United States Officer Air.Quality EPA-450/3-80-032a
Environmental Protection Planning and Standards November 1980
Agency Research Triangle Park NC 27711
Air
Benzene Fugitive Draft
Emissions EIS
Background Information
for Proposed Standards
-------
EPA-450/3-80-032a
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
November 1980
U.S. Invifomnental Protection Agency
Region 5, Library (PL-12J)
77 West Jackson Boulevard, 12th Plow
Chicago, II 60604-3590
-------
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-450/3-80-032a
U,S. Environmental Protection Agency
-------
ENVIRONMENTAL PROTECTION AGENCY
Background Information
and Draft
Environmental Impact Statement
for Benzene Fugitive Emissions
Prepared by:
(Date)
UUII r^. uuuuuwiii .
Director, Emission Standards and Engineering Division
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
1 The proposed national emission standard would limit fugitive
' emissions of benzene from existing and new petroleum refining and
chemical manufacturing units. The proposed standard implements
Section 112 of the Clean Air Act and is based on the Administrator s
determination of June 8, 1977, (42 FR 29332) that benzene presents
a significant risk to human health as a result of air emissions
from one or more stationary source categories, and is therefore a
hazardous air pollutant.
2 Copies of this document have been sent to the following: Federal
' Departments of Labor, Health and Human Services, Defense,
Transportation, Agriculture, Commerce, Interior, and Energy; the
National Science Foundation; and Council on Environmental Quality,
members of the State and Territorial Air Pollution Program Adminis-
trators; the Association of Local Air Pollution Control Officials;
EPA Regional Administrators; and to other interested parties.
3. The comment period for review of this document is 75 days and is
expected to begin on or about December 15, 1980.
4. For additional information contact:
Ms. Susan R. Hyatt
Standards Development Branch (MD-13)
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
telephone: (919) 541-5477
4. Copies of this document may be obtained from:
U.S. EPA Library (MD-35)
Research Triangle Park, NC 27711
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
111
-------
TABLE OF CONTENTS
Page
List of Tables xi
List of Figures . ., xvii
Chapter 1. Summary 1-1
1.1 Statutory Authority 1-1
1.2 Regulatory Alternatives 1-1
1.3 Environmental Impact 1-2
1.3.1 Air Quality Impact 1-4
1.3.2 Water, Solid Waste, and Energy Impacts for
New and Existing Sources 1-4
1.4 Economic Impact 1-5
1.4.1 Existing Sources 1-5
1.4.2 New Sources 1-5
Chapter 2. Introduction 2-1
Chapter 3. Sources of Benzene Fugitive Emissions in
Petroleum Refining and Organic Chemical
Operations . 3-1
3.1 Introduction 3-1
3.2 Sources of Benzene Emissions 3-1
3.2.1 Potential Leak Sources 3-2
3.2.2 Other Potential Sources 3-8
3.3 Magnitude of Benzene Emissions from Refining
and Organic Chemical Production Operations 3-12
3.4 References 3-15
Chapter 4. Emission Control Techniques 4-1
-------
TABLE OF CONTENTS (Continued)
Page
4.1 Introduction .................... 4-1
4.2 Leak Detection and Repair Programs ........ 4-2
4.2.1 Definition of a Leak ............ 4-4
4.2.2 Inspection Interval ............. 4-4
4.2.3 Allowable Repair Time ............ 4-6
4.2.4 Visual Inspections ............. 4-6
4.2.5 Other Leak Detection Techniques ....... 4-6
4.2.6 Repair .................... 4~8
4.2.7 Emission Control Effectiveness of
Leak Detection and Repair .......... 4-10
4.3 Preventive Programs ................ 4-11
4.3.1 Pumps .................... 4"n
4.3.2 Valves ................... 4'17
4.3.3 Safety /Relief Valves ............ 4-19
4.3.4 Open-Ended Valves .............. 4-22
4.3.5 Closed-Loop Sampling ............ 4-24
