United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
March 1980
Air
Benzene Fugitive Draft
Emissions — EIS
Background Information
for Proposed Standards
Preliminary
Draft
-------
NOTICE
This document has not been formally released by EPA and should not now be construed to represent
Agency policy. It is being circulated for comment on its technical accuracy and policy implications.
Benzene Fugitive Emissions
Background Information
for Proposed 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
March 1980
-------
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
(MD-35), U.S. Environmental Protection Agency, Research Triangle Park,
N.C. 27711, or from National Technical Information Services, 5285 Port
Royal Road, Springfield, Virginia 22161.
PUBLICATION NO.
EPA-450/3-78-
-------
DRAFT
Background Information Document
Benzene Fugitive Emissions
Type of Action:
Administrative
Prepared by
Date
Don R. Goodwin, Director
Emission Standards and Engineering Division
Environmental Protection Agency
Research Triangle Park, North Carolina 27711
Approved by
Date
David G. Hawkins
Assistant Administrator
Office of Air, Noise, and Radiation
Environmental Protection Agency
401 M Street, S.W.
Washington, D.C. 20460
Draft Document Submitted To EPA's Office
Of Federal Activities For Review On
Date
I '
Additional copies may be obtained or reviewed at:
Environmental Protection Agency (PM-213)
Library (MD-35)
Research Trinagle Park, North Carolina 27711
Public Information Reference Unit
Room 2922
401 M Street, S.W.
Washington, D.C. 20460
-------
TABLE OF CONTENTS
List of Tables.
.......
. . . .
........
Li st of Figures
. . . . . . .
.....
.....
Chapter 2.
Chapter 3.
Introduction
.............
Sources of Benzene Fugitive Emissions in
Petroleum Refineries and Organic Chemical
Manufacturing Operations. . . . . . . .
3.1 Introduction...........
.....
3.2 Sources of Benzene Emissions. .
......
3.2.1 Potential leak Sources
.....
3.2.2 Other Potential Sources. . . . . .
3.3 Magnitude of Benzene Emissions from
Refineries and SOCMI Production ' Operations .
..;
3.4 References. . . .
.........
. . . .
Chapter 4.
Emission Control Techniques
......
4.1 Introduction. . .
. . . . . . . . . . .
4.2 Monitoring and Maintenance Programs
4.2.1 Definition of a leak. .
. . . .
......
......
. . . . .
. . . .
. . . .
......
.....
. . . .
. . 0 .
.....
......
........
. . . .
4.2.2 Inspection Interval.
.....
.......
4.2.3 Allowable Repair Time. . .
4.2.4 Visual Insp~ctions
......
.........
......
.......
4.2.5 Other Monitoring Techniques
.....
4.2.6 Maintenance. . . . . . . . .
. . . .
........
4.2.7 Emission Control Effectiveness of
Monitoring and Maintenance. . .
i
......
Paqe
vi i
xii
2-1
3-1
3-1
3-1
3-2
3-8
3-11
3-15
4-1
4-1
4-2
4-2
4-4
4-6
4-6
4-6
4-9
4-10
-------
TABLE OF CONTENTS (Continued)
4.3 Preventive Programs
. . . ..
. . . . . . . .
. . . .
4.3.1 Pumps. . .
4.3.2 Valves
. . . .
.......
......
. . . .
........
. . . . . . .
4.3.3 Safety/Relief Valves
. . . .
........
4.3.4 Open-Ended Valves. . . . . .
4.3.5 Closed-Loop Sampling
. '. . .
. . . .
.......
.....
4.3.6 Accumulator Vessel Vents and Seal
Oil Degassing System Vents. . . . . . . . .
4.4 Process Modifications
........
. . . . . . .
4.5 References. . . ~ . . . .
.....
....~...
Chapter 5. Modification and Reconstruction
.....
. . . .
5.1 General Discussion of Modification and
Reconstruction Provisions. . . . . .
Page
4-13
4-13
4-15
4-18
4-23
4-23
4-25 .
4-26
4-27
5-1
.....
. . . 5-1
5-1
5.1.1 Modification
. . . .
. . . . . . . .
. . . .
5.1.2 Reconstruction
......
. . . . . . . . .
5.2 Applicability of Modification and
. Reconstruction Provisions. . . . . .
.......
5.2.1 Modification
......
......
. . . .
5.2.2 Reconstruction
. . . .
. . . .
.......
Chapter 6. Model Units and Regulatory Alternatives
6.1 Introduction. . . . . . . .
.....
......
. . . . . .
6.2 Model Unit Parameters
. . . .
.......
. . . .
ii
5-2
5-3
5-3
5-3
6-1
6-1
6-1
-------
TABLE OF CONTENTS (Continued)
6.3 Regulatory Alternatives
......
........
6.3.1 Regulatory Alternative I
6.3.2 Regulatory Alternative II
..........
..........
6.3.3 Regulatory Alternative III
6.3.4 Regulatory Alternative IV .
.........
. . . . . . . . .
6.3.5 Regulatory Alternative V
6.3.6 Regulatory Alternative VI
......
..........
. . . .
6.4 References. . . . . . . .
. . . .
.........
Chapter 7. Environmental Impact. .
7.1 Introduction. . . . . . .
. . . .
.....
........
7.2 Air Quality Impacts
........
........
7.2.1 Development of Benzene Emission Levels
7.2.2 Future Benzene Emissions. .
. . . .
.....
. . .
........
7.3 Water Pollution Impact. . . . .
7.4 Solid Waste Impact
......
.................
7.5 Energy Impact
. . . .
...........
........
7.6 Other Environmental Concerns. .
. . . . . .
7.6.1 Irreversible and Irretrievable
Commitment of Resources. . . .
. . . .
. . . . . . .
7.6.2 Environmental Impact of Delayed
Regulatory Action. . . . . . . . . .
7.7 References. . . . . . . . . . . . . .
. . . .
.......
i i i
Page
6-3
6-3
6-6
6-7
6-7
6-8
6-8
6-9
7-1
7-1
7-1
7-1
7-9
7-14
7-18
7-19
7-19
7-19
7-21
7-22
-------
Chapter 8.
TABLE OF CONTENTS (Continued)
Cost of Controls
.....
.......
.....
8.1 Introduction. . ~ . . .
.....
.........
8.2 Capital Cost Estimates.
......
........
8.3 Annualized Cost Estimates
'. . . . . .
. . . . . . .
8.3.1 Derivation of Annualized Cost Estimates. . .
8.3.2 Cost-Effectiveness
......
.......
8.4 Cost Comparison
. . . .
.......
. . . . . . .
8.5 References. . . . .
Chapter 9.
9.1
. . . .
. . . .
. . . . . . . .
Economic Impact
.....
. . . .
........
Industry Characterization
.....
. . . . . . . .
9.1.1 General Profile. . .
.....
. . . . . . .
9.1.2
Production of Benzene, Ethylene,
and Benzene Derivatives. . . . . . . . . . .
9.1.3 Methods of Manufacture. . . . .
. . . . . .
9.1.4 Uses of Benzene. .
......
. . . . . . .
9..1. 5 Price Hi story. . .
.....
. . . . . . . .
9.1.6 Market Factors that Affect the
Benzene Industry. . . . . . . . . . . . . .
9.1.7 Feedstock Substitutions for
Benzene Derivatives. . . . .
. . . .
. . . .
9.1.8 Future Trends. .
......
........
9.2 Microeconomic Impact
. . . .
........
. . . .
9.2.1
9.2.2
Introduction
. . . .
.....
. . . . . . .
Industry Structure
......
. . . . . . .
iv
Page.
8-1
8-1
8-1
8-7
8-7
8-21
8-37
8-42
9-1
9-1
9-1
9-1
9-11
9-16
9-25
9-25
. 9-25
9-27
9-30
9-30
9-30
-------
TABLE OF CONTENTS (Continued)
9.2.3 Demand Characteristics
.....0
.....
9.2.4 Supply Characteristics
000..
......
9.2.5 Economic Impact Methodology.
........
9.2.6 Model Plant Impact Analysis.
........
9.3 Macroeconomic Impact.
00.............
9.3.1 Summary. . . .
0........
o . . . . .
9.3.2 Inflationary Impacts
9.3.3 Energy Impacts
....00......
0....
. 0 0 .
......
9.3.4 Employment Impacts
0......
......
9.3.5 Fifth Year Annualized Costs
o 0 . .
.....
9.4 References.
. 0 0 0
000........
0000.
Appendix C. Emission Source Test Data
C.1 Introduction. .
0000....
. . . .
. . . .
....000.......
C.2 Data Summaries
000...
........0
o 0 0 "
C.2.1 Refinery Valve Repair Data
.........
C.2.2 Phillips Petroleum Company Data
.......
C.2.3 Shell Oil Company Data
. 0 . 0
.......
C.2.4 Union Oil Company, San Francisco
Refi nery Data- . . . . . . . . . . . . . . . .
C.3 References
0.0.0..
. 0 . .
.0.00000..
Appendix D. Emission Measurement and Continuous Monitoring. .
0.1 Emission Measurement Methods. . . . . . .
.....
D.2- Continuous Monitoring Systems and Devices
...00
v
---------/"
Page
9-34
9-39
9-44
9-48
9-55
9-55
9-55
9-57
9-57
9-57
9-58
C-1
C-1
C-1
C-1
C-2
C-2
C-2
C-17
0-1
D-1
D-4
-------
TABLE OF CONTENTS (Concluded)
0.3 Performance Test Method
. . . . . . .
0.4 References. . . . . . . . . .
.....
vi
. . . . . . .
......
Page
0-4
0-6
-------
LI ST OF TABLES
3-1 Estimated Vapor Emission Factors for Nonmethane
Hydrocarbons from Refinery and SOCMI Sources 0 0 0 0 0 0
3-2 Estimated Benzene Emissions from an Average Plant
. . . .
4-1 Percentage of Sources Predicted to be Leaking
in an Individual Component Survey 0 0 0 00 0
......
4-2 Percent of Total Mass Emissions Affected at
Various Action Levels 0 0 0 0 0 0 0 0 0 0 0
.......
4-3 Emission Correction Factors for Various Inspection
Intervals, Allowable Repair Times and Action Levels 0 0 0
6-1 Model Unit Equipment with >10 Percent Benzene 0 0
. . . .
6-2 Monitoring Intervals and Equipment Specifications
for Benzene Fugitive Regulatory Alternatives 0 0 0 0 0 0
7-1 Controll ed VOC Emission Factors for Regulatory
Al ternative II 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
7-2 Controlled VOC Emission Factors for Regu 1 a tory
Alternative III 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
7-3 Can troll ed VOC Emission Factors for Regul atory
Alternative IV 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
7-4 Controlled VOC Emission Factors for Regulatory
Alternative V 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
7-5 Calculation of Weighted Percent Benzene for
over Ten Percent Equipment in Model Units 0
.......
7-6 Total Benzene Emissions and Relative Control
Effectiveness for the Regulatory Alternatives
and Model Units 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
......
7-7 Total National Benzene Emissions from Refinery
and SOCMI Processes 0 0 0 0 0 0 0 0 0 0 0 0 0 0
.....
7-8 Numbers of Units Estimated to Meet 1980 Demand
for Benzene and Benzene Derivatives by Model Units
vii
Page
3-13
3-14
4-3
4-5
4-12
6-2
6-4
7-2
7-3
7-4
7-5
7-7
7-10
7-11
7-13
-------
LIST OF TABLES (Continued)
7-9 Number of Projected New Units and Replacements
between 1981 and 1990 . . . . . . . . . . . .
......
7-10 Estimated Benzene Fugitive Emissions from New
Units and Replacements Between 1981 and 1990
7-11 Energy Impact of Benzene Emission Reduction
for Regulatory Alternatives. . . . . . . . .
......
......
8-1 Model Unit Equip~ent with >10 Percent Benzene.
.....
8-2 Monitoring Intervals and Equipment Specifications
For Benzene Fugitive Regulatory Alternatives. . . . . .
8-3 Capi ta 1
Cost Data. . . . . . . . . . . . .
. . . . . . .
8-4 Capital Cost Estimates per Model Unit. . . .
......
8-5 Monitoring and Maintenance Labor-Hour Requirements
for Regulatory Alternative II . . . . . . . . . . . . . .
8-6 Monitoring and Maintenance Labor-Hour Requirements
for Regulatory Alternative III .............
8-7 Monitoring and Maintenance Labor-Hour Requirements
for Regulatory Alternative IV . . . . . . . . . . . . . .
8-8 Monitoring and Maintenance Labor-Hour Requirements
for Regulatory Alternative V .. . . . . . . . . . . . .
8-9 Recovered Product Credits. . . . . . . . . . .
.....
8-10 Initial Survey Start-Up Costs for
Regulatory Alternative II . . . . . . . . .
. . . . . . .
8-11 Annualized Control Cost Estimates per Model Unit
8-12 Benzene Emission Reductions. . .
. . . .
. . . . .
. . . . . . .
8-13 Cost-Effectiveness for Model Units (Existing Units) . . .
8-14 Cost-Effectiveness for Model Units (New Units)
. . . . .
viii
Page
7-15
7-17
7-20
8-2
8-3
8-5
8-8
8-16
8-17
8-18
8-19
8-20
8-22
8-26
8-34
8-35
8-36
-------
LIST OF TABLES (Continued)
8-15 Range of Control Costs for the Benzene Source
Categories for Existing and New Units. . . . . .
8-16 Costs for the Control of Total Benzene Emissions
from the Hael ic Anhydride Industry. . . . . . .
8-17 Costs for the Control of Total Benzene Emissions
from the Ethyl benzene-Styrene. Industry. . . . .
8-18 Total Costs for the Control of Benzene Emissions
from Producer Benzene Storage Tanks and Benzene
Fugitive Sources. . . . . . . . . . . . . . . .
8-19 Total Costs for the Control of Benzene Emissions
from Consumer Benzene Storage Tanks and Benzene
Fugitive Sources. . . . . . . . . . . . . . . .
9-1 Refineries and Synthetic Organic Chemical
Manufacturing Sites with Benzene Fugitive
Emission Potential. . . . . . . . . . .
Page
. . . . 8-38
. . . . 8-39
. . . . 8-39
. . . .
. . . .
........
9-2 Number of Companies and Plant Sites that
. Manufacture Benzene Derivatives. . . . . .
.......
9-3 Summary of Production and Capacity for
Benzene, Ethylene, and Benzene Derivatives
.......
9-4 Ethylene Usage
. . . 0
........
.........
9-5 Monochlorobenzene Usage.
9-6 Dichlorobenzenes Usage
................
9-7 Nitrobenzene Usage
.....
.................
....00........
9-8 Aniline Usage. . . .
.......
. . . .
. . . . . . .
9-9 Ethylbenzene Usage
9-10 Styrene Usage. . .
.........
.....
. . . . . . . .
.....
.....
...........
.....
9-11 Linear Alkybenzene Usage
......
9-12 Cyclohexane Usage. . . . .
........
.......
ix
8-41
8-41
9-2
9-11
9-12
9-18
9-19
9-19
9-20
9-20
9-21
9-21
9-22
9-22
-------
LIST OF TABLES (Continued)
9-13 Cumene Usage 0 . 0 0 . . . .
9-14 Maleic Anhydride Usage
. . 0 .
..'...
.....
o . . 0
. . . 0
9-15 Resorcinol Usage
. . 0 . .
.....
0000.0
9-16 Benzenesulfonic Acid Usage
.....
. . . .
9-17 Hydroquinone Usage
.....
.....
. . . . .
9-18 Price History for Benzene, Ethylene, and
Benzene Derivatives. 0 0 . . 0 0 . 0 . 0
. . . . .
. . . .
. . . .
......
. . . .
. . . . . . . .
9-19 Alternative Processes for the Manufacture of
Benzene Derivatives 00 0 ~ . 0'. 0 . . 0 .
. . . . . . .
9-20 Projected Annual Growth Rates for Demand of
Benzene, Ethylene, ~nd 'enzene Derivatives
.......
9-21 Concentration Ratios for Benzene, Ethylene, and
. Benzene Derivatives. . . 0 0 . . 0 . . 0 . 0 . . 0 . 0 0
Page
9-23
9-23
9-24
9-24
9-24
9-26
9-27
9-28
9-32
9-22 Qualitative Evaluation of Price Elasticity
of Demand. 0 . . . 0 0 0 . . 0 0 . 0 0 0 . . . . . . 0 0 9-40
9-23 Model Pl ant Annual Revenues - Model Plant A 0 . 0 0 0 0 0 9-46
9-24 Model Plant Annual Revenues - r~ode 1 Plant B . 0 0 . . 0 0 9-47
9-25 Model Plant Annual Revenues - Model Plant C 0 . . . . . 0 9-47
9-26 Total Capital Investment Required -
New Model Plants. . . 0 . 0 . . . . . . .
9-27 Percentage Price Increases
0......
. . . .
.......
. . . .
9-28 Cumulative Percentage Price Increases. .
9-29 Percentage Increase in New Plant Capital
Investment Required 0 0 . 0 0 . . . . 0 .
. . . .
. . . .
........
C-1 Refinery Valve Repair Data
o . 0 .
. . . ., 0 . .
x
. . . .
9-49
9-52
9-54
9-56
C-4
-------
LIST OF TABLES (Concluded)
C-2 Leak Data for the Phillips Petroleum Company, Sweeny
Refinery and Petrochemical Complex, Sweeny, Texas. . . .
C-3 Phillips Sweeny Refinery Ethylene Unit Block
Valve Repairs. . . . . . . . . . . . . . .
.......
C-4 Summary of Phillips Sweeny Block Valve Leak
and Repair Data. . . .. . . . . . . . . . .
......
C-5 Leak and Repair Data for Refinery Valves from
the Shell Oil Company, Martinez Manufacturing
Complex, Martinez, California. . . . . . . .
......
C-6 Leak and Repair Data for Refinery Valves from
the Union Oil Company San Francisco Refinery,
Rodeo, California. . . . . . . . . . . . . . . .
. . . .
C-7 Attempted Repair Data for Valves from the Union-
San Francisco Refinery. . . . . . . . . . . .
.....
C-8 Effects on Emissions of Repairing Valves in the
1,000 - 10,000 ppm Range. . . . . . . . . . .
. . . . .
xi
Page
C-6
C-8
C-ll
C-12
C-13
C-14
C-15
-------
LIST OF FIGURES
3-1 Simple Packed Seal
. . . .
. . . .
. . . . . . .
. . . .
3-2 Basic Single Mechanical Seal
3-3 Globe Valve with Packed Seal
......
........
0.......
. . . . . .
3-4 Diagram of a Spring-Loaded Relief Valve
.........
3-5 Liquid-Film Compressor Shaft Seal. . . . .
4-~ Double Mechanical Seal
.......
. . . . . . .
.....
.....
4-2 Seal less Canned Motor Pump. .
4-3 Diaphragm Valve. . .
0...0....
. . . .
. . . . . . . .
..........
4-4 Sealed Bellows Valve
. . . . . . .
...........
4-5 Rupture Disk Installation Upstream of a Relief Valve
4-6 Simplified Closed-Vent System with Dual Flares
. . . . .
4-7 Diagram of Two Closed-Loop Sampling Systems. .
. . . . .
9-1 Percentages of Total Benzene Producing Consumed
by Intermediate and Final Products. . . . . .
.....
xii
Page
3-3
3-4
3-6
3-7
3-9
4-14
4-16
4-17
4-19
4-20
4-22
4-24
9-17
-------
2.
INTRODUCTION
The Environmental Protection Agency proposed on October 10, 1979,
"Policies and Procedures for Identifying, Assessing, and Regulating Airborne
Substances Posing a Risk of Cancer" (44 FR 58642). All standards for
carcinogens regulated under section 112 of the Clean Air Act are being.
developed in accordance with those proposed, Policies and Procedures. The
following is a section quoted from the Policies and Procedures which describes
the procedures for establishing standards once the decision has been made
as to which pollutants are to be regulated.
"(2) The Proposed EPA Approach (44 FR 58642)
The standard-setting policy proposed today requires, as a minimum,
the use of "Best Available Technology" (BAT) to controf emissions from source
categories presenting significant risks to public health. The policy-_;would
also require additional controls, as necessary, to eliminate "unreasonable
residual risks" remaining after the use of BAT. This approach is a judgmental
one, designed to protect the public health with an ample margin of safety
from risks associated with exposure to airborn~ carcinogens. The implementing
procedure described below puts prime emphasis on public health, consistent
with section 112, but permits consideration of economic impacts and benefits
of the activity in setting standards for each source category. Uncertainties
in the assessments of risks, costs, and potential benefits, as well as the
distributional (equity) problems of various situations, would also be
considered in setting standards.
(a) Source Categories Regulated
The first step in establishing standards and requirements for pollutants
listed under section lT2 under this proposed policy is the determination of
which categories of sources emitting the pollutants will be regulated, and
in what order regulations will be developed. Although a pollutant may have
been listed because emissions from a particular source category pose a
2-1
-------
I---~------- ~---
significant risk, other source categories may also emit the pollutant in
lesser amounts. This may occur, for example, because the sources process
very little of the substance, because the substance is present in only
trace amounts in the sources' raw materials, or because sources have installed
adequate controls on their own initiative or in response to other regulatory
requirements.
The Administrator will therefore propose regulations only for those
source categories which may pose significant .risks to public health. The
determination of whether a source category emitting a listed pollutant poses
a significant risk will be made on essentially the same basis as the listing
decision, except that the more detailed exposure analysis and risk assessment
then available will be used in lieu of the preliminary information used in
the listing decision. As in the listing decision, the risk assessment will
be used to evaluate the existence of a significant risk where the exposure
analysis alone is insufficient, but will not be used as evidence that a
significant risk does not exist where the exposure analysis .indicates to
the contrary.
(b) Priorities for the Development of Standards - --
EPA anticipates that a substantial number of substances will be listed
as carcinogenic air pollutants under section 112 in the near future. It is
also likely that many of these substances will be emitted in significant
quantities from More than one source category. As a result, EPA will need
to develop emission standards and other requirements for a large number of
source categories emitting these substances. At least until generic standards
can be developed for large groups of these sources, the resources that
would be necessary to complete this task immediately far exceed those
available to EPA for this purpose. Today's proposal therefore provides for
the assignment of priorities to significant source categories for the develop-
ment of these regulations, through publicily stated criteria and announced
decisions.
Under today's proposal, source categories posing significant risks will
be assigned priority status (high, medium, or low) for further regulatory
action (beyond generic standards) based primarily on the resu1ts of the
~ .
quantitative risk assessments described above. Priorities will be assigned
according to the following criteria: (1) the magniture of expected and upper
2-2
-------
.
bound total excess cancer incidence associated with current and future
source 'emissions; (2) magnitude of cancer risks for the most exposed
individuals; (3) ease of expeditious standards development and implementation;
and (4) feasibility of significant improvements in controls. In addition,
significant sources of more than one carcinogen may be given priority over
single-pollutant sources,. based on the sum of risks from the emitted
substances.
A high priority will be assigned, for example, tO,a source category
constituting an important problem requiring immediate attention, or where
risks are somewhat lower but an appropriate regulatory solution is both
feasible and readily available. Source categories assigned medium priority
will generally be those that present lower risks and will be scheduled for
standard development as resources become available. Lower risk source
categories for which the extent of feasible control may be substantially
limited will be assigned low priority for regulation development. Assign-
ment to the low priority category will generally mean that active develop-
ment' of the regulations will not begin until there is some change in the
factors which led to the assignment, or until higher priority actions have
been completed.
(c) Regulatory Options Analysis
EPA will perform detailed analyses to identify alternative, technologically
feasible control options and the economic, energy, and environmental impacts
that would result from their application. Where substitution' is determined
to be a feasible option, the benefits of continued use of the substance
or process will be considered. These analyses will rely primarily on the
procedures and techniques employed by EPA for developing New Source Performance
Standards under section 111 of the Act.
The identification of feasible control options will initially survey
the existjng control devices at the sources within a particular category
to determine the best controls currently in use. The potential emission
points of the listed pollutant at a particular kind of facility will also be
identified, as will possible emissions of carcinogens other than the specific
one under study. EPA will, in addition, examine the applicability of .
available technologies which are not currently used by the industry to con-
trol the pollutant of concern (technology transfer) but which have been
2-3
-------
demonstrated in pilot tests or other industrial applications. Finally, the
availability and adequacy of substitutes Which would eliminate some or all
emissions of the pollutant will be assessed.
Once the technologically feasible control alternatives~ which may range
from no further control to a complete ban on emissions, have been identified,
the environmental, economic and'energy impacts of these options will be
determined. Considerations in these impact assessments ~11 include for
each option: the number of plant closures predicted and the direct impact
on employment and end product prices; the impact on growth and expansion
of the industry; the resulting changes in profitability; capital availability
for control equipment; the impacts from the availability of substitute
products and -foreign imports; the potential increases in national energy
consumption; and the impacts on other environmental media including increased
water pollution and solid waste disposal. On the basis of these assessments,
one of the control options identified will be designated as the BAT for the
control of emissions from the sources in the category. This level of control
will be that technology, which in the judgment of the Administrator, is the
most advanced level of control ,dequately demonstrated, considering economic,
energy, and environmental impacts. .
The control level designated BAT may be different for new and.existing
facilities in a category. For practical purposes, this level of control for
new sources will, as a minimum, be equivalent to that which would be selected
as the basis for a New Source Performance Standard (NSPS) under section 111.
The requirement of BAT for new sources would consider "economic feasibility"
and would not preclude new construction. .
The selection of BAT for existing sources may require consideration of
the technological problems associated with retrofit and related differences
in the economic, energy, and environmental impacts. In practice, BAT for
existing sources would consider ~conomic feasibility and would not exceed
the most advanced level of technology that at least most members of an industry
could afford without plant closures.
(d) Minimum Requirements for Existing Sources
Final section 112 standards will require existing sources in any
. .
regulated source category, as a minim~m, to limit their emissions to the
levels corresponding to the use of BAT. This requirement is based on the
2-4
-------
Administrator's' judgment that any risks that could be avoided through the
use of these feasible control measures are unreasonable. Whether BAT"controls
" "
are sufficient to protect public health will be determined by a subsequent
evaluation of the remaining risks. ".
(e) Determination of Unreasonable Residual Risk for Existing Sources
Following the identification of BAT for existing sources, the quantitative
risk assessment described earlier will be used to determine the risks
remaining after the application of BAT to the source category. If the residual
risks are not judged by the Administrator to beu"nreasonable, further controls
would not be required. If, however, there is a finding of unreasonable
residual risk, a more stringent alternative would be required. Among the
possible alternatives would be the immediate application of more restrictive
emission standards, including those based on more extensive use of substitutes,
and scheduled or phased reductions in permissible emissions. The alternative
selected would be that necessary, in the Administrator's judgment, to
eliminate the unreasonable residual risks. "
Given the differences in the degree of certainty in risk estimates, in
" "
the numbers of people exposed, in benefits, in the distribution of r~~~s and
- _.~ .
benefits, in the cost of controls, in the availability of substitutes, and
in other relevant factors, it is not possible to state any precise formula
for determining unreasonable residual risk. The determination will
necessarily be a matter of judgment for each category involved. Nevertheless,
the process followed and the various factors involved can be outlined.
The determination of unreasonable residual risk will be based primarily
on public health, and will require protection with an ample margin of safety.
To the extent possible, quantitative or qualitative estimates of various
factors will be made for purposes of comparison. Among these are (1) the
range of total expected cancer incidence and"other health effects in the
existing and future exposed populations through the anticipated operating
life of existing sources; (2) the range of health risks to the most exposed
individuals; (3) readily identifiable benefits of the substance or activity;
(4) the economic impacts of requiring additional control measures: (5) the
distribution of the benefits of the activity versus the risksJt causes;
and (6) other possible health and environmental effects resulting from the
increased use of substitutes."
2-5
-------
3.0 SOURCES OF BENZENE FUGITIVE EMISSIONS IN PETROLEUM
REFINERIES AND ORGANIC CHEMICAL MANUFACTURING OPERATIONS
3.1
INTRODUCTION
Valves, pumps, flanges, and other pieces of equipment are used
extensively in refineries and the synthetic organic chemical manufac-
turing industry (SOCMI) to move streams of organic compounds to and
from various process vessels. Since this type of equipment can develop
leaks, each individual piece is a potential source of organic compound
emissions whenever it handles a process stream containing such compounds.
When a piece of equipment handles a process stream containing benzene,
it is a potential source of benzene emissions.
This chapter will discuss the types of equipment that can be
sources of' benzene fugitive emissions. Estimates of uncontrolled
emission factors will also be presented, and the parameters which may
influence the emissions will be discussed.
3.2 SOURCES OF BENZENE EMISSIONS
Benzene fugitive e~ission sources are pieces of equipment handling
streams that could potentially contain benzene. These include sources
that develop leaks after some period of operation ,due to seal failure
as well as other sources that can emit benzene when used in specific
conditions in the production unit. T~e sources that develop leaks due
to seal failure are those using a sealing mechanism to limit the
escape of organic compounds to atmosphere. These include pumps,
pipeline valves, safety/relief valves, flanges, and compressors.
Other types of equipment are potential benzene fugitive emission
sources for reasons other than leaking seals. These types of equipment
might have the potential for benzene emissions, for example, because
they vent organic materials that contain benzene to atmosphere. These
types include process drains, sampling connections, open-ended valves,
3-1
-------
wastewater separators, cooling towers,
safety/relief valves, and process unit
3.2.1 Potential leak Sources
3.2.1.1 Pumps. Pumps are used extensively in petroleum refineries
and in SOCMI for the ~ovement of organic fluids. Centrifugal pumps
are used most often in these industries, although positive-displacement
pumps, reciprocating and rotary action pumps, and the specialized
c3n~ed, diaphrag~ pumps, and magnetically coupled pumps are used for
some applications. Except for the canned, diaphragm, and magnetically
coupled types, the pumps have a shaft that requires a seal to isolate
the pump's interior fluid from the atmosphere. Packed and mechanical
shaft seals are most commonly used. Proper installation and maintenance
are required for all seal types if they are to function properly and
,'etain their ability to seal. The possibility of a leak through this
seal makes the pump a potential source of benzene emissions.
3.2:1.1.1 Packed seal. Figure 3-1 is a diagram of a simple
packed seal. Packed seals can be used on both reciprocating and
rotary action pumps. This seal consists of a stuffing box in the pu~p
casing filled with specialized packing material that is compressed
with a packing gland to fit closely around the shaft.. To prevent
buildup of frictional heat, lubrication is required. A sufficient
amount of either the liquid being pumped or i:inother liquid that is
injected r.lUst be allowed to flow between the packing and the shaft to
provide the necessary lubrication. Degradation of this packing and/or
the shaft seal face after a period of usage can be expected to eventually
result in leakage of organic compounds to atmosphere.
3.2.1.1.2 Mechanical seal. Figure 3-2 is a diagram of a basic
single mechanical seal. The rotating seal-ring face and the stationary
elel:1ent face are lapped to a very high degree of flatness to maintain
contact throughout their entire mutual surface area. As with packing,
the faces must be lubricated; however, because of the seal'5 construction,
much less lubrication is needed. There are many variations to the
basic design, but all have the lapped seal face between a stationary
element and a rotating seal ring. Again, if the seal becomes imper-
fect due to wear, the organic compounds being pumped can leak between
the s~al faces and can be emitted to atmosphere.
product accumulator vessels,
turnarounds.
3-2
-------
Pump stuffing box
,
\
\
\
,
,
,
Fluid
end
""Packing gland
'"
,
------
-Seal face
--.
I
I
I
I
Packing
,
,
"
,
"
, ---_~Possible leak
,,-- area
Figure 3-1. Simple Packed Sea11
-------
G land gasket
\
\
Pump stuffing box
,
"
,
,
"
Fluid
end
W
I
+::>
I
I
I
I
I
t
I
I
Sp'ring'
\
\
\
\
\.
"
"
"
I
I
Shaft
packing
\
\
,
Sea I face
\ .
\
\
\
\
\
\ '.
Rotating
seal ring
Fi gure 3-2.
Basic Single Mechanical Seall
.....G land ri ng
...
.....Insert packing
Stationary
,,/element
"
..",...
--- Possible
,
leak
area
-------
3.2.1.2 Pipeline Valves. One of the most common pieces of
equipment in a refinery or a SOCMI production unit is the valve. The
types of valves commonly used are globe, gate, lug, ball, relief, and
check valves. All except the relief valve and check valve are acti-
vated by a valve stem, whose motion may be rotational or linear or
both, depending on the specific design. This stem requires a seal to.
isolate the valve interior fluid from atmosphere. The possibility of
a leak through this seal makes the valve a potential source of benzene
emissions.
The most common type of valve stem seal in use is the packed
seal. It consists of a stuffing box in the valve housing filled with
specialized packing material that is compressed with a packing gland
to fit closely around the stem. Figure 3-3 is a diagram of a globe
valve with a packed seal.
3.2.1.3 Safety/Relief Valves. Safety/relief valves are required
by engineering codes for applications where the pressure on a vessel
or a system may exceed the maximum allowed. A spring-loaded safety/
relief valve, which is shown in Figure 3-4, is typically used for this
service. The seal is a flat disk held in place on a seat by a spring
during normal system operation. The possibility of a leak through
this seal makes it a potential source of benzene emissions. The
potential causes for leaks are "simmering, II a condition caused by the
system pressure being close to the valve set pressure, improper
reseating following a relieving operation, and corrosion or degradation
of the valve seat.
3.2.1.4 Flanges. Flanges are bolted, gasket-sealed junctions
used in joining pipe or equipment components, such as vessels, pumps,
valves, and heat exchangers, that may require isolation or removal.
The possibility of a leak through the gasket seal makes flanges potential
sourc~s of benzene emissions.
Two primary causes of leakage are seal deformation,
stress on the adjoining piping or equipment, and opening
without replacement of the gasket.
3.2.1.5 Compressors. Compressors, like
centrifugal and positive displacement types.
due to thermal
of the flange
pumps, can be both
Compressors have a shaft
3-5
-------
HANDWHEEL
STEM
SEAT
PACKING NUT
DISK
BODY
Figure 3-3.
2
Globe Valve with Packed Seal
I .
I .
3-6
-------
SE.A T .
Figure 3-4.
; PROC.E5~ ~I CE.
.5P RI)JG
DISK
- NOZ.~LE
Diagram of a Spring-Loaded Relief Val vel
3-.7
-------
.
r~--- ~- --~----
I.
that requires a seal to isolate the compressor interior gas from
atmospher~. The possibility of a leak through this seal makes it a
potential source of benzene emissions. In addition to having seal
types like those used for pumps, centrifugal compressors can be equip-
ped with. a liquid-film seal as shown in Figure 3-5. The seal is a
film of oil that flows between the rotating shaft and the stationary
gland. The oil that leaves the compressor from the pressurized system
side is under the system internal gas pressure and is contaminated
with the gas. When this contaminated oil is returned to the open oil
reservoir, process gas and .entrained benzene can be released to atmosphere.
3.2.2 Other Potential Sources
3.2.2.1 Process Drains.
process units entails draining
from process equipment. These
spills, and water used to cool
drains are open to atmosphere,
to atmosphere.
3.2.2.2 Sampling Connections. The operation of process units is.
checked periodically by routine analysis of feedstocks and products.
To obtain representative samples for these analyses, sampling lines
must first be purged. If this flushing liquid is not returned to the
process, it could be drained onto the ground or into a process drain,
where it would evaporate and release benzene to atmosphere.
3.2.2.3 Open-Ended Valves. Some valves are installed in a
system so that they function with the downstream line open to
atmosphere. Examples are purge valves, drain valves, and vent valves.
A faulty valve seat or incompletely closed valve would result in
leakage through the valve and benzene emissions to atmosphere.
3.2.2.4 Wastewater Separators. Conta~inated wastewater can
originate from several sources including, but not limited to, leaks,
spills, pump and compressor seal cooling and flushing, sampling,
equipment cleaning, stripped sour water, desalter water effluent, and
rain runoff. Contaminated wastewater is collected in the process
drain system and directed to the wastewater treatment system where oil
is skimmed in a separator, and the v/astewa~er undergoes additional
The operation.of refinery and socrn
condensate water and flushing water
drains also receive liquid leakage,
pump glands. Because most of these
benzene in the wastewater can be emitted
3-8
-------
INNER
BUSHING
INTERNAL
GAS
PRESSURE
Q
CaNT AMINATED
OIL OUT
TO RESERVOIR
OIL OUT
ATMOSPHERE
Figure 3-5.
Liquid-Film Compressor Shaft Sea13
3-9
-------
treatment as required. If it is present, benzene will be emitted
wherever wastewater is exposed to atmosphere due to evaporation of
benze~e contained in the wastewater. As such, the primary emission
points include surfaces of forebays and separators. Data are not
available to characterize uncontrolled emission rates for wastewater
separators.
3.2.2.5 Product Accumulator Vessels. Product accumulator vessels
include overhead and bottoms receiver vessels utilized with fractionation
columns, and product separator vessels utilized in series with reactor
vesselstlto separate reaction products. Accumulator vessels can be
vented directly to atmosphere or indirectly to atmosphere through a
b1owdo~m drum,or vacuum system. When an accumulator vessel contains
benzene and vents to atmosphere, benzene emissions can occur.
3.2.2.6 Vacuum-Producing Systems. The vacuum-producing systems
attendant to vacuum distillation and other processes (including Sulfolane
aromatic extraction) are potential sources of atmospheric emissions of
benzene. Two types of vacuum-producing systems could he used for
these processes:
o Steam ejectors with contact condensers.
o Steam ejectors with surface condensers.
In the contact condenser, condensable organics and steam from the
vacuum still and the jet ejectors are condensed by intimately mixing
with cold water. The condensabl,e organics and water vapor flow to a
condenser hot well. Benzene in the hot well can be evaporated and
emitted to atmosphere. Any benzene that is not condensed in the
barometric condenser can also be emitted directly to atmosphere. In a
surface condenser, non-condensables and process steam from the vacuum
still, mixed with steam from the jets, are condensed by cooling water
in heat exchangers. These potential pollutants, therefore, do not
crnne in contact with the cooling water. Again, any benzene that is
not condensed may be emitted to atmosphere.
3.2.2.7 Cooling Towers. Cooling towers dissipate heat to ab~osphere
from the recirculating water that in turn is used to remove heat from
s~ch process equipment as reactors, condensers, and heat exchangers.
If a .leak in the process equipment occurs and if the equipment is
3-10
-------
operating at a pressure higher than that of the recirculating water,
process material can be entrained in the water stream. This material
can be evaporated and. released to atmosphere from the cooling tower,
making it a potential source of benzene emissions. Another source of
emissions is the use of benzene-contaminated process water as a cooling
water source. Uncontrolled emission data are not available for cooling
towers.
3.2.2.8 Process Unit Turnarounds. Process units, such as reactors
and fractionators, are periodically shut down and emptied for internal
inspection and maintenance. The process of unit shutdown, repair or
inspection, and start-up is termed a unit turnaround. Purging the
contents of a vessel to provide a safe interior atmosphere for workmen
is termed a vessel blowdown.
In a typical process unit turnaround, the liquid contents are
pumped from the vessel to some available storage facility. The vessel
is then depressurized, flushed with water, steam, or nitrogen, and
ventilated. Depending on the facility configuration, the vapor con-
tent of the vessel may be vented to a fuel gas system, flared, or
released directly to atmosphere. When vapors are released directly to
atmosphere with potential benzene emissions, it is through a knock-out
pot, which removes condensable benzene, and a blowdown stack, which is
usually remotely located to ensure that combustible mixtures will not
be released within the facility. Data are not available to characterize
uncontrolled emission rates for process unit turnarounds.
