possible health effects resulting from their use. In
all cases where estimated health risks are used, the significant
uncertainties associated with those numbers were weighed carefully in
reaching the final decision.
In EPA's judgment, benzene fugitive emission standards based on
this interpretation of Section 112 provide an ample margin of safety to
protect the public health.
In summary, BAT represents a level of technology that reduces
emissions to the greatest extent possible, given cost, energy,
environmental, and technological factors. As such, it is a reasonable
level of control that may, based on an evaluation of the health risks
remaining after its application and the costs of further reducing those
risks, provide an "ample margin of safety" to protect public health.
In view of the carcinogenic nature of benzene, which is regulated under
Section 112, application of BAT to fugitive emission sources of benzene
D-5
-------
posing signficant health risks, therefore, imposes the least stringent
control level possible that is judged to protect public health with an
ample margin of safety (assuming, of course, that the residual risk
remaining after application of BAT is not unreasonable).
In response to the commenter who recommended that EPA determine
acceptable risk levels in setting standards rather than require BAT as
a minimum, EPA continues to believe that the adoption of acceptable
specific risk levels or cost per life is inappropriate. Even for the
residual risk analysis, because of the shortcomings of quantitative
risk assessments, and the inability of cost-per-life calculations to
deal with unacceptably high risks to smaller groups, EPA does not
believe that cost-per-life calculations are appropriate as the principal
means of assessing residual risks.
D-6
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-450/3-80-032b
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Benzene Fugitive Emissions
for Promulgated Standards
- Background Information
5. REPORT DATE
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO
9. PERFORMING ORGANIZATION NAME AND ADDRESS
U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Emission Standards and Engineering Division
Research Triangle Park, North Carolina 27711
1O. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-02-3060 77-5J
12. SPONSORING AGENCY NAME AND ADDRESS
13. TYPE OF REPORT AND PERIOD COVERED
Director for Air Quality Planning and Standards
Office of Air, Noise, and Radiation
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
14. SPONSORING AGENCY CODE
EPA/200/04
1 s. SUPPLEMENTARY NOTES thi s document presents the background information used by the
Environmental Protection Agency in developing the promulgated national emission
standard for benzene fugitive emissions.
16. ABSTRACT
A national emission standard for the control of fugitive emissions of benzene
from the petroleum refining and organic chemical manufacturing industries is
being promulgated under Section 112 of the Clean Air Act (42 U.S.C. 7412, as
amended). This standard will limit fugitive emissions of benzene from existing
and new equipment in benzene service. This document summarizes the responses
to public comments received on the proposed standards and also summarizes
the basis for changes made in the standards since, proposal.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Air Pollution
Petroleum Refining
Organic Chemical Industry
Performance Standards
Air Pollution Control
13b
OnlimitecL Available to the public free
of charge from: U.S. EPA Library
(MD-35), Research Triangle Park, NC 27711
19. SECURITY CLASS (This Report)
Unlimited
21. NO. OF PAGES
225
20. SECURITY CLASS fThispagei
SECURITY CLA
Unlimited
22. PRICE
EPA Form 2220—1 (Rev. 4—77) PREVIOUS EDITION is OBSOLETE
-------
------- |