4.3.6 Accumulator Vessel Vents and Seal
Oil Degassing System Vents ......... 4-26
4.4 Process Modifications ............... 4-26
4.5 References ..................... 4"28
Chapter 5. Modification and Reconstruction ......... 5-1
5.1 General Discussion of Modification and
Reconstruction Provisions ............. 5-1
5.1.1 Modification
5.1.2 Reconstruction
-------
TABLE OF CONTENTS (Continued)
Page
5.2 Applicability of Modification and
Reconstruction Provisions 5-2
5.2.1 Modification 5-2
5.2.2 Reconstruction 5-3
Chapter 6. Model Units and Regulatory Alternatives 6-1
6.1 Introduction 6-1
6.2 Model Unit Parameters 6-1
6.3 Regulatory Alternatives 6-3
6.3.1 Regulatory Alternative I 6-3
6.3.2 Regulatory Alternative II 6-6
6.3.3 Regulatory Alternative III 6-7
6.3.4 Regulatory Alternative IV 6-7
6.3.5 Regulatory Alternative V 6-8
6.3.6 Regulatory Alternative VI 6-8
6.4 References 6-9
Chapter 7. Environmental Impact 7-1
7.1 Introduction 7-1
7.2 Air Quality Impacts 7-1
7.2.1 Development of Benzene Emission Levels . . . 7-1
7.2.2 Future Benzene Emissions 7-9
7.3 Water Pollution Impact 7-16
7.4 Solid Waste Impact 7-20
7.5 Energy Impact 7-21
vi
-------
TABLE OF CONTENTS (Continued)
Page
7.6 Other Environmental Concerns ............ 7~
7.6.1 Irreversible and Irretrievable
Commitment of Resources ........... /~^i
7.6.2 Environmental Impact of Delayed
Regulatory Action ..............
7-24
7.7 References .....................
O 1
Chapter 8. Cost of Controls .................
Q_1
8.1 Introduction ....................
8.2 Capital Cost Estimates ............... 8~1
8.3 Annualized Cost Estimates ............. 8"7
8.3.1 Derivation of Annualized Cost Estimates ... 8-7
8.3.2 Cost-Effectiveness ............. 8~21
8.4 Total Industry Impacts ............... 8"37
o 07
8.4.1 Existing Units ............ . . . .
8.4.2 New Units .................. 8"37
8.5 Cost Comparison ..................
8-45
8.6 References .....................
9-1
Chapter 9. Economic Impact .................
9.1 Industry Characterization ............. 9-1
9.1.1 General Profile ............... 9"1
9.1.2 Production of Benzene, Ethylene,
and Benzene Derivatives ........... y~i
9.1.3 Methods of Manufacture ........... 9~11
9.1.4 Uses of Benzene ............... 9"15
9.1.5 Price History ................ 9"25
vii
-------
TABLE OF CONTENTS (Continued)
Page
9.1.6 Market Factors that Affect the
Benzene Industry 9-25
9.1.7 Feedstock Substitutions for
Benzene Derivatives 9-25
9.1.8 Future Trends 9-27
9.2 Microeconomic Impact 9-30
9.2.1 Introduction 9-30
9.2.2 Industry Structure 9-30
9.2.3 Demand Characteristics 9-34
9.2.4 Supply Characteristics 9-39
9.2.5 Economic Impact Methodology 9-44
9.2.6 Model Unit Impact Analysis 9-48
9.3 Macroeconomic Impact 9-55
9.3.1 Summary 9-55
9.3.2 Inflationary Impacts 9-55
9.3.3 Energy Impacts 9-57
9.3.4 Employment Impacts 9-57
9.3.5 Fifth Year Annualized Costs 9-57
9.4 References 9-58
Appendix A. Evolution of the Background Information Document . A-l
Appendix B. Index to Environmental Considerations B-l
Appendix C. Emission Source Test Data C-l
C.I Introduction C-2
VI 1 1
-------
TABLE OF CONTENTS (Continued)
Page
C-2
C.2 Data Summaries ...................
C.2.1 Refinery Valve Maintenance Data ....... c~2
C.2. 2 Phillips Petroleum Company Data ....... c-5
C.2. 3 Shell Oil Company Data ........... c~5
C.2. 4 Union Oil Company, San Francisco
Refinery Data ................ L'5
C.2. 5 Benzene-Producing Units E and F ....... C-14
C.2. 6 Ethyl ene and Cumene Unit Data ........ c-19
C.2. 7 Exxon Chemical Lompany Data ......... c~19
C 2.8 Benzene - producing unit at Refinery "E",
Gulf Coast, U.S ................ u-iy
C.2. 9 Benzene - producing unit at the Amoco Texas
Refining Company, Texas City, Texas
C-24
C.3 References .....................