3.2.2.9 Safety/Relief Valve Discharges. Safety/relief devices
are designed to release a product material from distillation columns,
pressure vessels, reactors, and other pressurized systems during
emergency or upset conditions. Release of material containing benzene
makes this equipment an emission source. The frequency and duration
I . .
of releases, however, are dependent on the operating conditions of the
particular plant, and wide operational variations between plants can
occur.
3.3 MAGNITUDE OF BENZENE EMISSIONS FROM REFINERIES AND SOCMI
PRODUCTION OPERATIONS
Data on the measurement of benzene fugitive emissions from refinery
and SOCMI sources are limited. However, recent testing efforts have
3-11
-------
generated a great deal of information on VOC emissions from refinery
operations. Refinery benzene fugitive emissions are assumed to be
similar to the refinery VOC e~issions for light liquid service equipment
because of their similar vapor pressures. It is straightforward,
therefore, to estimate refinery benzene fugitive emissions from these
data. Since the majority of SOCt1I benzene fugitive e~issions originates
from equipment handling benzene and benzene-containing organic streams,
the e~ission factors developed from the refinery data should apply to
SOOll sources as well. Table 3-1 presents VOC emission factors for
refinery and SOCMI sources. Benzene emissions can be related to these
VOC emission factors, if it is assumed that the ~'leight percent benzene
in the gaseous emissions from a leaking piece of equipMent is equiva-
lent to the weight percent ,benzene in the product stream being handled
by the equipment. .
To illustrate the usage of these factors, benzene emissions were
. .
estimated for an example medium-sized production operation. Table 3-2
represents an average number of pieces of equipment handling benzene
in a chlorobenzene production unit, a reformate benzene extraction
unit, or a linear alkyl benzene production unit. The number of pieces
of equipment multiplied by the appropriate emission factor from Table 3-1
yields the total benzene, fugitive emissions for each type of equipment.
3-12
-------
Table 3-1. ESTIMATED VAPOR EMISSION FACTORS FOR
NONMETHANE HYDROCARBONS FROM REFINERY AND SOCMI SOURCES
Source Type
Emission Factor Estimatea
(Kg/hr - source)
Pumps
Pipeline Valves
a. Gas/Vapor Streams
b. Liquid Streams
Safety/Relief Va1vesb
Flanges
Compressors
Process Drains
Sampling Connections
Wastewater Separators
Vacuum-Producing Systems
Cooling Towers
Process Unit Turnarounds
Product Accumulator Vessel Vents
Safety/Relief Valve Discharges
0.12e
0.021
0.010
0.16
0.00026
0.44
0.032
0.01 5c
NA
NA
Negligible
NA
1.23d
NA
NA - No factor available.
aFrom Reference 4 except where otherwise noted.
bGas Service only.
cThis factor was derived by the following equation:
[sampling
~~m~~~~~O~~n- = loo6.~~lK~e~~~ery (R~~~m5) + 8~o~oC~~~i~~~~ns (R~~~m6)]
nections throughput refinery throughput
+ 0.0032 Kg/hour for one open-ended line = 0.015 Kg/hour
dFrom Reference 7
eEmission factor for refinery equipment in light liquid service.
3-13
-------
Table 3-2.
ESTIMATED BENZENE EMISSIONS FROM AN AVERAGE PLANT7
Equipment Typea
Number of
Pieces Handling
10 or More
Weight Percent
Benzene
Uncontrolled
Benzene e
Emission
(kg/hr)
Pumps
15
1.80
Process Valves
Gas
91
1.91
1.68
Liquid
168
9
1.44
Relief Valves
Open-Ended Valvesb
Gas
9 0.22
96 1. 27
26 0.39
600 0.16
15 0:48
1,003d 9.35
Li qu i d
Sample Connectionsc
Flanges
Ora ins
Totals
aWastewater separators, vacuum-producing systems, process unit
turnarounds, cooling towers and safety relief valve discharges
have been excluded due to lack of emission factors.
bUncontrolled benzene emissions include emissions for
open-ended lines: gas = 0.0242 kg/hr per open-ended
liquid = 0.0132 kg/hr per Qpen-ended valve.
valves and
valve and
CNumber of sample connections is 25 percent of the number of open-
ended valves.
dTotal equipment excludes sample connections since they are included
in the total number of open-ended valves.
eThis assumes 100 percent benzene in the equipment.
3-14
-------
3.4 REFERENCES
1.
Erikson, D.G. and V. Kalcevic. Emission Control Options for
the Synthetic Organic Chemicals Manufacturing Industry,
Fugitive Emissions Report. Hydroscience, Incorporated.
Knoxville, TN. For U.S. Environmental Protection Agency.
Research Triangle Park, NC. Draft Report for EPA Contract
No. 68-02-2577. February 1979. p. 11-3, 11-6.
2. . Edwards, J.A. Valves, Pipe and Fittings - A Special Staff
Report. Pollution Engineering. 6:22-30. December 1974.
3. Boland, R.F., et!l. Screening Study for Miscellaneous
Sources of Hydrocarbon Emissions in Petroleum Refineries.
Monsanto Research Corporation. Dayton, OH. For U.S.
Environmental Protection Agency. Research Triangle Park, NC.
Report No. EPA-450j3-76-041. December 1976.
4. Wetherold, R. and L. Provost. Emission Factors and Frequency
of Leak Occurrence for Fittings in Refinery Process Units.
Radian Corporation. Austin, TX. For U.S. Environmental
Protection Agency. Research Triangle Park, NC. Report
No. EPA-600j2-79-044. February 1979. p. 22.
5.
Burklin, C.E. Revision of Evaporative Hydrocarbon Emission
Factors. Radian Corporation. Austin, TX. For U.S.
Environmental Protection Agency. Research Triangle Park, NC.
Report No. EPA-450j3-76-039. August 1976. 80 p.
6.
Powell, D., et ale Development of Petroleum Refinery Plot
Plans. Pacific Environmental Services, Incorporated. Santa
. Monica, CA. For U.S. Environmental Protection Agency.
Research Triangle Park, NC. Report No. EPA-450j3-78-025~
June 1978. 180 p.
7.
Briggs, T. and V.P. Patel. Evaluation of Emissions from
Benzene-Related Petroleum Processing Operations. PEDCo
Environmental, Incorporated. Cincinnati, OH. For U.S.
Environmental Protection Agency. Research Triangle Park, NC.
Report No. EPA-450j3-79-022. October 1978.
3-15
-------
4.0 EMISSION CONTROL TECHNIQUES
4.1
I NTRODUCTI ON
As identified in Chapter 3, there are several categories of
potential sources of benzene fugitive emissions in SOCMI and refinery
operations. These sources include: (1) the cumulative total of small
continuous leaking emission sources caused by seal leakage in pumps,
valves, flanges, safety/relief dev1ces, and compressors, (2) continuous
emissions from the operation of vacuum-producing systems, drains,
wastewater separators, and cooling ~owers, and (3) intermittent emis-
sions from the operation of safety/relief devices, product accumulator
vessels vents, sampling connections, open-ended lines, and process
unit turnarounds.
Three basic control techniques can
fugitive emissions from these potential
are:
be applied to reduce benzene
sources. These techniques
. Monitoring and maintenance programs in which fugitive sources
are located and repaired at regular intervals.
. Preventive programs in which potential fugitive sources are
eliminated by either retrofitting with specified controls or replacement
with leakless equipment.
. Proces's modifications that reduce or el iminate benzene fugi tive
emissions by reducing or eliminating the use of benzene in production
operations.
This chapter will discuss these control techniques and their
effectiveness in reducing benzene fugitive emissions. Technical
aspects of retrofitting specified controls and leakless equipment
for the industry will also be discussed.
Three of the sources described in the previous chapter are not
included in this discussion of emission control techniques -- wastewater
4-1
-------
separators, cooling towers, and process unit. turnarounds. No reliable
data on enission rates and control techniques are available for cooling
towers and process unit turnarounds, so these sources will not be
included in this standard. Wastewater separators are also not included
in this standard, because the wide variation in the'types of systems
used in the industry would make it difficult to specify a control
technique, and because emission information for controlled and uncon-
trolled operation is not available. These sources may be addressed
in the future.
4.2 MONITORING AND MAINTENANCE PROGRAMS
The types of equipment that have the potential to be benzene
fugitive emission sources (i.e., pumps, valves, etc.) have been
identified and discussed in Chapter 3. When such a piece of equi~nent
develops a leak, the leak can be detected by using a portable VOC
detection instrument (performance criteria for the instrument and a
description of the leak testing methods are given in Appendix 0).
When the leak has been located, it can be repaired through maintenance
procedures, such as tightening the packing for valves.
Potential benzene fugitive emission sources at a given plant can
be monitored at regular intervals with the portable detector, and the
identified leaks can be repaired within a specified time limit. This
approach is referred to as a monitoring and maintenance program, and
it may be used to effect various control efficiencies depending on the
. .
action level (VOC concentration in parts per million by volume that
defines a leak) monitoring interval, and allowable repair til'1e specified.
Recently developed data can be used to predict the potential
number of leaks from the various equipment types.1 For example,
Table 4-1 presents data on the percentage of pieces of equipment that
are predicted .to be found leaking at various action levels during an
initial source screening survey.
4.2.1 Definition of a leak
In order to develop a monitoring plan for equipment leaks, an .
. .
equipment leak must first be defined. .The choice of the action level
for defining a leak is influenced by several considerations. First,
the ~ercent of total mass emissions that can potentially be controlled
4-2
-------
Table 4-1 PERCENTAGE OF SOURCES PREDICTED TO BE LEAKING
IN AN INDIVIDUAL COMPONENT SURVEy1
Equipment Predicted Percent of Sources Leakingb
Typea >100,000ppmv >50,000ppmv >10 ,000ppmv >1000ppmv
Pumps 6 .9 23 . 41
Pipeline Valves
a. Gasc 4 5 10 22
b. Liquid 2 4 12 25
Safety/Relief Devices ." 1 2 8 21
Pipeline Flanges 0 0 0 2
Compressors 6 ' 10 33 68
Process Drains 0 .1 4 10
aData were not available for open-ended lines, sampling connections,
wastewater separators, vacuum-producing systems, and cooling towers.
This type of information would not be appropriate for process unit
turnarounds, product accumulator vessel vents, and relief valve
over-pressure.
bThe technical feasibility of repairing leaks in the 1000 - 10,000 ppm
range and achieving an overall emission reduction has not been demon-
strated in field testing.
cValves in gas service contain process fluid in the gaseous state.
4-3
-------
by the monitoring and maintenance program can be affected by varying
the action level. Table 4-2 gives the percent of total mass emissions
affected at various action levels for a number of equipment types.
From the table, it can be seen that; in general, a low action level
results in larger potential emission reductions. However, the choice
of an appropriate leak definition is limited by the ability to r~pair
leaking compon~nts.
. The ability to repair leaking equipment from above 10,000 ppm to
below 10,000 ppm has been demonstrated in field testing.2 This repair
ability has not been demonstrated for a 1,000 ppm action level, however.
Available data do not support the conclusion that repairing leaks in
the 1,000 to 10,000 ppm range would result in an overall reduction in
emissions.
The nature of repair techniques for pipeline valves, for instance,
is such that attempts to repair leaks below a certain level by tightening
the packing gland may result in an increase in emissions. In practice,
valve packing material can become hard and brittle after extended use.
As the packing loses its resiliency, the valve packing gland must be
tightened to prevent loss of product to atmosphere. Excessive tightening,
however, may cause cracks in the packing, thus exacerbating the leak
rate.
4.2.2 Inspection Interval
A monitoring plan may include annual, quarterly, monthly, or even
weekly inspections. . The length of time between inspections should
depend on the expected occurrence and recurrence of leaks after a
piece of equipment has been checked and/or repaired. Th1s interval
can be related to the type of equipment and service conditions, and
different intervals can be specified for different.pieces of equipl:1ent.
In the refinery VOC leak Control Technique Guideline (CTG)
document3, the recommended mo~itoring intervals are: annual -- pump
seals, pipeline valves in liquid service, and process drains; quarterly --
compressor seals, pipeline valves in gas service, and safety/rel.ief
valves .in gas service; ~eekly -- visual inspection of pump seals; and
no individual mqnitoriDg -- pipeline flanges and other connections,
and safety/relief valves in liquid service. The choice of the interval
4-4
-------
Table 4-2. PERCENT OF TOTAL MASS EMISSIONS
AFFECTED AT VARIOUS ACTION LEVELSl .
Percent of Mass Emissions Affected
at this Action Leve1a
Source Type 100,000 ppmv 50,000 ppmv 10,000 ppmv 1 ,000 ppmvb
Pumps 56 68 87 97
Valves
Gasc 85 92 98 99
Liquid 49 62 84 96
Safety/Relief Devices 19 33 69 91
Compressors 28 48 84 98
Drains 0 8 46 82 I
aThese figures relate the action level to the percentage of total mass
emissions that can be expected from sources with concentrations at
the source greater than the action level. If these sources were
instantaneously repaired .to a zero leak rate and no new leaks occurred,
then emissions could be expected to be reduced by this maximum theo-
retical efficiency.
bThe technical feasibility of repairing leaks in the 1000 to 10,000 ppm
range and achieving an overall emission reduction has not been demon-
strated in field testing.
cValves in gas service contain process fluid in the gaseous state.
4-5
-------
affects the emission reduction achievable, since more frequent inspection
will r~sult in earlier detection and repair of leaking sources.
4.2.3 Allowable Repair Time
If a leak is detected, the equipment should be repaired within a
certain time period. The allowable repair time should reflect an
interest in eliminating a source of benzene emissions, but it should
also allow th~ plant operator sufficient time to obtain necessary
repair parts and maintain some degree of flexibility in overall plant
maintenance scheduling. Once again, the determination of this allowable
repair time will affect emission reductions by influencing the length
. .
of tilTle that leaking sources are allowed to continue to emit benzene.
. .
A portion of the components with concentrations in excess of the
action level will not be able to be repaired within the specified
. '
allowable time period. Some pieces of equipment, for example, may not
be repairable on-line and may need to be isolated from the process for
repair. Unless a spare is available, in the case of pumps or compressors,
or unless a valve can be blocked off from the process line, it may be
necessary to partially or completely shut down a process unit to repair
these sources. Such leaks can be reported to the appropriate regulatory
agency, and ~rrangements c,n be made for the equipment to be repaired
, .'
as soon as is practicable, or during the next scheduled turnaround.
4.2.4. Visual Inspecti~ns
Visual inspections can be performed to detect evidence of liquid
leakage from plant equipme~t. When such evidence is observed, the
operator can use a portable VOC detection instrument to measure the
VOC concentration of the source. All liquid leaks will not necessarily
result in a reading of greater than the action level.3 In a specific
application, visual inspections can be used to detect the failure of
. .
the outer seal of a pump double mechanical seal system. Observation
of liquid leaking along the shaft indicates an outer seal failure and
. .
sign'll s the need for seal repair.
4.2.5 Other f10nitoring Techniques
Other monitoring techniques have been proposed to supplement the
individual component survey. These techniques include unit area
surveys (walkthroughs) and fixed-point monitoring systems. In theory,
4-6
-------
these techniques allow the operator to
that must be individually surveyed and
requirements.
4.2.5.1 Unit Area Survey. A unit area survey entails measuring
the ambient VaG concentration within a given distance, for example,
one meter of all equipment located on ground and other accessible
levels within a processing area. These measurements are performed
with a portable VaG detection instrument utilizing a strip chart
recorder.
The instrument operator walks a predetermined path to assure
total available coverage of a unit on both the upwind and downwind
sides of the equipment, noting on the chart record the location in a
unit where any elevated VaG concentrations are detected. If an
elevated VaG concentration is recorded, the components in that area
can.be screened individually to locate the specific leaking equipment.
It is estimated that 5a percent of all significant leaks in a
unit are detected by the walkthrough survey, provided that there are
only a few pieces of leaking equipment, thus reducing the VaG back-
ground concentration sufficiently to allow for reliable detection.4
The major advantages of the unit area survey are that leaks from
accessible leak sources near the ground can be located quickly and
that the leak detection manpower requirements can be lower than those
for the individual component survey. Some of the shortcomings of this
method are that VaG emissions from adjacent units can cause false leak
indications; high or intermittent winds (local meteorological conditions)
can increase dispersion of VaG, causing leaks to be undetected; elevated
equipment leaks are not detected; and additional effort is necessary
to locate the specific leaking equipment (i.e., individual checks in
areas where high concentrations are found).
4.2.5.2 Fixed-Point Monitoring Systems. . The basic concept of
the fixed-point monitoring system is that sampling-point devices can
be installed at specific sites within a process area to monitor for
leaks automatically. The ambient benzene concentration can be remotely
and centrally indicated to the operator, who can respond appropriately
when elevated levels are recorded. The monitoring sites would not
reduce the number of components
hence reduce monitoring labor
4-7
-------
I
I .
include the entire area of the plant, but only areas where equipment
handling benzene is located.
The approaches to leak detection with fixed-point monitors differ
in the number and placement of the sample points and in the manner in
which the sample is taken and analyzed. One approach is to estahlish
the sample points near specific pieces of equipment, such as process
pumps and compressors. A second approach is to establish the sample
points in a grid pattern throughout the process area. When an
elevated concentration is noted, the operator performs an individual
component survey on equipment in that area to locate the leaking
component. In addition to these variations in the location of the
sampling points, different types of systems can be used. For example,
the sampling can be done on-site or the samples can be collected at
the site and then analyzed at a central location (an automatic
sequential system).
One feature of the fixed-point monitor approach is that the
location of the monitor and the type of sampling and analysis can be
tailored to meet the requirements of individual plant sites and VOC.
Fixed-point monitors have the capability to sample for benzene specif-
ically by gas chromatography -- flame ionization detection (GC/FID) or
infrared (IR) analysis. This allows the sources leaking benzene to be
located more easily. However, this approach will require the use of a
. .
portable VOC detection device to locate the leak, particularly if a
gridded process-area monitoring approach is used. Leak detection
efficiency for fixed-point monitoring systems is estimated to be 33
percentS for facilities with a small number of recurring leaks, pro-
vided that major leaking equipment has been repaired, reducing the
benzene hackground concentration sufficiently to allow reliable
detection.
The fixed-point monitoring approach can also be utilized to
monitor the operation of individual pieces of equipment and detect
control device failures that would result in leaks. In specific
applications, the barrier fluid system of a double pump seal can be
monitored with a pressure sensing device to signal inner seal failure.
Also, in cases where rupture disks are installed upstream of pressure
4-8
-------
relief devices, leakage through the disk can be monitored if bubbles,
pressure gauges and/or excess flow valves are installed between the
disk and the valve.
4.2.6 Maintenance
When leaks are located by the monitoring methods described in
this section, the leaking component can then be repaired or r~placed.
Many components can be serviced on-line. This is generally regarded
as routine maintenance to keep operating equipment functioning properly.
Equipment failure, as indicated by a leak not eliminated by servicing,
requires isolation of the faulty equipment for either repair or
replacement.
4.2.6.1 Pumps. Most critical service process pumps are backed
up with a spare so that they can be isolated for repair. Of those
pumps that are not backed up with spares, some can be corrected by
on-line repairs (e.g., tightening the packing). However, most leaks
that need correction require that the pump be removed from service for
seal repair. When the seal leak is small, there can be situations in
which removing the pump from service can result in larger temporary
emissions than the emissions that would occur if the pump remained in
service until shut down for other process reasons.
4.2.6.2 Valves. Most valve leaks can be reduced on-line by
tightening the packing gland for valves with packed seals or by lubri-
cation for plug valves, for example. Based on field observations, one
refinery study assumed that 75 percent of leaking valves r.ould be
. d 1. 6
repalre on- lne.
Valves that need to be repacked or replaced to reduce leakage
must be isolated from the process. Control valves, 6 to 8 percent of
the total valves in refinery and SOCMI benzene service,4 can usually
be isolated. Block valves, which are used to isolate or by-pass
equipment, normally cannot be isolated. One refiner estimates that 10
percent of the block valves can be isolated.7
When leaking valves can be corrected on-line, repair servicing
can be done immediately after detection of the leak. When the leaks
can be corrected only by a total or partial shutdown, the temporary
emissions could be larger than the continuous emissions that would
4-9
-------
result from not shutti~g down the unit until it was time for a shutdown
for other reasons. Simple maintenance procedures, such as packing
gland tightening and grease injection, can be applied to reduce
e~'1issions from leaking valves lintila shutdown is scheduled. Leaks
that cannot be repaired on-line can be repaired by drilling into the
valve housing and injecting an epoxy sealing compound. This practice
is growing in acceptance, especially for safety concerns.
4.2.6.3 Flanges. One refinery field study noted that most
flange leaks could be sealed effectively on-line by simply tightening
the flange bolts.6 For a flange leak that requires off-line gasket
seal replacement, a total or partial shutdown of the unit would probably
be required because most flanges cannot be isolated.
For ~any of these cases, there are temporary flange repair methods
that can be used. Unless a leak is major and cannot .be temporarily
corrected, the temporary emission from shutting down a unit would
probably be larger than the continuous emissions that would result
from not shutting down the unit until time for a shutdown for other
reasons.
4.2.6.4 Compressors. Compressors usually are in critical service
but often spares are not provided. Consequently, the compressors
would need to be bypassed, possibly by a partial or complete unit shut-
down, so that repairs can be made. In most cases the shutdown for repair
of the leaking seal and the subsequent startup will involv~ flaring
the process stream until operations are stabilized. This can result
in the temporary emissions being larger than the continuous emissions
that \'lOuld occur until the unit was shut down for other reasons.
4.2.7 Emission Control Effectiveness of ~1onitoring and r~aintenance
The control efficiency achieved by a monitoring and maintenance
program is dependent on several factors, including the action level,
the inspection frequency, and the allowable repair time.
Data are presented in Table 4-2 that show the expected fraction
of total emissions from each type of source contributed by those
sources with VOC concentrations greater than given action levels. If
a leak detection and maintenance program resulted in repair of an
such sources to 0 ppm, elimination of all sources over the action
4-10
-------
level between inspections, and instantaneous repair of those sources
found at each inspection, then emissions could be expected to be
reduced by the amount reported in Table 4-2. However, since these
conditions are not met in practice, the fraction of emissions from
sources with VOC concentrations over the action level represents the
theoretical maximum reduction efficiency. The approach to estimation
of emission reduction presented here is to reduce this theoretical
maximum control efficiency by accounting quantitatively for those
factors outlined above.
This approach can be expressed mathematically by the following
. 8
equatlon:
Reduction efficiency = A x B x C x 0
Where:
A =
Theoretical Maximum Control Efficiency = fraction of
total mass emissions from sources with VOC concentra-
tions greater than the action level (from Table 4-2).
B =
Leak Occurrence and Recurrence Correction Factor = .
correction factor to account for sources which start to
leak between inspections (occurrence) and for sources
which are found to be leaking, are repaired and start
to 1eak"again before the next inspection (recurrence).
C =
Non-Instantaneous Repair Correction Factor = correction
factor to account for emissions which occur between
detection of a leak and subsequent repair, since repair
is not instantaneous.
Imperfect Repair Correction Factor = correction factor
to account for the fact that some sources which are
repaired are not reduced to zero emission levels. For
computational purposes, all sources which are repaired
are assumed to be reduced to a 1000 ppm emission level.
An implicit assumption here is that the leak detection program detects
all of the sources with VOC concentration~ greater than the action
level that are present at the time of the inspection. As an example
of this technique, Table 4-3 gives values for the "B," "C" and "0"
correction factors for various possible inspection intervals, allowable
repair times, and action levels.
o =
4-11
-------
Table 4-3. E~1ISSION CORRECTION FACTORS FOR VARIOUS INSPECTION
INTERVALS, ALLOWABLE REPAIR TIMES AND ACTION LEVELSa (Reference 8)
.,,..-
Leak Occurrence and Non-Instantaneous Imperfect ~{epair
Recurrence Correction Repair Correction Correction
Factorb FactorC Factord
Allowable Repair
Inspection Interval Time (Days) Action Level (ppmv)
Source Yearly Quarterly ~.1onth 1 y 15 5 1 100,000 50,000 10,000 I,OOOe
Pumps 0.800 0.900 0.950 0:979 0.993 0.999 0.969 0.961 0.923 0.876
Valves
Gasf 0.800 0.900 0.950 0.979 0.993 0.999 0.997 0.996 0.993 0.986
Liquid 0.800 '0.900 0.950 0.979 0.993 0.999 0.984 0.975 0.944 0.898
Safety/Relief Devicesg 0.800 0.900 0.950 0.979 0.993 0.999 0.989 0.987 0.976 0.951
~ Compressors 0.800 0.900 0.950 0.979 0.993 0.999 0,983 0.984 0.970 0,946
I
I-'
N
Drains 0.800 0.900 0.950 0.979 0.993 0.999 0.864 0.906' 0.868
aNote that these correction factors taken individually do not corr-espond exactly to the overall emission
reduction obtainable by a monitoring and maintenance program. The overall effectiveness of the program is
determined by the product of all correction factors. .
bValues are assumed and account for sources that start to leak between inspections (occurrence) and for
sources that are found to be leaking, are repaired, and start to leak again before the next inspection
(recurrence) .
cAccounts for emissions that occur between detection of a leak and subsequent repair.
dAccounts for the fact that some sources that are repaired are not reduced to 0 ~pm emission levels.
average repair factors at 1000 ppmv are assumed.
eThe technical feasibility of repairing leaks in the 1000-10,000 ppm range and achieving an overall emission
reduction has not been demonstrated in field testing.
The
fValves in gas serVlce carry process fluids in the gaseous state.
gGas service only.
-------
4.3 PREVENTIVE PROGRAMS
Another approach to reducing benzene fugitive emissions from
chemical and refinery operations is to replace components with equip-
ment which does not leak. This approach is referred to as a preventive
program. This section will discuss the kinds of equipment that could
be applied in such a program and the advantages and disadvantages of
this equipment.
. 4. 3. 1 Pumps
As discussed in Chapter 3, pumps can be potential benzene fugitive
emission sources because of leakage through the drive-shaft sealing
mechanism. This kind of leakage can be reduced to a negligible level
through the installation of improved shaft sealing mechanisms, such as
double mechanical seals, or it can be eliminated entirely by installing
sealless pumps.
4.3.1.1 Double Mechanical Seals. By design, double mechanical
seals have a chamber between the two seal faces that either is flushed
with a circulating sealing fluid that allows control of the conditions
under which the seal operates, or is flooded with a static fluid whose
pressure is higher than the process fluid pressure (Figure 4-1).
Field screening has shown that the magnitude of emissions from new
pumps equipped with double mechanical seals is negligible. Double
seals do fail after extended periods of use, however, and can develop
leaks.9
Mechanical seals, single or double, have limits on their
applicability. They can be used only on shafts with a rotary motion.
Also, the maximum service temperature is usually limited to less than
260°C.4 In spite of these limitations, double mechanical seals can be
used in most new pump applications.4 Double mechanical seals can also
be retrofitted on many pumps that were designed for single mechanical
seals and packed seals. In most cases, the retrofit involves the
engineering required for selection of a suitable double seal assembly,
purchase of the seal, and installation. For some existing packed and
single mechanical seal pumps, however, the entire pump may have to be
replaced because the existing pump casing will not adequately house
a double mechanical seal a~sembly. Data from industrial sources
.,
4-13
-------
Possible leak into
sealing fluid
\
\
Sealing-liquid
( inlet
(
I
I
I
I
Sealing.liquid
outlet
I
. I
I
. I
I
Fluid
end-.
\\ \ I I /
\1 I
\ , I'
\ 'I ,I
, 1.:.-------..1
I I
\ ~-------)
\ '~._-'1'''
\ I I
, I I
'I I
.+:>
I
......
.+:>
I <'Seal face \
I \
L. ... )
'Inner seal assembly
"
t Seal face: \
t . J
'Outer seal assembly""
Figure 4-1. Doubl~ Mechanical Sea14
-------
indicate that this situation may occur for about 10 percent of all
. b . 11
pumps 1n enzene serV1ce.
4.3.1.2 Seal less Pumps. The seal less or canned-motor pump is
designed so that the pump casing and rotor housing are interconnected.
As shown in Figure 4-2, the impeller, motor rotor, and bearings are
completely enclosed and all seals are eliminated. A small portion of
process fluid is pumped through the bearings and rotor to provide
lubrication and cooling.
Standard single-stage canned-motor pumps are available for flows
up to 700 gallons per minute and heads up to 250 feet. Two-stage
units are also available for heads up to 600 feet. Canned-motor pumps
are widely used in applications where leakage is a problem.12
The main design limitation of these pumps is that only clean
process fluids may be pumped without excessive bearing wear. Since
the process liquid is the bearing lubricant, abrasive solids cannot be
tolerated. Also, there is no potential for retrofitting mechanical
or packed seal pumps for seal less operation. Use of these pumps in
existing plants would require that existing pumps be replaced.
4.3.2 Valves
As in the case of pumps, val ves can be sources of benzene fugitive
emissions because of leakage through the packing used to isolate
process fluids from atmosphere (see Chapter 3). If the valve stem can
be isolated from the process fluid, however, this emission source can
be eliminated. There are two types of valve designs in which the
stem is so isolated, and potential for leakage around the stem is
thus eliminated. These valve types are the diaphragm valve
and the sealed bellows valve.
4.3.2.1 Diaphragm Valves. The general configuration of a
diaphragm valve is shown in Figure 4-3. The process fluid is isolated
from the valve stem by a flexible elastomer diaphragm, thus eliminating
the potential for leakage around the stem. The position of the diaphragm
is regulated by a plunger, which is in turn controlled by the stem.
The stem may be actuated manually or automatically by standard hydraulic,
pneumatic, or electric actuators.
4-15
-------
-+=-
I
-'
C"\
SUCTION
DISCHARGE t
COOLANT CIRCULATING TUBE
IMPELLER
BEARINGS
Figure 4-2.
Seal-less Canned Motor Pump
-------
HANDWHEEL
FLEXIBLE
DIAPHRAGM
SADDLE
SHAPED
SEAT
Figure 4-3. Diaphragm Valve
4-17
-------
These valves have excellent corrosion resistance characteristics
and are reported to perform well in control valve situations with
minimal l'laintenance.13 The design problems associated with diaphragm
valves are the temperature and pressure limitations of the elastomer
used for the diaphragm. It has been found that temperature extremes
tend to damage or destroy the diaphragm in the valve. Also, operating
pressure constraints may limit the application of diaphragm valves to
1 t. 14
ow pressure opera 10ns.
4.3.2.2 Sealed Bellows Valves. The basic design of a sealed
bellows valve is shown in Figure 4-4. The stem in this type of valve
is isolated from the process fluid by a metal bellows. The bellows is
generally welded to the bonnet and disk of the valve, thereby effecting
isolation of the stem.
There are two ~ain disadvantages to these valves. First, they
. are only available in globe and gate valve configurations. Second,
the crevices of the bellows may be subject to corrosion under severe
conditions if the bellows alloy. is not carefully selected.
The main advantage of these valves is that they can be designed
to withstand high temperatures and pressures so that leak-free service
can be provided at operating conditions heyond the 1 imits of diaphragm
val ves .
4.3.3 Safety/Relief Valves
4.3.3.1 Rupture Disks. A rupture disk can be used upstream of a
safety/relief valve so that under normal conditions it seals the
system tightly but will break when its set pressure is exceeded, at
which time the safety/relief valve will relieve the pressure. Figure 4-5
is a diagram of a ruptur~ disk and safety/relief valve installation.
The installation is arranged to prevent disk fragments from lodging in
the valve and preventing the valve from being reseated if the disk
ruptures. It is important that no pressure be allowed to build in the
pocket hetween the disk and the safety/relief valve; otherwise, the
disk will not function properly. A pressure gauge and bleed valve can
be used to prevent pressure buildup. With the use of a pressure
gauge, it can be determined whether the disk is properly sealing the
.4-18
-------
.
YOKE
.
Figure 4-4. Sealed Bellows Valve
4-19
-------
To
atmospheric
vent
Figure 4-5.
- --Tension-adjustment
thimble
----Spring
RUPTURE DISK
FROM SYSTEM
Rupture Disk Installation Upstream of a Relief Valve4
4-20
-------
system against leaks. It is also necessary to install a block valve
upstream of the rupture disk so that the disk can be' isolated and
repaired on-line without shutting down the unit.
Use of a rupture disk upstream of a safety/relief valve would
eliminate leaks due to improper seating and simmering of the relief
valve. Also, the disk can protect the safety/relief valve against
system materials that could be corrosive and thereby cause seal
degradation.
4.3.3.2 Closed-Vent Systems. A closed-vent system can be used
to collect and dispose of gaseous benzene emissions resulting from the
relieving of safety/relief valves. Emissions from safety/relief
valves overpressure are typically intermittent, and their flow rates
during major upsets can be large. The usual method of disposing of
these gases, if collected in a closed vent, would be by flaring.
Figure 4-6 is a diagram of a dual-flare system. The smaller flare
operates more efficiently with routine smaller exhaust. The larger
flare is normally on standby to handle large emergency exhausts. It
is important to note that this type of control system would control
intermittent large releases of benzene as well as the continuous small
emissions from relief valve leakage.
By connecting safety/relief valve discharges to a closed-vent
and flare system, their emissions can be effectively controlled.
The effectiveness of benzene d~struction will depend on the flare
design and turn-down capability. The major technical difficulties
with flares occur in manifolding. These problems may be especially
important in existing plants if emissions from safety/relief valves
are manifolded to an existing flare that was not designed for the
additional flow. In new plant situations, the flare can be designed
for expected flow rates and frequency of safety/relief valve discharges.
Finally, off-gases from some chemical processes could not be flared
due to the presence of other hazardous compounds such as chlorine, which
would not be destroyed by flaring.
At present, no conclusive data are available on
Calculations and limited test data show efficiencies
99 percent.15-19 The presence of saturated organics
flare efficiency.
ranging from 60 to
or aromatic compounds
4-21
-------
MA\ t-J FLARE..
HEADERS?
. t
~-m{
CL~ GA5
~ELIEF VAL'-J'E- Oi~CHA~ES
~ COMPR.ES.so;:z. S.=A L 01 L
OE.G~I?-.JG VE..NTS .
Figure 4-6.
ELEVATE-D' J=LARE..'
C:,ROUt-.JD i=L/-'.RS
;>--
Simplified Closed-Vent System with Dual Flares4
4-22
-------
may decrease efficiency since such compounds tend to be thermally
stable.
4.3.4 Open-Ended Valves
Caps, plugs, and double block and bleed valves are devices for
closing off the ends of valves and pipes. When installed downstream
of an open-ended valve, they are effective in preventing leaks through
the seat of the valve from reaching atmosphere. Open-ended valves,
about 20 percent of the total valves handling benzene,20 are used mostly
in intermittent service for sampling, venting, or draining. If a cap
or plug is used downstream of a valve when it is not in use, benzene
emissions can be reduced. No test data are available to support a
control efficiency for these devices. However, the control efficiency
will depend on such factors as frequency of valve use, valve seat
leakage, and material that may be trapped in the pocket between the
valve and cap or plug and lost on removal of the cap or plug. For the
purposes of emission calculations, 100 percent control efficiency of
these emissions has been assumed. The installation of a cap, plug, or
second valve does not prevent the leakage that may occur through the
valve stem seal. The attachment of a second valve downstream of the
open-ended valve provides a double block and bleed arrangement. In
this system, it is important that the upstream valve be closed first.
Otherwise, product will remain in the line between the valves, and
expansion of this product will cause leakage through the valve stem
seals.
4.3.5 Closed-Loop Sampling
A frequent operation in most refinery and organic chemical
production operations is to withdraw a sample of material from the
process for analysis. To ensure that the sample is representative,
purging of the sample lines and/or sample container is often required.
If this purging is done to atmosphere or to open drains, or if there
are incidental handling losses, benzene emissions can result. A
closed-loop sampling system is designed so that the purged VOC is
returned to the system or sent to a closed disposal system and so that
the handling losses are minimized. Figure 4-7 gives two examples of
closed-loop sampling systems where the purged VOC is flushed from a
4-23
-------
PROCE.S5. L' J--.1E..
~
)1" SAMPLE.
U . C.O~TA\~E..R
. PROCESS LI}..JE
5AM PLE...
CO?-J T"i~ E.R.
Figure 4-7.
4
Diagram of Two Closed-Loop Sa~pling Systems
4-24
-------
point of higher pressure to one of lower pressure in the system and
where sample-line dead space is minimized. Reduction of emissions
from the use of closed-loop sampling is dependent on many highly
variable factors, such as frequency of sampling and amount of purge
required. For emission calculations, it has been assumed that closed-
loop sampling systems will provide 100 percent control efficiency.
4.3.6 Accumulator Vessel Vents and Seal Oil Degassing System Vents
Benzene emissions from accumulator vessel vents and seal oil
degassing system vents can be controlled by a closed-vent system. The
flow rates of these gaseous emissions are of a much smaller magnitude
than those of safety/relief devices, however. These emissions could,
therefore, be vented to a closed combustion device, such as a process
heater or a boiler or to a vapor recovery device. The operating param-
eters of the combustion device will affect the overall control efficiency
of the closed-vent/combustion device system. Combustion tempera~ure and
residence time are the critical parameters influencing benzene destruction
efficiency, and theoretical kinetic calculations indicate that a combustion
temperature of 760°C and a 0.5 second residence time will result in
100 percent benzene destruction efficiency.21
Organic compounds which, if combusted, would produce noxious or
corrosive gases (e.g., chlorobenzene) may be present in some benzene-.
containing vapor streams. In these situations, benzene emissions
from accumulator vessel vents and seal oil degassing vents should be
controlled by a closed-vent/vapor recovery device system. The overall
benzene control efficiency of a. closed-vent/vapor recovery device
system is dependent on the benzene collection efficiency of the vapor
recovery device. Vapor recovery devices, such as adsorbers, absorbers,
and condensers have been shown to range in benzene collection efficiency
from 90 to 99 percent depending on the parameters of the gas stream
in which the benzene is contained and the type of vapor recovery device
utilized. Therefore, the overall efficiency of closed-vent/vapor
recovery device systems can be expected to be in the 90 to 99 percent
range.
4-25
-------
4.4 PROCESS MODIFICATIONS
In some instances, benzene fugitive emissions could be eliminated
by process modification. For some chemical processes, feedstocks
other than benzene can be used, thus eliminating benzene from the
process. The following are some examples of this kind of substitution.
. f1aleic aflhydride can be produced by oxidation of n-butane
rather than by oxidation of benzene.22,23
. Cyclohexane can be extracted from refinery products rather
than by hydrogenation of benzene.24
I
I '
4-26
-------
4.5 REFERENCES
Wetherold, R.G. and L. Provost.
of Leak Occurrence for Fittings
Radian Corporation. Austin, TX.
Agency, Research Triangle Park,
February, 1979.
2. Trip report from K. C. Hustvedt to J. F. Durham summarizing test
data gathered at Phillips Petroleum Company's Sweeny, Texas,
refinery. March 14, 1979.
1.
Emission Factors and Frequency
in Refinery Process Units.