Appendix D. Emission Measurement and Continuous Monitoring . . D-l
D.I Emission Measurement Methods ............ D"
D.2 Continuous Monitoring Systems and Devices ..... D-5
D.3 Performance Test Method ..............
D-8
D.4 References ..................
Appendix E. Methodology for Estimating Leukemia Mortality
and Maximum Lifetime Risk from Exposure to Benzene
Fugitive Emissions from Petroleum Refineries and
Organic Chemical Plants .................
E.I Introduction ....................
E.2 Summary and Overview of Health Effects .......
IX
-------
TABLE OF CONTENTS (Concluded)
Page
E.2.3 Health Effects at Environmental Exposure
Levels,
E-4
E.3 Population Density Around Petroleum Refineries
and Organic Chemical Plants E-5
E.4 Population Exposures, Mortalities, and Risks .... E-6
E.4.1 Summary of Methodology for Calculating
Deaths E-6
E.4.2 Estimates of Leukemia Deaths E-7
E.4.3 Example of Leukemia Death Calculation .... E-9
E.4.4 Estimate of Leukemia Risk E-ll
E.4.5 Validity of Estimates E-14
E.5 References E-26
Appendix F. Estimates of Benzene Emissions and Control Cost of
Product Accumulator Vessels F-1
F.I Introduction F~2
F.2 Estimate of Benzene Emissions F-2
F.2.1 Uncontrolled Estimates F-2
F.2.2 Controlled Estimates F-2
F.2.3 Nationwide Estimates F-4
F.3 Control Cost Estimates . F~4
-------
LIST OF TABLES
Page
Metric Conversion Table ................. Xv
1.1 Environmental and Economic Impacts of Regulatory
Alternatives .......................
3-1 Estimated Emission Factors for Nonmethane Hydrocarbons
from Refining and Organic Chemical Industry Sources ... J-lJ
3-2 Estimated Benzene Emissions from an Average Plant . . . . 3-14
4-1 Percentage of Sources Predicted to be Leaking
in an Individual Component Survey ............ *-*
4-2 Percent of Total Mass Emissions Affected at
Various Action Levels ..................
4-3 Emission Correction Factors for Various Inspection
Intervals, Allowable Repair Times and Action Levels . . . 4-.
6-1 Model Unit Equipment Containing Greater Than
10 Percent Benzene ..................... b"^
6-2 Monitoring Intervals and Equipment Specifications
for Benzene Fugitive Regulatory Alternatives
7-1 Controlled VOC Emission Factors for Regulatory
Alternative II
7-2 Controlled VOC Emission Factors for Regulatory
Alternative III ..................... 7"3
7-3 Controlled VOC Emission Factors for Regulatory
Alternative IV ..................... '"4
7-4 Controlled VOC Emission Factors for Regulatory
Alternative V ...................... 7"b
7-5 Calculation of Weighted Percent Benzene for Emission
Sources in Model Units ............ 7~'
7-6 Benzene Emissions (Kg/hr) by Source for the
Regulatory Alternatives - Model Unit A .......... /-i
7-7 Benzene Emissions (Kg/hr) by Source for the
Regulatory Alternatives - Model Unit B
-------
LIST OF TABLES (Continued)
Page
7-8 Benzene Emissions (Kg/hr) by Source for the
Regulatory Alternatives - Model Unit C 7-12
7-9 Total National Benzene Emissions from Refining and
Organic Chemical Processes in 1980 7-13
7-10 Numbers of Units Estimated to Meet 1980 Demand
for Benzene and Benzene Derivatives by Model Units . . . 7-15
7-11 Cumulative Annual Number of Projected New Units and
Replacements between 1981 and 1990 7-17
7-12 Cumulative Annual Estimated Benzene Fugitive Emissions
From New Units and Replacements Between 1981 and 1990 . . 7-19
7-13 Energy Impact of Benzene Emission Reduction
for Regulatory Alternatives 7-22
8-1 Model Unit Equipment Containing>10 Percent Benzene . . . 8-2
8-2 Monitoring Intervals and Equipment Specifications
For Benzene Fugitive Regulatory Alternatives 8-3
8-3 Capital Cost Data 8-5
8-4 Capital Cost Estimates per Model Unit 8-8
8-5 Monitoring and Maintenance Labor-Hour Requirements
for Regulatory Alternative II 8-16
8-6 Monitoring and Maintenance Labor-Hour Requirements
for Regulatory Alternative III 8-17
8-7 Monitoring and Maintenance Labor-Hour Requirements
for Regulatory Alternative IV 8-18
8-8 Monitoring and Maintenance Labor-Hour Requirements
for Regulatory Alternative V 8-19
8-9 Recovered Product Credits 8-20
8-10 Initial Survey Start-Up Costs for
Regulatory Alternatives II-IV 8-22
8-11 Annualized Control Cost Estimates per Model Unit .... 8-26
XII
-------
LIST OF TABLES (Continued)
Page
8-12 Benzene Emission Reductions 8-34
8-13 Cost-Effectiveness for Existing Model Units 8-35
8-14 Cost-Effectiveness for New Model Units 8-36
8-15 Nationwide Costs for the Existing Industry 8-38
8-16 Nationwide Costs for New Units (Fifth Year Impact). . . . 8-39
8-17 Range of Control Costs for the Benzene Source
Categories for Existing and New Units 8-41
8-18 Costs for the Control of Total Benzene Emissions
from the Maelic Anhydride Industry 8-42
8-19 Costs for the Control of Total Benzene Emissions
from the Ethylbenzene-Styrene Industry 8-42
8-20 Total Costs for the Control of Benzene Emissions
from Producer Benzene Storage Tanks and Benzene
Fugitive Sources 8-44
8-21 Total Costs for the Control of Benzene Emissions
from Consumer Benzene Storage Tanks and Benzene
Fugitive Sources 8-44
9-1 Refineries and Organic Chemical Manufacturing
Sites with Benzene Fugitive Emission Potential 9-2
9-2 Number of Companies and Plant Sites that
Manufacture Benzene Derivatives 9-11
9-3 Summary of Production and Capacity for
Benzene, Ethylene, and Benzene Derivatives 9-12
9-4 Ethylene Usage 9-18
9-5 Monochlorobenzene Usage 9-19
9-6 Dichlorobenzenes Usage 9-19
9-7 Nitrobenzene Usage 9-20
9-8 Aniline Usage 9-20
XI 11
-------
LIST OF TABLES (Continued)
Page
f
9-9 Ethyl benzene Usage 9-21
9-10 Styrene Usage 9-21
9-11 Linear Alkybenzene Usage 9-22
9-12 Cyclohexane Usage 9-22
9-13 Cumene Usage 9-23
9-14 Maleic Anhydride Usage 9-23
9-15 Resorcinol Usage 9-24
9-16 Benzenesulfonic Acid Usage 9-24
9-17 Hydroquinone Usage 9-24
9-18 Price History for Benzene, Ethylene, and
Benzene Derivatives 9-26
9-19 Alternative Processes for the Manufacture of
Benzene Derivatives 9-27
9-20 Projected Annual Growth Rates for Demand of
Benzene, Ethylene, and Benzene Derivatives 9-28
9-21 Concentration Ratios for Benzene, Ethylene, and
Benzene Derivatives 9-32
9-22 Qualitative Evaluation of Price Elasticity
of Demand 9-40
9-23 Model Unit Annual Revenues - Model Unit A 9-46
9-24 Model Unit Annual Revenues - Model Unit B 9-47
9-25 Model Unit Annual Revenues - Model Unit C 9-47
9-26 Total Capital Investment Required -
New Model Units 9-49
9-27 Percentage Price Increases 9-52
9-28 Cumulative Percentage Price Increases 9-54
xiv
-------
LIST OF TABLES
C-8 Effects on Emissions of Repairing Valves in the
1,000 - 10,000 ppm Range
C-9 Frequency of Leaks from Fugitive Emission Sources in
Organic Chemical Units E and F
Page
9-29 Percentage Increase in New Plant Capital
Investment Required
C-l Refinery Valve Maintenance Data c'3
C-2 Leak Data for the Phillips Petroleum Company, Sweeny
Refinery and Natural Gas Liquids Processing 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-16
C-10 Screening Data For Cumene Units ............. c-20
C-ll Screening Data For Ethylene Units ............ °-21
C-12 Screening Data For Cyclohexane Unit at Exxon Chemical
Company, Baytown, Texas .................