For U.S. Environmental Protection
NC. Report Number EPA-600/2-79-044.
3.
Hustvedt, K.C., R.A. Quaney, and W.E. Kelly. Control of Volatile
Organic Compound Leaks from Petroleum Refinery Equipment. U.S.
Environmental Protection Agency. Research Triangle Park, NC.
Report Number EPA-450/2-78-036. June 1978.
4.
Erikson, D.G. and V. Kalcevic. Emissions Control Options for the
Synthetic Organic Chemicals Manufacturing Industry, Fugitive
Emissions Report. Hydroscience, Inc. Knoxville, TN. For U.S.
Environmental Protection Agency. Research Triangle Park, NC.
Draft Report for EPA Contract Number 68-02-2577. February 1979.
Hustvedt, K.C. and R.C. Weber. Detection of Volatile Organic
Compound Emissions from Equipment Leaks. Paper presented at
71st Annual Air Pollution Control Association Meeting, Houston,
Texas, June 25-30, 1978.
5.
6.
Emissions from Leaking Valves, Flanges, Pump and Compressor
Seals, and Other Equipment in Oil Refineries. Report Number LE-
78-001. State of California Air Resources Board. April 24, 1978.
7. J. Johnson, Exxon Co., letter to Robert T. Walsh, EPA.
1977 .
July 28,
8. Tichenor, B.A., K.C. Hustvedt, and R.C. Weber. Controlling
Petroleum Refinery Fugitive Emissions Via Leak Detection and
Repair. Symposium on Atmospheric Emissions from Petroleum
Refineries. Austin, TX. November 6, 1979.
9.
L. Kronenberger, Exxon Company, letter to Don R. Goodwin, EPA.
February 2, 1977.
Edwards, J.E. Valves, Pip~ and Fittings -- A Special Staff
Report. Pollution Engineering. 6:24. December 1974.
10.
11.
"Benzene Emission Control Costs in Selected Segments of the
Chemical Industry," prepared for the Manufacturing Chemists
Association by Booz, Allen and Hamilton, Inc. June 12, 1978.
Perry, John H. Chemical Engineers Handbook. Robert Perry,
Cecil Chilton, Sidney Kirkpatric, eds. McGraw-Hill Book
Company. New York. 1963. p. 6-7.
12.
4-27
-------
! .
!
. 23.
13.
K. Hanover, Rhom & Haas, conversation with J. f1cInnis, Pacific
Environmental Services, Incorporated. September 5, 1979.
Sienknecht, Paul J., Dow Chemical, letter to Jerel R. ~1cInnis,
Pacific Environmental Services, Incorporated. December 20, 1979.
14.
15.
MemoranduM from Seeman, W.R., Hydroscience, to White, R.,
Hydroscience. May 19, 1978.
16.
Memorandum from Seeman, W.R., Hydroscience, to Kalcevic, V.,
Hydroscience. September 29, 1978. .
17.
Letter from Bergman, H., EPA Region 6, to Strader, W.C., Ethyl
Corporation. August 30, 1977.
18.
Straite, J. Flaring for Gaseous Control in the Petroleun
Industry. National Air Oil. Philadelphia, Pennsylvania.
Presented at Air Pollution Control Association. Pittsburgh.
June 26-30, 1978. .
19.
Telecon. Redd, 0., North American Manufacturers, with Mascone, D.C.,
EPA. October 4, 1978.
20.
Development of Petroleum Refinery Plot Plans. Pacific Environmental
Services, Incorporated. Report Number EPA-450/3-78-025. June 1978.
21.
22.
Patrick, David R., EPA, Memo to Jack R. Farmer, EPA.
July 18, 1979.
Industrial Process' Profiles for Environmental Use: Chapter 6.
The Industrial Organic Chemicals Industry. Research Triangle
Institute. Research Triangle Park, NC. Radian Corporation.
Austin, TX. For U.S.' Environmental Protection Agency. Cincinnati,
OH. Publication Number' EPA-600/2-27~023f. February 1977.
p. 6-125, 6-826. .
Gunn, T.C. and K. Ring. C~H Marketing Research Report on Benzene.
Chemical Economics Handbook. Stanford Research Institute. Menlo
Park, CA. t1ay 1977.
24.
Blackford, J.L. CEH Product Review on Cyclohexane. Chemical'
Economics Handbook. Stanford Research Institute. Menlo Park,
CA. February 1977.
4-2!3
-------
5.0 MODIFICATION AND RECONSTRUCTION
The Environmental Protection Agency will propose and promulgate
general provisions for 40 Code of Federal Regulation (CFR) Part 61
that will be similar to the general provisions of 40 CFR Part 60.
Provisions similar to 40 CFR 60.14 and 60.15 will be included to
establish that an "existing source" can become a "new source" if it is
deemed modified or reconstructed. An "existing source", as defined in
40 CFR 60.2(aa), is a facility of the type for which standards have
been promulgated and the construction or modification of which
was begun prior to the proposal date of the applicable standards.
The following discussion examines the applicability of modification/
reconstruction provisions to refinery and SOCMI operations that involve
benzene fugitive emissions.
5.1 GENERAL DISCUSSION OF MODIFICATION AND RECONSTRUCTION PROVISIONS
5.1.1 Modification
"Modification" will be defined in the provisions to be added to
40 CFR Part 61 as any physical or operational change of an existing
source which increases the ~mission rate of any pollutant to which a
standard applies. Exceptions to this definition will be provided
which will be similar to those presented in paragraphs (e) and (f) of
Section 60.14.
Paragraph (e) lists physical and operational changes which will
not be considered modifications. These changes include: (1) routine
maintenance, repair, and replacement; (2) an increase in the production
rate not requiring a capital expenditure as defined in Section 60.2(bb);
(3) an increase in the hours of operation; (4) use of an alternative
fuel or raw material if prior to the standard, the existing facility
was designed to accommodate that alternative fuel or raw material; and
(5) the addition or use of any system or device whose primary function
is the reduction of air pollutants, except whe~ an emission control
5-1
-------
i-
system is removed or replaced by a system considered to be less
efficient.
Paragraph (f) provides for superceding any conflicting provisions
of this section. Upon modification, an existing source would become a
new source for each pollutant to which a standard applies and for
which there is an increase in the emission rate to atmosphere. The
provisions also state that the addition of new source to a plant site
through any lIlechanism--new construction, modification, or
reconstruction--does not make any other so~rce within the plant site
subject to the applicable standards.
5.1.2' Recons truct i on
, .
~nder the provisions to be added to Part 61, an existing source
would become a new source upon reconstruction, irrespective of any
change in emision rate. Generally, reconstruction is considered to
occur upon the replacement of components if the fixed capital cost of
the new components exceeds 50 percent of the fixed capital cost that
would be required, to construct a comparable entirely new source, and
it is ,economically and technically feasible for the source to comply
with the applicable standards. The final judgment on whether a replacement
constitutes reconstruction and when it is technologically and economically
feasible to comply with the applicable standards will be made by the
Administrator. The Administrator's final determinations \'Ii11 be made
on the following bases: (1) comparison of the fixed capital costs of
the replacement components and a newly construction comparable source;
(2) the estimated life of the source after the replacements compared to
the life of a comparable entirely new source; (3) the extent to which
the components being replaced cause or contribute to the emissions
from the source; and (4) any economic or technical limitations on
compliance with applicable standards which are inherent in the proposed
replacements.
The purpose of this provision will be to insure that an owner or
operator does not perpetuate an existing source by replacing all but
vestigial components, $upport structures~ frames, housing, etc.,
rather than totally replacing it in order to avoid being subject to
applicable new source ~tandards.
5-2
-------
5.2 APPLICABILITY OF MODIFICATION AND RECONSTRUCTION PROVISIONS
5.2.1 Modification
The replacement of a potential benzene fugitive emission source
such as a pump or valve commonly occurs in refineries and organic
chemical plants. These replacements may occur either to increase the
capacity of existing sources, or to convert production from one chemical
to another chemical, or to replace worn-out or obsolete equipment. If
a component is replaced with an equivalent component, the benzene
fugitive emissions from the source should not increase because the
number of potential sources at the same benzene concentration (handling
the same process stream) remains unchanged. If an existing component
is replaced with a component with a higher leak rate (i .e., a mechanical
seal replaced with a packed seal), however, benzene fugitive emissions
would increase.
In some cases a production unit in the organic chemical industry
can be converted from the production of one ~hemical to the production
of a second chemical. If either the number of fugitive emission
sources or the percentage of benzene in the process streams of the
second chemical increases d~ring this conversion, the level of benzene
emissions from the unit could be expected to increase. However,
controls could be added in order to maintain the same emission rates
that existed prior to the conversion.
In many cases, there may be a desire to increase the capacity of
an existing plant. This may be achieved by replacing certain process
equipment (pumps, heat exchangers, reactors, etc.) with similar equipment
bu~of larger capacity. If this replacement does not increase the
number of fugitive emission sources handling benzene, the level of
benzene fugitive emissions would not be expected to increase. However,
if the number of sources were increased due to this replacement, then
benzene emissions would increase.
5.2.2 Reconstruction
When an owner or operator rebuilds components of an existing
production unit, the repuilt component may become subject to applicable
standards under the provisions to be added to Part 61. Under these
5-3
-------
provisions, reconstruction will be
fixed capital cost of rebuilding a
cost of an entirely new component.
considered to have occurred if the
component exceeds 50 percent of the
5-4
-------
6.0 MODEL UNITS AND REGULATORY ALTERNATIVES
6.1
INTRODUCTION
This chapter defines model units and alternative methods for
regulating benzene fugitive emissions from these units. The model
units characterize a range of existing processes that are used to
produce benzene as a finished product, that use benzene in the
production of other organic chemicals, or that use or produce benzene
or benzene-containing streams in the manufacture of organic chemicals.
Although these model unit parameters may vary from the actual parameters
that exist at a particular facility, they are the most useful means of
determining and comparing the environmental and economic impacts of
regulatory alternatives.
Regulatory alternatives are also defined in this chapter. These
regulatory alternatives represent comprehensive programs for reducing
benzene fugitive emissions 6nd provide varying degrees of emission
reduction. The regulatory ~lternatives will be applied in the analysis
of both new and existing model units.
6.2 MQDEL UNIT PARAMETERS
Refinery and SOCMI production units vary considerably in size,
configuration, age, and complexity. Because of variations among
production units, model unit parameters were selected to represent
processes with varying numbers of potential leak sources. These model
units are not necessarily related to production capacity, but approxi-
mate various levels of process complexity. The technical parameters
of the three model units selected are shown in Table 6-l.
The model unit parameters displayed in the table were developed
through analyses of process flow diagrams, material balances, and
modular equipment counts for various benzene-related production opera-
tions. Since no equipment count data are available for vacuum-producing
6-1
-------
Table 6-1.
MODEL UNIT EQUIPMENT WITH >10 PERCENT BENZENE
Number of Components per Model Unit
Source Typea Ab BC Cd
Pumps 5 15 25
Process Valves, Gase 30 91 151
Process Valves, Liquide 56 168 280
Relief Valves, Gas 3 9 16
Open-Ended Valves, Gas 33 7 12
Open-Ended Valves, Gas 23 72 119
Drains 5 15 25
Sample Connections 9 26 44
aVacuum-producing systems and product accumulator vessel vents, des-
cribed in Sections 3.2.2 and 3.2.3, are not included since no equip-
ment count data are available for these sources.
bRepresents an average inventory of equipment for production of
benzene from toluene, ethyl benzene, styrene, cumene, cyclohexane,
benzene sulfonic acid, resorcinol, maleic anhydride, or 1 ethylene
production unit.
cRepresents an average inventory of equipment for production of
benzene from extraction of reformate, chlorobenzenes, linear
alkylbenzenes, or 2 or 3 ethylene production units.
dRepresents an average inventory of equipment for production of
benzene from extraction of pyrolysis gasoline, nitrobenzene,
hydroquinone, or 4 or 5 ethylene production units. .
eNine percent of all process valves are automatic control valves.
6-2
-------
systems and product accumulator vessel vents, these sources are not
included in the model units. These model units represent average
inventories of equipment handling process streams containing greater
than 10 weight percent benzene for various refinery and SOCMI production
operations. Model A represents an average inventory of such equipment
for units involved in the production of benzene from toluene, ethyl benzene,
styrene, cumene, cyclohexane, benzene sulfonic acid, resorcinol, or
maleic anhydride; Model B represents units involved in the production
of benzene from extraction of reformate, chlorobenzene or linear
alkylbenzene; Model C represents production of benzene from extraction
of pyrolysis gasoline, nitrobenzene or hydroquinone. Ethylene produc-
tion may be represented by either Model A, B, or C, depending on the
number of ethylene production units at the plant site. One ethylene
unit would be represented by Model A, two or three ethylene units
would be represented by Model B, and four or five ethylene units would
be represented by Model C. New units would be characte~ized by the
same set of model units.
It is estimated that 62 percent of existing refinery and SOCMI
benzene-related production units would be represented by Model A,
31 percent by Model B, and 7 percent by Model C. It is expected that
new units would follow the same distribution~
6.3 REGULATORY ALTERNATIVES
The regulatory alternatives in this section represent feasible
methods of controlling benzene fugitive emissions from petroleum
refinery and SOCMI process units. Each regulatory alternative presents
a comprehensive program for reduction of emissions from the sources
listed in Table 6-2 by combining the individual control techniques
described in Chapter 4. Table 6-2 summarizes the requirements of the
regulatory alternatives.
6.3.1 Regulatory Alternative I
Regul atory Al ternative I represents a basel ine regul atory al ternative.
The baseline regulatory alternative describes the industry in the
absence of additional regulatory requirements, and it provides the
basis for incremental comparison of the impacts of the other regulatory
alternatives.
6-3
-------
Table 6-2. MONITORING INTERVALS AND EQUIPMENT SPECIFICATIONS
FOR BENZENE FUGITIVE REGULATORY ALTERNATIVES
--..-.. -.. -,;:.::~.-
Regulatory Alternativesa
lIb III IV
sourcec.d Inspection Equipment Inspection Equipment . Inspection Equipment Inspection Equipment
Interval Specification I nterva I Specification I nterva I Specification Interval Specification
1. Pumps Yearlye None Monthlye None Nonee Double seal sf None Double seal sf
and contro" ed and contro" ed
degassing vent degassing vent
2. Process Valves
a. Gas Service Quarterly None Monthly None Monthly None None D1a~hragm or
sea ed e"ows
va hes
b. Liquid Service Yearly None Monthly None Monthly None None Dia~hragm or
sea ed bellows
valves
0"\ 3. Safety/Relief Quarterly None None Rupture disks None Rupture disks None Rupture disks
I Valves (Gas or tie into or tie into or tie into
~ Service) existing flare existing flare existin!!, flare
4. Open-Ended Valves
a. Gas Service Quarterlyg Caps, blinds, Monthlyg Caps, blinds, Monthlyg Caps, bli nds, None Diaphragm or
or pl ugs or pl ugs or plugs sea 1 ed bellows
valve plus caps,
blinds, or plugs
b. Liquid Service Yearlyg Caps, blinds, Monthlyg Caps, blinds, Monthlyg Caps, blinds, None Diaphragm or
or pl ugs or plugs. or plugs sealed bellows
valve plus caps,
blinds, or plugs
5. Drains Yearly None Monthly None Monthly None None Enclosed active
drain systems
6. Sampl; ng None None None Closed-loop None Cl osed-l oop None Closed-loop
Connect ions sampl i ng sampling sampl i ng
7. .Compressors Quarterly None Monthly None None Double seals None Doubl e sea Is
8. Product Accumu- None None None Tie into None Tie into None Tie into
1 a tor Vessel closed control closed control closed control
system system system
-------
. .~
Table 6-2. MONITORING INTERVALS AND EQUIPMENT SPECIFICATIONS
FOR BENZENE FUGITIVE REGULATORY ALTERNATIVES
NOTES:
aRegulatory Alternative I (baseline) includes no new regulatory
specifications and, hence, is not included in this table. Regu-
latory Alternative VI is not included since it would require that
no benzene be emitted from refinery or SOCMI sources.
bAlternative II is equivalent to controls recommended in the
refinery CTG for fugitive VOC emissions.
CLiquid service safety/relief valves, flanges, wastewater separators,
vacuum-producing systems, process unit turnarounds, and cooling.
towers are not routinely monitored. Wastewater separators,
vacuum-producing systems, process unit turnarounds, and cooling
towers are not included in this table since there are no available
control technologies for these sources.' .
dFor all alternatives, the sources would handle organic streams
with over 10 percent benzene by weight.
eFor pumps, instrument monitoring would be supplemented with weekly
visual inspections for liquid leakage. If liquid is noted to be
leaking from the pump seal, the pump seal will be repaired.
fA pressure sensing device should be installed between the double
seals and should be monitored to detect seal failure.
gInspection applies to the valves.
6-5
-------
As discussed in Chapter 4, a number of factors influence the
baseline emission level. Examination of benzene control programs at
individual' plants reveals a range of existing control levels. Many
plants rely on normal maintenance procedures to control fugitive
emissions from leaks. Other plants may have developed a leak detection
and repair program in response to OSHA regulation requirements, State
or local agency regulations~ or emission offset provisions. To
characterize baseline conditions, however, a general description of
the entire industry is desirable, rather than a description of
site-specific or geographic-specific conditions. Baseline conditions,
therefore, will be assumed to reflect normal existing plant maintenance
procedures. These conditions are reflected in the lias is" emission
factors from Table 3-1, which are used in the environmental impact
analysis of the baseline regulatory alternative in Chapter 7.
6.3.2 Regulatory Alternative II
A higher level of benzene fugitive emission control could be
achieved with Regulatory Alternative II than with the baseline level.
This Regulatory Alternative would require periodic leak detection and
repair for most sources, and the installation of specified equipment
for other sources. The requirements of this regulatory alternative
are based upon the recommendations of the refinery VOC leak control
techniques guideline (CTG) document.1
Quarterly monitoring for leaks from safety/relief valves, process
valves and open-ended valves in gas service, and compressors would be
required. Pumps, drains and valves would be required to be. monitored
annually for leaks. Weekly visual inspections of pump seals would be
required; visual detection of a liquid leak would direct that monitoring
be initiated to determine if the action level were being exceeded, and
that the pump seal be subsequently repaired, if necessary. Safety/relief
valve monitoring would also be required after over pressure relieving
to detect improper reseating. Finally, open-ended valves and lines
would be required to be sealed with a cap, blind, plug, or another
valve.
6-6
-------
6.3.3 Regulatory Alternative III
Regulatory Alternative III would provide for more restrictive
control than Regulatory Alternative II by increasing the frequency of
leak detection and repair for some sources and requiring the installation
of specified control equipment for other sources. Installation of
closed-loop sampling systems would be required; rupture disks would be
required on gas service safety/relief valves that vent to atmosphere;
degassing vents on pump seal oil reservoirs would be required to be
vented to a closed system; accumulator vessels would be required to be
vented to a closed system; and open-ended valves would be required to
be sealed with a cap, blind, plug, or another valve. Based on a
preliminary cost analysis,2 each of these equipment specifications is
expected to have similar costs for the amount of benzene emissions
reduced.
Monthly monitoring for detection of leaks from pumps, drains;
compressors, and valves would also be required in this regulatory
alternative. The purpose of the increased frequency of monitoring is
to reduce emissions from residual leaking sources (i.e., those sources
that are found leaking and are repaired and recur before the next
inspection, and those sources that begin leaking between inspections).
Weekly visual inspections of pump seals would be required as discussed
for Regulatory Alternative II.
6.3.4 Regulatory Alternative IV .
Regulatory Alternative IV includes equipment specifications that
are expected to have greater costs for the amount of benzene emisions
reduced than for those included in Regulatory Alternative III. Double
seals would be required on pumps and compressors in this regulatory
alternative in addition to the equipment requirements for the sampling
systems, gas service safety/relief valves, degassing vents, accumulator
vessel vents, and open-ended valves specified in Regulatory
Alternative III. Diaphragm and sealed-bellows valves are not included
because the expected cost for the amount of benzene emissions reduced,
based on a preliminary cost analysis, was much greater than that for
double seals. In addition to these equipment specifications, drains
and valves would be required to be monitored for leaks each month, as
in Regulatory Alternative III.
6-7
-------
6.3.5 Regulatory Alternative V
This regulatory alternative would require leakless emission
control equipment for the sources listed in Table 6-2. In addition to
the equipment specifications discussed for Regulatory Alternative IV,
this regulatory alternative would require installation of diaphragm or
sealed-bellows type valves, "and woul~ require drains to be enclosed.
All of these sources would, therefore, be controlled to the maximum
degree, and leaks would virtually be eliminated from these sources.
6.3.6 Regulatory Alternative VI
This regulatory alternative would require the elimination of all
benzene fugitive emissions from the affected industry. Although the
equipment specifications required in Regulatory Alternative V would
virtually eliminate such emissions from equipment handling greater
than 10 weight percent benzene, there would still be some emissions
from equipment handling less than 10 weight percent benzene.
Three approaches to totally eliminating benzene fugitive emissions
were considered. These are: (I) to require the use of leakless technology
for all equipment handling benzene-containing streams, (2) to require
the use of substitute fe~dstocks, thus eliminating the use of benzene,
and (3) to prohibit the production or consumption of benzene.
The use of 1eakless technology could eliminate most benzene
fugitive emissions. However, there would still be some benzene emissions
from spills and occasional equipment failure.
The use of substitute feedstocks could be effective for some
operations; for example, n-butane could be used in the production of
maleic anhydride instead of using benzene. This approach could not be
used for all benzene-consuming processes, however, since there are no
substitutes for benzene in some cases.
The only approach that could totally eliminate benzene fugitive
emissions is the prohibition of all benzene-producing and consuming
. .
processes. This appro~ch, however, would lead to the shutdown of all
refineries and a number of chemical plants because of the presence of
benzene in refinery feedstocks and the lack of available substitutes
for benzene in many chemical plant operations.
6-8
-------
6.4 REFERENCES
1. Hustvedt, K.C., R.A. Quaney, and W.E. Kelly. Control of
Volatile Organic Compound leaks from Petroleum Refinery Equipment.
U.S. Environmental Protection Agency. Research Triangle Park, N.C.
Report Number EPA 450/2-78-036, June 1978. 72 p.
2. Memo with attachments from Umlauf, G.E., Pacific Envirpnmental
Services, Inc., to EPA Docket (No. A-79-27). December 1979.
9 p.
6-9
-------
7.0 ENVIRONMENTAL IMPACT
7.1
INTRODUCTION
The environmental impacts for the regulatory alternatives presented
in Chapter 6 are discussed in this chapter. Both beneficial and
adverse impacts are assessed for air and water pollution, solid waste,
and energy use. Included are the derivations of controlled benzene
emission factors for the affected equipment. Total benzene emissions,
incremental benzene emission reductions for the regulatory alternatives,
and projected future benzene emissions are presented. Other environmental
concerns are discussed, including irreversible and irretrievable
commitment of resources as well as the environmental impact of delayed
,regulatory action.
7.2 AIR QUALITY IMPACTS
7.2.1 Development of Benzene Emission Levels
In order to estimate the impacts of the regulatory alternatives
on benzene emission levels, emission factors for the model plants were
determined for each regulatory alternative. Controlled VOC emission
. .
factors for Alternatives II, III, IV, and V are presented in Tables 7-1
through 7-4. .
Alternative I represents baseline emissions and includes no new
regulatory specifications. Regulatory Alternative VI represents no
allowable benzene fugitive emissions from refinery and SOCMI sources.
As discussed in Chapter 6, neither of these alternatives is being
considered as a viable control option~
The factors used to estimate the emissions were calculated for
each type of equipment using the methodology presented in Chapter 4.
The controlled emission factors are calculated by multiplying the
uncontrolled factors for each type of equipment by a set of correction
factors, which account for imperfect repair, non-instantaneous repair,
and the occurrence or recurrence of leaks between inspections.
7-1
-------
Table 7-1. CONTROLLEO vot EMISSION FACTORS FOR REGULATORY ALTERNATIVE II
Uncontro 11 ed Contro 11 ed
Emissi~n C9rrection Factors Control Emission
Inspection Factor AC Bd Ce Of Efficiency Factorg
Source Intervala (Kg/hr) (AxBxCxO) (Kg/hr)
Pumps Yearly 0.12h 0.87 0.80 0.98 0.92 0.63 0.044
Valvesi
Gas Quarterly 0.021 0.98 0.90 0.98 0.99 0.86 0.003
Liquid Yearly O.OlOh 0.84 0.80 0.98 0.94 0.62 0.004
Gas Service -
Pressure Relief
Devices -Quarterly 0.16 0.69 0.90 0.98 0.98 0.60 0.064
Dra ins Yearly 0.032 0.46 0.80 0.98 0.91 0.33 0.021
Compressors Quarterly 0.44 0.84 0.90 0.98 0.97 0.72 0.123
aFrom Table 6-2.
bFrom Table 3-1.
cTheoretical maximum control efficiency -- From Table 4-2.'
d \
Leak Occurrence and recurrence correction factor -- AS2umed to be 0.80 for yearly inspection, 0.90 for
quarterly inspection, and 0.95 for monthly inspection.
eNon-instantaneous repair correction factor -- for a lS-day maximum allowable repair time~ the 7.5-day
average repair time yields a 0.98 yearly correction factor: [365 - (15/2)] + 365 = 0.98. -
fImperfect repair correction factor -- From Table 4-3, calculated as 1 - (f + F), where f = average 1 2 I
emission rate for sources at 1000 ppm and F = average emission rate for sources greater than 10,000 ppm.' I
I
gControlled emission factor = uncontrolled emission factor x [1 - (A X B X C X D)]. I
hEmission factor for light liquid streams is used. (Reference 1) I
iEmission factors for in-line and open-ended valv~s are identical. since emissions from the open end would be._--
essentially eliminated by a cap,- plug,_blind orsecond_valve:.____-------------------- --
[
7-2
-------
Table 7-2. CONTROLLED vat EMISSION FACTORS FOR REGULATORY ALTERNATIVE III
Uncontrolled Controlled
Emissi~n Correction Factors. Control Emission
Inspection Factor AC Sd Ce Df Efficiency Factorg
Source Intervala (Kg/hr) (AxBxCxD) (Kg/hr)
Pumps h 0.12j 0.87 0.95 0.98 0.92 0.75 0.030
Monthly
k
Valves
Gas Monthly 0.021 0.98 0.95 0.98 0.99 0.90 0.002
Liquid Monthly O.OlOj 0.84 0.95 ' 0.98 0.94 0.74 0.003
Gas Service
Pressure Relief i
Devices None 0.16 NA NA NA NA 0.0
Dra ins Monthly 0.032 0.46 0.95 0.98 .0.91 0.39 0.020
Compressors Monthly 0.44 0.84 0.95 0.98 0.97 0.76 0.106
aFrom Table 6-2.
bFrom Table 3-1.
cTheoreticaJ maximum control efficiency -- F~om Table 4~2.1
dLeak Occurrence and recurrence co~rection factor -- As~umed to be 0.80 for yearly inspection, 0.90 for
quarterly inspection, and 0.95 for monthly inspection. .
eNon-instantaneous repair correction factor -- for a 15-day maximum allowable repair timeA the 7;5-day
average repair time yields a 0.98 yearly correction factor: [365 - (15/2)] + 365 = 0.98.~ .
fImperfect repair correction factor -- From Table 4-3, calculated as 1 - (f + Fl, where f = average
emission rate for sources at 1000 ppm and F = average emission rate for sources greater th~n 10,000 ppm.1 ,2
gContro1led emission factor = uncontrolled emission factor x (1 ~ (A X B X C X D)].
~For pumps, instrument monitoring should be supplemented with weekly visual inspections for liquid leakage.
lContro1 equipment for ~his source is specifice~ in Table 6-2, and 100 percent control efficiency is assumed.
Therefore, the correctlon factors are not app11cab1e, and there are essentially no fugitive emissions from
the source.
jEmission factor for light liquid streams is used. (Reference 1)
kEmission factoTs for in-line and open-end~d valve$ are idp.ntical.
essentially eliminated by a cap, blind, plug or second valve.
I
I
since emissions from the open end would be i
-' ... "".l
7-3
-------
Table 7-3. CONTROLLED VOC EMISSION FACTORS FOR REGULATORY ALTERNATIVE IV
Uncontrolled Contro 11 ed
Emission Correction Factors .',,', Control Emission
Inspection Factorb AC Bd Ce Df Efficiency Factorg
Source Interva1a (Kg/hi") (AxBxCxD) (Kg/hr)
Pumps Noneh 0.12 i NA NA NA NA 0.0
I
,
Va1vesj :
. .
Gas Monthly 0.021 0.98 0.95 0.98 0.99 0.90 0.002
Liquid Monthly O.OlOi 0.84 0.95 0.98 0.94 0.74 0.003
Gas Service
Pressure Relief h
Devices None 0.16 NA . NA NA NA 0.0
Ora i ns Monthly 0.032 0.46 0.95 0.98 0.91 0.39 0.020
Compressors None h 0.44 NA NA NA NA 0.0
aFrom Table 6-2.
bFrom Table 3-1.
cTheoretica1 maximum control efficiency -- From Table 4~2.1
d' . .
Leak Occurrence and recurrence correction factor -- As~umed to be 0.80 for yearly inspection, 0.90 for
quarterly inspection, and 0.95 for monthly inspection.
eNon-instantaneous repair correction factor -- for a 15-day maximum allowable repair time~ the 7.5-day
average repair time yields a 0.98 yearly correction factor: [365 - (15/2)] .f 365 = 0.98.t:
flmperfect repair correction factor -- From Table 4-3, calculated as 1 ~ (f + F), where f = average 1 2
emission rate for sources at 1000 ppm and F = average emission rate for sources greater than 10,000 ppm. '
gContro11ed emission factor ='uncontro11ed emission factor' x [1 - (A X B X C X D)].
hContro1 equipment for ~his source is specified in Table 6-2, and 100 percent control efficiency is assumed.
Therefore, the correctlon factors are not applicable, and there are essentially no fugitive emissions from
the source. .
iEmission factor for light liquid streams is used. (Reference 1)
jEmission factors for in-line and open-ended valves are identical. since emissions from the ope~ end would be
essentially eliminated by a cap, blind, plug or second valve.
7-4
-------
Table 7-4. CONTROLLED vot EMISSION FACTORS FOR REGULATORY ALTERNATIVE V
Uncontro 1 i ed Controlled
Emissi~ Correction Factors ..'. Control Emission
Inspecti~n Factor AC ad Ce Of Efficiency Factorg
Source Interval (Kg/hr) (AxaxCxD) (Kg/hr)
h i
Pumps None 0.12 NA NA NA NA 0.0
Valvesj Noneh
Gas 0.021 'NA NA NA NA 0.0
Liquid ' Noneh 0.010 i NA NA NA NA 0.0
Gas Service
Pressure Relief "Noneh
Devices 0.16 NA NA NA NA 0.0
Ora i ns h 0.032 NA NA NA NA 0.0
None
Compressors h 0.44 NA NA NA NA 0.0
None --
aFrom Table 6-2.
bFrom Table 3-1.
cTheoretical maximum control effi~iency -- From Table 4~2.1
dLeak Occurrence and recurrence correction factor -- As~umed to be 0.80 for yearly inspection, 0.90 for
quarterly inspection, and 0.95 for monthly inspection. '
eNon-instantaneous repair correction factor -- for a 15-day maximum allowable repair timeA the 7.5-day
average repair time yields a 0.98 yearly correction factor: [365 - (15/2)] + 365 = O.98.~
fImperfect repair correction factor -- From Table 4-3, calculated as 1 ~ (f + F). where f = average
emission rate for sources at 1000 ppm and F = average emission rate for sources greater than 10,000 ppm.l.2
gControlled emission factor = uncontrolled emission factor x [1 - (A X a X C X D)].
hControl equipment for this source is specified in Table 6-2, and '100 percent cont~ol efficiency is assumed.
Therefore, the correction ,factors are not ,applicable, and there are essentially no fugitive emissions from
the source.., '
iEmission factor for light liquid streams is used. (Reference 1)
jEmission factors for in-line and,open-ended valves are identical, since emissions from the open end would be
essentially eliminated by a cap, plug, blind or second valve.
7-5
-------
!- ------------------------------
For each regulatory alternative and model plant, the number of
components handling greater than 10 percent benzene (Table 6-1) was
multiplied by the controlled VOC emission factor for each source
(Tables 7-1 through 7-4). SumMing the VOC emissions from these sources
yielded a total VOC emission estimate for the over 10 percent equipment
in each model unit. Since all sources do not handle pure benzene, it
was necessary to convert the VOC emission figure to benzene emissions..
This was accomplished by first estimating an average percent benzene
for over 10 percent benzene equipment in each process by examining
flow diagrams and material balances. Table 7-5 shows the resulting
average percent benzene for each process. Knowing the existing
distribution of processes within each model unit category (also shown
in Table 7-5), it was then possible to calculate a weighted average
percent benzene for over 10 percent benzene equipment in each model
unit. From the table, these weighted averages were calculated to be
64, 54, and 75 percent for model units A, B, and C, respectively.
Application of these factors,to the VOC emissions yielded a benzene
emission estimate for over 10 percent benzene equipment in each model
unit and for each regulatory alternative. An example calculation is
given below for over 10 percent benzene pumps in Model Unit A under
Regulatory Alternative II.
Benzene emissions from
>10 percent benzene (Bz)
Pumps in Model Unit A
Under Regulatory
Alternative II (Mg/yr)
>10 percent Bz Pumps
= 5 in Model Unit A
(from Table 6-1)
x 0.044
Kg vac per hour
per pump for
Regulatory
Alternative II
(from Table 7-1)
x
0.635
Wei ghted factor
of percent benzene
in Model Plant A
x
8400
Hours
per
year
x
0.001
~1g
per
kg
= 1.17 Mg Benzene per year from
>10 percent Bz Pumps in Model Unit A
7-6
-------
Table 7-5. CALCULATION OF WEIGHTED
PERCENT BENZENE FOR OVER
TEN PERCENT EQUIPMENT IN MODEL UNITSa
Percent Average Percent Benzene
Number of of Total for Over '10 Percent Weighted
Units Model Equipment in Each Percent
(1980) b Units Processc 8enzene
MODEL UNIT A
Ethylene (1 Unit) 41 28.3 33 9.34
Ethyl benzene 22 15.2 71 10.79
Benzene (Dealkylation) 20 13.8 75 10.35
Styrene 19 13.1 30 3.93
Cumene 16 11.0 100 11.00
Cyclohexane 12 8.3 100 8.30
Maleic Anhydride 10 6.9 100 6.90
Benzenesulfonic Acid 5 3.4 85 2.89
TOTAL 145 100.0 63.50
MODEL UN IT B
Benzene (Reformer) 49 68.0 45 30.60
Chlorobenzene 12 16.7 100 16.70
Ethylene (2 or 3 Units) 7 9.7 33 3.20
Alkylbenzenes 4 ~ 65 3.64
TOTAL 72 100.0 54.14
HODEL UN IT C
Benzene (Pyrolysis Gas) 12 50.0 57 28.50
Nitrobenzene 10 41.6 100 41.60
Hydroquinone 1 4.2 75 3.15
Ethylene (4 or 5 Units) -1 ~ 33 1.39
TOTAL 24 100.0 74.64
7-7"
-------
Table 7-5. CALCULATION OF WEIGHTED
PERCENT BENZENE FOR OVER
TEN PERCENT EQUIPMENT IN MODEL UNITSa (Concluded) ,
NOTES
aThe Resorcinol process is' not included in this analysis. Resorcinol
is produced at only one facility in the country, and the process
would have only one feed pump and one control valve in benzene
- service. Although this equipment would be covered by the regu-
lation, this process is much smaller than the smallest model unit
and, hence, would not significantly impact the analysis. .
bInc1udes plants under construction expected to be completed in
1 980.
cThe average' percentage of benzene for allover 10 percent equip-
ment in each product process is derived from observations of flow
diagrams, material balances, and equipment counts for each
manufacturing process. For each of the products listed, the
percentage of benzene represents a weighted average of the number
of pieces of equipment and the concentrations of benzene in each
product stream:
Weighted ~ Number of Pi eces
Percent =, of Equipment
~enzene Total Pieces
Concentration )
x in Each Stream
of Equipment
Example:
If there are 5 pumps at 100 percent benzene and 5 pumps
at 50 percent benzene, then the weighted average of
percent benzene for the equipment can be expressed as
Weighted
Percent =
Benzene
(5 x 100) + (5 x 50)' = 75 Percent
10
7-8
-------
1-
Using a similar technique, it was also possible to estimate
benzene emissions from equipment handling streams with less than
10 percent benzene. VOC emissions from these sources were calculated
using the same VOC emission factors as were used for over 10 percent
equipment. The average percent benzene for the under 10 percent
sou.rces was estimated to. give the benzene emi ss ions from under
10 percent sources for each process. A weighted average of the
processes was then applied to give the "residual" benzene emissions
. from the under 10 percent equipment for each model unit. These
emissions were calculated to be 2.94, 4.64, and 4.03 megagrams per
year for Model Units A, B, and C, respectively. It should be noted
that these emissions would not be affected by this regulation and,
therefore, are included .in the benzene emission totals for each of the
regulatory alternatives.
Table 7-6 presents the to~al benzene emission estimates for each
model unit and each regulatory alternative. The table also compares
the relative control effectiveness of Regulatory Alternatives II
through V with Alternative I, which represents baseline emissions. .
Applying Alternative II reduces emissions from the baseline level by
approximately 57 percent. For Alternatives III through V, benzene
emissions are reduced by approximately 73, 77, and 90 percent,
respectively. Table 7-7 presents total national benzene emissions for
each alternative based on the number of model units expected to be in
operation in 1980 and the benzene emissions per model unit from Table 7-6.
7.2.2 Future Benzene Emissions
In order to assess potential future impacts of
alternatives, benzene emissions were projected over
(1981-1990). To estimate future benzene emissions,
were made:
the regulatory
a ten-year period
two assumptions
1.
The capacity utilization of all benzene products remains the
same over the projected ten-year period.
2.
An increase in demand is accounted for by
units, expanding the capacity of existing
renovating existing units.
buil ding new
units, or by
7-9
-------
Table 7-6. TOTAL BENZENE EMISSIONS AND RELATIVE CONTROL EFFECTIVENESS
FOR THE REGULATORY ALTERNATIVES AND MODEL UNITSa
~todel Unit A Model Unit B Model Unit C
Percent percentc percentc
Benzene Contro 1 c Benzene Control Benzene Control
RegulatorYb Emissions Relative to Emissions Relative to Emissions Relative to
Alternative (Mgfyr) Alternative I (Mgfyr) Alternative I (Mgfyrl Alternative I
18.62 44.71 97.07
II 8.67 53 19.33 57 38.23 61
III 5.86 69 12.13 73 21.23 78
IV 5.06 73 10.09 77 16.53 83
V 2.94d 84 4.64d 90 4.03d .96
aTotal benzene emissions include emissions from greater than 10 percent benzene equipment and residual
benzene (less than 10 percent).
bThe regulatory alternatives are summarized in Table 6-2. Alternative I represents baseline emissions.
cPel"cent decrease in benzene emissions from Alternative 1.
dEmissions are from less than 10 percent benzene equipment.
,
,
,
I.