E-l Estimated Leukemia Deaths from Benzene Fugitive
Emissions from Petroleum Refineries and Organic
Chemical Plants Under Current Control Conditions ..... t-ib
E-2 Example Calculation of Leukemia Deaths, Plant 4 ..... E-25
xv
-------
LIST OF TABLES (Concluded)
Page
F-l Calculation of Uncontrolled Benzene Emissions
from Accumulator Vessels by Model Unit ......... F-3
F-2 National Benzene Emissions from Accumulator Vessels
in 1980 for Regulatory Alternative I .......... F-5
F-3 Effect of Adding Accumulator Vessel Emissions
on Baseline Risk .................... F-6
F-4 Capital and Annual i zed Cost for Vent Systems for
Accumulator Vessels by Model Unit ........... F-7
F-5 National Capital and Annual Costs for Controlling
Existing Accumulator Vessels ............. F-8
F-6 Nationwide Costs for the Existing Industry for
Regulatory Alternatives III, IV, and V ........ F-9
xvi
-------
LIST OF FIGURES
Page
3-3
3-1 Simple Packed Seal
3-2 Basic Single Mechanical Seal 3"4
3-3 Globe Valve with Packed Seal 3'6
3-4 Diagram of a Spring-Loaded Relief Valve 3~7
3-5 Liquid-Film Compressor Shaft Seal 3'9
4 14
4-1 Double Mechanical Seal
4-2 Seal less Canned Motor Pump
4-18
4-3 Diaphragm Valve
4-20
4-4 Sealed Bellows Valve
4-5 Rupture Disk Installation Upstream of a Safety/Relief ^^
Valve
4-6 Simplified Closed-Vent System with Dual Flares 4-23
4-7 Diagram of Two Closed-Loop Sampling Systems 4-25
9-1 Percentages of Total Benzene Production Consumed
by Intermediate and Final Products
xvii
-------
METRIC CONVERSION TABLE
EPA policy is to express all measurements in agency documents in
metric units. Listed below are metric units used in this report with
conversion factors to obtain equivalent English units. A list of
prefixes to metric units is also presented.
To Convert
Metric Unit
centimeter (cm)
meter (m)
liter (1)
2
cubic meter (m )
cubic meter (m )
kilogram (kg)
megagram (Mg)
gigagram (Gg)
gigagram (Gg)
joule (J)
gigajoule (GJ)
Degree Celsius (°C)
Cubic meters/second
(m3/sec)
Multiply By
Conversion Factor
0.39
3.28
0.26
264.2
6.29
2.2
1.1
2.2
1102
9.48 x 10"4
9.48 x 105
(°C x 1.8) + 32
4.40
To Obtain
English Unit
inch (in.)
feet (ft.)
U.S. gallon (gal)
U.S. gallon (gal)
barrel (oil) (bbl)
pound (Ib)
ton
million pounds (10 Ibs)
ton
British thermal unit (Btu)
British thermal unit (Btu)
Degree Fahrenheit (°F)
gallons/minute
(gal/min)
Prefix
tera
giga
mega
kilo
centi
milli
micro
PREFIXES
Symbol
T
G
M
k
c
m
Multiplication
Factor
10
12
10
10
10
3
-2
r3
,-6
XVTM
-------
1.0 SUMMARY
1.1 STATUTORY AUTHORITY
National emission standards for hazardous air pollutants are
established in accordance with Section 112(b)(l)(B) of the Clean Air
Act (42 U.S.C. 7412), as amended. Emission standards under Section 112
apply to new and existing sources of a substance that has been listed
as a hazardous air pollutant. This study examines fugitive emission
sources in benzene service (containing 10 or more percent by weight
benzene) in petroleum refining and organic chemical manufacturing
industries.
1.2 REGULATORY ALTERNATIVES
Six regulatory alternatives were developed by employing various
combinations of the available control techniques in the affected
industry. Reflecting increasing levels of emission reduction, these
alternatives range from requiring no new controls to eliminating all
benzene fugitive emissions.
Regulatory Alternative I represents a baseline emission level
that describes the industry in the absence of regulations for the
control of benzene fugitive emissions. This alternative provides the
basis for incremental comparison of the other regulatory alternatives.
Regulatory Alternative II would reflect emission controls equivalent
to the recommendations suggested by the Control Techniques Guideline
(CTG) document for refinery volatile organic compound emissions.
Periodic leak detection and repair would be required for most sources
as well as the installation of specified equipment for other sources.
This alternative would result in a 57 percent reduction in benzene
fugitive emissions from the baseline level.