7-10
-------
Table 7-7. TOTAL NATIONAL BENZENE EMISSIONS FROM REFINERY AND
SOCtll PROCESSESa
Emission. Reduction
From Alternative I
Regulatory Emissions
Alternative (Ng/yr)b
IC 8250
II 3570
III 2230
IV 1860
V 860d
Mg/yr
Percent
4680 57
6020 73
6390 77
7390 90
aAny process that has equipment handling over 10 percent benzene streams is
'included in this analysis. The emission totals include the emissions from
over 10 percent and under 10 percent equipment for each of these processes.,
bCalculated for each regulatory alternative as the summation of the product
of the' annual emissions (Mg/yr) from each model unit and the number of
existing refinery and chemical units that are represented by that model
unit. The number of units expected to be operating in 1980 is as follows:
Model Unit A ~ 145
Model Unit B -- 72
Model Unit C -- 24
cRepresents baseline emissions (the industry in the absence of new regu-
lations) .
dResidual from less than '10 percent benzene equipment.
7-11
, i
-------
In order to calculate benzene emission levels over the 10-year
period, a weighted growth rate for each model unit was calculated
based on the number of existing units in the base year of the analysis
(1980), and the average annual growth rate for each production process.
For each model unit, Table 7-8 lists the number of plants estimated in
operation in 1980 and the average annual growth rates for each product.
The weighted average growth rates for each model unit are approximately
6,2, and 10 percent for Model Units A, B, and C, respectively.
In the determination of future impacts of benzene fugitive emissions
from refineries and chemical plants, a distinction is made between new
unit growth and growth as a result of unit replacement. For each
model unit, the weighted average annual growth rate calculated in '
Table 7-8 represents a net increase attributable to overall industrial
growth. The number of units constructed in any year equals this net
increase plus the number of units constructed to replace ones that
cease production due to obsolescence, deterioration, or other factors.
For each model unit, the number of new units (N) constructed in
any period (X years) is calculated from the number of existing units
(E) in 1980 and the projected average annual growth rate (i) for the
model unit, using a rearrangement of the formula for simple interest
compounded annually, as follows:
N = E(l + i)x - E
As ~n example, to calcul~te the projected number of new Model A units
constructed between 1981 and 1985, the above formula is applied as
fOll ows:
N1981-1985 = 145 (1 + .0576)5 - 145
= £L new pl ants.
Assuming an average unit life of 20 years,
(R) is calcul~ted by th~ following:
- N
R - rf - N,
where ~ = the number of new units and
rf = a replacement factor as follows:
rf = 1 - 1
(1 + i)20.
the number of repl acements,
7-12
-------
Table 7-8. NUMBERS OF UNITS ESTIMATED TO MEET 1980
DEMAND FOR BENZENE AND BENZENE
DERIVATIVES BY MODEL UNITS
Estimated
Number of
Units in Percent
"Year 0" of Total
Model A (1980) Units
Ethylene (1 unit)
Ethyl benzene
Benzene (Dealkylation)
Styrene
Cumene
Cyclohexane
Maleic Anhydride
Benzenesulfonic Acid
Total
Model B
Benzene (Reformer)
Chlorobenzene
Ethylene (2 or 3 units)
Alkyl benzenes
Total
Model C
Benzene (Pyrolysis Gas)
Nitrobenzene
Hydroquinone
Ethylene (4 or 5 units)
Total
41
22
20
19
16
12
10
_5
49
12
7
4
72
12
10
1
. 1
24
28.3
15.2
13.8
13.1
11.0
8.3
6.9
3.4
100.0
68.0
16.7
9.7
5.6
100.0
50.0
41.6
4.2
4.2
100.0
Average Annual,
Growth RateJ
(Percent)
5.5
6.0
3.7
6.0
7.5
5.0
11.0
0.0
1.5
1.5
5.5
2.0
14.1
6.0
0.0
5.5
Weighted Annual
Growth;- Rate
(Percent;"
1.56
0.91
0.51
0.79
0.82
0.41
0.76
0.00
5.76
1.03
0.25
0.53
0.11
T92
7.05
2.50
0.00
0.23
9778"
*Does not include replacement of old units.
7-13
-------
The results are presented in Table 7-9, which shows the projected
growth of new and replacement units over a 10-year period (1981-1990).
Using the benzene emission values for each regulatory alternative
and model unit from Table 7-6, the expected benzene emissions that
will be contributed by new units and replacements can now be calculated
by Multiplying the emissions (Table 7-6) by the number of new units
and replacements estimated to be operating between 1981 and 1990
(Table 7-9). For each alternative, emissions from Model Units A, 8,
and C are summed to obtain the total benzene emissions for the regulatory
alternative. Table 7-10 presents anticipated benzene fugitive emissions
for new units and replacements for each alternative over the ten-year
period.
7.3 WATER POLLUTION IMPACT
Implementation of any of the regulatory alternatives would result
in slight positive benefits to water quality, depending on the specific
control requirements of the alternative. The regulatory alternatives
would not cause the organic streams being handled by affected equipment
to contact water. Neither would benzene emissions be physically
removed (as in the case of wet scrubber control for particulates).
Rather, emissions are expected to be contained. Therefore, implementing
any of the benzene fugitive emissions alternatives would not adversely
impact water quality. At best, the quality of runoff water might
improve slightly due to the improved containment of benzene and other
volatile organic compounds.
Specifically, provisions of Regulatory Alternative II would
require leak detection and repair for some liquid service equipment.
Repair of this equipment would require that process liquids be drained'
or flushed, thus generating a small negative wastewater impact.
Alternative III would require that sample purge material be
returned to the system or contained. This requirement would result in
a small positive wastewater impact, since in some plants, this
material is currently routed to a drain system.
7-14
-------
l I
I
Table 7-9. NUMBER OF PROJECTED NEW UNITS AND REPLACEMENTS
BETWEEN 1981 AND 1990a
Model Unit
Year A B C
1981
Newb 8 1 2
ReplacementC 4 0 0
1982
b 17 3 5
New
ReplacementC 8 1 1
1983
b 27 4 8
New
ReplacementC 13 2 1
1984
Newb 36 6 11
c
Replacement 17 3 2
1985
Newb 47 7 14
ReplacementC 23 4 3
1990
b 109 15 37
New
ReplacementC 53 7 7
7-15
-------
NOTES FOR TABLE 7-9
aprojections of new and replacement unit growth over a 10-year
period were based on the following:
A
B
C
Number of
Existing Plants
(E) in 1980
145
72
24
Projected
Growth Rate (i)
(% per year)
5.76
1.92
9.78
Replacement
Factor
(rf)
0.674
0.688d
0.845
Model
. Unit
bprojected number of new plants (N) was determined from the following:
N = E(l + i)x - E
where
E = number of existing plants in 1980
i = projected average annual growth rate
x = year from 1980 (1,2,3...10)
CProjections of replacements were based on the replacement factor (rf):
rf = 1 -
1
(1+i )20,
assuming an average unit life of 20 years.
(r) is calculated from the following:
N
R = - - N,
rf
Number of replacements
where N = number of new plants
and rf = replacement factor.
dThe replacement factor of 0.688 for Model Unit B reflects a 6 percent
annual growth rate over the last 20 years (1960-1980) rather than the
1.92 percent annual growth rate over the next 10 years.
7-16
-------
Table 7-10. ESTIMATED BENZENE FUGITIVE EMISSIONS FROM
NEW UNITS AND REPLACEMENTS BETWEEN
1981 and 1990 (Mgfyear)
Regulatory Alternative
Year I II III IV V
1981
New 388 165 101 84 36
Replacement 74 35 23 20 12
1982
New 936 397 242 199 84
Replacement 291 127 80 67 32
1983
New 1458 617 377 309 130
Replacement 429 190 122 102 52
1984
New 2006 849 517 425 178
Replacement 645 282 178 149 72
1985
New 2547 1078 658 540 227
Replacement 898 391 247 206 98
1990
New 6292 2649 1606 1315 539
Replacement 1979 862 544 455 217
7-17
-------
Under Regulatory Alternative IV or V~ a double mechanical seal!
. degassing vent arrangement would reduce product leakage from pumps and
thus result in a slight positive impact on water quality. Implementation
of Regulatory Alternative IV could also result in a negative impact on
water quality from the operation of possible control devices which
"capture" the fugitive VOC's from the degassing vent. If a carbon
adsorption device were used~ for example~ a wastewater stream containing
suspended solids and small quantities of dissolved organics would be
produced during the carbon regeneration process if the carbon is
regenerated at the unit. ' Th~ use of a refrigeration process as the.
control device would possibly result in a condensate containing dissolved
organics. The wastewater flow. rates would be quite small since the
amount of VOC being removed is small ~ and this wastewater would generally
be suitable for treatment in existing wastewater treatment systems.
7.4 SOLID WASTE IMPACT
The regulatory alternatives will contain benzene in the vapor and
liquid states. This contained material is not expected to generate a
solid waste. The solid wastes associated with the alternatives are
replaced mechanical seals~ packing~ rupture disks~ and valves. In
Regulatory Alternative II~ capping open-ended valves would result
in no solid waste impacts. Implementation of Regulatory Alterna-
tive III~ which requires the installation of rupture disks on
relief valves~ a closed-loop sampling system, and seals on open-ended
valves and lines would have a negligible impact. For Regulatory
Alternative IV, double seals would be retrofitted on affected pumps
and compressors. Existing packing materials and single mechanical
,seals may not be reusable and hence would be. discarded. As the double
seals wear out, they also would be replaced and the old seals would be
discarded. This material would have a very minor impact on the quantity
of solid waste generated by the plant, however, since existing single
seals would have worn out and been replaced also. Therefore, Regulatory
Alternative IV would have a negligible negative impact on solid waste.
In addition to the equi~ment specifications of Regulatory Alternative IV,
Regulatory Alternative V would require that diaphragm or sealed-bellows
type valves be installed and that drains be enclosed. The solid wastes
7-18
-------
generated by
material, as
unrecyc1ab1e
the replacement of single mechanical seals and packing
in Regulatory Alternative IV, and the disposal of
valves would have an insignificant impact on solid waste.
7.5 ENERGY IMPACT
The controls necessary for the implementation of Regulatory
Alternatives II through V would require no significant increase in
energy consumption. The application of double seals, however, would
require a minimal increase in energy usage over single seal operation
because of the slight increase in seal/shaft friction and because of
the energy required to operate the fluid flush system. Since the
product emissions do have an energy value, a net positive energy
impact is expected.
The average energy value of total VOC fugitive emissions from
refineries and the SOCMI is estimated as 8.62 x 106 jou1es/kg.14
Table 7-11 presents the energy savings over a five-year period that
result from the VOC fugitive emission reductions associated with
Regu1tory Alternatives II through V. Since Regulatory Alternative V
represents the most stringent option, it achieves the greatest emission
reduction by reducing uncontrolled fugitive emissions by 72,700 Mg
over a five-year period. These "recovered" VOC emissions have a total
energy value of 627 terajoules based on a heat value of 8.62 x 106
joules/kg. Assuming an energy value of 6.12 x 109 joules per barrel
of crude oil, the energy value of the recovered VOC fugitive emi~sions
is approximately 102,400 barrels of crude oil for the period 1981
through 1985 under Regulatory Alternative V. This represents an
average annual savings of 20,480 barrels of crude 0;1 over the five-year
period.
7.6 OTHER ENVIRONMENTAL CONCERNS
7.6.1 Irreversible and Irretrievable Commitment of Resources
Implementation of any of the regulatory alternatives is not
expected to result in any irreversible or irretrievable commitment of
resources. As previously noted, the regulatory alternatives should
help to save crude oil due to the energy savings associated with the
reductions in emissions. Materials used in double mechanical pump
seal mechanisms, such as tungsten carbide, will be committed, but the
7-19
-------
Table 7-11. ENERGY IMPACT OF BENZENE EMISSION
. REDUCTION FOR REGULATORY ALTERNATIVES
Reduction from Baseline Energy Value of Emission Crude Oil Equivalent ~f
Emissions Under Regu1a- Reductions Under Regu1a- Emission Reductions
tory A1ternativesa,b tory Alt~rnativesC , (barrels - bb1)
(Mg) (10 Btu)
Year II III IV V II III IV V II III IV V
I 1981 7.93 10.2 10.8 12.5 68.4 87.9 93.1 108 11 .2 1 4.4 15.2 17.6
1982 8.59 lLO 11 .7 13.6 74.0 94.8 101 117 12.1 1 5.5 16.5 19. 1
1983 9.16 1i:8 12.5 14.5 79.0 102 108 125 12.9 16.7 . 17.6 20.4 ,
1984 9.82 12.6 13.4 15.5 84.6 109 116 134 13.8 1 7.8 19.0 21. 9
..J 1985 10.5 13.5 14.3 16.6 90.5 116 123 143 14.8 1 9.0 20.1 23.4
"
.:J
5-year
Total 46.0 59.1 62.7 72.7 397 510 541 627 64.8 83.4 88.4 102
.>
aAlternattve I represents baseline emissions.
bEmission reduction calculated from VOC emissions per model unit and regulatory alternative as well as the total number of
units projected to be in operation between 1981 and 1985.
CEnergy value of benzene is based on 8.62xl06 joules (from conversion of 17,986 Btu/lb given in Ref. 14, p. 3-143).
dBased on 6.12x109 jou1es/barrell crude oil.
-------
amount of the material lost will be very slight, and although the
material is valuable, it is not particularly scarce. Other materials
used to manufacture piping, valves, rupture disks and line caps are
not scarce and will not be committed in significant quantities for any
of the regulatory alternatives.
7.6.2 Environmental Impact of Delayed Regulatory Action
As indicated above, implementation of any regulatory alternative
would only have minor impacts on water and solid wastes. Consequently,
delaying regulatory action will have essentially no impact on these
problems. However, a delay in implementing the alternatives will have
a greater impact on air pollution and associated energy impacts. The
air and energy impacts of delayed standards are shown in Table 7-11.
The emission reductions and associated energy savings shown would be
lost at the rates shown for each of the five years.
7-21
-------
7.7 REFERENCES
1. Wetherold, R.G. and L. P. Provost. Emission Factors and Frequency
of Leak Occurrence for Fittings in Reftnery Process Units.
Radian Corp. Austin, TX. For U.S. Environmental Protection
Agency. Research Triangle Park, N.C. Report No. EPA-600/2-79-044.
February 1979. .
2. Tichenor, B.A., K.C. Hustvedt, and R.C. Weber. Controlling
Petroleum Refinery Fugitive Emissions Via Leak Detection and
Repair. Draft. Symposium on Atmospheric Emissions from
Petroleum Refineries, Austin, TX. November 6, 1979.
3. Soder, S.L. CEH Product Review on Styrene. Chemical Economics
Handbook. Stanford Research Institute. Menlo Park, CA.
January 1977.
4. CEH Product Review on Cyclohexane. Chemical Economics Handbook.
Stanford Research Institute. Menlo Park, CA. February 1977..
5. Gunn, T.C., and K. Ring. CEH Marketing Research Report on
Benzene. Chemical ECQnomics Handbook. St~nford Research
Institute. Menlo Park, CA. May 1977.
6. CEH Product Review on Chlorobenzenes. Chemical Economics
Handbook. Stanford Research Institute. Menlo Park, CA.
July 1977.
7. CEH Product Review on Ethylene. Chemical Economics Handbook.
Stanford Research Institute. Menlo Park, CA. January 1978.
8. Cogswell, S.A. CEH Product Review on Resorcinol. Chemical
Economics Handbook. Stanford Research Institute. Menlo Park,
CA. October 1978.
9.
CEH Product Review on Aniline and Nitrobenzene. Chemical
Economics Handbook. Stanford Research Institute. Menlo
Park, CA. January 1979.
CEH Product Review on Linear and Branched Alkylbenzenes.
Chemical Economics Handbook. Stanford Research Institute.
Menlo Park, CA. January 1979.
Al-Sayyari, S.A., and K. Ring. CEH Product Review on Cumene.
Chemical Economics Handbook. Stanford Research Institute.
Menlo Park, CA. March 1979.
Ring, K., and S.A. Al-Sayyari. CEH Product Review on Ethyl-
benzene. Chemical Economics Handbook. Stanford Research
Institute. Menlo Park, CA.. March 1979.
Greene, R. U.S. Benzene Markets to Face Slower Growth.
Chemical Engineering. 85(3):62-64. January 30, 1978.
Perry, J.H. Chemical Engineer's Handbook. Fourth edition.
New York, McGraw-Hill Book Co., 1963. p. 3-143.
10.
11.
12.
13.
14.
7-22
-------
8.0 COST OF CONTROLS
8.1
I NTRODUCTI ON
In order to project the economic impact of the regulatory alternatives
on the petroleum refining and SOCMI industries, it is first necessary to
calculate the control costs of the regulatory alternatives. Installed
capital costs and annualized costs are estimated in this section for
each model unit and each regulatory alternative. In order to assure a
common cost basis, cost data from the various sources have been corrected
to 1979 values by means of appropriate cost inflation indicators from
the "Economic Indicators" sections of Chemical Engineering. For
background to this analysis, it is helpful to review information
presented in Chapter 6 describing the model units and regulatory
alternatives. For example, Table 8-1 restates the model unit equipment
handling over 10 percent benzene, and Table 8-2 reviews the inspection
intervals and equipment specifications for each regulatory alternative.
8.2 CAPITAL COST ESTIMATES
Capital cost expenditures are required for all of the regulatory
alternatives and model units. These costs will be incurred for the
purchase of monitoring instruments and control equipment. Two monitoring
devices will be purchased at each unit, regardless of the regulatory
alternative chosen. This minimum is required to allow for backup if
one unit is inoperative. Additional capital costs depend on the
number of pieces of affected facilities (potential leak sources) in
the model unit and the types of control equipment specified for the
regulatory alternative. To calculate these additional costs, data
presented in Table 8-3 were accumulated for monitoring and control
equipment.
Using the model
capital cost data in
unit parameters given in Table 8-1 and the
Table 8-3, capital costs for each model unit
8-1
-------
Ta'ble 8-1. MODEL UNIT
EQUIPMENT WITH >10 PERCENT BENZENE
Number of Components per Model Unit
Source Type ~1odel Aa b Model CC
. Model B
Pumps 5 15 25
Process d
Valves, Gas
Block 27 83 138
Control 3 8 13
Process Valves, L" odd
lqUl
Block 51 153 256.
Control 5 15 24
Relief Valves, Gas 3 9 16
Open-Ended Valves, Gas 33 7 12
Open-Ended Valves, Liquid 23 72 119
Drains 5 15 25
Sample Connections 9 26 44
aRepresents an average inve.ntory of equipment for production of
ethyl benzene, styrene, curi,ene, cycl ohexane, benzene sultoni c
acid, resorcinol, maleic anhydride, or 1 ethylene production'
. unit.' .
bRepresents an average inventory of equipment for production of
chlorobenzenes, linear c:.lkylbenzenes, or 2 or 3 ethylene produc-
tion units. '.
CRepresents an average inventory of equipment for production of
benzene, nitrobenzene, hydroquinone, or 4 or 5 ethylene produc-
tion units.'
dFrom Hydroscience, 6 percent of all valves are control valves;
69 percent of all valves are process valves; so, 6 + 69 or 8.7
percent of process valves are control valves (Ref. 2).
8;..2
-------
Table 8-2. MONITORING INTERVALS AND EQUIPMENT SPECIFICATIONS
FOR BENZENE FUGITIVE REGULATORY ALTERNATIVES
_._._~-- --- --- --- -
-..----------
Regulatory Alternativesa
lIb III IV V
SourceC,d Inspection Equipment Inspection Equipment Inspection Equi pment Inspection Equipment
Interval Specification Interval Specification I nterva 1 Specification Interval Specification
1. Pumps Yearlye None Monthlye None Nonee Double sealsf None Double sealsf
and controlled and controlled
degassing vent degassing vent
2. Process Valves
a. Gas Service Quarterly None Monthly None Monthly None None Di aphragm or
sealed bellows
valves
b. Liquid Service Yearly None Monthly None Monthly None None Diaphragm or
sealed bellows
valves
00 3. Safety/ReI ief Quarterly None None Rupture di sks None Rupture disks None Rupture di sks
I. Valves (Gas or tie into or tie into or tie into
W
Service) existing flare existing flare existin~ flare
4. Open-Ended Valves
a. Gas Service Quarterlyg Caps. blinds, Month lyg Caps, bl inds. Monthlyg Caps, blinds, None Di aphragm or
or plugs or plugs or pI ugs sealed bellows
valve plus caps,
blinds, or plugs
b. Liquid Service Yearlyg Caps, bl inds, Monthlyg Caps, blinds, Monthlyg Caps, bl inds, None Diaphragm or
or pI ugs or plugs or pI ugs sea I ed bellows
valve plus caps,
blinds, or plugs
5. Drains Yearly None Monthly None Monthly None None Enclosed active
drain systems
6. Sampling None None None Closed-loop None Closed-loop None Closed-loop
Connect ions sampling sampl i ng sampling
7. Compressors Quarterly None Monthly None None Double seals None Double seals
8. Product Accumu- None None None Tie into None Tie into None Tie into
lator Vessel closed control closed control closed control
system system system
-------
Table 8-2. MONITORING INTERVALS AND EQUIPMENT SPECIFICATIONS
FOR FUGITIVE BENZENE REGULATORY ALTERNATIVES (Concluded)
NOTES:
aRegulatory Alternative I (baseline) includes no new regulatory
specifications and, hence, is not included in this table. Regu-
latory Alternative VI is not included since it would require that
no benzene be emitted from refinery or SOCMI sources.
bAlternative II is equivalent to controls recommended in the
refinery CTG for fugitive VOC emissions.
CLiquid service safety/relief valves, flanges, wastewater separators,
vacuum-producing systems, process unit turnarounds, and cooling. .
towers are not routinely monitored. Wastewater separators,
vacuum-producing systems, process unit turnarounds, and cooling
towers are not included in this table since there are no available
c~ntrol technologies for these sources.
dFor all alternatives, the sources would handle organic streams
with over 10 percent benzene by weight.
eFor pumps, instrument monitoring would be "supplemented with weekly
visual inspections for liquid leakage. If liquid is noted to be
leaking from the pump seal, the pump seal will be repaired.
fA pressure-sensing device should be installed between the double
seals and should be monitored to detect seal failure.
gInspection applies to the valve.
8-4
-------
I
I
Table 8-3. CAPITAL COST DATA
(May 1979 Dollars)
Item
Reference
Capital Cost
1. Monitoring Instrument
2. Caps for Open-Ended Lines
3. Double Mechanical Seals
4. Seal Oil Recirculation
System for Double Seals
5. Degassing Vents
6. Rupture Disks for
Relief Valves
7. Closed-Loop Sampling
Connections
8. Sealed Bellows Valves
2 x 4,250 = $8,500
$ 50/cap
$590/pump (new)
$870/pump (retrofit)
$1,530/pump
1. 61 m of 5.1 cm
carbon steel pipe
( i ns ta 11 ed )
$2,700
2. 3 x 5.1 cm valves
( i ns ta 11 ed)
$940
3. 1 x 5.1 cm flame
arrestor
(installed)
$450
Total = $4,090/pumpd
$1,800/relief valve
(new)
$3,230jrelief valve
(retrofit)
$480 (new or retrofit)
$3,700/valve
(retrofit )
$2,500/valve (new)
1
a
2, 3, 4
b
2, 3, 4
b
2, 3, 4
c
2, 3, 4
c
2, 3, 4
c
2, 3, 4
2, 3, 4,'13, l4a,e
2, 3, 4a,d
5
8-5
-------
Table 8-3.
CAPITAL COST DATA (Concluded)
(May 1979 Dollars)
Item
Capi ta 1 . Cost
Reference.
9.
10.
Replacement Pump
$l,Ooo/drain
$3,000/pump
6f
7g
Sealed Drain Covers
NOTES:
aplant cost indices were used.
bPump and compressor cost indices were used.
CPiping, valve and fitting cost indices were used.
d .
No retrofit penalty.
eCost of rupture disks includes rupture disk, block valve, and
replacement of relief valve.
fConsists of sealed drain cover ahd sealed pump drain line.
gIn retrofitting double seals, it has been assumed that 10 percent
of the pumps will have to be replaced. Thus, the following
numbers would have to be replaced: .
. Model A - 1 Pump
Model B -2 Pumps
Model C -- 3 Pumps
8-6
-------
size are estimated. Table 8-4 includes capital costs in May 1979
dollar values for new and existing equipment handling more than 10
percent benzene by weight in their organic streams. Regulatory
Alternative I represents baseline emission control for units that are
assumed to need no additional. controls and, as such, will incur no
capital costs. Regulatory Alternative II includes capital costs for
the purchase. of monitoring instruments and the installation of caps on
open-ended valve lines. Capital costs per model unit (existing and
new) for Regulatory Alternative II are $10,300 (A), $13,800 (B), and
$17,300 (C). In addition to the controls for Regulatory Alternative II,
Regulatory Alternative III provides for the installation of rupture
disks on gas-service relief valves that vent to the atmosphere and the
instailation of closed-loop sampling connections. Capital costs would
increase to the following for existing units: $24,300 (A), $55,400
(B), and $90,100 (C). For new units capital costs would be $20,000
(A), $42,500 (B), and $67,200 (C). The high cost of retrofitting
rupture disks on relief valves explains the difference in capital
costs between existing and new units. Regulatory Alternative IV
specifies that pumps and compressors be equipped with double
mechanical seals for which degassing vents on pump seal reservoirs
are installed. These costs added to the ones for Regulatory
Alternative III include the following for existing units:
$56,900 (A), $152,900 (B), and $252,500 (C). For new units capital
costs would be $51,200 (A), $135,800 (B), and $222,600 (C). Alterna-
tive V, which requires leak-less emission control equipment and thus
maximum control, includes additional control costs for installing
diaphragm or sealed-bellows type valves as well as sealing drain
covers. Capital costs for each existing model unit would be increased
to the following: $504,100 (A), $1,512,200 (B), and $2,520,200 (C).
Capital costs for new model units would be $350,200 (A), $1.052,300
(B), and $1,754,100 (C).
8.3 ANNUALIZED COST ESTIMATES
8.3.1 Derivation of Annualized Cost Estimates
Annualized cost estimates are given in six categories:
8-7
-------
Table 8-4. CAPITAL COST ESTIMATES PER MODEL UNIT
(Thousands of May 1979 Dollars)
Regulatory Alternative
Capital Cost Item I II III IV V
Model Unit A (Existing)
1. Monitoring Instrument 8.5 8.5 8.5
2. Caps for Open-Ended
Valve Lines 1.8 1.8 1.8 1.8
3. Double Mechanical Seals 4.4 4.4
4. Seal Oil Recirculation
System for Double
Mechanical Seals 7.7 .7 .7
5. Vents for Seal Oil
Degassing Reservoirs 20.5 20.5
6. Replacement Pumps 3.0
7. Rupture Disks for
Relief Valves 9.7 9.7 9.7
8. Closed Loop Sampling
Connections 4.3 4.3 4.3
9. Sealed Bellows Valves 447.7
10. Hard Piping and Drain
Covers 5.0
TOTAL 0.0 10.3 24.3 56.9 504.1
8-8
-------
Table 8-4. CAPITAL COST ESTIMATES PER MODEL UNIT (Continued)
(Thousands of May 1979 Dollars)
Regulatory Alternative
Capital Cost Item I II III IV V
Model Unit A (New)
1. Monitoring Instrument 8.5 8.5 8.5
2. Caps for Open-Ended
Valve Lines 1.8 1.8 1.8 1.8
3. Double Mechanical Seals 3.0 3.0
4. Seal Oil Recirculation
System for Double
Mechanical Seal s. 7.7 7.7
5. Vents for Seal Oil
Degassing Reservoirs 20.5 20.5
6. Rupture Disks for
Relief Valves 5.4 5.4 5.4
7. Closed Loop Sampling
Connections 4.3 4.3 4.3
8. Sealed Bellows Valves 302.5
9. Hard Piping and Drain
Covers 5.0
TOTAL 0.0 10.3 20.0 51.2 350.2
8-9 '.
Ii
I
-------
Table 8-4. CAPITAL COST ESTIMATES PER MODEL UNIT
(Thousands of May 1979 Dollars)
Regulatory Alternative
Capital Cost Item I II III IV V
Model Unit B (Existing)
1. Monitoring Instrument 8.5 8.5 8.5
2. Caps for Open-Ended
Val ve Lines 5.3 5.3 5.3 5.3
3. Double Mechanical Seals 13.1 13.1
4. Seal Oil Recirculation
System for Double
Mechanical Seals 23.0 23.0
5. Vents for Seal Oil
Degassing Reservoirs 61.4 61.4
6. Replacement Pumps 6.0
7. Rupture Disks for
Relief Valves 29.1 29..1 29.1
8. Closed Loop Sampling
Connections 12.5 12.5 12.5
9. Sealed Bellows Valves 1346.8
10. Hard Piping and Drain
Covers 15.0
TOTAL 0.0 13.8 55.4 152.9 1512.2
8-10
-------
T----~
Table 8-4. CAPITAL COST ESTIMATES PER MODEL UNIT (Continued)
(Thousands of May 1979 Dollars)
Regulatory Alternative
Capital Cost Item I II III IV V
Model Unit B (New)
1. Monitoring Instrument 8.5 8.5 8.5
2. Caps for Open-Ended 5.3 5.3
Valve Lines 5.3 5.3
3. Double Mechanical Seals 8.9 8.9
4. Seal Oil Recirculation
System for Double
Mechanical Seals 23.0 23.0
5. Vents for Seal Oil
Degassing Reservoirs 61.4 61.4
6. Rupture Disks for
Relief Valves 16.2 16.2 16.2
7. Closed Loop Sampling
Connections 12.5 12.5 12.'5
8. Sealed Bellows Valves 910.0
9. Hard Piping and Drain
Covers 15.0
TOTAL 0.0 13.8 42.5 135.8 1052.3
8-11
-------
Table 8-4. CAPITAL COST ESTIMATES PER MODEL UNIT
(Thousands of May 1979 Dollars)
Regulatory Alternative
Capital Cost Item I II III IV V
Hodel Unit C (Existing)
1. Monitoring Instrument 8.5 8.5 8.5
2. Caps for Open-Ended
Valve Lines 8.8 8.8 8.8 8.8
3. Double Mechanical Seals 21.8 21.8
4. Seal Oil Recirculation
System for Double
Mechanical Seals 38.3 38.3
5. Vents for Seal Oil
Degassing Reservoirs 102.3 102.3
6. Replacement Pumps 9.0
7. Rupture Oisks for
Relief Valves 51. 7 51. 7 51. 7
3. Closed Loop Sampling
Connections 21.1 21.1 21.1
9. Sealed Bellows Valves 2242.2
10. Hard Piping and Drain
Covers 25.0
TOTAL 0.0 17.3 90.1 252.5 2520.0
8-12
-------
(
Table 8-4. CAPITAL COST ESTIMATES PER MODEL UNIT (Concluded)
(Thousands of May 1979 Dollars)
Regulatory Alternative
Capital Cost Item I II III IV V
Model Unit C (New)
1. Monitoring Instrument 8.5 8.5 8.5
2. Caps for Open-Ended
Valve Lines 8.8 8.8 8.8 8.8
3. Double Mechanical Seals 14.8 14.8
4. Seal Oil Recirculation
System for Double
Mechanical Seals 38.3 38.3
5. Vents for Seal Oil
Degassing Reservoirs 102.3 102.3
6. Rupture Disks for
Relief Valves 28.8 28.8 28.8
7. Closed Loop Sampling
Connections 21.1 21.1 21.1
8. Sealed Bellows Valves 1515.0
9. Hard Piping and Drain
Covers 25.0
TOTAL 0.0 17.3 67.2 222.6 1 7 54 . 1
8-13
-------
I.
(1) monitoring instrum~nt .annualized capital charges, and material,
maintenance, and calibration expenses,
(2) emissions control equipment maintenance and capital charges,
(3) leak detection labor, .
(4) repair labor,
(5) administration and support, and
(6) initial control program startup.
Annualized capital charges include depreciation, interest, property
taxes, and insurance. Depreciation and interest are computed by the
use of a Capital Recovery Factor (CRF), based on the lifetime of the
equipment and the annual interest rate. Property taxes and insurance
are also included and are estimated at.4 percent of the total capital
cost. These items are calculated by means of the formula:
. C = C1 + C2
where C = total annualized costs; C1 = annual depreciation and interest
charges; and C2 = property taxes and insurance. Now C1 and C2 are
described as:
C1 = Cc x (CRF) and
where: C = capital cost of the equipment
c
C2 = Cc x (4%)
i (1+i)n
CRF = (1+i)n-1
where:
i = annual interest rate (10%)
Annualized leak detection and repair
means of the formula:
n = lifetime of equipment, years
(n=6 for the monitoring
instrument, and n=10 for
control equipment)
labor costs are derived by
L = L1 + L2
where L = total annual lea~ detection and repair labor costs; L1 =
annual leak detection cost; and L2 = annual repair cost. Now L1 and
L2 are described as:
8-14
-------
Q,
L1 = (AxBxOxE)xN and L2 = F2xG2xE
where: A = Number of model unit components affected
B = Monitoring time, hours (leak detection)
o = Times monitored per year
E = Labor cost, $/hr = $15.50/hr
F = Estimated number of leaks per year
G = Repair time, hours (maintenance)
N = Number of workers involved in monitoring = 2
Annualized administrative and support costs are estimated at 40
percent of the leak detection and repair labor costs.
Finally, the cost of repairing leaks found during an initial unit
survey is also computed. This cost is amortized by employing a Capital
Recovery Factor using the control equipment lifetime (10 years), and
an annual interest rate of 10 percent.
Tables 8-5 through 8-8 give annual leak detection and repair
labor costs (in May 1979 values) for Regulatory Alternatives II through
V. Total annual leak detection labor costs range from $30 (Unit A,
Regulatory Alternative V) to $4,850 (Unit C, Regulatory Alternative III).
Total annual repair labor costs range from $0 (Units A, B, and C,
Regulatory Alternative V) to $5,750 (Unit C, Regulatory Alternative III).
Leak detection and repair labor costs do not follow a linear relation- .
ship for increasing levels of benzene emission control. Lower monitoring
(leak detection) costs for Regulatory Alternative V in comparison with
the other Regulatory Alternatives result from the fact that there
would be virtually no leaks if Regulatory Alternative V were applied.
The only monitoring performed is the weekly visual inspection of
pumps. Since this Regulatory Alternative represents leakless
emission control, no repair labor costs would be incurred.
Estimates of credits from the recovery of refinery and SOCMI
benzene emissions have been made, based on the market price of benzene
as of May 1979 ($370/Mg or $1.30/gal).9 Table 8-9 presents the recovered
product credits derived from the market price of the recovered product
and the quantity of total VOC emissions reduced as a result of each
regulatory alternative. Recovered product credits range from $4,740
(Unit A, Regulatory Alternative II) to $35,700 (Unit C, Regulatory
8-15
-------
~-' 1 8 ~ "f'''.-r'l'").\'~ ""0 "'.~I-"'~'A"C" 'A"'OR "OUR RI'"QU.RE''''E''T-
'CIO e, -:J., J';viUIi.li\ll\\.l hi"" ,",f'dhlc.l\ J'; C. I.. 0 ;-n c. 1 1'1 j'{ ~
FOR REGULATORY ALTERNATIVE II
co
f
,....
0"1
I
I ~~ONITORING
~ ", . ' MAINTENANCE
Number of
Components Monitoring Estimated, Maintenance
Per Hodel Labor-Hours Number of Labor-Hours
Unit t~on i tori ng Times Required Leaksb Repair Required
Sou rce ~ Type of Timea Monitored Time
Type Monitoring (Minutes) Per Year A B C A B C (Hours) A B C
I . , Instrument 5 1 0.8 2.5 4.2
Pumps 5 15 25 1 1 1 80 80 80 80
Vi~1J'a 1 0.5 , 52 2.2 6.5 10.8
Valves
Gas 34 100 167 Instrument 1 4 4.5 13.3 22.3 1 4 7. 1.13 1 5 8
Liquid 87 264 439 Instrument 1 1 2.9 8.8 14.6 2 6 .-11 1.13 2 7 12.
Safety/ . ,
Relief 4 11 19 Instrument S 4 4.3 11.7 20.3 c c c 0 0 0 0
Valves
prains 5' 15 25 Instrument 1 ' 1 0.2 :.0.5 0.8 1 1 1 4 4 4 4
Total Monitoring 14 9
Hours = ~
x $15.50/hour
Total Monitoring 230
"Dollars =
43.3 73.0
Total Maintanence
Hours =
. x $15.50/hour
Total Maintenance
. Dollars =
87
96 104
--
--
670 . 1130
J.:350 1490 1610
---
alnstru~ent monitoring requires a two-man team.
bRecurrence factors of 0.6 and 0.4 have been applied for monthly and quarterly instrument inspections:
cThese leaks repaired by routine maintenance at no incremental increase in manpower ~equire~ents. '.
Safety/ relief devices are normally reset during routine maintenance without a leak det~ction and repalr
~~:gra~. (Refer~~ce 8).
-------
. .
.
.
Table 8-6. MONITORING AND MAINTENANCE LABOR-HOUR REQUIREMENTS
FOR REGULATORY ALTERNATIVE III
co
I
~
-.....I
~10NITORING
., t~AINTENANCE
Number of
Components Monitoring Estimated Maintenance
Per ~1ode 1 Labor-Hours Number of Labor-Hours
Unit r.1onitoring Times Required Leaksb Repair Required
Source Type of Timea Monitored Time
Type A B C t.1oni tori ng (Minutes) Per Year A B C A B C (Hours) A B C
,
"
Instrument 5 12 10.0 30.0 50.0
Pumps 5 15 25 1 .2 4 80 80 160 320
Visual 0.5 52 2.2 6'.5 10.8
Valves
, Gas 34 100 167 Instrument 12 2 11
j 1 13.6 40.0 66.8 2 6 10 1.13 7
Li qui d 87 264 439 Instrument 1 12 34.8 105.6 175.6 6 19 32 1.13 7 21 36
Safety/
Re lief 4 11 19 Instrument- 8 0 0.0 0.0 0.0' -- -- -- 0 0 0 0
Valves
Drains 5 15 25 Instrument 1 12 2.0 6.0 10.0 .1 1 1 4 4 4 4
Total Monitoring 62.6 188.1 313.2
Hours = . ----
x $15.50/hour
Total Monitoring 970 2920 4850
Dollars = - - -
Total Maintanence
Hours =
x $15.50/hour
Total Maintenance
. Dollars =
93
192,
371
--
1440 2980 5750
--
aInstrument monitoring requires a two-man team,~
bRecurrence factors of 0.6 and 0.4 have been applied for monthly and quarterly instrument inspections.'
-------
Table 8-7. MONiTORING AND MAINTENANCE lABOR-HOU~ REQUIREMENTS
, FOR REGULATORY ALTERNATIVE ~
(X)
I
.......
(X)
I I I --
~10NITORING
I MAINTENANCE
I Number of
I
Compont?nts Monitoring Estimated Maintenance
Per Hode1 Labor-Hours Number bof Labor-Hours
Unit ~,1onitoring Times Required Leaks Repair Required
Source ATBTC Type 'of TimeQ Monitored Time
Type t'ioni tori ng (Minutes) Per Year A B C A B C (Hours) A B C
~nstr~~:nt . .' -- ~ -
5 ..., 0 0 0 a .