1-1
-------
Pumps and valves in liquid service would be monitored annually for
leaks. Safety/relief valves, pipeline valves, open-ended valves, and
compressors in gas service would be monitored for leaks on a quarterly
basis. Open-ended valves would be sealed with a cap, blind, plug, or
a second valve.
Regulatory Alternative III would result in a 73 percent reduction
in emissions by increasing the frequency of leak detection and repair
and by requiring additional equipment. Monthly leak detection and
repair would be required for pumps, valves, and compressors. Open-ended
valves would be sealed with a cap, blind, plug, or a second valve.
Rupture disks would be installed on safety/relief valves, or as an
alternative, they would be vented to a flare.
Regulatory Alternative IV would increase control efficiency to
77 percent by requiring additional specified equipment. In addition
to the leak detection and repair and equipment requirements of
Alternative III, Regulatory Alternative IV would require the use of
mechanical seal systems on pumps and compressors in benzene service.
In addition, degassing vents on pump seal oil reservoirs would be
required to be vented to a closed system.
Regulatory Alternative V would effect a 90 percent control
efficiency by requiring the installation of mechanical seals on pumps
and compressors, the installation of diaphragm or sealed bellows
valves on valves, and the sealing of open-ended valves. Closed-purge
sampling systems would be installed and rupture disks would be installed
on safety/relief valves.
Regulatory Alternative VI would result in a 100 percent control
efficiency for benzene fugitive emissions by prohibiting all production
and use of benzene in the affected industry.
1.3 ENVIRONMENTAL IMPACT
Included in the evaluation of environmental impacts were estimates
of air quality, water, noise, and solid waste impacts. Table 1-1
summarizes the environmental impact assessments for each regulatory
alternative.
1-2
-------
Table 1-1. ENVIRONMENTAL AND ECONOMIC IMPACTS OF REGULATORY ALTERNATIVES
I
CO
Alternative
I (no
II
III
IV
V
VI
.
_ '
Air
Impact
action) 0
+2**
+2**
+3**
+3**
+4**
__
Water
Impact
0
+1**
+1**
+1**
+1**
+1**
Solid Waste
Impact
0
0
0
0
0
0
Energy
Impact
0
+1**
+1**
+1**
+1**
_2**
Noise
Impact
0
0
0
0
0
0
Economic
Impact
0
+1**
0
_!**
_2**
_4**
. .
Key: + Beneficial Impact
- Adverse Impact
0 No impact
1 Negligible Impact
2 Small Impact
3 Moderate Impact
4 Large Impact
* Short-Term Impact
** Long-Term Impact
*** Irreversible
-------
1.3.1 Air Quality Impact
1.3.1.1 Existing sources. For the existing industry including
about 240 production units, total nationwide benzene fugitive emissions
are estimated to be 8,300 megagrams per year (Mg/Yr) for the baseline
alternative. Regulatory Alternative II would reduce these emissions
57 percent, from 8,300 Mg/Yr to 3,600 Mg/Yr. Alternative III would
reduce emissions to 2,200 Mg/yr, yielding a 73 percent reduction.
Alternative IV would yield a 77 percent reduction in benzene fugitive
emissions to a level of 1,900 Mg/yr. Alternative V would reduce
emissions to 900 Mg/yr, yielding a 90 percent reduction in emissions
from the baseline.
1.3.1.2 New sources. For new sources through 1985 including
68 production units, total nationwide benzene fugitive emissions are
estimated to be about 2,500 Mg/yr for Regulatory Alternative I (baseline).
Regulatory Alternative II would reduce these emissions to about
1,100 Mg/yr, a 56 percent reduction through 1985. Alternative III
would reduce emissions from about 2,500 Mg/yr to about 700 Mg/yr, a
72 percent reduction through 1985. Alternative IV would reduce emissions
to about 500 Mg/yr, resulting in an 80 percent reduction through 1985.
Alternative V would reduce emissions to about 200 Mg/yr, yielding a
92 percent reduction from the baseline alternative through 1985.
1.3.2 Water. Solid Uaste. and Energy Impacts for New and Existing
Sources
Since none of these alternatives would require any additional
water discharges, there would be no negative impact on water quality.
There is potential for a positive benefit to water quality, however,
due to decreased amounts of organic materials entering drains, sewers,
and waste water discharges because of better leak control. This
benefit would increase with the stringency of the alternative because
each successive alternative requires additional leak control measures.