,-
Pumps 5 15 25 ' a 0 0, 80 0 ,0 .....0:'
, Visual' 2..2 6.5 10.8 " ......_-
0.5 52
..
Valves 66.8
Gas 34 100 167 Instrument 1 12 " 13.6 40.0 ,'2 6 10 1.13 2 7 11
Liquid 87 264 439 Instrument 1 12 34.8 105.6 17,5.6 6 19 32 1.13 7 21 36
Safety/ " "
Re 1 i ef 4 11 19 . Instrument 8 a 0 0 0 " 0 0 0 0 0 0 0
Valves
Drains 5 15 25 Instrument 1 12 2.0 ' 6.0 10,0 1 1 1 4 4 4 4
\
Total Monitoring 52.6 158.1 263;2, Tota' Maintanence 13 32 51
- -
'Hours = Hours =
x $15.50/hour x $15. 50/hour,
Total Monitoring 820. 2450 4080 Total Maintenance . ,200 500 790
"
Dollars = - - - Dollars = - - "
aInstrument monitoring requires a two-man team.
bRecurrence factors of 0'.6 and'O~4 have been applied for monthly and quarterly.instrument inspections,
, ,
-------
Tab 1 e 8,.£. yI:ONITOR.ING AND ~1A!NTtNANCE LABOR-HOUR REQUIREMENTS
. FOR REGULATORY ALTERNATIVE V
co
I
I--'
1.0 .
NONITORING
'.- MAINTENANCE
Number of .
Components Monitoring Estimated Maintenance
Per rv!ode 1 Labor~Hours Number of Labor-Hours
Unit ~.1onitoring Times Required Leaks Repair Required
Source Type .of Time Monitored Time
Type A B C t~on i tori ng (Minutes) Per Year A B C A B C (Hours) A B C
Instrument 5 0 0 a a ..
, -
Pumps 5 15 25 a a a 80 a ,0 0: '
Visual 0.5 52 2.2 6.5 10.8
Valves
Gas 34 100 167 Instrument 1 a .. a a a a a 0 1.13 a a a
Li qui d 87 264 439 Instrument 1 a a a a 0 0 0 1.13 0 0 a
I
Safetyl
Relief 4 11 19 . Instrument B 0 0 0 0 0 a 0 0 0 0 0
Valves
Drains 5 15 25 Instrument 1 0 0 .0 a a a a 4 a a a
Total Monitoring 2.2 6.5 10.8 Total Maintanence a a a
'Hours = .. - - Hour~ = . - -
x $I5.50/hour... x 15.50/hour
Total Monitoring 30 lOa' 170 Total Maintenance 0 0 0
Dollars = - - Dollars =. - -
-------
Table 8-9.
RECOVERED PRODUCT CREDITS
co
I
N
a
Model Unit A Model Unit 8 Model Unit C
Emission Emi ss ion Emission
Reduction Recovered Reduction Recovered Reduction Recovered
From Recover~ Produ
-------
Alternative V) per year, the credits being a function of model unit
size and the regulatory alternative.
The annualized costs associated with the initial screening survey
and the resultant leak repairs are detailed for Regulatory Alternatives II,
. III, and IV per model unit in Table 8-10. Total repair labor costs
ranging from $280 (Model A, Regulatory Alternative IV) to $8,730
(Model C, Regulatory Alternatives II and III) are multiplied bya CRF
of 0.25 (10 years, 10 percent) to yield annual repair labor costs
ranging from $70 (Model A, Regulatory Alternative IV) to $2,180 (Model C,
Regulatory Alternatives II and III).
Annualized costs for implementing Regulatory Alternatives I
through V for three model unit sizes are given in Table 8-11. Each
model unit is classified as either existing or new. Regulatory
Alternative I is assumed to need no additional controls, thus there
are no annualized costs. Net annualized costs (including recovery
credits) for the existing units range from a savings of $3,700 (Model
B, Regulatory Alternative II) to $609,000 (Model C, Regulatory Alternative V).
For new units, which do not incur the high retrofitting costs of
existing units, the range is from a savings of $4,800 (Model B, Regulatory
Alternative II) to $412,900 (Model C, Regulatory Alternative V).
8.3.2 Cost-Effectiveness
A cost-effectiveness analysis was performed to determine which
regulatory alternative reduces the greatest benzene emissions at the
least cost. Total annual costs due to each regulatory alternative
were divided by the annual benzene emission reduction achieved under
that Regulatory Alternative to generate a cost-effectiveness figure.
Table 8-12 lists benzene emission reductions per model unit for each
regulatory alternative, while Tables 8-13 and 8-14 present the cost-
effectiveness for the existing and new model units, respectively.
Baseline or uncontrolled benzene emissions, represented by Regulatory
Alternative I, per model unit are estimated to be 18.62 Mg/year (A),
44.71 Mg/year (B), and 97.07 Mg/year (C). Under Regulatory Alternative II,
benzene emissions from Model Unit A are estimated to decrease 53
percent from the baseline emissions (from 18.62 Mg/year to 8.67 Mg/year).
Similarly, benzene emission reductions of 57 and 61 percent for Model
8-21
-------
Table 8-10. INITIAL SURVEY START-UP COSTS
FOR REGULATORY ALTERNATIVE II
B
Percent of
Sources
Leaking in
Initial
Survey
Estimated
Number of Leaks
Number of Components
per Model Unit
Source Type
A
C
A
B
Repair
Time
(Hours)
Repair Labor Hours
c
A
B
C
Pumps 5 15 25 23 1 3 6 80 80 240 480
Valves
Gas 34 100 ..J67. 10 3 10 17 1.13 3 11 19
OJ Liquid 87 264 439 12 10 '32 53 1.13 ' 11 36 60
I
N
N
Safety/relief
Devices 4 11 19 8 a a a 0 0 0 0
Drains 5 15 . 25 4 1 1 1 4 4 4 4
Total Hours = 98 291 563
x $15.50/hour
TOTAL = 1520 4510 8730
x CRF = 0.25
\. (10 year, 10 Percent)
Annualizedb = 380 1130 2180
-------
Table 8-10. INITIAL SURVEY START-UP COSTS
FOR REGULATORY ALTERNATIVE III (Continued)
Number of Percent
Components Sources Estimated
Per Unit Leaki ng Ii n Number of Leaks Repair Repair Labor Hours
Initial Time
Source Type A' B C Survey A B C (Hours) A B C
. Pumps 5 15 25 23 1 3 6 80 80 240 480
'Valves
Gas 34 100 167 10 3 10 17 1.13 3 11 19
Liquid 87 264 439 12 10 32 53 1.13 11 36 60
ex>
, . Drains 5 15 25 4 1 1. 1. .4 ~- .4 . 4 4
N
W
Total Hours
98 291 563
- -
x $15.50/hour
= 1520 4510 8730
---
x CRF = 0.25 (10 year,
10 Percent)
= 380 1130 2180
--
=
TOTAL
Annualizedb
-------
Table 8-10. !NITIAL SURVEY START-UP COSTS
FOR REGULATORY ALTERNATIVE .IV (Continued)
Number of Percent
Components Sources Estimated
Per Unit Leaking in Number of Leaks Repair Repair Labor Hours
Initi a 1 Time
Source Type A B C Survey A B C (Hours) A B C
Valves
Gas 3"4 100 167 10 3 10 17 1.13 3 , 11 19
Q:) 87 264
I Liquid 439 12 10 32 53 1.13 11 36 60
N
~
. Drains 5 15 25 4 .. .1 1. ...1. 4. .. ....4 4 4
Tota 1 ;Hburs
" =
18 51
- -
x $lS.SOjhour
= 280 790 1290
x CRF = 0.25 (10 years
10 Percent)
lQ 200 320
83
TOTAL.
Annualizedb.
=
-------
Table 8-10.
INITIAL SURVEY START-UP COSTS
(Conc 1 uded)
NOTES:
aLeaks are repaired by routine maintenance at no incremental
increase in manpower. requirements. Safety/relief devices are
normally reset during routine maintenance without a leak
detection and repair program.
bSince there are no one-time start-up costs, these numbers can
be capitalized using the Capital Recovery Factor (CRF) method.
8-25
-------
Table 8- 11. ANNUALIZED CONTROL COST
ESTIMATES PER MODEL UNITa
(Thousands of May 1979 Dollars)
Hodel Unit: A (Existing)
Cost Itetll II III IV V
Annualized Capital Charges
I. Control Equipment
b 2.0 2.0 2.0 0.0
a. Instfument
b. Caps 0.3 0.3 0.3 0.3
c. Oouble Seals
8 Seals 1.7 1.7
8 Installationd 0.2 0.2
d. Seal Oil System 1.2 1.2
e. Vents-Pumps and Compressors 3.4 3.4
f. Rep I acemen t Pump!! 0.5
g. Rupture Oi sksf
8 Oisks 0.3 0.3 0.3
8 Instal1ation 1.5 1.5 1.5
h. Closed-Loop Samplinggh 0.7 0.7 0.7
i. Sealed-Bello~s Valves 72.9
j. Hard Pipingg.1 0.8
2. Initial Leak Repairj .0.4 0.4 0.1 0.0
Operating Costs
I. Maintenance Charges
a. Instrument 2.7 2.7 2.7 0.0
b. Caps 0.1 0.1 0.1 0.1
c. Oouble Seals 0.2 0.2
d. Seal Oil Sys tem 0.3 0.3
e. Vents-Pumps and Compressors 0.8 0.8
f. Replacement Pumps 0.1
g. Rupture Disks 0.4 0.4 0.4
h. Closed-Loop Sampling 0.2 0.2 0.2
i. Sealed-Bellows Valves 17.9
j. Hard Piping 0.2
2. Miscellaneous (Taxes. Insurance.
Administration)
a. Ins trumen t 0.3 0.3 0.3 0.0
b. Caps 0.1 0.1 0.1 0.1
c. Double Seals 0.2 0.2
d. Sea I Oi I Sys tern 0.4 0.4
e. Vents-Pumps and Compressors 1.0 1.0
f. Replacement Pumps 0.2
g. Rupture Oi sks' 0.5 0.5 0.5
h. Closed-LoOp Sampling 0.2 0.2 0.2
i. Sealed-Bellows Valves 22.4
j. Hard Piping 0.3
3. Labor
a. Leak Detection Labor 0.2 1.0 0.8 0.0
b. Repair Labor 1.4 1.4 0.2 0.0
c. Administrative and Support 0.6 1.0 0.4 0.0
Total Before Credit 0.0 8.1 13.1 20.2 129.0
Recovered Product Creditk,l 0.0 (4.7) (5.7) (6.2) (7.1)
Net Annualized Costl 0.0 3.4 7.4 14.0 121.9
8-26
-------
Table 8-11. ANNUALIZED CONTROL COST
ESTIMATES PER MODEL UNITa (Continued)
(Thousands of May 1979 Dollars)
Hodel Unit:
A (New)
Cost Item
Annualized Capital Charges
I. Control Equipment
b
a. InstEument
b. Caps
c. Double Seals
. Sea Is
. Installationd
d. Seal 011 System
e. Vents-Pumps and Compressors
f. Rupture Oisksf
. Oi s k s
. Installation
g. Closed-Loop Samplinggh
h. Sealed-Bello~s Valves
i. Hard Pipingg,l
2. Initial Leak RepairJ
Operating Costs
I. Maintenance Charges
a. Instrument
. b. Caps
c. Double Seals
d. Sea 1 011 Sys tern
e. Vents-Pumps and Compressors
f. Rupture Oi sks
g. Closed-Loop Sampling
h. Sealed-Bellows Valves
i. Hard Piping
2. Miscellaneous (Taxes, Insurance,
Administration)
a. Instrument
b. Caps
c. Double Seals
d. Seal 011 System
e. Vents-Pumps and Compressors
f. Rupture Oi sks
g. Closed-Loop Sampling
h. Sealed-Bellows Valves
i. Hard Piping
3. Labor
a. Leak Detection Labor
b. Repair Labor
c. Administrative and Support
Total Before Credit
Recovered Product Credi tk.l
Net Annualized Costl
0.0
0.0
7.7 11.6
(4.7) (5.7)
0.0
82.5
8-27
II
2.0
0.3
2.7
0.1
0.3
0.1
0.3
0.1
0.3
0.2
0.2
1.4
0.6
1.0
1.4
1.0
3.0
5.9
III
2.0
0.3
0.3
0.8
0.7
0.0
2.7
0.1
0.2
0.2
0.8
0.2
0.4
18.2
(6.2)
12.0
IV
V
2.0
0.3
1.0
0.2
1.2
3.4
0.0
0.3
1.0
0.2
1.2
3.4
0.3
0.8
0.7
0.3
0.8
0.7
49.3
0.8
0.0
0.0
2.7
0.1
0.1
0.3
0.8
0.2
0.2
0.0
0.1
0.1
0.3
0.8
0.2
0.2
12.1
0.2
0.3
0.1
0.2
0.4
1.0
0.3
0.2
0.0
0.1
0.2
0.4
1.0
0.3
0.2
15.1
0.3
0.0
0.0
0.0
89.6
(7.1)
-------
Table 8- 11. ANNUALIZED CONTROL COST
ESTIMATES PER MODEL UNITa (Continued)
(Thousands of May 1979 Dollars)
Hodel Unit: B (Existing)
Cost Item 11 111 IV V
Annualized Capital Charges
I. Control Equipment
a. InstEumentb 2.0 2.0 2.0 0.0
b. Caps 0.8 0.8 0.8 0.8
c. Double Seals
. Seals 5.0 5.0
. Installationd 0.7 0.7
d. Sea I Oi I Sys tem 3.7 3.7
e. Vents-Pumps and Compressors g.g g.g
f. Replacement pumpse 1.0
g. Rupture Oi sksf
. Oisks 1.1 1.1 1.1
. Instdllation 4.3 4.3 4.3
h. Closed-loop Samplinggh 2.1 2.1 2.1
1. Sealed-Bello~s Valves
j. Hard Pipingg,l 219.5
2. Initial leak Repa i rj 1.1 1.1 0.2 0.0
Operating Costs
1. Maintenance Charges
a. Ins trumen t 2.7 2.7 2.7 O.~
b. Caps 0.2 0.2 0.2 0.2
c. Ooubl e Sea Is 0.5 0.5
d. Seal Oil 'System 0.9 O.g
e. Vents-Pumps and Compressors 2.5 2.5
f. Replacement Pumps 0.2
9. Rupture Disks 1.2 1.2 1.2
h. Closed-loop Sampling 0.5 0.5 0.5
1. Sealed-Bellows Valves 53.9
j. Hard Pipin9 0.6
2. Miscellaneous (Taxes, Insurance.
Administration)
a. Instrument 0.3 0.3 0.3 0.0
b. Caps 0.3 0.3 0.3 0.3
c. 004ble S~als 0.7 0.7
d. Se~ I Oil Sys tem 1.2 1.2
e. Vents-Pumps and Compressors 3.1 3.1
f. Replacement Pumps 0.3
g. Rupture Oisks 1.4 1.4 1.4
h. Closed-loop Sampling 0.6 0.6 0.6
1. Sealed-Bellows Valves 67.3
j. Hard Pi pi ng 0.8
3. labor
a. leak Detection labor 0.7 2.9 2.5 0.0
b. Repair labor 1.5 3.0 0.5 0.0
c. Adminis~rative and Support 0.9 2.4 1.2 0.0
Total Before Credit 0.0 10.5 26.9 50.1 386.7
Recovered Product Creditk,l 0.0 (14.2) (17.2) (18.6) (21.4)
Net Annualized Cost 1 0.0 (3.7) 9.7 31.5 365.3
8-28
-------
Table 8-11. ANNUALIZED CONTROL COST
ESTIMATES PER MODEL UNITa (Continued)
(Thousands of May 1979 Dollars)
Hodel Unit:
B (New)
Cost Item
Annualized Capital Charges
1. Control Equipment
a. Instfumentb
b. Caps
c. Double Seals
. Sea 15
. Installation
d. Seal Oil systemd
e. Vents-Pumps and Compressors
f. Rupture Oi sksf
. Disks
. Installation
g. Closed-Loop Samp1inggh
h. Sealed-Bellows Valves
i. Hard Pipingg,i
2. Initial Leak RepairJ
Operatin9 Costs
1. Maintenance Charges
a. Instrument
b. Caps
c. Oouble Seals
d. Seal Oil System
e. Vents-Pumps and Compressors
f. Rupture Disks
9. Closed-Loop Sampling
h. Sealed-Bellows Valves
i. Hard Piping
2. Miscellaneous (Taxes. Insurance,
Administration)
a. Ins trument
b. Caps
c. Double Seals
d. Seal Oil System
e. Vents-Pumps and Compressors
f. Rupture Disks
g. Closed-Loop Sampling
h. Sealed-Bellows Valves
i. Hard Piping
3. Labor
~. Leak Oetectio~ Labor
b. Repair Labor
~. Administrative and Support
Total Before Credit
Recovered Product Creditk.1
Net Annualized Cost1
0.0
0.0
(14.2) (17.2)
(4.8) 5.5
0.0
8-29
II
III
2.0
0.8
0.0
2.7
0.2
0.3
0.3
0.7
1.5
0.9
9.4 22.7
2.0
0.8
1.1
2.4
2.1
0.0
2.7
0.2
0.6
0.5
0.3
0.3
0.8
0.6
2.9
3.0
2.4
(18.6) (21.4)
25.7 247.7
IV
V
2.0
0.8
3.0
0.6
3.7
9.9
0.0
0.8
3.0
0.6
3.7
9.9
1.1
2.4
2.1
1.1
2.4
2.1
148.3
2.4
0.0
0.0
2.7
0.2
0.4
0.9
2.5
0.6
0.5
0.0
0.2
0.4
0.9
2.5
0.6
0.5
36.4
0.6
0.3
0.3
0.4
1.2
3.1
0.8
0.6
0.0
0.3
0.4
1.2
3.1
0.8
0.6
45.5
0.8
2.5
0.5
1.2
0.0
0.0
0.0
44.3 269.1
-------
Table 8-11. 'ANNUALIZED CONTROL COST
ESTIMATES PER MODEL UNITa (Continued)
(Thousands of May 1979 Dollars)
Hodel Unit: C (E~isting)
Cost Item II III IV V
Annualized Capital Charges
1. Control Equipment
a. Instfumentb 2,0 2.0 2.0 0.0
h. Caps 1.4 1.4 1.4 ,1.4
c. Double Seals
. Seals 8.3 8.3
. Instal1ationd 1.2 1.2
d. Sea 1 Oil Sys tem, 6.2 6.2
e. Vents-Pumps and Compressors 16.7 16.7
f. Replacement Pumpse 1.4
g. Rupture Oi sksf
. ~isks 1.9 1.9 1.9
. Installation 7.8 7.8 7.8
h. Closed-loop Samplinggh 3.4 3.4 3.4
i. Sealed-Bello~s Valves 365.5
j. liard Pipingg,l 4.0
2. Initial leak Repairj 2.2 2.2 0.3 0.0
Operating Costs
I. Maintenance Charges
a. Instrument 2.7 2.7 2.7 0.0
b. Caps 0.4 0.4 0.4 0.4
c. Oouble Seals 0.9 0.9
d. Seal Oil System 1.6 1.6
e. Vents-Pumps and Compressors 4.1 4.1
f. Replacement Pumps 0.4
g. Rupture Di sks 2.0 2.0 2.0
h. Closed-loop Sampling 0.8 0.8 0.8
i. Sealed-Bellows Valves 89.7
j. Hard Piping 1.0
2. Miscellaneous (Ta~es. Insurance.
Admini stration)
a. Instrument 0.3 0.3 0.3 0.0
b. Caps 0.4 0.4 0.4 0.4
c. Double Seals 1.,1 1.1
d. Seal Oil System 1.9 1.9
e. Vents-Pumps and Compressors 5.1 5.1
f. Replacement Pumps 0.5
g. Rupture Disks 2.5 2.5 2.5
h. Closed-loop Sampling 1.1 1.1 1.1
i. Sealed-Bellows Valves 112.1
j. Hard Piping 1.3
3. labor
a. leak Detection labor 1.1 4.9 4.1 0.0
b. Repa i r labor 1.6 5.8 0.8 0.0
c. Administrative and Support 1.1 4.3 2.0 0.0
Total Before Credit 13.2 43.9 81.0 644.7
Recovered Product Creditk,l (23.7) (28.7) (31.0) (35.7)
Net Annualized Costl (10.5) 15.2 50.0 609.0
8-30
-------
Table 8-11. ANNUALIZED CONTROL COST
ESTIMATES PER MODEL UN ITa, (Concluded)
(Thousands of May 1979 Dollars)
Hodel Unit: C (New)
Cost Item 11 III IV V
Annualized Capital Charges
1. Control Equipment
b 2.0 2.0 2.0 0.0
a. InstEument
h. Caps 1.4 1.4 1.4 1.4
c. Double Seals
. Seals 5.0 5.0
. Instal1ationd 1.0 1.0
d. Seal Oil System 6.2 6.2
e. Vents-Pumps and Compressors 16.7 16.7
f. Rupture Di sksf
. Disks 1.9 1.9 1.9
. . Installation 4.2 4.2 4.2
g. Closed-Loop Samplinggh 3.4 3.4 3.4
h. Sealed-Be11o~s Valves 246.9
1. Hard Pipingg,1 4.0
2. Initial Leak RepairJ 0.0 0.0 0.0 0.0
Operating Costs
1. Maintenance Charges
a. Instrument 2.7 2.7 2.7 0.0
b. Caps 0.4 0.4 0.4 0.4
c. Double Seals 0.6 0.6
d. Sea 1 Oil Sys tem 1.6 1.6
e. Vents-Pumps and Compressors 4.1 4.1
1. Rupture Di sks 1.1 1.1 1.1
g. Closed-Loop Sampling 0.8 0.8 0.8
h. Sealed-Bellows Valves 60.6
1. Hard Piping 1.0
2. Miscellaneous (Taxes, Insurance,
Admi n i s tra t i on)
a. Instrument 0.3 0.3 0.3 0.0
h. Caps 0.4 0.4 0.4 0.4
c. Double Seals 0.7 0.7
d. Sea 1 Oil Sys tem 1.9 1.9
e. Vents-Pumps and Compressors 5.1 5.1
1. Rupture Disks 1.4 1.4 1.4
g. Closed-Loop Sampling 1.1 1.1 1.1
h. Sealed-Bellows Valves 75.8
i. Hard Piping 1.3
3. Labor
a. Leak Detection Labor 1.1 4.9 4.1 0.0
b. Repair Labor 1.6 5.8 0.8 0.0
c. Administrative and Support 1.1 4.3 2.0 0.0
Total Before Credit 11.0 36.1 70.9 448.6
Recovered Product Creditk,l (23.7) (28.7) (31.0) (35.7)
Net Annualized Costl (12.7) 7.4 39.9 412.9
8-31
-------
Table 8-11. ANNUALIZED COST ESTIMATES PER MODEL UNIT (Continued)
NOTES:
aCost Factors Used in Computing Annual Costs:
(Ref. 2, pp. IV-3 through IV-10)
1.
Instrument
a.
Capital Charges
i. Cost = 2 x $4,250
ii. Operating Life = 6 Years
iii. Annual Interest = 10 Percent
i v.
CRF = 0.23
v. Miscellaneous = 0.04
b.
Materials and Maintenance
i. Cost = $2,700 Per Year
2.
Control Equipment
a. Capital Charges
b. Operating Life = 10 Years (2 yrs. for double seal &
c. Annual Interest = 10 Percent
d. CRF = 0.16 (0.58 for double seal and rupture disk)
e. Miscellaneous = 0.04
. . .
f. Maintenance = 0.05
rupture disk)
3.
Administration and Support = 40 Percent of Leak
Detection and Repair labor Cost .
bCost is for Century System~ Corporation's Organic Vapor Analyzer
(Model OVA-l08). '
cUsed to seal open-ended lines.
dUsed as auxilliary for double seal..
8-32
-------
Table 8-11. ANNUALIZED COST ESTIMATES PER MODEL UNIT (Concluded)
NOTES:
(Concluded)
eIn retrofitting double seals, it ~as been assumed that 10 percent
of the pumps will need to be replaced. Thus, the following
numbers of pumps would be replaced:
Model A -- 1 pump
Model B -- 2 pumps
Model C -- 3 pumps
fCost includes rupture disk, block valve, and replacement of
relief valve.
gCost is for new or retrofitted installation.
hUsed for control valves.
iConsists of sealed cover on drain with line leading to pump.
jWhere equipment standards are applied, as in Regulatory Alternatives
IV and V, the amount of leak detection and repair labor decreases.
k .
Based on an average price of $370/Mg (Ref. 9).
lNumbers in parentheses represent savings (net credit).
8~3
-------
Table 8-12. BENZENE EMISSION REDUCTIONS
-------
Table 8-13.
COST-EFFECTIVENESS FOR MODEL UNITS (EXISTING UNITS)
ex>
I
W
(J'1
Model Unit A Model Unit B Model Unit C
Regulatory Alternative I II III IV V I II III IV V I II III IV V
Total Capital Cost ($1.000) 0.0 10.3 24.3 56.9 504.1 0.0 13.8 55.4 152.9 1512.2 0.0 17.3 90.1 252.5 2520.5
Total Annualized Cost ($1.000) 0.0 8.1 13.1 20.2 129.0 0.0 10.5 26.9 50.1 386.7 0.0 13.2 43.9 81.0 644.7
Total Annual Credit ($1.000) 0.0 (4.7) (5.7) (6.2) (7.1) 0.0 (14.2) (17.2) (18.6) (21.4) 0.0 (23.7) (28.7) (31.0) (35.7)
Net Annualized Cost ($1,000) 0.0 3.4 7.4 14.0 121.9 0.0 (3.7) 9.7 31.5 365.3 0.0 (10.5) 15.2 50.0 609.0
Total Benzene Reduction ---- 10.0 12.B 13.6 15.7 ---- 25.4 32.6 34.6 40.1 ---- 5B.8 75.B aD.S .$!.O
(Mg/Year)
Cost-Effectiveness (Net Annualized ---- . 0.34 0.58 1.03 7.76 ---- 0.0 o.~ 0.91 9.11 ---- 0.0 0.20 0.62 6.55
$l,ooO/Mg Benzene)
-------
CP
I
W
C"I
Table 8-14.
COST -EFFECTIVENESS FOR MODE.L UNITS (NEW UNITS)
Model Unit A Mode 1 'Unit B Model Unit C
Regulatory Alternative I II 1II IV V I II 1II IV V I II III IV . V
Total Capital Cost ($1,000) 0.0 10.3 20.0 51.2 350.2 0.0 13.8 42.5 135.8 1052,3 0.0 17.3 67.2 222.6 1754.1
Total Annualized Cost ($1.000) 0.0 7.7 11.6 18.2 89.6 0.0 9.4 22.7 44.3 .269.1 0.0 11.0 36.1 70.9 448.6
Total Annual Credit ($1.000) 0.0 (4.7) (5.7) (6.2) (7.1) 0.0 (14.2) (17.2) (18.6) (21.4) 0.0 (23.7) (28.7) (31.0) ( 35 .7)
Net Annualized Cost ($1.000) 0.0 3.0 5.9 12.0 82.5 0.0 (4.8) 5.5 25.7 247.7 0.0 (12.7) 7.4 39.9 41~.~
Total 8enzene Reduction ---- 10.0 12.8 13.6 15.7 ---- 25.4 32.6 34..6 40.1 ..-..- 58.8 75.8 80.5 93.0
(Mg/Year)
tost-Effectiveness (Net Annualized ...--- 0.30 0.46 0.88 5.25 ---- 0.0 0.17 0.74 6.18 ---- 0.0 0.10 0.50 4.44
$1.000/Mg Benzene)
-------
Units Band C, respectively, are estimated. For Regulatory Alternative III,
emissions are expected to decrease 69, 73, and 78 percent for Model
,
Units A, B, and C, respectively. Reductions of 73, 77, and 83 percent
are estimated for Model Units A, B, and C, respectively, for Regulatory
Alternative IV. Regulatory Alternative V, the most stringent control
level, is estimated to reduce emissions from the baseline level by 84
percent for Model A, 90 percent for Model B, and 96 percent for Model C.
For existing units, the annualized cost-effectiveness varies from no
cost (Models Band C, Regulatory Alternative II) to $9,110/Mg of
benzene (Model B, Regulatory Alternative V). For new units, the
cost-effectiveness ranges from no cost (Models Band C, Regulatory
Alternative II) to $6,180/Mg benzene (Model B, Regulatory Alternative
V).
8.4 COST COMPARISON
The costs of controlling benzene fugitive emissions from petroleum
refineries and chemical plants are compared with other benzene source
categories in order to consider the total impact of all regulations
(e.g., OSHA,"air and water quality, solid waste) on the entire benzene
industry. Control cost data for comparison with the benzene fugitive
sources are derived from capital and annual control cost estimates
that are presented in the proposed NESHAP documents for the above
source categories.10-12 Table 8-15 summarizes capital and annual
contro1 cost data for the refinery-SOCMI benzene industry (fugitive
emissions), the maleic anhydride and ethyl benzene-styrene industries
(process emissions), and benzene storage tanks from consumers and
producers. In each category, ranges of control costs are given for
existing and/or new units, reflecting differences in model units and
control technologies. In general, the range of capital and annualized
costs for the benzene fugitive sources (refineries and SOCMI) is much
wider than the range for the other sources. Control costs for the
maleic anhydride plants (existing) appear to be the highest, followed
by costs for controlling benzene fugitive sources.
Tables 8-16 and 8-17 present total control costs associated with
all emissions (i.e., process, fugitive, and storage) for existing and
new units representative of the maleic anhydride and ethyl benzene-styrene
8-37
-------
Table 8-15. RANGE OF CONTROL COSTS FOR THE BENZENE SOURCE
CATEGORIES FOR EXISTING AND NEW UNITS
(Thousands of Dollars)
Total Capital Cost a
Total Annual Cost
Source Category Existing New Existing New
Refinery-SOCMI 10-2502 10-1754 (11 )-609 (13) -41 3
(Benzene Fugitives)
Maleic Anhydridel0 1160-1440 354-600
(Process Emissions)
Ethy1benzene-Styrenel1 268-555 (150)-45
(Process Emissions)
Benzene Storage 12
(Producers) 7-290 6-290 0.2-68 0-69
Benzene Storage 12 (0.8)-6~
(Consumers) 16-249 15-249 2-68.
a credits.
Includes recovered
bNumbers in parentheses represent savings.
8-38
-------
Table 8-16. COSTS FOR THE CONTROL OF TOTAL BcrNfzNE EMISSIONS
FROM THE MAELIC ANHYDRIDE INDUSTRY}', ,a
(Thousands of Dollars)
Cost Item
Existing
New
Capital Cost
Annual Costb
1193-3018
359-725
31-1241
4-54
aIncludes control costs for benzene fugitive emissions
from refineries and chemical plants.
blncludes recovered credits.
Table 8-17. COSTS FOR THE CONTROL OF TOTAL BENIE~~ ~MISSIONS
FROM THE ETHYLBENZENE-STYRENE INDUSTRyIJ, ,
(Thousands of Dollars)
Cost Item
Existing
New
Capital Cost
Annual Costb
301-2133
( 1 50) - 114
31-1241
4-54
alncludes control costs for benzene fugitive
from refineries arid chemical plants.
blncludes recovered credits.
cNumber in parentheses represents savings.
8-39
d
J
emissions
-------
industries, respectively. Total control cost data are given in
Tables 8-18 for controlling benzene emissions from producer storage
tanks and fugitive sources, while similar costs are incurred for
benzene emission controls from consumer storage tanks and fugitive
sources, as shown in Table 8-19.
8-40
-------
Table 8-18. TOTAL COSTS FOR THE CONTROL OF BENZENE EMISSIONS
FROM PRODUCER BENZENE STORAGE TANKS AND BENZENE FUGITIVE SOURCES12,a
(Thousands of Dollars)
Cost Item Existing New
Capital Cost 17 -2792 16-2044
Annual Costb (11)-677 (13)-482
aInc1udes control costs for benzene fugitive emissions
from refineries and chemical plants.
bInc1udes recovered credits.
cNumbers in parentheses represent savings.
Table 8-19. TOTAL COSTS 'FOR THE CONTROL OF BENZENE EMISSIONS fROM
CONSUMER BENZENE STORAGE TANKS AND BENZENE FUGITIVE SOURCES! ,a
(Thousands of Dollars)
Cost Item
Existing
New
Capital Cost
Annual Costb
26-2751
(9)-677
25-2003
(14)-475
aInc1udes control costs for benzene fugitive emissions
from refineries and chemical plants.
bInc1udes recovered credits.
cNumbersin parentheses represent savings.
8-41
! I
I!
-------
8.5 REFERENCES
1.
Letter from Guy C. Arney, Century Systems Corporation, to James C.
Serne, PES, Incorporated. October 17, 1979. Cost data for VOC
monitoring instrument. .
2.
Erickson, D.G., and V. Kalcevic. Emissions Control Options for
the Synthetic Organic Chemicals Manufacturing Industry. Fugitive
Emissions Report. Hydroscience, Incorporated. Knoxville, TN.
Prepared for U.S. Environmental Protection Agency, Emission
Standards and Engineering Division. Research Triangle Park, NC.
EPA Contract No. 68-02-2577. February 1979.
3.
Economic Indicators (for January 1979).
86(9):7. April 23, 1979.
Economic Indicators (for April and May 1979).
86(16):7. July 30, 1979. .
Chemical Engineering.
Chemical Engineering.
4.
5.
Telecon. McInnis, J.R., PES, Incorporated, with Hetrick, C.,
Crane Chempum Division, Warrington, PA. August 24, 1979.
6'.
PES estimate.
7.
Letter from Brian Crutchfield, Duriron Company, Incorporated,
to J.R. McInnis, PES, Incorporated. August 31, 1979. Cost
data for replacement pump.
8.
Letter from J. Johnson, Exxon Company, to R.T. Walsh, EPA, CPB.
July 28, 1977. Response to EPA draft document, "Control of
Hydrocarbon from Miscellaneous Refinery Sources." '
9.
Current Prices of Chemicals'and Related Materials.
Marketing Reporter. 215(21):43. May 21, 1979.
U.S. Environmental Protection Agency. Draft Preamble for National
Emission Standard for Benzene Emissions from Maleic Anhydride
Plants. Emission Standards and Engineering Division. Research
Triangle Park, NC. p. 19-20. .
Chemical
10.
11.
U.S. Environmental Protection Agency. Draft Preamble for National
Emission Standard for Benzene Emissions from Ethylbenzene -
Styrene Plants. Emission Standards and Engineering Division.
Research Triangle Park, NC. p. 18.
Letter and attached cost tables from D.C. Ailor, Energy Systems
Group of TRW, Incorporated, ta D.W. Markwardt, EPA, CPB. July 24,
1979.
12.
8-42
-------
-I
13. Peters, M.S., and K. D. Timmerhaus. Plant Design and Economics
for Chemical Engineers. Second edition. New York, McGraw-Hill,
1968. p. 451-452.
14. Hustvedt, K.C., R. A. Quaney» and W. E. Kelly. Control of Volatile
Organic Compound Leaks from Petroleum Refinery Equipment. U.S.
Environmental Protection Agency, Office of Air Quality Planning
and Standards. Research Triangle Park, N.C. Report No.
EPA-450/2-78-036.June 1978. p. 4-6.
8-43
-------
9.0 ECONOMIC IMPACT
9.1
INDUSTRY CHARACTERIZATION
9.1.1 General Profile
Throughout the United States there are 74 petroleum refining and
SOCMI companies operating 134 plant sites that manufacture benzene and
derivatives of benzene.1-15,32-33 Table 9-1 lists the companies
alphabetically and shows their plant locations and capacities as
obtained from the most recent data available. In addition, the table
includes 6 new plants under construction on 1 or more units, 14 exist-
ing plants undergoing expansion, 6 on standby or not currently in
operation, and 1 plant in the engineering phase of construction.
Table 9-1 lists 32 companies that currently produce pure benzene
at 50 plant sites by 74 units (e.g., Su1fo1ane, UDEX). Benzene is
also produced as an impure by-product in the manufacture of ethylene,
which is produced by 28 companies at 47 sites and 59 units. There are
50 companies that manufacture benzene derivatives at 78 production
sites. These derivatives, which consume benzene as a feedstock, are
listed in Table 9-2.
Benzene, ethylene, and benzene-derivative production is fairly
concentrated geographically. Over 85 percent of the total U.S. benzene
capacity is located in two states and one territory: Texas (61 percent),
Louisiana (19 percent), and Puerto Rico (7 percent).
9.1.2 Production of Benzene, Ethylene, and Benzene Derivatives
Total 1977 U.S. production of benzene by petroleum refineries and
SOCMI production units was estimated to be 5260 gigagrams (Gg).16,26
Total ethylene production was higher at 10,600 Gg, while total produc-
tion of benzene derivatives was approximately 8700 Gg. Table 9-3
summarizes 1977 production and capacity data for benzene, ethylene,
and benzene derivatives.
9-1
-------
Table 9-1. REFINERIES AND SYNTHETIC ORGANIC
CHEMICAL MANUFACTURING SITES
WITH BENZENE FUGITIVE EMISSION POTENTIALI-15,32-33
Benzene-Related Capaci tyb
Product~
Plant City/State At Site ~
1. Allied Chemical Geismar, LA Et 340
2. Allied Chemical Moundsvi11e, WV .NiBz 25
3. Allied Chemical Solvay, NY C1Bz 13
4. American Cyanamid Bound Brook, NJ NiBz 48
5. American Cyanamid Willow Island, WV NiBzc 34
6. Amerada Hess St. Croix, VI Bz 217
7. American Hoechst Baton Rouge, LA EtBz 526
St NDg
8. American Hoechst Bayport, TX EtBzd 469
Std 409
9. American Petrofina Port Arthur, TX Bz 67
(of Texas)
10. American Petrofina Big Spring, TX Bz 194
(Cosden Oil) Cyx 35
EtBze 20
St 41
11. American Petrofina Groves, TX Et 9
(Cosden Oi1/Petrogas)
12. American Petrofina/ Beaumont, TX Bz 73
Union Oil of CA Cyx 88
13. Amoco Chemicals Chocolate Bayou, TX Et 909
14. Ashland Oil Ashland, KY Bz 214
Cu 181
Cyx NDg
15. Ashland Oil N~a1, WV MAN 27
16. Ashland Oil Nprth Tonawanda, NY Bz 77
17. Atlantic Richfield Beaver Valley, PA St 200
( Ko bu ta)
18. Atlantic Richfield Channel view, TX Bzc 107
Et (2 units) 1179
9-2
i
[
i'
-------
Table 9-1. REFINERIES AND SYNTHETIC ORGANIC
CHEMICAL MANUFACTURING SITES
WITH BENZENE FUGITIVE EMISSION POTENTIAL (CONTINUED)
Benzene-Related Capaci t)b
Pl ant City/State Productg
At Site (Gg/yr
19. Atlantic Richfield Wilmington, CA Bz 40
Et 45
20. Atlantic Richfield Houston, TX Bzc 140
(ARCO/Polymers) Et 227
EtBz 61
St 54
21. Atlantic Richfield Port Arthur, TX EtBz 114
(ARCO/Polymers)
22. Charter Houston, TX Bz 17
Internati onal EtBz 16
23. Chemetics International Geismar, LA NiBz 173
24. Chemplex Clinton, 10 Et 227
25. Cities Service Lake Charles, LA Bz 83
Et (2 units) 400
26. Clark Oil Blue Island, IL Cu 50
27. Coastal States Gas Corpus Christi, TX Bz 234
Cue 64
28. Commonwealth Oil Penuelas, PR Bz 618
Cyx 117
EtBze 73
29. Continental Oil Ba Himore, MD LAB 122
30. Continental Oil Lake Charles, LA Et 302
31. Core-Lube Danville, IL BSA NDg
32. Corpus Christi Corpus Christi, TX Bzd 100
Petrochemicals Etd 544
33. Cos-Mar, Inc. Carrvill e, LA EtBz 690
St 590
34. Crown Central Pasadena, TX Bz 77
35. Oenka (Petrotex) Houston, TX MAN 23
36. Dow Chemical Bay City, tn Bz 100
Et 86
9-3
-------
Tabl e 9-1. REFINERIES AND SYNTHETIC ORGANIC
CHEMICAL MANUFACTURING SITES
WITH BENZENE FUGITIVE EMISSION POTENTIAL (CONTINUED)
Benzene-Related ~
Product~
Plant City/State At Site G r
37. Dow Chemical Freeport, TX Bz 167
Et (5 units) 1136
EtBz 794
St 658
38. Dow Chemical Midland, MI C1Bz 129
EtBze 249
St 181
39. Dow Chemical Orange, TX Et 375
40. Dow Chemical Plaquemine, lA Bzd 200
Et (2 units) 545
41. Dupont Beaumont, TX NiBz 159
42. Dupont Gibbstown,. NJ NiBz 110
43. Dupont Orange, TX Et 374
44. Eastman Kodak longview, TX Et 580
45. El Paso Natural Gas Odessa, TX Et NDg
EtBz 125
St 68
46. El Paso Products/ Odessa, TX Et 236
Rexene Polyolefins StC 47
47. Exxon Baton Rouge, LA Bz 234
Et 816
EtBz NDg
St NDg
48. Exxon Bay town, TX Bz 200
Cyx 147
Etc 36
49. First Chemical Pascagoula, MS NiBz 152
50. Georgia-Pacific Houston, TX Cu 340
51. Getty Oil Delaware City, DE Bz 37
~.
,
9-4
-------
Table 9-1. REFINERIES AND SYNTHETIC ORGANIC
CHEMICAL MANUFACTURING SITES
WITH BENZENE FUGITIVE EMISSION POTENTIAL (CONTINUED)
Benzene-Related Capacityb
Product~
Plant City/State At Site (Gg/yr)
52. Getty Oil El Dorado, KA Bz 43
Cu 61
53. B.F. Goodrich Calvert City, KY Et 136
54. Goodyear Tire & Rubber Bayport, TX Hqn 5
55. Gulf Coast Olefins Taft, LA EtC 218
56. Gulf Oil All i ance, LA Bz 224
57. Gulf Oil Donaldsonville, LA EtBz 313
St 272
58. Gulf Oil Philadelphia, PA Bz 124
Cu 209
59. Gulf Oil Chemicals Cedar Bayou, Tx Et (2 units) 719
60. Gul f Oil Chemicals Po rt Arthu r, TX Bzc 134
Cu 204
Cyx 106
Et (2 units) 558
61. Hercules McGregor, TX C1Bzf 0.05
62. Howell San Antonio, TX Bz NDg
63. ICC Industries Niagara Falls, NY C1Bz 11
64. Independent Refining Winnie, TX Bz 10
Corp.
65. Jim Walter Resources Birmingham, AL BSA NDg
66. Kerr-McGee Corp. Corpus Christi, TX Bz 53
67. Koppers Bridgeville, PA MAN 15
68. Koppers Cicero, IL MAN 5
69. Koppers Petrolia, PA Rcnol 16
70. t1arathon Oil Texas City, TX Bz 23
Cue 95
71. Mobay Chemical New Martinsville, WV NiBz 61
9-5
-------
Table 9-1. REFINERIES AND SYNTHETIC ORGANIC
CHEMICAL MANUFACTURING SITES
WITH BENZENE FUGITIVE EMISSION POTENTIAL (CONTINUED)
Benzene-Related . b
Product~ Capac i ty .
plant City/State At Site (Gg/yr)-
72. Mobil Oil Beaumont, TX Bz 200
Et 410
73. r~onsanto Alvin, TX Cu 340
(Chocolate Bayou) EtC 285
EtBz 27
LAB 102
74. Monsanto Sauget, IL C1Bz 80
NiBz. 5
75. Monsanto St. Louis, MO MAN 48
76. Monsanto Texas City, TX Bz 284
Et 45 .
EtBz 744
St 680
77. Montrose Chemical Henderson, NV C1Bz . 32
78. National Distillers Tuscola, IL Et 181
(U.S.!.)
79. Nease Chemical State Colleg~, PA BSAe NDg
80. Northern Petrochemical '~orris, IL Et 400
81. Olin Corporation Brandenburg ,. KY Et 50
82. Oxirane Channel view, Tx EtBz 525
St 454
83. Pennzoil (Atlas) Shreveport, LA Bzc 49
84. Phill ips Petroleum. Borger, TX Cyx 104
EtBz NDg
85. Phi 11 ips Petroleum Pasadena, TX Et 13
86. Phi 11 ips Petroleum Sweeny, TX Bz 33
Cyx. 250
Et (3 units) 973
9-6
-------
Tabl e 9-1. REFINERIES AND SYNTHETIC ORGANIC
CHEMICAL MANUFACTURING SITES
WITH BENZENE FUGITIVE EMISSION POTENTIAL (CONTINUED)
Benzene-Related Capaci tyb
Product~
Plant City/State At Site (Gq/yr )-
87. Phillips Puerto Rico Guayania, PR Bz 367
CyxC 212
88. Puerto Rico Olefins Penuelas, PR Et 454
89. PPG Natrium, WV Cl Bz NOg
90. PPG New Martinsville, WV C1Bz 64
91. Quintana-Howell Corpus Christi, TX Bzc 23
92. Reichhold Chemicals Elizabeth, NJ MAN 14
93. Reichhold Chemicals Morri s, IL MAN NOg
94. Reichhold Chemicals Tuscaloosa, AL BSA NOg
95. Rubicon Geismar, LA NiBz 170
96. Shell Chemi cal Houston, TX Et 590
97. Shell Oil Oeer Park, TX Bzc 301
Cu 326
Et 681
98. Shell Chemi ca 1 Norco, LA d 133
BZd
Et 681
99. Shell Oil Odessa, TX Bz 40
100. Shell Oil Wood River, IL Bz 150
101. Specialty Organics Irwindale, CA C1Bz 2
102. Standard Chlorine Delaware City, DE C1Bz 125
103. Standard Chlorine Kearny, NJ C1Bz 7
104. Standard Oil (CA)/ El Segundo, CA Bz 77
Chevron Chemical Cu 45
105. Standard Oil (CA) Pascagoula, HS Bz NOg
Chevron
106. Standard Oil (CA) Ri c hmond, CA Bz NDg
Chevron
107. Standard Oil (IN) Alvin, TX Et (2 units) 907
9-7
-------
Table 9~1. REFINERIES AND SYNTHETIC ORGANIC
CHEMICAL MANUFACTURING SITES .
WITH BENZENE FUGITIVE. EMISSION POTENTIAL (CONTINUED)
Benzene-Related Capacityb
Product~
Plant City/State At Site ~
108. Standard Oil (IN)/ Joliet., IL MAN 27
Amoco
109. Standard Oil (IN)/ Texas City, TX Bz 284
Amoco Cu 14
EtBz 286
St 381
110. Standard Oil (OH)/ Marcus Hook, PA Bz 27
B P Oi 1
111. Stauffer Chemical Henderson, NV BSA 4
112. Sun Oil Corpus Christi, TX Bz 127
Cu 113
Et 9
EtBz 61
St 54
113. Sun Oil Marcus Hook, PA Bz 97
114. Sun Oil Toledo, OH Bzc 164
115. Sun 0; 1 Tulsa, OK Bz 80
CyxC 83
116. Sun-Ql;n Claymont, DE Et 109
117. Tenneco Chalmette, LA Bz 33
EtBz 16
,
118. Tenneco FOlrds, NJ MAN 12
119. Texaco Port Arthur, TX Bz 150
CyxC 117
Et 454
120. Texaco Wes tv ill e, NJ Bz 117
Cu 64
121. Texaco/Jefferson Bellaire, TX Et 240
Chemical
122. Texaco/Jefferson Port Neches, TX Et 238
Chem;ca1
9-8
-------
Table 9-1. REFINERIES ANO SYNTHETIC ORGANIC
CHEMICAL MANUFACTURINQ SITES
WITH BENZENE FUGITIVE EMISSION POTENTIAL (CONTINUEO)
Benzene-Related Capacityb
Product~
Plant City/State At Site (Gg/yr)
123. Union Carbide Institute, WV EtBz N09
LAB 64
St N09
124. Union Carbide Penuel as , PR Bz N09
Cu 290
Et 454
125. Union Carbide Seadrift, TX Et 546
EtBz 154
St 136
126. Union Carbide Taft, LA Bzc 234
Et 500
127. Union Carbide Texas'City, TX Et 546
128. Union Carbide Torrance, CA Et 73
129. Union Oil of CA Lemont, IL Bz 57
130. Union Pacific/ Corpus Christi, TX Bz 33
Champlin Cud N09
Cyx 65
131. U. S. Steel Neville Island, PA MAN 38
132. USS Chemicals Houston, TX Et 227
133. Vertac/Transvaal Jacksonvi 11 e, AR C1Bz N09
134. Witco Chemical Carson, CA LAB 20
aBSA = Benzenesulfonic Acid Hqn = Hydroquinone
Bz ~ = Benzene LAB = Linear Alkylbenzene
C1Bz = Chlorobenzene MAN = Maleic Anhydride
Cu = Cumene NiBz = Nitrobenzene
Cyx = Cyclohexane Rcnol = Resorcinol
Et = Ethylene St = Styrene
EtBz = Ethylbenzene
9-9
I'
,[
-------
Table 9-1. REFINERIES AND SYNTHETIC ORGANIC
CHEMICAL MANUFACTURING SITES
WITH BENZENE FUGITIVE EMISSION POTENTIAL {CONCLUDED}
bAnnual capacities for each product were obtained from the following
sources {effective date of capacity in parentheses}:
BSA - Ref. 3 {January 1977}
Bz - Refs. 3 {January 1977}, 14.
C1Bz - Refs. 4 (January 1977), 13, 14
Cu - Ref. 9 (January 1979), 13, 14 .
Cyx ~ Ref. 2 (November 1976), 3 (January 1977)
Et - Refs. 5 {1977 year-end), 15 (June 1979), 11,
EtBz - Ref. 10 (January 1979) .
Hqn - Capacity estimate from industry (1979)
lAB - Ref. 8 {June 1978} .
MAN - Ref. 3 {January 1977}
NiBz - Refs. 7, 32
Renol - Ref. 6
St - Refs. 1 (1977 year-end), 14
13, 14, 33
cProduct unit under expansion
dproduct unit under construction
eproduct unit on standby or not currently in use
fproduct unit in engineering phase
gNo data available
9-10
-------
~"
Table 9-2. NUMBER OF COMPANIES AND PLANT SITES
THAT MANUFACTURE BENZENE DERIVATIVES
Number of Number of Sites
Benzene Derivative Companies (and Units)
Chlorobenzene 8 10
Nitrobenzene 8 10
Ethyl benzene 15 18
Styrene 10 14
Linear Alkylbenzene 4 4
Cyclohexane 9 11
Cumene 12 13
Maleic Anhydride 8 10
Resorcinol 1 1
Benzenesulfonic Acid 4 4
Hydroquinone 1 1
, I
, I
I
9.1.3 Methods of Manufacture
9.1.3.1 Benzene. Benzene is primarily manufactured by five
methods. In most instances, benzene producers obtain the material
from which benzene is made from their own refining or manufacturing
operations. In other cases, a benzene producer may buy benzene-
containing material from another source. Four of the methods
(extraction from catalytic reformate, toluene dealkylation, toluene
disproportionation, and processing benzene from pyrolysis gasoline)
use refinery products as the feedstock; coke-oven light oil, from
which benzene can also be extracted, is a by-product of converting
coal into coke for steel manufacturing. Of these methods, extraction
from catalytic reformate accounted for over half of the 1976 benzene
supply.
9.1.3.2 Ethylene. Almost all commercial ethylene is produced by
pyrolysis of natural-gas liquids and petroleum fractions. Although
significant amounts of ethylene were once extracted from by-product
refinery streams (40 percent of the U.S. production in 1956), only
9-11
I
,
'i
-------
Table 9-3. SUMMARY OF PRODUCTION AND CAPACITY FOR.
BENZENE, ETHYLENE, AND BENZENE DERIVATIVES1-10
Total U.S. Total U.S.
Capacity Production Capacity
Product (Gg 1977) (Gg 1977) Utilization
Benzene 7,008 5,256a 75 Percent
Ethylene 15,100 10,600 70 Percent
Chlorobenzenes 440 234 53.Percent
Nitrobenzene 441 252 61 Percent
Ethyl benzene 5,070 2,829 56 Percent
Styrene 3,741 2,694a 72 Percent
Linear
Alkylbenzenes 308 239 78 Percent
Cyclohexane 1,395 983 70 Percent
Cumene 1,653 1,281 77 Percent
t'1a 1 ei c Anhydri de 236 132 56 Percent
Resorcinol 16b 14a 86 percentC
Benzenesulfonic 6lb 48b . b
Acid 79 Percent
Hydroquinone 3b d d
N.D. N.D.
aEstimated from the percent capacity utilization for the product.
bRepresents 1977 capacity for one company.
cBased on benzene consumption estimated for 1976.
dN.D. designates no data availa~le.
9-12
-------
about 2 percent of the current ethylene production is derived from
this source. Most of the plants that are extracting ethylene from
refinery s~reams also produce ethylene by pyrolysis.17
Several alternative pyrolysis processes, primarily utilizing
feedstocks not currently in common use, are either being commercially
attempted on a limited scale or are in the developmental stage with
expectations of limited commercial application between 1980 and 1985.
Although these processes are all expected to be commercially proven
within five years, wide application will depend on demonstrated favor-
able process economics. No significant impact on total olefins
production is anticipated from these developmental processes for at
least 10 years.17 .
The primary difference between the domestic and foreign olefins
industries has been in the feedstocks used for pyrolysis. In Japan
and Europe natural-gas liquids have historically been scarce and
naphtha has been the predominant feedstock.17
9.1.3.3 Chlorobenzene. All domestic chlorobenzene production is
based on direct chlorination of benzene. The principal chlorobenzene
product is monochlorobenzene with smaller amounts of ortho- and para-
dichlorobenzene being co-produced.18
9.1.3.4 Nitrobenzene. Nitrobenzene is produced by the direct
nitration of benzene with a mixture of nitric acid, sulfuric acid, and
water.19,20 The reaction vessels are specially built cast iron or
steel kettles fitted with efficient agitators. The kettles are
jacketed and generally contain internal cooling coils for proper
temperature control of the strongly exothermic reaction. Typically, a
batch process is employed; however, newer plants use a continuous
process.
9.1. 3.5 Ethyl benzene/Styrene. More than 95 percent of domes tic
ethyl benzene production is by benzene alkylation with ethylene. The
remainder is recovered by distillation from mixed xylene streams that
result from naphtha reforming or cracking in petroleum refineries.
More than 99 percent of the ethylbenzene produced is used as an inter-
mediate for making styrene, often in an integrated ethyl benzene-styrene
plant. Except for a new plant brought on-stream in July 1977 by
9-13
-------
[.
Oxirane Corporation, all domestic styrene is produced by catalytic
dehydrogenation of ethyl benzene. The Oxirane ethyl benzene oxidation
process is also used in Spain and Japan; however, most foreign styrene
production is by dehydrogenation of ethylbenzene.2l .
9.1.3.6 Linear Alkylbenzene. Two major processes are used to
manufacture linear alkylbenzene (LAB) in the United States. Approxi-
mately 67 percent is manufactured by three companies using the paraffin
chlorination process, and approximately 33 percent is manufactured by
one company using the olefin (paraffin dehydrogenation) process. The
only significant foreign process not used in the United States uses as
feedstock the lin~ar alphaolefins produced by Shell's wax cracking
process (Shell Nederland Chemie NV, Pernis, The Netherlands). These
linear alpha olefins are alkylated with benzene at several locations
to produce LAB, but the LAB from linear alpha olefins produces a
detergent with a slightly different balance of detergent properties.22
9.1.3.7 Cyclohexane. Two processes are used commercially to
manufacture cyclohexane: catalytic hydrogenation of benzene, which
accounts for approximately 85 percent of the cyclohexane capacity in
the United States; and separation from petroleum liquids, which con-
stitutes the remaining 15 percent.23
9.1.3.8 Maleic Anhydride. The two major processes used to
manufacture maleic anhydride (M~N) in the United States are benzene
oxidation and. butane oxidation. Most major U.S. producers are employ-
ing the latter method. A small amount of MAN is recovered as a
by-product of phthalic anhydride production. The only significant
foreign process for MAN production not used in the United States
starts with a butene mixture feedstock. This process is operated in
France and Japan.24
9.1.3.9 Cumene. .All commercial cumene is produced by alkylating
benzene in the vapor phase with propylene in the presence of a phosphoric
acid catalyst. An excess of benzene is maintained to suppress dialkyla-
tion, oligomerization, and other side reactions.20 Essentially all .
cumene produced is consumed in the manufacture of phenol and acetone.9
9-14
-------
-r-- - - I
I
9.1.3.10 Resorcinol. All commercial resorcinol is produced by
the benzene sulfonation process, in which benzene is sulfonated to
m-disulfonic acid and treated with sodium sulfite to form sodium salt.
After the salt is fused with sodium hydroxide, dissolved in water, and
acidified with sulfuric acid, resorcinol is obtained by solvent extrac-
tion. Other processes have been developed, but such operations have
. 6
not been proved commercially successful.
9.1.3.11 Benzenesulfonic Acid (BSAt. BSA can be produced by
three methods: sulfonation with sulfuric acid, oleum, or sulfur
trioxide.20 Sulfonation with sulfuric acid can be accomplished by a
batch or continuous process. In the batch process, benzene and sulfuric
acid monohydrate are added to a sulfonator, agitated, and heated. In
the continuous process, sulfuric acid is steadily fed to the sulfonator
simultaneously with benzene, which has been previously fed through a
vaporizer-superheater. The mixture flows through the reactor and is
discharged from the bottom.
Sulfonation with oleum is accomplished by charging liquid benzene
to a pre-sulfonator and feeding 9.5 percent oleum over a period of
time. The mixture is pumped to vapor-feed sulfonators where benzene
vapor is added until a desired residual-acid level is attained. The
BSA mixture flows from the bottom of the reactor to storage.
In the third sulfonation method, benzene reacts with sulfur
trioxide in liquid sulfur dioxide, which is evaporated until a certain
temperature is attained. Benzene is then added, the temperature is
raised, and sulfur dioxide is removed by an air stream.
9.1.3.12 Hydroquinone. In the manufacture of hydroquinone,
benzene (or recycled cumene) is alkylated with propylene to yield four
intermediate products: p-diisopropylbenzene (p-DIPB), o-DIPB, m-DIPB,
and triisopropylbenzene (TIPB). The p-DIPB is purified and oxidized'
to a dihydroperoxide. Addition of sulfuric acid effectively splits
the intermediate into hydroquinone and acetone. The other intermediates
are reacted with benzene to yield cumene, which is recycled back to
the alkylation process.25 This'methoq is used by the only U.S. producer
of hydroquinone.
9-15
I, I
-------
Hydroquinone can also be manufactured by oxidizing aniline to
form p-benzoquinone, which is reduced, filtered to remove 'iron oxide,
and distilled. The distilled product is dissolved in dilute sulfuric
acid with a decolorizing agent and then filtered. A small amount of
sodium hydrosulfite is added to the filtrate from which the hydroquinone
crys ta 11 i zes. 20
9.1.4 Uses of Benzene
The companies that produce benzene often consume it in the manufacture
of another product. According to the International Trade Commission,
captive consumption accounted for 55 percent of production in 1976 and
54 percent of production in 1977.26 Captive consumption is relatively..
dependent on benzene price, since the prices of products made from
benzene usually follow the same trends as benzene prices.
.. .
As a feedstock material, henzene presents a complex picture
because of the diverse number of chemicals derived from it. Benzene
derivatives find their largest uses in consumer goods, which account
for 25 percent of the benzene produced in the U.S. This area comprises
packaging, toys, sporting goods, disposables, novelties, and other
small items. The major benzene derivatives used for these products
include styrenics such as polystyrene, epoxy resins, acrylonitrile-
butadiene-styrene (ABS) and styrene acrylonitrile (SAN). The other
major end-uses are household goods and transportation, each taking 17
percent of the benzene con$umption. Household goods -- furniture,
appliances, carpeting -- use nylon fibers and resins, ABS, polystyrene,
phenolics, and epoxies, am9ng others. Plastics, fibers, elastomers,
and rubber are used in boats and airplanes, as well as in trucks and
27
automobiles.
Figure 9-1 depicts the usage and percentages of total henzene
production that is consumed by intermediate and final products. The
products that demand over half: of the benzene production are ethyl-
benzene and styrene. Tables 9-4 through 9-17 list the usage of ethylene
and benzene derivatives. As shown in the tables, some of these products
are consumed almost entirely (often captively) as intermediates in the
manufacture of another product. In such cases, the use and consumption
of the second-generation product have been included for reference.
9-16
-------
...
--~-
BENZENE
1.0
.
~
-....J ,
POLYSTYRENE (28%)
- molded plastic
- packaging' , '
STYRENE COPOLYMER RESINS (9%)
~ construction '
',"; . automobiles'
- appliances
SBR elastomers (5%)
- tires' ,
ETHYLBENZENE/STYRENE (51%)
CUMENE/PHENOL (17%)
CYCLOHEXANE (15%)
ANILINE (4%)
Figure 9-1.
r:H~~~~;d ~~~=;i~;:)'
I
-------
Table 9-4.
ETHYLENE USAGE17
End Use
Percent of
Consumption
(1976)
Low-Density Polyethylene
High-Density Polyethylene
27.4
14.7
Ethylene Oxide
Vinyl Chloride
18.6
12.0
Ethylebenzene, Styrene
9.1
3.5
Ethyl Alcohol
Aliphatic Alcohols
Acetaldehyde
2.5
2.6
Vinyi Acetate
Ethyl Chloride
2.2
Alpha Olefins .
Other
1.4
1.5
4.5
100.0
I
I .
I
9-18
-------
Table 9-5. MONOCHLOROBENZENE USAGE18
End Use
Percent of
Consumption
(1977)
Solvents
Nitrochlorobenzene
(Agricultural Products)
DDT, Silicones, etc.
Diphenyl Oxide
Rubber Intermediates
30
35
15
10
10
Table 9-6.
DICHLOROBENZENES USAGE18
End Use
Percent of
Consumption
(1976) .
o-Dichlorobenzene
3,4,Dichloroaniline, etc.
TDl* Process Solvent
Solvents
Dye Manufacture
Pesticides, etc.
65
15
10
5
5
~-Di~hlorobenzene
Space Deodorant
,
Intermediate for Pesticides.
90
10
*Toluene Diisocyanate
9 -19
-------
Table 9-7. NITROBENZENE USAGE7
End Use
Percent of
Consumption
(1978 )
Anil i ne
Solvent, Dichloroaniline
98
2
Table 9-8. ANILINE USAGE7
End Use
Percent of
Consumption
(1978) .
MDI*
Rubber Chemicals
Dyes
Hydroquinone
Drugs, Pesticides
52
29
4
3
12
*p,p'-Methylene Diphenyldiisocyanate
9-20
-------
Table 9-9.
ETHYLBENZENE USAGE20
End Use
Percent of
Consumption
(I 976)
Styrene
Solvent
99
1
Table 9-10.
STYRENE USAGE20
End Use
Percent of
Consumption
(1976)
Po 1 ys tyrene
Styrene Copolymer Resins
Styrene-Butadiene Elastomers
Unsaturated Polyester Resins
Miscellaneous
Exports
54
17
9
6
1
13
9-21
[!
1
I'
-------
Table 9-11. LINEAR ALKYLBENZENE USAGE8
End Use
Percent of
Consumption
. (1977)
Linear Alkylbenzene
Sulfonates*
Export
90
10
*Detergent Surfactant.
Table 9-12. CYCLOHEXANE USAGE22
End Use
Percent of
Consumption
(1977)
Adipic Acid
Exports
Caprolactam
1,6,-Hexamethylenediamine
(HMDA)
Miscellaneous
53
18
23
3
3
9-22
-------
r
:1
II
Table 9-13. CUMENE USAGE9
End Use
Percent of
Consumption
(1976)
Phenol and Acetone
a-Methyl styrene, Solvent
99
1
Table 9-14. MALEIC ANHYDRIDE USAGE24
End Use
Percent of
Consumption
(1975 )
Unsaturated Poly~ster Resins
Fumaric Acid
Agricultural Chemicals
Alkyd Resins
Lubricating Additives
Copolymers
Reactive plasticizers
Maleic Acid
Chlorendic Anhydride and
Acid
Surface-Active Agents
Other
51.1
6.4
10.0
1.3
7.8
5.3
3.6
3.8
1.1
2.9
6.7
9-23
-------
Table 9-15. RESORCINOL USAGE6
End Use
Percent of
Consumption
(1977)
Rubber Products
Wood Adhesive Resins
Mi sce 11 aneous
59.6
25.5
14.9
Table 9-16.
BENZENESULFONIC ACID USAGE3
End Use
Percent of
Consumption
Phenol
Dyes
N.D.*
N.D.*
*N.D. designates no data available.
I'
Table 9-17.
HYDROQUINONE USAGE3
End Use
Percent of
Consumption
Rubber Antioxidant
Photographic Developer
Dye Intermediates
*N.D. designates no data available.
N.D.*
N.D.*
N.D.*
9-24
i .
-------
9.1.5 Price History
In spite of price controls through August 1976, benzene prices as
well as benzene derivative prices rose at a greater rate than prices
of .most other chemicals in response to the higher cost of crude oil.
Even with a decline in demand in 1975, benzene prices continued a
general upward movement.3 Table 9-18 gives the price history of
benzene, ethylene, and benzene derivatives since 1974. Market experts
forecast a continuing upward trend in benzene prices.29 Increases in
the cost of crude oil along with other market functions (discussed in
Section 9.1.6 below) are responsible for a large portion of benzene
price increases.
9.1.6 Market Factors that Affect the Benzene Industry.
Benzene is contained in materials that have other uses. Therefore,
the chemical uses of benzene must be profitable enough to justify
recovering benzene from these materials. Whether the benzene will be
produced depends on a number of factors, such as the value of the
material in which benzene is contained (reformate, pyrolysis gasoline,
toluene, coke-oven light oil), the value of benzene before it is
recovered, processing costs, operating costs, and the value of benzene
in relation to benzene substitutes.
9.1.7 Feedstock Substitutions for Benzene Derivatives
The price of benzene has increased considerably since the
beginning of 1979. As the cost of benzene rises, greater incentive is
provided for developing and using alternative feedstocks for the
production of benzene derivatives. For some products, such as nitro-
benzene, there are no feedstock substitutes; consequently, increased
benzene costs must be passed through to the customer. For other
products, however, alternative processes using feedstocks other than
benzene are available, and it is possible that production by such
alternative processes will increase.
Table 9-19 lists some alternative processes by which benzene
derivatives may be produced in the United States. Most of the sub-
stitute feedstocks used in these processes are derived from petroleum
or natural gas, however, so that the economics of the alternative
processes are still tied to the cost of crude oil and natural gas.
9-25
-------
Table 9-18.- PRICE HISTORY FOR BENZENE,
ETHYLENE, AND BENZENE DERIVATIVES 28
Unit Price History (Dollars/Kilogram)
Product May
1979 1978 1977
Benzene 0~37 0~22 0.24
Ethylene 0.31 0.29 0.27
Chlorobenzenes 0.30 0.27 0.27
Nitrobenzene a 0.51 0.51
N.D.
Ani,l i neb 0.79 0.75 0.75
Ethylbenzenec a a 0.15
N.D. N.D.
Styrened 0.63 0.43 0.46
Linear Alkylbenzene 0.75 0.65 0.58
Cyclohexane 0.46 0.27 0.29
Cumene 0.40 0.31 0.31
Maleic Anhydride 0.88 0.68 0.82
Benzenesulfonic Acid/Phenole 0.68 0.46 0.59
Resorcinol 3.09 2.87 2.72
Hydroquinone 3.39 3.39 3.30
1976 1975 1974
0.23 0.23 0.23
0.25 0.19 0.17
a 0.29 0.25
N.D.
0.51 0.42 0.21
0.71 0.71 0.24
a- a 0.13
N.D. N.D.
a a a
N.D. N.D. N.D.
0.56 0.25
0.27 0.27
0.31. 0.35
0.25
0.27
N.D.a
0.35
0.82
0.82
a
N.D.
a a
N.D. N.D.
a
N.D.
a a
N.D. N.D.
a
N.D.
N.D;a
a
N.D.
aN.D. designates no data available.
bSince 97 to '98 percent of nitrobenzene production goes into aniline
manufacture, aniline prices hav,e b~en provided for comparison.
cPrices are not generally available for ethyl benzene because greater
than 99 percent of production is capitvely consumed.
dSince most ethyl benzene produc~ion is consumed in styrene manufacture,
styrene prices have been provided for comparison.
eSince benzenesulfonic acid (BSA) is used in the production of phenol and
not as a final product, no price data are available for BSA. Prices
given are for phenol.
9-26
-------
Table 9-19. ALTERNATIVE PROCESSES FOR THE
MANUFACTURE OF BENZENE DERIVATIVES
Benzene Derivative
Alternative Processes
Reference
Number
Cumene
Cyclohexane
Ethyl benzene
Separation from petroleum liquids
Separation from petroleum liquids
Extraction from mixed xylene streams
2
20
Maleic Anhydride
Oxidation of n-butane; by-product of
phthalic anhydride production (xylene
derivative) ..
3,20
Styrene
Propylene oxide coproduct; extraction
from pyrolysis gasoline; production
from toluene and ethylene via stilbene
1
9.1.8 Future Trends
9.1.8.1 Projected Growth Rates. Table 9-20 depicts projected
growth rates for benzene, ethylene, and benzene derivatives through
1983. Demand for benzene produced from all sources (extraction from
catalytic reformate, toluene dealkylation, toluene disproportionation,
and processing benzene from pyrolysis gas) is expected to grow at 5 to
5.5 percent per year through 1985. The gap between production and
capacity is expected to narrow from the 1977 value of 32 percent to
about 20 percent by the end of this period.27,29,30 Production capac-
ity is expected to reach 8359 Gg by 1985.27,31 During this forecast
period, demand for benzene from catalytic reforming is likely to
increase over 1977 requirements at about 1 percent annually, while
toluene dealkylation production is expected to grow at approximately
3.4 percent per year. Benzene from pyrolysis gas extraction and
dealkylation is expected to increase 14 percent per year, providing
the largest new source of the aromatic.27
Styrene manufacture is expected to continue to be the biggest
consumer of benzene, requiring over 50 percent of the benzene market,
followed by cumene/phenol and cyclohexane. Benzene production capac-
ity is expected to be satisfactory through 1982-83, with only minimal
needs for additional capacity beyond this point until 1986-87.27
9-27
-------
Table 9-20. PROJECTED ANNUAL GROWTH RATES FOR DEMAND OF
BENZENE, ETHYLENE, AND BENZENE DERIVATIVES1-10, 17-19, 21-24, 27
Product
Projected
Average Annual
Percent Growth
(1977 - 1983)
Benzene
5.5a
Ethylene
5.5
1.5
Chlorobenzenes
Nitrobenzen~s
EthylebenzenejStyrene
6.0
6.0
Linear Alkylbenzene
2.0
5.0
Cyclohexane
Cunene
7.5
Maleic Anhydride
Resorcinol
11.0
b
N.D.
b
N.D.
b
N.D.
Benzenesulfonic Acid
Hyd roqu i none
aGrowth rate for benzene is estimated at 5 to 5.5 percent per
year through 1985.
bN.D. designates no data available.
9-28.
-------
Although capacity may pose no problem, benzene may grow scarce as
the demand for unleaded gasoline increases into the early 1980's.
Although the total demand for gasoline is growing slowly or not at
all, the unleaded portion is increasing rapidly. With increased
unleaded gas production, refiners need a higher clear-pool octane.
The result is an increase in the amounts of aromatics needed in gaso-
line (see Section 9.1.6 above).29,31
9.1.8.2 Replacement Rate of Equipment. The replacement rate of
benzene-manufacturing equipment is low since companies tend to refur-
bish their equipment on a continuous basis rather than replace it.
This practice is characteristic of refinery and SOCMI operations.
9.1.8.3 Planned Expansions of Capacity. Table 9-1 includes
plants that are undergoing expansions of capacity through 1980. An
additional capacity of 1119 Gg per year is estimated for all benzene-
producing companies, 1407 Gg per year for all ethylene plants, and
1432 Gg per year for all plants that manufacture benzene derivatives.
Expansions of capacity will take place primarily at present plant
locations by means of purchase or construction of new equipment.
Capacity expansions are expected to be located mainly in Texas and
Louisiana.
9-29
-------
9.2 MICROECONOMIC IMPACT
9.2.1
Introduction
In the following sections the microeconomic impacts of applying the
regulatory alternatives are detailed. Such impacts are discussed in terms
of both the potential price, as well as capital availability, impacts of
each alternative.
With regard to the maximum price increases, regulatory alternatives II,
III, IV, and V could cause average benzene derivative prices to rise .04,
.13, .37, and 4.12 percent, respectively. Concerning the burden imposed by
the capital control costs, the .average percentage increases in capital
" investment required of new plants are .07, .20, .59, and 4.47 for regulatory
alternatives II through V.
The conclusions noted above are based upon observations regarding the.
market structure and competitive nature of the industry, as well as the
demand and supply outlook for benzene and its derivatives. Specific con-
clusions concerning potential price and investment impacts, resulted from
assessing the responses of individual model plants to the capital control
and annualized costs presented in Chapter 8.
In the sections which follow the industry structure (Section 9.2.2),
demand characteristics (Section 9.2.3), and supply characteristics (Section
9.2.4) are examined so that industry responses to control costs may be
assessed. Section 9.2.5 details the methodology employed in deriving the
conclusions noted in Section 9.2.6, while Section 9.3 addresses potential
macroeconomic impacts.
9.2.2 Industry Structure
As noted in Section 9.1.1, there are currently 74 domestic producers of
benzene and the eleven derivatives of benzene dealt with in this analysis.
Benzene in its pure form is produced by 32 companies, and benzene, obtained
as a by-product in the manufacture of ethylene, is produced by 28 companies.
The eleven derivatives of benzene are produced by 50 companies. Production
capacity for benzene, ethylene, and the benzene derivatives is concentrated
in Texas and Louisiana.
Many of the firms that produce benzene and benzene derivatives are
diversified, and involved in industrial sectors other than organic chemi-
cals. Some of the firms rank among the largest in the world. Oil companies
9-30
-------
playa particularly important role in the production of benzene and some
of its derivatives. One industry source estimates that oil companies pre-
sently account for about 85 percent of domestic production capacity for
benzene.34
As is the case with most chemicals, the production of benzene and its
deri vat i ves tends to be domi nated by a re 1. at i ve ly sma 11 number of fi rms. The.
degree of dominance in the various product areas is illustrated by the con-
centration ratios presented in Table 9-£1. For each chemical, the table
shows the precentages of production capacity accounted for by the top two and
top four companies (for benzene and ethylene, percentages for the top eight
are also given, because of the greater number of firms involved). In the
case of benzene, the top two companies account for 18 percent of production
capacity, while the top four companies account for 34 percent. The eight-
firm ratio is 58 percent. For ethylene, which has a total of 28 producers,
the two- and four-firm ratios are 23 and 43 percent, respectively. The ratio
at the eight-firm level is 68 percent. With respect to the eleven benzene
derivatives, the numbers of producers range from one to fifteen. In the
cases of hydroquinone and resorcinol, there is only one p~oducer involved
in each, and the concentration ratios are therefore 100 percent. For the
other derivatives, the two-firm ratios range from 34 percent for cumene (12
firms) to 73 percent for linear alkyl benzene (4 firms). The four-firm ratios
range from 59 percent for ethyl benzene (15 firms) to 100 percent for linear
alkylbenzene.
Many of the firms that produce benzene and benzene derivatives are
vertically integrated organizations which capitvely consume much or even all
of their output in the manufacture of other products. For example, it is
estimated that approximately 54 percent of the benzene produced domestically
in 1977 was consumed captively.26 In product areas where the degree of
captive use is high, the merchant market for the chemical. in question tends
to be dominated by a small number of high volume sellers.
The existing market structures for benzene, ethylene, and benzene deri-
vatives are reinforced by the existence of significant barriers to entry.
Over the years, there has been a trend in the chemical industry towards
constructing larger and larger plants in order to take advantage of both
new technologies as well ~s economies of scale. Important economies can be
9-31
'.'
-------
Table 9-21. CONCENTRATION RATIOSa FOR
BENZENE, ETHYLENE, AND BENZENE DERIVATIVES
# of Concentration ratio (%)
Chemical firms 2 firm 4 firm
Benzeneb 32 18 34
EthyleneC 28 23 43
Benzenesulfonic Acid 4 d 100
Chlorobenzene 8 56 87
Cumene 12 . 34 63
Cyclohexane 9 54 12
Ethyl benzene 15 35 59
Hydroqui none 1 100
Linear Alkylbenzene 4 73 100
Ma 1 ei c Anhydri de . 8 41 67
Nitrobenzene 8 47 82
Resorcinol 1 100
Styrene. 10 45 74
aRatios calculated from production capacity data presented in Table 9-1.
The two (four) firm ratios indicate the percentage of total productive
capacity controlled by 'the two (four) largest producers of each chemical.
bRatio for top 8 firms is 58 percent.
CRatio for top 8 firms is 68 percent.
dproductive ~apacity data not available.
I
I
9-32
-------
I
r~--- -
I
I
, I
, ,
I
realized in areas such as the purchase of raw materials, spreading overhead,
and achieving lower capital requirements per unit of capacity. The effect of
these economies is evidenced by' the fact that as plant sizes have increased~
unit costs have declined. What this situation means to a new firm hoping to
enter a merchant market is that in order to compete with the same unit costs~
it must be able to achieve the same scale of production~ thus requiring
the considerable financial resources needed to construct a large facility.
Secondly, in order to justify such an investment, the firm must be able
to capture a sizeable share of the market. The barriers to entry can be
heightened further by the presence of captive consumption, which reduces
the potential of the merchant market.
It should be noted that despite the trend toward larger plants, the
chemical industry does have a number of smaller plants which operate effec-
tively in their respective product markets. But, in general, these tend to
be older plants, which may be fully depreciated. Often, such small plants
provide chemicals for captive use. Given the importance of assured supplies
of raw materials in the chemical industry, many firms would be reluctant to
abandon internal sources of supply to gain favorable, but relatively small
decreases in costs.35
When new firms do begin production of a chemical, it is frequently
for purpose of vertical integration. In these cases, the manufacture of a
chemical is initiated in order to supply the firm's internal demand for an
intermediate chemical. Another possible avenue for the entry of new firms
is through technological change, which may provide an immediate cost advan-
tage. This has not been that important in recent years, however.35
The market structures for benzene and benzene deri vat i ves may be cha rac-
terized as oligopolistic, that is, a small number of participating firms
and/or high concentration ratios, with recognition on the part of partici-
pating firms that their decisions are interdependent. In general, price
.changes tend to be initiated by a price leader (or leaders); however, the
price increases may be withdrawn if the other participants in the product
market do not follow the lead. Prices are usually set as simple percentage
markups over costs, or on the basis of target rates of return.36
In many industries, price is the most important factor in competition,
although in the chemical industry its impor~ance tends to be reduced by three
, I
I :
, I
9-33
I
,I
-------
- 4 .- -- ,--
factors: 1) joint product cost accounting; 2) low price elasticity of
demand; and 3) the consumer's interest in an uninterrupted supply.37 In
the case of the first factor, the basic point is that ,it is often diffi-
cult for chemical firms to assign costs to any single product because of
the complex, interrelated nature of chemical production processes, in which
bQth variable and fixed costs must be apportioned between several different
products manufactured at a single plant. Given this type of situation, the
price of a particular product can become almost an arbitrary matter as long
as the combined prices of the joint products yield the firm's desired return.
The second factor, low price elasticity of demand, reduces the importance of
price competition since it implies that product demand tends to be insensi-
tive to changes in price. The basic reason for low price elasticity of
demand is the absence of substitutes to which customers could switch in the
face of price increases. As will be discussed in Section 9.2.3.13, the price
elasticity of demand for benzene and benzene derivatives tends to be low.
With regard to the third factor, the consumer's interest in an uninterrupted
sup~ly, it generally is the case that consumers of chemical products are more
sensitive to interruptions in supply than increases in price. This tendency
is reflected in the fact that most chemicals are sold under long-term con-
tracts which insure the customer of a secure supply. In periods when there
are supply shortages, producers generally grant preferential treatment to
their long-term customers. This policy has the effect of discouraging
customers from switching suppliers when small differentials in price occur.
9.2.3 Demand Characteristics
9.2.3.1 Benzene. At the present time, the largest uses for benzene
are ethylbenzene/styrene (51%), cumene/phenol (17%), cyclohexane(15%), and
nitrobenzene/aniline (4%) as noted in Figure 9-1. The major end uses for
these benzene derivatives are in packaging, consumer goods, toys, and dispo-
sables, which account for about 25 percent of the benzene produced domesti-
cally. Other major areas of end use are transportation and household goods,
each of which accounts for about 17 percent of benzene consumption.27 In
1977, about 54 percent of the benzene produced domestically was captively
consumed.
Between 1950 and 1974, the consumption of benzene increased cyclically
, at an average annual rate of about 9 percent. In 1975, consumption plummeted
9-34
-------
f
by more than 30 percent, and while 1976 witnessed a substantial recovery,
consumption was still below the peak levels attained in 1973 and 1974.38
Continued growth in demand is expected, but at a lower rate compared to
earlier years. It is estimated that the demand for benzene from all sources
(extraction from catalytic reformate, toluene dealkylation, toluene dispro-
portionation, and processing from pyrolysis gas) will grow at a rate of from
5 to 5.5 percent per year between 1977 and 1985.39 Ethylbenzene/styrene is
expected to continue to be the largest use of benzene, followed by cumene/
phenol and cyclohexane.
Production capacity is expected to remain adequate through 1982-83;
however, benzene may become scarce as the demand for unleaded gasoline
increases. As noted in Section 9.1.8.1, the total demand for gasoline is
growing slowly or not at all, while the demand for unleaded gasoline is
increasing rapidly. With increased production of unleaded gasoline, the
amount of aromatics needed by refiners will increase.
9.2.3.2 Benzenesulfonic Acid. Benzenesulfonic acid is primarily used
in the production of phenol, with lesser amounts consumed in the manufacture
of dyes, or as catalysts. No data on past consumption or future demand are
available. According to one source,38 the product will continue to be used
in the commercial production of phenol, however no appreciable increase in
consumption for this use is anticipated. Dye and catalyst use are also
expected to remain small.
9.2.3.3 Chlorob~nzene. The principal chlorobenzene product is mono-
chlorobenzene, with smaller amounts of ortho-and para-dichlorobenzene being
co-produced. The consumption of monochlorobenzene declined from a high of
174 Gg in 1960 to an estimated 159-163 Gg in 1976. This drop was due to:
1) the replacement of monochlorobenzene by cumene as an intermediate in the
production of phenol, and 2) the decline in the production of the pesticide
DOT. The demand for monochlorobenzene is expected to increase, but at an
annual rate of no more than two percent through 1981.4 For chlorobenzenes
as a whole, demand is projected to grow at a rate of 1.5 percent per year
through 1983 (refer to Section 9.1.8).
9.2.3.4 Cumene. Nearly all of the cumene produced is used in the
manufacture of phenol and acetone. Roughly half of domestic production is
sold on the merchant market. During the 1960's, the production of cumene
9-35
I
I, !
-------
increased at a rate of about 25 percent per year as it replaced benzene
sulfonation and chlorination as an intermediate step in the production of
phenol. Because most of the domestically-produced phenol is currently made
from cumene, the producti on growth rate has slackened somewhat, and it is
anticipated that production will tend to approximate the growth trend for
phenol.9 It is anticipated that the demand for cumene will increase at
an average annual rate of 7.5 percent through 1983 (refer to Section 9.1.8).
9.2.3.5 Cyclohexane. Most of the cyclohexane consumed in the U.S. is
used in the production of nylon fibers and resins (these are the principal
end uses for adipic acid, caprolactam, and 1,6-hexamethylenediamine, which
are made from cyclohexane). Between 1971 and 1974, total consumption of
cyclohexane increased at an average annual rate of 5.9 percent, from 814 Gg
to 966 Gg. In 1975, consumption fell by 12.7 percent to 844 Gg. The follow-
ing year, there was a reversal, with consumption increasing by 13.4 percent
to 957 Gg.2 The demand for this chemical is projected to grow at an
average annual rate of 5.0 percent (refer to Section 9.1.8).
9.2.3.6 Ethyl benzene. Virtually all ethylbenzene is consumed captively
in the production of styrene. Over the period 1960-1972, consumption of
ethylbenzene increased at an average annual rate of 10.8 percent. From 1972
through 1974, domestic consumption remained essentially constant, due to
shortages of benzene in 1973 and early 1974, and a decline in the demand for
styrene in late 1974. Mirroring the general recession, consumption in 1975
fell by 21.5 percent from the 1974 level. In the following year, there was
an increase of 34.9 percent.l0 It is estimated that demand for ethyl ben-
zene wi 11 grow at an average annua 1 rate of 6 percent through 1983 (refer
to Section 9.1.8).
9.2.3.7 ~droquinone. Hydroquinone is produced in the U.S. by only one
firm, and is used primarily as a rubber antioxidant. It is also used as a
photographic developer, dye intermediate, and in other specialty applications
which capitalize upon its antioxidant properties. No historical data on
demand are available. Projections of future demand are also unavailable.
Some sources have indicated that the product has good growth prospects;
however; no domestic firms other than Goodyear (the only domestic producer
at present) have shown any interest. It is reported that Goodyear plans
to increase the output of its plant.38
9-36
-------
9.2.3.8 h~~_~~_A]~~l~en~~n~. The dominant use for linear alkyl benzenes
(LAB) is as a raw material in the production of linear alkyl benzene sulfonates
(LAS). LAS are currently the principal surfactants for home laundry and dish-
washing detergents. In 1977, the domestic consumption of LAB for production
of LAS amounted to 216 Gg. This was 6.4 percent more than the 203 Gg consumed
in 1975. The future demand for LAB is completely dependent upon the fortunes
of LAS.8 One projection calls for a growth in demand on the order of 2
percent per year through 1983 (refer to Section 9.1.8).
9.2.3.9 ~alei~~nhydrije. The largest use for maleic anhydride is in
the. production of unsaturated polyester resins, with secondary uses in the
production of fumaric acid and agricultural chemicals. Most of the output
of domestic producers is sold on the merchant market. Between 1968 and 1973,
U.S. consumption of maleic anhydride increased at an average annual rate of
over 9 percent, goi ng from 83 Gg to 128 Gg. Most of thi s growth vias due to
increased use for the manufacture of unsaturated polyester res ins. I n the
following year, growth dropped to 7 percent because of feedstock shortages
and the onset of a general business recession. In 1975, when the full impact
. .
of the recession was felt, consumption declined by 28 percent to a level of
98 Gg.40 Since that time, consumption has risen, and it is estimated that
demand will grow at a rate of 11 percent per year through 1983 (refer to
Section 9.1.8). The demand for maleic anhydride will continue to be keyed
to its use in the manufacture of unsaturated polyester resins. According to
one industry source, major gains from unsaturated polyester fibers may be
developing in the auto markets. Depending on their timing and the condition
of the economy, auto market gains could give unsaturated polyester resins a
major boost in the next several years.41
9.2.3.10 Nitrobenzene. Approximately 98 percent of the nitrobenzene
-------
consumed in the U.S. is used in the manufacture of aniline. In turn, the
major uses of aniline are in the manufacture of p,pl-methylene diphenyldi-
isocyanate (MOl) (52%) and the manufacture of rubber-processing chemicals
(28-29%). Most of the nitrobenzene used for aniline is captively consumed.
From 1975 to 1978, the production of nitrobenzene increased at an annual rate
of over 5 percent, going from 439 Gg to 510 Gg.7 It is estimated that
demand will increase at an average annual rate of 6 percent through 1983
(refer to Section 9.1.8).
9-37
-------
[-
----<--
! '
9~2.3.11 Resorcinol. Resorcinol is produced in the u.s. by only one
--- _.- -"---
company, and is consumed mainly in the production of resorcinol-fonmaldehyd~
resins. which are used as high-performance adhesives in the rubber 'and wood
products industries. From 1964 to 1974, domestic consumption of resorcinol
increased from 5 Gg to 11 Gg, yielding an average annual growth of approxi-
mately 7 percent. In 1975, the recession brought a sharp drop in consumption
which was not overcome until 1977.6 Future growth in demand will depend
greatly upon events in the tire industry, the main consumer of resorcinol.
Projections of future demand are not available.
9.2.3.12 Styrene. At the present time, nearly all of the styrene
produced in the U.S. is consumed in the manufacture of polymers (polystyrene,
54%; styrene copolymer resins, 17%;sytrene-butadiene elastomers, 9%, unsatu-
rated polyester resins, 6%)'. Over the years 1960-1972" the consumption of
styrene increased at an average annual rate of 10 to 11 percent. In 1973 and
1974, growth rates declined substantially, and in 1975, consumption declined
18 percent from the level of the previous year. The lower growth rates in
1973 and 1974 were a reflection of lower styrene production gro~~h rates
. ,
brought about by shortages of benzene. The drop in consumption in 1975 was
due to the general economic recession. By late 1975, market conditions began
to improve, and demand rebounded in 1976 to slightly exceed the levels exper-
ienced in 1972-1974. It is anticipated th~t polystyrene will continue to
be the dominant end-use for styrene. Growth in this area, however, can be
expected to slacken since many markets for polystyrene are maturing, and the
product is facing increasing competition from other materials such as paper
and other resins.1 It is estimated that the demand for styrene will grow
at an average annual rate of 6 percent through 1983 (refer to Section 9.1.8).
9.2.3.13 Price Elasticity of Demand for Benzene and Benzene Derivatives.
Price elasticities of demand for benzene and the benzene derivatives must
be evaluated in order to determine the ,sensitivities of the markets for these
products to price increases which could result from the implementation of the
regulatory alternatives outlined in Chapter 6. Data on price elasticity is
an important input to the model plant analysis (presented in Section 9.2.6),
since it provides a basis for determining the ability of firms to pass the
costs of regulation to the consumers of benzene and its derivatives.
9-38
-------
The price elasticities of demand for benzene and the benzene derivatives
were assessed through the use of a qualitative approach involving the exami-
nation of four different screening factors. This methodology is based on
one which was originally employed in a study prepared by Energy Resources
Co., Inc. for the EPA's Office of Solid Waste.35 For each of the chemi-
cals, the following were examined:
o
Historical and projected growth in
especially during periods of price
low price elasticity of demand.
level of captive consumption - This is indicative of the degree to
which the product is insulated from the competitive pressures of
the merchant market. A relatively high level suggests that the
chemical is less subject to price considerations.
Potential for substitution - The availability of direct or indirect
substitutes increases the price elasticity of demand.
level of foreign competition - A high level of import competition
increases the extent to which consumers can switch to foreign
supplies when the prices of domestic products increase.
demand - High rates of growth,
increases, are indicative of a
o
o
o
As part of the evaluative procedure, recognition was given to the fact
that all of the chemical products considered are used mainly as primary
feedstocks or intermediates. Because of heavy investments in existing pro-
cesses, consumers of primary feedstocks and intermediates have a stake in the
continued availability of these inputs. Generally, there is a tendency on
the part of such consumers to be more sensitive to supply interruptions than
they are to price increases. To the extent that this is the case, the price
elasticity of demand will tend to be reduced. The sensitivity of chemical
consumers to price increases can also be reduced by non-price factors such
as product quality, effectiveness in use, and stability of supply.
The results of the qualitative analysis are presented in Table 9-22. In
all cases, the price elasticit;>, of demand is determined to be low.
9.2.4 ~':!PN Characteristics
At present it appears:the
ductive capacity, augmented by
I I
I I
existing benzene and benzene derivative pro-
projected additions to capacity (Table 7-9)
9-39
-------
<.0
I
~
o
Table 9-22. QUALITATIVE EVALUATION
OF PRICE ELASTICITY OF DEMAND
(Continued)
Eva 1 uat; ve factors
Product Demand arowtha Fore; gn Price elasticity
Captive consumDtion Potential for substitution comDetiti on of demand
Benzene 1950-1974. 9% per year. Substantial. Estimated Main uses are ethylbenzenel Imports small. low
Decline of more than 30% tobe.more than 50% of styrene. cumene/phenol. &
in 1975. Recovery followed. production. cyc1ohexane. No good sub-
Projected 5-5.5% per year stitutes for benzene as
over period 1977-1985. raw material for these.
Continuing upward move- Some indirect substitution
ment of pri ces. possible-e.g. substitution
of pOlyester fibers for
nylon fibers.
Benzenesu1fonic Acid No data available on No data available. but Main use is an intermediate None low
historical or projec- captive consumption in production of phenol.
ted demand growth. appears to be important. Only one domestic producer
Small amount of growth by thi s route. Domi nant
anti ci pated. route to phenol is via
cumene.
Chlorobenzene Principal chlorobenzene Much of monochloro- No direct substitutes Imports low
product is monoch10ro- benzene production available in current negli9ib1e.
benzene. From 1960- consumed captivel,. applications. Some
1976. demand for mono- Ortho-& para-dich oro- indirect substitution
ch10robenzene declined benzenes mainly sold. at end-product level
from 174 Gg to 159-163 may be possible in
Gg. For ch1orobenzenes future.
as whole. growth pro-
jected at 1.5% per year
over period 1977-1985.
Prices have been re1a-
tively stable in recent
years.
..
. .
-------
Table 9-22. QUALITATIVE EVALUATION
OF PRICE ELASTICITY OF DEMAND
(Continued)
'0
I
..~
Evaluative factors
Demand arowtha Forei gn Pri ce e 1 as ti city
Product CaDtive consumDtfon Potential for substitution comDeti t ion of demand
Cumene During 1960's, about 25% About 50% consumed Other routes to phenol Substanti al amounts Low
per year. With matura- captive1y. exist, but cumene is of cumene are
tion of phenol markets, preferred. imported.
growth has slowed some-
what. Growth projected
at 7.5% per year over
period 1977-1983. Up-
ward movement in prices.
Cyc10hexane 1971-1974, 5.9% per year. Mainly sold on Limited substitution Imports negligible. Low
Decline in consumption merchant market. possible. Adipic acid
in 1975, followed by can be made from phenol,
reversa 1 in 1976. Growth but this is not an
projected at 5% per year economic substitute.
over period 1977-1983. Several other routes are
Prices continuing to available in production
move upward. of HMO. Acry1 i c and
polyester fibers can be
substituted for nylon.
Ethyl benzene 1960-1972, 10.8% per Virtually all pro- No substitute for ethyl- Imports negligible. Low
year. Consumption con- ductfon captfvely benzene in production of
stant from 1972 to 1974. consumed for styrene. styrene.
Big drop in 1975, fol-
lowed by reversal in
1976. Future growth
projected at 6% per
year through 1983.
Pri ce not a factor
sfnce most ethyl benzene
consumed captively.
;~
-------
Table H2. QUALITATIVE EVALUATION
OF PRICE ELASTICITY OF DEMAND
(Continued)
1..0
I
+>-
N
Evaluative factors
Demand Qrowtha Forei gn Price elasticity
Product Caotive consumotion Potential for substitution comoetition of demand
Hydroquinone No historical data Both capti ve use No substitutes. Not a factor Low
available. Projec- & sale on merchant
tions also unavail- market. Breakdown
able. May have growth unknown.
prospects. Only one
domestic producer at
present.
Linear Alkylbenzene 1975-1977, slightly Both capti ve use Ma in use i sin Imports Low
more than 3% per year. and sale on merchant production of linear negligil3le.
Estimated 9rowth of 2% market. Breakdown alkylbenzene sulfon-
per year over period unknown. ates. No substitu-
1977-1983. Pri ce has tion possible in this
risen somewhat in use.
recent years.
Maleic Anhydride 1968-1973, over 9% per Prod'uced mainly for No direct substitutes Imports Low
year. In 1974, growth merchant market. in production of un- negligible.
slowed to 7%. In 1975, saturated polyester
consumption dropped resins, fumaric acid,
28%. Recovery since and agricultural
then. Growth projec- chemicals.
ted at 11% per year
over period 1977-1983.
Price has increased
greatly since 1974,
and is continuing upward.
..
-------
Tab 1 e 9-22. QUALIT ATl VE EVALUATl ON
OF PRICE ELASTICITY OF DEMAND
(Concluded)
\.0
I
+:>
w
Evaluative factors
Demand orowth a ~orel gn Price elasticity
Product Caoti ve consumotion Potential for substitution comoet1ti on of demand
Nitrobenzene 1975-1978, 5% per year. Most consumed captive1y Aniline can be produced Imports Low
Projected growth of 6% in manufacture of from ammolysis of chloro- negli~ible.
per year over period aniline. benzene, but this source
1977-1983. Pri ce has dependent on surplus of
been gradually increasing. chlorine & sales of by-
product ammonium chloride.
Resorci no 1 1964-1974, about 7% per Degree of captive use No substitutes. Imports Low
year. In 1975, sharp not known. Appears negligible.
drop in consumption, that sale on the mer-
which was not overcome chant market is impor-
until 1977. Projection tanto
of future demand not
available. Pri ce has
been advancing.
Styrene 1960-1972, 10-11% per Captive use estimated No good substitutes for Imports LtM
year. Drop in growth to be around 60%. styrene available at negligih1e.
rates in 1973 & 1974. present time.
In 1975, consumption
declined 18% from pre-
vious year. Reversal
of downward trend in
1976. Growth projected
at 6% per year through
1983.
aSee discussion in Sections 9.2.3.1 to 9.2.3.12 for sources of these data.
-------
I
I .
will be capable of satisfying the growth in demand as discussed in the pre-
vious section. Therefore, while the price of benzene and its derivatives
may rise due to increments in crude oil prices, there is no reason to suspect
additional price pressure due to supply based market disruptions over the
forecast period.
It is, however, important to note two elements which will have a strong
influence upon the future proportions of benzene derived from the major
petroleum sources. First, the demand for the octane enhancing qualities of
petroleum reformate will increase with the rising demand for unleaded gaso-
line. It has been projected42 that the unleaded proportion of the total
gasoline supply will reach about 75 percent by 1985. Second, the supply of
benzene from pyrolysis gasoline (a by-product of ethylene production) will
increase as new, larger ethylene plants, using heavier feedstocks, come
on-stream.42 While the former may restrict the availability of petroleum
. .
reformate as a benzene feedstock, the latter should increase the benzene
producing capabilities of ethylene sources. The combined effect of these
market developments will serve to increase that portion of the total supply
of benzene attributed to pyrolysis gasoline, while the portion contributed
through petroleum reformate will decline.
9.2.5 Economic Impact Methodology
9.2.5.1 Model Plants and Economic Impacts. In the following sections
the economic implications of applying the regulatory alternatives upon the
industry, as represented by the model plants defined in Chapter 6, are
discussed. Impacts are estimated in terms of the increased capital require-
ments of, and the potential price increases associated with, each model plant
under the various regulatory alternatives.
Concerning capital availability impacts, the extent to which the capital
control costs may increase the total investment required for new plants, has
been estimated. With regard to prices, separate analyses have been completed
for both existing and new plants, thus recognizing the difference in net
annualized control costs for both types of plants. While the methodology
employed in the determination of economic impacts is detailed below, the
results of this analysis are presented in Section 9.2.6.
9.2.5~2 Price Impacts Under Full Cost Pricing. In the estimation of
maximum potential product price increases resulting from the alternatives, the
I .
9-44
-------
I
full pass through of control costs, to the consumers of benzene and its
deri vat i ves, has been assumed for all regul atory alternati ves. This full
cost pricing assumption is supported by the low price elasticity of demand
conclusions detailed in Section 9.2.3.13. The methodology employed in the
estimation of price impacts is detailed below while the price impacts them-
selves are presented in Section 9.2.6.1.
Price increases for each chemical produced at each model plant have been
estimated by expressing the net annualized costs of control for each model
plant and regulatory alternative (Table 8-11) as a percentage of the annual
total revenue of each model plant. These percentages are therefore indicators
of the percentage increase in model plant revenues (and thus product prices)
required if the net earnings (post control) of each plant are to remain
unaffected. While the net annualized cost estimates are presented in Section
8.2.3, the total annual revenues of each model plant have been estimated
based on annual output and price estimates as shown in Tables 9-23 through
9-25, and described below.
In the estimation of output, the capacity of each model plant has been
taken as the mean capacity of existing plants. The capacities noted in
Tables 9-23 through 9-25 are therefore the mean capacities of those plants
summarized in Table 9-1. Since Table 9-1 does not distinguish bewteen the
three major sources of benzene, the mean benzene capacity of these units
has been derived from separate industry surveys.3,43
With regard to new vs. existing plant capacities, it has been assumed
that the mean capacities of new plants will equal the mean capacities of
those plants currently in operation. While economies of scale have in the
past and may in the future favor the construction of larger plants, this
assumption, in effect, yields conservative estimates of new plant annual
revenues.
The annual output of each plant has been estimated based on the assump-
tion that the model plants will operate at a rate equivalent to the total
capacity utilization rate of all similar plants within the industry. The
derivation of capacity utilization rates for each chemical is displayed in
Table 9-3 while capacity utilization rates for the individual benzene sources
have been derived from industry projections.27
9-45
-------
I .
I
I
Table 9-23. MODEL PLANT
ANNUAL REVENUES - MODEL PLANT A
{May 1979 Dollars}
Chemical Capacity Capaci tyb Output Pricec Total
{Gg/yr} Utilization (Kg/yr) (!ffiU Revenue
Benzene
{toluene dealkylation} 153a .64e 97,920,000 .37 $ 36,230,000
Ethyl benzene/Styrene 282d .72 203,000,000 .63 $127,890,000
Cumene 160 .77 123,200,000 .40 $ 49,280,000
Cyclohexane 120 .70 84,000,000 .46 $ 38,640,000
Benzenesulfonic Acid 4 .79 3,160,000 .68 $ 2,149,000
Resorcinol 16 . .86 13,760,000 3.09 $ 42,518,000
Maleic Anhydride 23 .56 12,880,000 .88 $ 11,334,000
Ethylene
(1 unit) 323 .70 226,100,000 .31 $ 70,091,000
a
Reference .43, pp. 2-6.
bDerived from Table 9-3.
cAs presented in Table 9-18.
dcapacity presented is for styrene.
e
Reference 27, p. 64.
9-46
-------
Table 9-24. MODEL PLANT
ANNUAL REVENUES - MODEL PLANT B
(May 1979 Dollars)
Chemical Capacity Capac i tyb Output P' c To ta 1
nce
(Gg/yr) Utilization (Kg/yr) ( $/ Kg) Revenue
Benzene 80a .83d
(extraction) 66,400,000 .37 $ 24,600,000
Chlorobenzene 46 .53 24,400,000 .30 $ 7,300,000
Linear Alkylbenzene 77 .78 60,000,000 .75 $ 45,000,000
Ethyl ene
(2- 3 units) 755 .70 528,500,000 .31 $163,800,000
a
Reference 43, pp. 2-6.
bDerived from Table 9-3.
cAs presented in Table 9-18.
d
Reference 27, p. 64.
Table 9-25. MODEL PLANT
ANNUAL REVENUES - MODEL PLANT C
(May 1979 Dollars)
Chemi ca 1 Capacity Capaci tyb Output Pricec To ta 1
(Gg/yr) Utilization (Kg/yr) ($/Kg) Revenue
Benzene .76e
(pyro 1ys is gas) 107a 81,320,000 .37 $ 30,100,000
Nitrobenzene 63d .61 38,430,000 .79 $ 30,400,000
Hydroquinone 5 .69 3,450,000. 3.39 $ 11,700,000
Ethylene
(4- 5 units) 1136 .70 795,200,000 .31 $246,500,000
[
I
II
aReference 3.
bDerived from Table 9-3, except for hydroquinone which is
capacity utilization of all plants in Table 9-3.
cAs presented in Table 9-18.
dCapacity presented is for Aniline.
eReference 27, p. 64.
based upon the mean
9-47
Ii
-------
The total annual revenue of each model plant has been estimated accord-
ing to the output of each plant (in kilograms) and the price per kilogram of
each chemical, as noted in Table-9-18. The total revenues calculated in
Tables 9-23 through 9-25 are those used as the base for the estimation of
maximum product price increases.
9.2.5.3 Capital Availability and Model Plant Investment. All of the
previously discussed regulatory alternatives require capital expenditures for
both monitoring instruments and control equipment. To allow the assessment
of the burden of these additional capital expenditures, upon those firms
considering investment in new plants, the capital control cost estimates for
each model plant have been compared to the total investment represented by
each model plant.
Estimates of total plant investments (including process units, construc-
tion and start-up costs, and working capital) were obtained through several
sou rces i nc 1 ud i ng: trade jou rna 1 summa ri es of const ruct i on.act i vity; i ndu st ry
representatives; and vendor/licensor descriptions of process and budgetary
economics.
In all cases the investment estimates obtained from the above noted
sources were for plant capacities other than the mean model plant capacities
noted in 9-23 through 9-25. For this reason, observed investments were
adjusted according to the power capacity rule:44
Cost of A - [Capacity AJ.7
Cost of B - Capacity B
In additi~n, all investment totals have been expressed in May 1979 dollars,
through adjustment acco'rding to the Chemical EngineeriM "Plant Cost Index".45
Table 9-26 presents a summary of the investment totals used in the
estimation of capital availability impacts. For each plant the total capital
investment represents the required investment in both plant and working
capital, where working capital requirements are estimated as 15 percent of
the investment in plant.44
9.2.6 ~1odel 'Plant Impact Analysis
9.2.6.1 l'riceJ_mpact~. With the e~ception of regulatory alternative V
(i.e., leakless emission control 1quipme~t) the full cost pricing policies
9-48
-------
:I~------- --
i
Table 9-26. TOTAL CAPITAL
INVESTMENT REQUIRED - NEW MODEL PLANTS
(Millions of May 1979 Dollars)
(Continued)
I
I Chemical Capacity Investment Investment In Total Capital
I (GQ/yr) In PlantC WorkinQ Capital Inves tment
Model Pl ant A
Benzene
(toluene dealkylation) 153 $ 14.6 $ 2.2 $ 16.8
Ethylbenzene/Styrene a 282 56.7 8.5 65.2
Cumene 160 21.2 3.2 24.4
Cyclohexane 120 9.9 1.5 11.4
Benzenesulfonic Acid 4 3.8 .6 4.4
Resorcinol 16 9.7 1.5 11.2
Maleic Anhydride 23 20.8 3. 1 23.9
Ethylene
(1 unit) 323 384.5 57.7 442.2
Model Plant B
Benzene
(extraction) 80 31.4 4.7. 36. 1
Chlorobenzene 46 11. 7 1.8 13.5
Linear Alkylbenzene 77 13.8 2.0 15.8
Ethylene
( 2 - 3 un i ts ) 775 709.6 106.4 816.0
9-49
-------
Table 9-26. TOTAL CAPITAL
. INVESTMENT REQUIRED - NEW MODEL PLANTS
(Millions of May 1979 Dollars)
(Concluded)
Chemical Capacity Investment Investment In Total Capital
(Gg/yr) In P1antC Working Capital Investment
Model Plant C
Benzene
(pyrolysis gasoline) 107 $ 6.8 $ 1.0 $ 7.8
Nitrobenzene/Aniline 63 32.5 4.9 37.4
Hydroq ui none 5 22.0 3.3 25.3
Ethylene
( 4- 5 un i ts) 1136 946.8 142.0 1,088.8
aCapacity in items of styrene, investment totals are for an integrated
ethylbenzene/styrene facility in which all ethyl benzene is consumed in
the manufacture of styrene.
bcapacity in terms of aniline, invest~ent totals are for an integrated nitro-
benzene/aniline facility in which all nitrobenzene is consumed in the manu-
facture of aniline.
COriginal plant 'investment observations were obtained from the following sources:
Benzene (toluene dealkylation) - Ref. 46 .
Benzene (extraction) - Ref. 52
Benzene (pyrolysis gasoline) - Ref. 55
Benzenesulfonic Acid - Ref. 49
Chlorobenzene - Ref. 53
Cumene - Ref. 47
Cyclohexane - Ref. 48
Ethylbenzene/Styrene - Ref. 56
Ethylene - Ref. 51
Hydroquinone - Ref. 57
Linear Alkylbenzene - Ref. 54
Maleic Anhydride - Ref. 50
Nitrobenzene/Aniline - Ref. 52
Resorcinol - Ref.49
9-50
-------
pursued by manufacturers of benzene and its derivatives will have minimal
impacts upon the prices of those chemicals. In the tables which follow the
maximum price changes resulting from the alternatives are summarized. Since
the alternatives impact derivative chemicals (e.g., styrene, cumene) as
well as chemicals used as inputs in the manufacture of derivatives (e.g.,
benzene, ethylene) the full cost pricing assumption requires that two forms
of price impacts. be distinguished, that is,
.
Price increases attributable to the impacts of the alternatives
upon individual chemicals and;
Price increases attributable to
tives upon individual chemicals
the pass-through of benzene and
derivatives.
both the impacts of the alterna-
as well as those resulting from
ethylene price increases to
.
I
: I
I
,
I
: I
I
,
I
[ I
i
I
I
Maximum price increases which can be attributed to the pass-through of
control costs to individual chemicals are summarized in Table 9-27. As
noted in Section 9.2.5.2, the percentage price increase (based upon May,
1979 prices) for each chemical under each regulatory alternative has been
estimated through the expression of net annualized costs (Table 8-11) as a
percentage of the appropriate model plant revenue as determined in Tables
9-23 through 9-25.
The relatively low price increases noted in Table 9-27 can be attri- .
buted to both low annual control costs in conjunction with relatively high
product recovery credits. However, under regulatory alternative V, control
costs seriously outweight the value of recovered product to the extent that
imposition of this alternative could, under full cost pricing, increase
the price of several chemicals by more than 5 percent. For all chemicals
included, the potential price increases for the products of new plants are
slightly lower than those for existing plants. This is so, since the net
annualized control costs for new plants are lower.
Price increases attributable to both impacts upon individual chemicals,
as well as the pass through of benzene and ethylene price increases to the
manufactures of derivatives, are referred to as "cumulative price increases".
The cumulative price increases are perhaps of greatest concern since they
9-51
-------
Table 9-27. PERCENTAGE PRICE INCREASES
(May 1979 Prices)
Existing Plants New Plants
Chemical Regulatory Alternative Regulatory Alternative
II III IV V II. III IV V
Model Plant A
Benzene (toluene dea1ky1ation) .01 .02 .04 .34 .01 .02 .03 .23
Ethy1benzene/Styrenea <.01 .01 .01 . 10 <.01 <.01 .01 .06
Cumene . .01 .02 .03 .25 .01 .01 .02 .17
.Cyc1ohexane .01 .02 .04 .32 .01 .02 .03 .21
Benzenesulfonic Acid/Phenolb .16 .34 .65 5.67 .14 .27 .56 3.84
Resorcinol .01 .02 .03 .29 .01 .01 .03 .19
Maleic Anhydride .03 .07 .12 1.08 .03 .05 .11 .73
Ethylene (1 unit) <.01 .01 .02 . 17 <.01 .01 .02 . 12
1.0
I Model Plant B
(J1
N
Benzene (extraction) ( . 02) .04 .13 1.48 ( .02) .02 .10 1.01
Chlorobenzene ( . 05) .13 .43 5.00 ( .07) .08 .35 3.39
Linear A1kylbenzene (.01) .02 .07 .81 ( .01) .01 .06 .55
Ethylene (2-3 units) «.01) .01 .02 .22 . «.01) <.01 .02 . 15
. Model Plant C
Benzene (pyrolysis gasoline) (.03) .05 . .17 2.02 ( . 04) .02 .13 1.37
Nitrobenzene/Ani1inec ( .03) .05 . 16 2.00 ( . 04) .02 .13 1.36
Hydroqui none (.09) .13 .43 5.21 (.11) .06 .34 3.53
Ethylene (4-5. units) ( < . 01) .01 .02 .25 (.01 ) <.01 .02 .17
aprice increase for Styrene
bprice increase for Phenol
cPrice increase for Aniline
-------
',-
,
, I
I;
I
represent the maximum price increases of benzene derivatives which may result
from the full cost pricing policies of manufacturers. Accordingly, cumula-
tive price increases of any benzene derivative can be traced to the costs to
control fugitive benzene emissions during the manufacture of both benzene as
well as the benzene derivative.
Cumulative price increases have been summarized in Table 9-28. They
have been determined through the addition of the appropriate percentage price
increases noted in Table 9-27. For example the projected. increase in cumene
prices under regulatory alternative III (.07%) was calculated through the
addition of the increase in benzene price from model plant C, alternative III
(.05%) and the increase in cumene price from model plant A, alternative III
(.02%).
The percentage increases noted in Table 9-28 are conservative for three
major reasons. Primarily, the addition of percentage increases implies that
benzene is the single input in the manufacture of benzene derivatives. In
reality, the contribution of inputs that will be unaffected by the alternatives.
will minimize the effect of benzene price increases upon the final prices
of benzene derivatives. Second, the highest increases in benzene prices
(generally that from Pyrolysis Gasoline - model plant C) were used in the
calculation of cumulative price increases. In reality, the manufacture
of derivatives through the use of benzene from model plants A and B (i.e.,
Toluene Dealkylation and Solvent Extraction) will entail the pass through
of lower price increases. This is especially true for regulatory alterna-
tives IV and V. Finally, the recovered product credits used in the deter-
mination of net annualized costs are based upon current market prices.
Recognizing the past and projected future trends in petroleum and petroleum
based product prices, it is quite possible that the value of product re-
covered under each regulatory alternative will, over the forecast period,
increase at a rate higher than the rate of increase in annualized control
costs, thus essentially reducing the net annualized cost of each alternative.
9.2.6.2 Capital Availability Impacts. Each of the previously discussed
regulatory alternatives requires capital expenditures for monitoring instruments
and other control equipment. The need for such equipment requires that poten-
tial investors in new plants must obtain additional capital financing above that
which would be required in the absence of regulation. In those cases where
,
I
: I
9-53
-------
.
I -~- -----._-
Tab 1e 9-28.
CUMULATIVE PERCENTAGE PRICE INCREASES
.(May 1979 Pri ces)
Chemical Regulatory Alternative
II III IV V
a .03 .07 .20 2.37
Ethyl benzene/Styrene
Cumene .02 .07 .20 2.27
Cyc10hexane .02 .07 .21 . 2.34
Benzenesu1fonic Acid/Pheno1b .17 .39 .82 . 7.69
Resorcinol .02 .07 .20 2.31
Maleic Anhydride .04 . 12 .29 3.10
Ch1orobenzene ( . 04) . 18 .60 7.02
I
! . Linear Alky1benzene .00 .07 .24 2.83
Nitrobenzene/Anilinec (.02) . 10 .33 4.02
Hydroquinone ( . 08) .18 .60 7.23
ap' increase for Styrene
. nce
bp' increase for Phenol
nce
cp' increase for Aniline
rl ce
9-54
-------
I-I
. I
I
I
the additional capital control costs represent a significant increase in total
capital requirements, potential investors could, as a result of difficulties
in obtaining additional financing, abandon plans for new plant construction.
In order to estimate the extent to which the imposition of the regula-
tory alternatives could cause significant increases in the capital require-
ments of new plants, the capital control costs of each regulatory alternative
have been expressed as a percentage of the total investment required of each
model plant. Estimates of model plant investment requirements have been made
through the sources and methodology described in Section 9.2.5.3, while the
capital control costs associated with each model plant and regulatory alter-
native are presented in Table 8-4.
Table 9-29 summarizes the additional percentages of invested capital
required of new model plants. With regard to the increased capital require-
ments of regulatory alternatives II, III and IV, these increases are generally
low and cannot be considered potential obstructions to new plant construction.
Regulatory alternative V (i.e., leak less emission control equipment) however,
entails much larger increases in required investments and its imposition
could preclude future plant construction. This is especially true in the
case of benzene production by way of pyrolysis gasoline. In this case the
capital control costs under regulatory alternative V, could increase the
total capital requirements for such units by 22.49 percent.
I!
9.3 MACROECONOMIC IMPACT
I
I
9.3.1 Summary
From the analysis detailed in Section 9.2, it can be concluded that regu-
latory alternatives II, III and IV, will have very little effect either on pro-
duct prices or the capital investment requirements of related plants. In addi-
tion, since market conditions, price inelastic demand, and low maximum potential
price increases will allow the full pass through of control costs~ it can be
conluded that the profitability or market positions of individual manufacturers
will not be altered. Regulatory alternative V, however, holds the possibility
of significantly increased prices (for the products of existing and new plants)
and increased capital investment requirements for new plants.
9.3.2 Inflationary Impacts
Since the maximum potential
III, and IV) for benzene and its
price increases (regulatory alternatives II,
derivatives are low (Table 9-28), price
I
I.
9-55
-------
i
I
,
Table 9-29. PERCENTAGE INCREASE IN
NEW PLANT CAPiTAL INVESTMENT REQUIRED
Chemical RegulatorY Alternative
II III IV V
Model Plant A
Benzene (toluene dea1kylation) .06 .12 .30 2.08
Ethyl benzene/Styrene .02 .04 .09 .54
Cumene .04 .08 .21 1.44
Cyclohexane .09 .18 .45 3.07
Ben~enesu1fonic Acid .23 .45 1. 16 7.96
Resorcinol .09 . 18 .46 3. 13
Maleic Anhydride .04 .08 .21 1.47
Ethylene (1 unit) <.01 <.01 .01 .08
Model Plant B
Benzene (extraction) .04 . 12 .38 2.91
Ch 1 0 roben zene . 10 .31 1.01 7.79
Linear Alkylbenzene .09 .27 .86 6.66
Ethylene (2-3 units) <.01 .01 .02 . 13
Model Plant C
Benzene (pyrolysis gasoline) .22 .86 2.85 22.49
Nitrobenzene/Aniline .05 . 18 .60 4.69
Hydroqui none .07 .27 .88 6.93
Ethylene (4-5 units) <.01 .01 .02 . 16
9-56
-------
,
,i
I
increases at the level of consumer products will be imperceptible. There-
fore, the application of these regulatory alternatives will not contribute
to inflation as measured by the Consumer Price Index.
9.3.3 Energy Impacts
As noted in Sectipn 7.5, each regulatory alternative requires passive
controls on equipment handling benzene streams. Under these conditions, no
increase in energy consumption is expected.
9.3.4 Employment Impacts
Full cost pricing, on the part of individual manufacturers, will insure
that the profitability of existing plants will not be affected by the stan-
dard. In addition, the very low increases in capital investment requirements
for new plants (regulatory alternatives II, III and IV) will not inhibit the
construction of new facilities. Under these conditions, the alternatives will
not result in either plant closures or reductions in output, and thus the
level of employment in the industry will be unaffected.
9.3.5 Fifth Year Annualized Costs
Annualized costs in the fifth year following promulgation (1985), have
been estimated for each regulatory alternative, and under no alternative do
such costs exceed the $100 million criterion specified in E.O. 12044. Fifth
year annualized cost totals have been estimated by summing the annualized
costs (Table 8-11) applicable to existing, new, and replacement plants
presented in Table 7-9. Fifth year annualized cost totals for regulatory
alternatives III, IV and V are $2.7, $7.3, and $74.1 million dollars, while
regulatory alternative II would, if implemented, result in annualized cost
reductions due to relatively high product recovery.
I
I
I
II
;
I
. !
I
I
i I
I
I
9-57
-------
9.4 REFERENCES
14.
1.
Soder, S.l. CEHProduct Review on Styrene. Chemical
Economics Handbook. Stanford Research Institute. Menlo
Park, CA. January 1977. 27 p.
2.
Blackford, J.l. CEH Product Review on 'Cyclohexane.
Economics Handbook. Stanford'Research Institute.
CA. February 1977. 29 p.
Chemical
Menlo Park,
3. Gunn, T.C., and K. Ring. CEH Marketing Research Report on
Benzene. Chemical Economics Handbook. Stanford Research
Institute. Menlo Park, CA. May 1977. 66 p.
4. Klapproth, E.M. CEH Product Review on Chlorobenzenes.
Chemical Economics Handbook. Stanford Research Institute.
Menlo Park, CA. ~uly 1977. 10 p. .
5.
Soder, S.l. et al.
Economics Handbook.
CA. January 1978.
CEH Product Review on Ethylene. Chemical
Stanford Research Institute. Menlo Park,
108 p.
6.
Cogswell, S.A. . CEH Product Review on Resorcinol.
Economics Handbook. Stanford Research Institute.
Park, CA. October 1978. 11 p.
Chemica 1
Menlo
7.
Klapproth, E.M. CEH Product Review on Aniline and Nitrobenzene.
Chemical Economics Handbook. Stanford Research Institute. Menlo
Park, CA. January 1979. 10 p.
Bradley, R.F. CEH Product Review on linear and Branched
Al kyl benzenes. Chemical Economics Handbook. Stanford
Research Institute. Menlo Park, CA. January 1979. 16 p.
8.
9.
Al-Sayyari, S.A., and K. Ring. CEH Product Review on
Cumene. Chemical Economics Handbook. Stanford Research
Institute. Menlo Park, CA. March 1979. 16 p.
10.
Ring, K., and S.A. Al-Sayyari. CEH Product Review on
Ethyl benzene. Chemical Economics Handbook. Stanford
Research Institute. Menlo Park, CA. March 1979. 14 p.
Hydrocarbon Processing, Section 2: World-Wide HPI Con-
struction Boxscore. June 1978. p. 4-17.
11.
12.
Hydrocarbon Processing, Section 2:
struction Boxscore. October 1978.
World-Wide HPI Con-
p. 3-18.
World-Wide HPI Con-
p. 3-16.
13.
Hydrocarbon Processing, Section 2:
struction Boxscore. February 1979.
Hydrocarbon Processing, Section 2: World-Wide HPI Con-
struction Boxscore. June 1979. p. 3-13.
9-58
-------
',- !
I I
I
, I
I
i
I
I
I 20.
i
I
'I
I:
"
I'
II
"
!,
21.
f
I
I
,
"
!
I
II
I
i
I
I
'.
15.
Wett, T.
Growth.
Ethylene Report - Ethylene Capacity Outruns Demand
Oil and Gas Journal. 12(36):59-64. September 1979.
16.
Chemical Economics Handbook of Current Indicators - Supplemental
Data. Stanford Research Institute. Menlo Park, CA. April 1979.
p. 225-226.
17.
Standifer, R.L. Emission Control Options for the Synthetic
Organic Chemicals Manufacturing Industry, Ethylene Product
Report. Hydroscience, Inc., Knoxville, TN. For U.S. Environ-
mental Protection Agency, Emission Standards and Engineering
Division. Research Triangle Park, NC. July 1978. 200 p.
Dylewski, S.W. Emission Control Options for the Synthetic
Organic Chemicals Manufacturing Industry, Ch1orobenzene
Product Report. Hydroscience, Inc., Knoxville, TN. For
U.S. Environmental Protection Agency, Emission Standards
and Engineering Division. Research Triangle Park, NC.
August 1978. 90 p.
18.
19.
Emission Control Options for the Synthetic Organic Chemical
Manufacturing Industry, Nitrobenzene Product Report. Hydro-
science, Inc., Knoxville, TN. For U.S. Environmental
Protection Agency, Emission Standards and Engineering
Division. Research Triangle Park, NC. January 1979. 65 p.
Industrial Process Profiles for Environmental Use: Chapter 6.
The Industrial Organic Chemicals Industry. Research Triangle
Institute. Research Triangle Park, NC. Radian Corporation.
Austin, TX. For U.S. Environmental Protection Agency. .
Cincinnati~ OH. Publication No. EPA-600j2-27-023f. February
1977. 1014 p.
Hobbs, F.D., and J.A. Key. Emission Control Options for the
Synthetic Organic Chemicals Manufacturing Industry, Ethy1benzene
and Styrene Product Report. Hydroscience, Inc., Knoxville, TN.
For U.S. Environmental Protection Agency, Emission Standards and
Engineering Division. Research Triangle Park, NC. May 1978.
73 p.
22.
Peterson, C.A. Emission Control Options for the Synthetic
Organic Chemicals Manufacturing Industry, Linear A1kylbenzene
Product Report. Hydroscience, Inc., Knoxville, TN. For U.S.
Environmental Protection Agency, Emission Standards and
Engineering Division. Research Triangle Park, NC. September
1978. 137 p.
23.
Blackburn, J.W. Emission Control Options for the Synthetic
Organic Chemicals Manufacturing Industry, Cyc10hexane Product
Report. Hydroscience, Inc., Knoxville, TN. For U.S. Environ-
mental Protection Agency, Emission Standards and Engineering
Division. Research Triangle Park, NC. May 1977. 78 p.
9-59
-------
,----
37.
38.
39.
. .' 4
24.
lawson, J. F. Emission Control Options for the Synthetic Organic
Chemicals Manufacturing Industry, Maleic Anhydride Product Report.
Hydroscience, Inc., Knoxville, TN. For U.S. Environmental Protection
Agency, Emission Standards and Engineering Division. Research Triangle
Park, NC. March 1978. 120 p.. .
25. letter from J. A. Pearson, Goodyear Tire and Rubber Co., to D. R.
Goodwin, EPA, ESED. July 9, 1979. With attached article: Olzinger,
A.. H. New Route to Hydroquinone. Chemical Engineering. June 9,1975.
p. 50-51.
26.
Synthetic Organic Chemicals, United States Prodution and Sales, 1977.
U.S. International Trade Commission. U.S. Government Printing Office.
Washington, DC. USITC Publication No. 920. 1978.
27.
U.S. Benzene Markets to Face Slower Growth.
85(3):62-64. January 30, 1978.
Chemical Engineering.
28.
Current Prices of Chemicals and Related Materials. Chemical Marketing
Reporter. (Issues from the first week of January, July and December,
1974-1979 and May 21, 1979.)
29.
Chemicals and Gasoline Compete for Aromatics.
April 18, 1979.
Chemical Week. 124(16):31.
30.
Ethylene Oversupply Could last Until 1980.
-News. ~(14):10. April 4, 1977.
Chemical and Engineering
31.
Aromatics Seen Entering Slow-Growth Era as Energy, Government Strictures
Hobble Trade. Chemical Marketing Reporter. 213(24):11. June 12, 1978.
32.
Chemical Engineering. 86(23):136. October 22,
CE ConstructiQn Alert.
1979.
33.
34.
Hydrocarbon Processing, Section 2:
October 1979. p. 13.
lurie, M. Oil and Chemicals: Era of Peateful Coexistence? Chemical
Week. 125(16):70-92. October 1979.
World-Wide HPI Construction Boxscore.
35.
Energy Resources Co., Inc. Economic Impact Analysis of Anticipated
Hazardous Waste Regulations on the Industrial Organic Chemicals, Pesti-
cides, and Explosives Industries. U.S. Environmental Protectio~ Agency.
Washington, DC. Final Report SW-158c. 1978. p. 85.
36.
Reference 35, p. 87.
Reference 35, p. 87-88.
~ef~rence 3, p. 35.
Refer to Section 9.1.8.1.
9-60
-------
1
{,
",
.:"
{
!
~. I
!
'I
'I
tl
~
I!
~ '
tl
"
,I,
i I
;.1
~ i
, ,
.!
,\
I
I
I
I
II
[ ,
I
'[
I
I
I
II
I
I,
40.
Blackford, J. L. CEH Product Review on Maleic Anhydride. Chemical
Economics Handbook. Stanford Research Institute. Menlo Park, CA.
July, 1976. 35 p.
Chemical Weeks.
J_23(22) :21-22.
41.
As Polyester Goes, 50 Goes Maleic.
November, 1978.
Is Benzene Losing to Gas Tanks? Chemical Week. J23(4):26. July,
1978.
42.
4.3.
Evaluation of Emissions from Benzene-Related Petroleum Processing
Operations, PEDCo Environmental, Inc., EPA Contract No. 68-02-2603.
October 1978, p. 2-6.
44.
Part 1, Research-Project Evaluations.
ber, 1976. p. 137.
Hydrocarbon Processing.
Decem-
45.
CE Cost Indexes Maintain 13-Year Ascent.
189. May 8, 1978.
Chemical Engineering.
.?5 (11 ) :
46.
Economics of the Petrochemical Industry.
1978. p. 32.
June
Hyplan Consulting Group.
47.
Chemical Engineering. 86(23):136.
October 22,
CE Construction Alert.
1979.
ail and Gas Journal. 64(44):46.
48.
October 31, 1966.
49.
64 ( 1 ) :
How Synthetic Phenol Processes Compare.
83-88. January 1966.
Oil and Gas Journal.
50.
Hydrocarbon Processing, Section 2:
score. ~(2):18. February 1977.
Ethylene: Makke It or Buy It? Chemical Engineering Progress.
17. December 1978.
World-Wide HPI Construction Box-
J4(12) :
51.
52. Worldwide Construction. Oil and Gas Journal. }5(41):127. October
1977.
53. Chemical Week. 9~(5) :23. August 1964.
54. Worldwide Construction. Oil and Gas Journal. .!?~(3) :79-81. January
1967.
55.
Hydrocarbon Processing.
1977 . p. 132.
1977 Petrochemical Handbook Issue.
November
56.
Chemical Engineering.
September
CE Const ruct i on Alert.
1974.
~(20): 105.
57.
CE Construction Alert.
.?JJ7) :123.
Chemical Engineering.
April 1970.
9-61
-------
APPENDIX C - EMISSION SOURCE TEST DATA
C .1
INTRODUCTION
A survey was conducted by EPA to collect data from refineries
throughout the United States on leak detection and repairs of equip-
ment emitting volatile organic compounds (VOC's) in excess of a defined
action level. This appendix summarizes the results of data from four
sources: One reference summarizing valve repair data for a U.S.
refinery; Phillips Petroleum Company, Sweeny, Texas; Shell Oil Company,
Martinez, California; and Union Oil Company of California, San Francisco
Refinery, Rodeo, California. For most data sources, the number of
pieces of equipment checked and the number of equipment having leaks
in excess of a defined action level (usually 10,000 ppm) were determined.
When available, repair data were collected so that the number of
equipment repairs and non-repairs could be summarized.
, I
i'
: i
C.2 DATA SUMMARIES
C.2.1 Refinery Valve Repair Data 1
Table C-1 presents the results of repair data for gas service
and light liquid valves from a U.S. petroleum refinery. The number of
valve repairs attempted and the number of valves actually repaired
below the action level of 10,000 ppmv are shown as well as the per-
centage reduction in leak rates (lb/hr). Repair data for leaks less
than 10,000 ppmv are also given in Table C-1, showing the number of
attempted valve repairs and the number of valves with increased emis-
sions after repairs.
II
i i
II
Maintenance for the valves was classified as
directed. Directed maintenance valves were those
until hydrocarbon emissions were below the action
di rected or un- .
that were tightened
level. Undirected
C-l
-------
!-
.
maintenance valves were tightened, but no emission measurements were
taken while the valves were being repaired. No infonmation was
available on the. testing method used, the number of valves checked
or the number of valves found leaking.
C.2.2 Phillips Petroleum Company Data 2
Equipment leak testing was perfonmed at various units in the
Phillips Petroleum Company Sweeny Refinery and Petrochemical Complex,
Sweeny, Texas, in March, 1979. All tests were conducted using the
Century Instrument Company's Organic Vapor Analyzer (Model OVA - 108)
with readings being recorded as the maximum concentration at the seal
interface. The number of pieces of equipment leaking at or above
10,000 ppm is shown in Table C-2 by equipment type and unit tested.
Table C-3 presents block valve repair data from the ethylene unit of
Phillips Sweeny Refinery, and Table C-4 summarizes the repair data.
C.2.3 Shell Oil Company Data 3
Data were obtained from the refinery valve emissions study at
Shell Oil Company's Martinez Manufacturing Complex, Martinez,
California. Over 9,000 valves were checked for hydrocarbon leaks.
An action level of 10,000 ppm or greater was used to define a valve
leak. A summary of leaking valves and repair status is presented
in Table C-5. No data were available for detection and repair of
leaks from other refinery equipment or for testing methods used in
the study. .
C.2.4 .Union Oil Company, San Francisco Refinery Data 4
Leak detection.and repair data are presented in Tables C-6
through C-8 for refinery valves at Union Oil Company's San Francisco
facility in Rodeo, California. All emission measurements were taken
1 cm froo the valves, using a VOC detector (Model OVA - 108). Valves
C-2
-------
~- -~--------_._----
leaking at or above 3,000 and 10,000 ppm by volume hydrocarbon \"ere
identified, and repairs were attempted. Table C-6 summarizes the
number of valves "successfully repaired" that resulted in emissions
below each action level and the number of valves with increased
emissions after repair. In cases where enissions were reduced,
packing adjustments were made on the valves. Table C-7 presents
the number of valves before repair at different action levels and
the number of valves having increased emissions after repair.
The effects on emissions of repairing valves in the 1,000 through
10,000 ppm range are shown in Table C-8. Emissions from refinery
valves within 12 different units and lines are represented. Total
emissions (lb/hr) fran valves after repair increased 5 percent.
I
I
I!
I
I
I
I
I
II
II
I'
C-3
-------
Table C-l.
REFiNERY VALVE REPAIR DATA
Action Level: ~10,000 ppm
Distance from Source: 0 cm
Instrument: Bacharach "TLV Sniffer"
I. Hydrocarbon Leaks Greater Than or Equal to 10,000 ppmv
Number of Percent of
Valve Numbe-r of Emission
Repairs Valves Repaired Reduction
Source Type Attempted (to <10,000 ppm) (Leak Rate)
Gas Service Valves
Undirected
Maintenancea 8 7 53
Directed b 7 6 92
Maintenance
Light Liquid Valves
Undirected
Maintenancea 4 2 79
Directed
Maintenanceb . 4 2 92
aValves repaired after undirected maintenance were tightened, but
no emission measurements were taken during repair.
bValves repaired after directed maintenance were tightened until
emissions were below 10,000 ppm~
C-4
-------
--_---.J_------
Table C-1.
REFINERY VALVE REPAIR DATA (Concluded)
Action Level: ~10,000 ppm
Distance from Source: 0 cm
Instrument: Bacharach "TLV Sniffer"
II. Hydrocarbon Leaks Less Than 10,000 ppmv
Number of Percent of
Valve Number of ValVes Emission
Repairs With Increased Rate
Source Type Attempted Emissions Rate Reduction
Gas Service Valves
Undirected
Maintenancea 2 1 66
Directed b 3 a 54
Maintenance
Light Liquid Valves
Undirected
Maintenancea 7 1 56
I
I
Directed
Maintenanceb 10 2 76
aValves repaired after undirected maintenance were tightened, but
no emission measurements were taken ~uring repair.
bValves repaired after directed maintenance were tightened until
emissions were below 10,000 ppm.
C-5
-------
Table C-2. LEAK DATA FOR THE PHILLIPS PETROLEUM COMPANY, SWEENY
REFINERY AND PETROCHEMICAL COMPLEX, SWEENY, TEXAS
Action Level: ~10,000 ppm
Instrument: nOVA-108n VOC detector
Distance from Source: Maximum concentration at seal interface
Percent of
Number of Number of Equipment
Equipment. Leaking Type
1. Equipment Type Checked Equipment Leaking
Valves. 2,564 222 8.7
Pumps 190 41 21.6
Compressor Seals 33 1 3.0
Drains 150 9 6.0
Control Valves 68 13 19.1
Open-Ended Lines 420 39 9.3
TOTALS 3,425 325
Continued ....
C-6
-------
Table C-2. LEAK DATA FOR THE PHILLIPS PETROLEUM COMPANY, SWEENY
REFINERY AND PETROCHEMICAL COMPLEX, SWEENY, TEXAS (Concluded)
Action Level: ~10,OOO ppm
Instrument: IOVA-108" VOC detector
Distance from Source: Maximum concentration at seal interface
II. Units Tested
Number of Number of Percent
Unit Unit Equipment Leaking Leaking of
Number Name Checked Equipment Unit Tested
4 FCCU Gas 297 18 6.1
Concentration
9 Crude 443 8 1.8
Distillation
lOB NGL Manufacturing 178 13 7.3
11 High End Point 847 61 7.2
Reformer
12 Ethylene 1,096 170 15.5
Manufacturing
15 Hexane 564 55 9.8
Isomerization
TOTALS 3,425 325
C-7
-------
Table C-3. PHILLIPS SWEENY REFINERY ETHYLENE UNIT BLOCK VALVE REPAIRS
Action Level: ~10,000 ppm -Instrument: "OVA-108" VOC detector
Distance from Source: Maximum concentration at seal -interface
Undirected Directed Maintenance Readings
Tag Initial Date Maintenance Maintenance
Number Readinga Screened Attempted - Readi ng 1 2 3 Comments
32 ~10,000 03/06/79 No ~10,000 1,100 Only checked one valve
with tag -... meter lines
~10,000 03/06/79 No ?10,000 2'10,000 ~10,000 Only checked one valve
with tag
~10,000 03/06/79 No >10,000 ~10,OOO ?10,000 Only checked one valve
with tag
~10,000 03/06/79 Yes 2,000 100
I n 28 ~10,000 03/06/79 Yes 2,000
I
I OJ
I 16 ~10,000 03/06/79 Yes ~1O,000 ?10,000 ~10,000 700 Repaired when valve was
backseated
10 ~10,000 03/06/79 Yes 100
7 ~10,000 03/06/79 Yes ~10,000 ~10,000 Bolts all the way down
4 ~10,000 03/06/79 Ye~ 200
367 ~10,QOO 03/05/79 Yes 500
366 ~10,000 03/05/79 No Bolts need replacing
364 ~10, 000 03/05/79 Yes NCb
362 ~10,000 03/05/79 Yes ~10,000 ?10,OOO Leak at gland, not stem --
corrosion preventing good
seating of gland
-------
Table C-3.
PHILLIPS SWEENY REFINERY ETHYLENE UNIT BLOCK VALVE REPAIRS (Continued) -
Undirected Directed Maintenance Readings
Tag Initial Date Haintenance Maintenance
Number Readinga Screened Attempted Reading 1 2 3 Comments
360 >10,000 03/05/79 Yes 2,000
359 ~10,000 03/05/79 Yes 4,000
None . ~1O,000 03/05/79 No »10,000 »10,000 Mistagged originally so no
initial repair attempted --
tightened bolts -- needs
new packing
358 ~10,000 03/05/79 Yes NCb
(j
I 361 ~1O, 000 03/05/79 Yes ?10,000 ~10,000 Leak reduced but needs new
~
pack i ng
None ~1O,000 03/05/79 No ~10,000 ~10,000 Near No. 361 - needs new
packing
356 ~10,000 03/05/79 Yes NCb Was not leaking before
maintenance (mistagged)
354 >10,000 03/05/79 Yes 900
352 ?1O, 000 03/05/79 Yes NCb
65 ~10,000 03/06/79 Yes 3,000
64 ~1O, 000 03/06/79 Yes 1,000
(gland) ~10,000 ?10 ,000 7,000 Leak detected by soap
solution - missed by
instrument operator
-------
Table C-3. PHILLIPS SWEENY REFINERY ETHYLENE UNIT BLOCK VALVE REPAIRS (Concluded)
Undirected Directed Maintenance Readings
Tag Initial Date Maintenance Ma i ntena n.ce
Number Readinga Screened .Attempted Reading 1 2 3 Comments
315 ~10,000 03/06/79 Yes 3,000
311 NCb 03/06/79 Yes NCb Drai n sti 11 ~10,000
316 ~10,000 03/06/79 Yes ~10,000 2,000
313 ~10,000 03/06/79 Yes. ~10,OOO ~10,000 All the way down on
packing
312 ~10,000 03/06/79 Yes ~1O , 000 ~10,000 5,000 All the way down on
n packing
I
-'
0 314 ~10,000 03/06/79 No ~10,000 Bad bolts - need
replacing
aAll readings are in parts per million by volume calibrat~d to hexane using OVA-lOB detector.
. bNC = No change detected in reading above ambient level.
-------
Table C-4.
SUMMARY OF PHILLIPS SWEENY BLOCK VALVE LEAK AND REPAIR DATA
Action Level: ~ 10,000 ppm
Instrument: /lOVA-lOB/I VOG detector
Distance from Source: Maximum concentration at seal interface
'-
Leaking valvesa
Attempted repairs
29
27
(Percent of leaking valves
with attempted repairs)
Valves reparied after b
undirected maintenance
(93%)
15
(Percent of attempted repairs
successful after undirected
maintenance)
Valves repaired after
directed maintenancec
(56 %)
5
(Percent of attempted repairs
successful after'directed
maintenance)
Valves not repaired
(to < 10,000 ppm)
(42%)
9
(Percent of leaking valves pot
repaired)
(31%)
aLeaking valves are valve stems found emitting volatile organic
compounds (VOGs) at or aQove 10,000 ppmv. '
bValves repaired through undirected maintenance were tightened
until VOG emissions were below 10,OPO ppmv.
cValves reparied through qirected maintenance were tightened, but
no emission measurements were taken during repair.
C-11
-------
Table C-5. LEAK AND REPAIR DATA FOR REFINERY VALVES FROM THE SHELL.
OIL COMPANY, MARTINEZ MANUFACTURING COMPLEX, MARTINEZ, CALIFORNIAa
Equipment: Valves
Action Level: ~ 10,000 ppm
Instrument: "OVA-108" VOC detector
Distance from Source: 1 cm
Valves checked
Valves leaking
{Percent of valves checked)
In-service valve repairs
attempted
Successful repairsb
(Percent of repairs attempted)
Unsuccessful repairsc
(Percent of repairs attempted)
9,277
293
(,3% t.
230
199
(87%)
31
(13% ).
aTw6 sets of emission measurements taken before and after
repair during March and April 1979. .
bSuccessful repairs resulted in valves leaking below 10,000 ppm.
cUnsuccessful repairs are valves still leaking at or above
10,000 ppm after repair.
C-12
-------
Table C-6. LEAK AND REPAIR DATA FOR REFINERY VALVES FROM THE
UNION OIL COMPANY SAN FRANCISCO REFINERY, RODEO, CALIFORNIA
Equipment: Valves
Total Checked: 5,815
Instrument: IIOVA-10811
Distance from Source:
VOC detector
1 em
Action Level (ppm)
~3,000
~10,000
Leaking valves 300 215
In-service repairs attempted 158 125
Successful repairs* 107 74
(Percent of repairs attempted
that were successful)' (68%) (59%)
Valves with increased emissions
after repair 17 7
(Percent of repairs attempted
with increased emissions) (11%) (6%)
*Successful repairs resulted in valves leaking below 10,000 ppm.
C-13
-------
Table C-7. ATTEMPTED REPAIR DATA FOR VALVES
FROM THE UNION-SAN FRANCISCO REFINERY
Instrument: 10VA-lOB" VOC detector
Distance from Source: 1 cm
I .
Action level
(ppm)
Total Number
of Valves
Before Repair
T ota 1 Number.
. of Valves
With Increased
Emissions
After Repair
Percent of
Valves With
Increased
Emissions
o - 999 0 0
1,000 - 9,999 33 10 30
~10,000 125 7 6
C-14
-------
Tab 1 e C-B. EFFECTS ON EMISSIONS OF REPAIRING VALVES IN THE
1,000 - 10,000 PPM RANGE
(Union-San Francisco Refinery, 4/10/79 Data)
Instrument: IOVA-10B" VOC detector
Distance from Source: 1 cm
BEFORE REPAIR AFTER REPAIR
ppm 1 b/hr ppm 1 b/hr
5,000 0.039 10,000 0.062
3,000 0.027 300 0.005
5,000 0.039 100 0.003
5,000 0.039 1,000 0.013
4,000 0.033 7,000 0.049
8,000 0.053 1,000 0.013
4,000 0.033 400 0.007
5,000 0.039 100,000 0.297
4,000 0.033 1,500 0.017
1,000 0.013 1,000 0.013
7,000 0.049 4,000 0.033
9,000 0.058 4,500 0.036
5,000 0.039 400 0.007
, I 4,000 0.033 2,000 0.021
'I
3,000 0.027 2,000 0.021
2,000 0.021 400 0.007
5,000 0.039 1,000 0.013
9,000 0.058 700 0.010
8,000 0.054 10,000 0.062
3,000 0.027 30,000 0.131
i 4,000 0.033 10,000 0.062
I
5,000 0.039 10,000 0.062
I
II
1
C-15
'.1
-------
Table C-8. EFFECTS ON EMISSIONS OF REPAIRING VALVES IN THE
1,000 - 10,000 PPM RANGE
(Union-San Francisco Refinery, 4/10/79 Data)
(Concluded)
BEFORE REPAIR AFTER REPAIR
ppm 1 b/hr ppm 1 b/hr
5,000 0.039 8,000 0.053
7,000 0.049 1,500 0.017
8,000 0.053 2,000 0.021
4,000 0.033 1,500 0.017
6,000 0.044 10,000 0.062
4,000 0.033 9,000 0.058
4,000 0.033 1,500 0.017
3,000 0.027 100 0.003
TOTALS 1.136 1.192 (+5%)
C-16
-------
r')
II
I
I!
.1
, i
Ii
....
C.3 REFERENCES
1.
"Emi ss ion Factors and Frequency of Leak Occurrence for
Fittings in Refinery Process Units," EPA Report Number
600/2-79-044, February, 1979.
2.
Radian Corporation. "The Ass.essm,ent of Enyi,ronrnental
Emissions from Oi,l : Refi:ni;n9, Appendi,x B - De1;a.i:led
Results, II Ora.ft Fi.nal 'Report, , EPA, Contract Number
68~02~2147. August, 1979.
3.
Equ i pment Summa ry from Ph ill i ps Petrol eum Company,
Sweeny, Texas, March 14, 1979.
Valve Repair Summary and Memo from R. M. Thompson,
Shell Oil Company, Martinez Manufacturing Complex,
Martinez, California, to Milton Feldstein, Bay Area
Qual i ty Management Di strict, April 26, 1979.
4.
5.
Valve Repair Summary and Memo from F. R. Bottomley,
Union Oil Company, Rodeo, California, to Milton Feldstein,
Bay Area Quality Management District, April 10, 1979.
C-17
-------
I]
I
II
I'
II
APPENDIX D - EMISSION MEASUREMENT AND CONTINUOUS MONITORING
D.l
EMISSION MEASUREMENT METHODS
"
I
To develop data in support of standards' for the control of fugitive
emissions. EPA conducted leak surveys at six petroleum refineries and
three syntheti c organi c chemi ca 1 manufacturtng pl ants'. The resulti ng
leak determination procedures contained in Reference Method 113 were
developed during the course of this test program.
Prior to the first test. available methods for measurement of
fugitive leaks were reviewed. with emphasis on methods that would
provide data on emission rates from each source. 10 measure emission
rates. each individual piece of equipment must be enclosed in a temporary
cover for emission containment. After containment. the leak rate can be
determined using concentration change and flow measurements. This
procedure has been used in several studies.O.2) and has been demonstrated
to be a feasible method for research purposes. It was not selected for
this study because direct measurement of emission rates from leaks is a
time-consuming and expensive procedure. and is not feasible or practical
for routine testing.
Procedures that yield qualitative or semi-quantitative ,indications
of leak rates were then reviewed. There are essentially two alternatives:
leak detection by spraying each component leak source wi'th a soap solution
and observing whether or not bubbles were formed; and. the use of a
portable analyzer to survey for the presence of increased organic compound
concentration in the vicinity of a leak source. Visual. audible. or
olefactory inspections are too subjective to be used as indicators of
leakage in these applications. The use of a portable analyzer was
selected as a basis for the method because it would have been difficult
to establish a leak definition based on bubble formation rates. Also.
the temperature of the component. physical configuration. and relative
movement of parts often interfere with bubble formation.
II
II
,
, ,
I
I
I
[I
!
II
,I
I
I
I
II
L. ~
0-1
-------
Once the basic detection principle was selected, it was then necessary
to define the procedures for use of the portable analyzer. Prior to
performance of the first field test, a procedure was reported that
conducted surveys at a distance of 5 cm from the components. (3) This
information'was used to formulate the test plan for initial testi.ng. (4)
In addition, measurements were made at distances of 25 cm and 40 cm on
three perpendicular lines around individual sources. Of the three
distances, the most repeatable indicator of th.e presence of a leak was a
measurement at 5 cm, with a leak definition concentration of 100 or 1000
ppmv. The localized meteorologfcal condtttons. affe.cted dispersion
significantly at greater distances. Also, tt was more difficult to
define a leak at greater distances because. of the. small changes from
ambient concentrations.06serYed~ Surveys- were conducted at 5 cm from
the source during the next tfiree factltty tests..
The procedure was distributed for comment in a draft control techniques
guideline documents. Many commentors felt that a measurement distance
of 5 cm could not be accurately repeated during screening tests. Since.
the concentrati on profi 1 e ts rapidly cfuingtng between 0 and about 10 cm
from the source, a small variance from 5 cm could stgniftcantly effect
the concentration measurement. In response to these. comments, the
\
procedures were changed so that measurements were made at the surface of
the interface, or essentially 0 cm. This cfi.ange requtred that the leak
definition level 6e increased. Addtti.onal test tog at two refineries and
three chemi ca 1 pl ants was performed by measuring vol ati.l e organi.c concen-
trations at the tnterface surface.
A compHcation that thts change 'introduces ts that a. very small
mass emission rate leak C'p'in-hole leaR"l can D.e totally' captured by the
instrument and a high concentration result will Be obtatned. This has
occurred occasionally in EPA tests and a solution to thts problem tYas
not been found.
The calibration basis for the analyzer was evaluated. It was
recognized that there are a number of potential vapor stream components
and compositions that can be expected. Stnceall analyzer ty~es do not
respond equally to diff~rent compounds, it was necessary to establish a
reference calibration material. Based on the expected compounds and the
D-2
-------
,I
I!
i
I
I;
I
limited information available on instrument response factors~ hexane was
chosen as the reference calibration gas for EPA test programs. At the 5
cm measurement distance~ calibrations were conducted at approximately
100 or 1000 ppmv levels. After the measurement distance was changed~
calibrations at lO~OOO ppmv levels were required. Commentors pointed
out that,hexane standards at this concentration were not readily available
commercially. Consequently~ modifications were incorporated in the
method to allow a1:t~rnate .standard pr.epar~tion procedures or alternate ..
calibration gases.
The alternative of specifying a different ca1i.bration material for
each type stream and normal i zatton factors for each. i.ns.trument type was
not intensively investigated. There are at least four instrument types
available that might be used in this procedure~ and there are a large
number of potenti a 1 stream composi'tions possi'b:l e.. The amount of pri or
knowledge necessary to develop and subsequently use such factors would
make the interpretation of results prohibitively complicated. Based on
EPA test results~ the number of concentration measurements in the range
where a variability of 2 or 3 would change thedecision as to whether or
not a leak exists is small in comparison to the total number of potential
1 eak sources.
An alternative approach to leak detectfon was evaluated by EPA
during field testing. The approach used was an area survey~ or walk-
through~ using a portable analyzer. The unit area was surveyed by
walking through the unit positioning the instrument probe within 1 meter
of all valves and pumps. The concentratfon readings were recorded on a
portable strip chart recorder. After completfon of the walkthrough~ the
local wind conditions were used with the chart da,ta to locate the
approximate source of any i'ncreas'ed ambient concentrations. Thi s pro-
cedure was found to yield mfxed results. In some. cases~ th.e majorHy of
leaks located by individual component testing could be located by walk-
through surveys. In other tests, prevailfng dispersi'on conditions and
lQcal elevated ambient concentrations compli:cated or prevented the
interpretation of the res'ults. AddHionally, tt was not possible to
develop a .genera 1 criteria speci:fytng how' much. of an am5.tent i.ncrease
0-3
-------
. at a distance of 1 meter is indicative of a 10000 ppm concentration at
the leak source. Because of the potential variability in results from
site to site, routine walkthrough surveys were not selected as a reference
or alternate test procedure.'
0.2 CONTINUOUS MONITORING SYSTEMS AND DEVICES
Since the leak determination procedure is not a typical emission
measurement technique, there are no continuous monitor~ng approaches
that are directly applicaBle. Continual survei.llance is achieved by
repeated monftorfng or screening of all affect~d potential leak sQurces.
A continuous monitoring system or device could serve a~ an indicator
that a 1 eak has developed fiebreen inspectton .tnterva 1s.. EPA perf o ",e.d a
limited evaluation of. fixed~point monitoring systems for tneir effecti.ve~
ness in 1 eaR detectfon. Tf\e systems consist.ed of 5otf\. remote sensing
devices with a central readout and a central analyzer system .(gas cnromatograph)
with remotely collected samples. The results of these tests indtcated that
fixed point 'systems were not capable of sensing a~l leaRs'tfjat were found
by individual component testi~g. Tfifs is to fi.e'ex~ecte:d since these systems
are significantly affected DY- local dispersfon conditions and would require
either many individual point 10cattons, or very low detection sensitiviti.es
in order to achieve similar results to those obtatned using an tndividual
component survey.
. .
It is recommended tfiat ftxed-point monitoring systems not. be required'
since general speciftcations cannot De fonnulated to assure equivalent
results, and eacli installation would ftave to tie. evaluated tndtviduall.y-.
0.3 PERFORMANCE TEST METHOD
The recommended fugittve benzene emission detection procedure is
Method 113. This method:fnco~poratesthe~se of a portable analyzer to
detect the presence of volatile organic vapors at the ~urfac~ of the.
interface where direct leaRageto tne atmosphere could occur.. Tile
genera 1 approach. of this technique assumes that if an organic 1 eaR
exists, there will De an increased vapor concentration tn the vicinity
of the leak, and that the measured concentration t5. generall,Y' proportional
to.the m~ss emission rate of theorgantc co~pound.
D-4
-------
The test procedure does not detect benzene specifically, instead,
the volatile organic compound conientration is measured. There is
commercially available one type of portable analyzer that has the
capability of measuring benzene by chromatographic techniques. However,
the addition of the requirement that benzene be measured specifically
would require more time and more extensive testing support. Measurement
of benzene would not yield additional information since the affected
facilities are those in which benzehe is transported and a measure of
organic vapor leakage is indicative of a benzene leak.
Method 113 does not include a specification of th~ instrument
calibration basis or a definition of a leak in terms of concentration.
Based on the results of EPA field tests, hexane is recommended as the
reference calfbration basis for reffnery and SOCMl fugitive benzene
emission sources.
There are at least four types of detection principles currently
available in commercial porta51e instruments. These are flame ioniza-
tion, catalytic oxidation, fnfrared aDs'orpHon (NDI:RI a,nd photoioniza-
tion. Two types (flame ionization and cata1yttc oxidaHonI are known to
be ava 11 ab1 e fn factory mutual certiffed vers tons for use in hazardous
atmospheres.
The recommended test procedure fnc1udes a set of design and operating
speci fi cati ons and evaluation procedures 5y' whtch, an analyzer's performance
can be evaluated. These parameters we,re se,lected Qas'ed on the allowable
tolerances for data collection, and not on the perform~nce of individual
instruments. Based on manufacturers' 1 i'terature speci ficati ons, many
commercially available ana1yze~s ca~ meet these requirements.
The estimated purchase cost for an analyzer ranges from about $lQOO
to $5000 depending on the'type and optional equipment. The cost of an
annua 1 monitodng program per unft, inc 1 udi;ng semtannua 1 instrument
tests and reporting is' estimated to 5e from $3,000 to $4,500. This
estimate is based on EPA contractor costs' experienced during previous
test programs. Performance of monHorfng by p1 ant personnel may result
in lower costs. The above estimates do not include any costs associ.ated
with leak repair after detection.
0-5
-------
0.4 REFERENCES
1. "JOINT DISTRICT, FEDERAL, and STATE PROJECT for the Evaluation
of Refinery Emissions", ,Los Angeles County Air Pollution Control District,
Nine Reports, 1957-1958.
2. "Emission Factors and Frequency of Leak Occurrence Fittings in
Refinery Process Units" Radia~ Corporation Contract No. 68-02-2147 and
No. 68-02-2665, EPA Report No. 600/2-79-044, February 1979.
3. ,Telephone CO"lI1lUnic~tton: Paul Harrison, Meteorology Research,
Inc. to K. C. Hustvedt, EPA~ December 22, 1911.
, 4~ EMB Report No. 77-CAT-6~ "Miscellaneous Refinery Equi~ent VOC
Sources at Arco, Watson Refinery, and Newnal1 R~fining Company" ESED,
EPA, December, 1979.
5. "Control ()f Volatile Organic Compound le~ks' from petrol eum
Refi nery Equipment," OAQPS GuideHne Series, EPA-450t2..,.]8-036, June, 1978.,
j\
. ,.
0-6
------- |