There would be no significant solid waste or noise impact as a
result of implementing any of the regulatory alternatives. Additionally,
since the controls required to implement the alternatives are passive
in nature, there would be no significant negative energy impact. In
fact, there would be a slight energy benefit from the conservation of
raw materials and products that results from the control of leaks.
1-4
-------
1.4 ECONOMIC IMPACT
Industry total capital and annualized costs, including recovery
credits, were estimated for Regulatory Alternatives II, III, IV,
and V. These estimates for new and existing sources are based on
second quarter 1979 dollars. Table 1-1 summarizes the economic impacts
that result from these costs for each of the regulatory alternatives.
1.4.1 Existing Sources
Regulatory Alternative II would require a total capital investment
of $2.9 million and would result in an annualized savings of $25 thousand.
Alternative III would require a capital investment of $9.7 million and
an annualized cost of $2.1 million. Alternative IV would require a
capital investment of $25.3 million and an annualized cost of $5.5 mnllion.
Alternative V would require a capital investment of $242 million and
an annualized cost of $58.6 million. It should be noted that these
costs are for the entire industry and apply to 241 production units.
The annualized costs required to implement Alternatives II, III,
IV, and V could cause the average prices of benzene derivatives to
rise by 0.04, 0.13, 0.37, and 4.1 percent, respectively, if full
pass-through of the costs is assumed.
1.4.2 New Sources
The costs of implementing the alternatives are lower for new
sources than for existing ones because no retrofitting expenses are
involved. Regulatory Alternative II would require a fifth-year capital
investment of $820 thousand and would result in an annualized savings
of $70 thousand in the fifth year after implementing this alternative.
Alternative III would require a fifth-year capital investment of
$2.2 million and an annualized cost of $420 thousand after five years.
Alternative IV would require a fifth-year capital investment of
$6.5 million and an annualized cost of $1.3 million after five years.
Alternative V would require a capital investment of $48.4 million and
an annualized cost of $11.4 million after five years. These cost
estimates apply to 68 new production units built in the fifth year
after implementing each alternative.
1-5
-------
The annualized costs of implementing Regulatory Alternatives II
through V for new sources (assuming full cost pass-through) could
cause the average price of benzene derivatives to rise from
0.03 percent for Alternative II to 3.3 percent for Alternative V.
1-6
-------
2. INTRODUCTION
2.1 BACKGROUND
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
control 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 airborne 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 112 under this proposed
policy is the determination of which categories of sources
emitting the pollutants will be regulated, and in what order
2-1
-------
regulations wi'l be developed. Although a pollutant may
have been listed because emissions from a particular source
category pose a 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 initia-
tive 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 indicate 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 development of these
regulations, through publicly 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)
on the basis of: (1) magnitude of projected 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
2-2
-------
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 development 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 existing 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 control the
pollutant of concern (technology transfer) but which have
been 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 will include tor
each option: the number of plant closures predicted and the
direct impact on employment and end product prices; the
2-3
-------
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 environ-
mental 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 adequately demonstrated, considering economic,
energy, and environmental impacts.
The control level designated "best available technology"
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 III. The requirement of "best available
control technology" 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 economic 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 minimum, to
limit their emissions to the levels corresponding to the use
of "best available technology." This requirement is based
on the 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 be unreasonable, further
2-4
-------
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 alterna-
tives 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 risks 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 risks it 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
REFINING AND ORGANIC CHEMICAL OPERATIONS
3.1 INTRODUCTION
Valves, pumps, flanges, and other pieces of equipment are used
extensively in the refining and organic chemical industries 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 emission 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. The 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, agitators, 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, product accumulator vessels,
safety/relief valves, and process unit turnarounds.
3.2.1 Potential Leak Sources
3.2.1.1 Pumps. Pumps are used extensively in the refining and
organic chemical industries for the movement of organic fluids.
Centrifugal pumps are used most often in these industries, although
positive-displacement pumps, reciprocating and rotary action pumps,
and the specialized canned, diaphragm 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 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 retain 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 pump
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 another liquid that is
injected must 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
element 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's 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 seal faces and can be emitted to atmosphere.
3-2
-------
Pump stuffing box
CO
I
\
\
\
\
V. \
xxtxxx
/
X
r"
X
s
Fluid
end
xPacking gland
_X_\_A_X 4Seal face
*~-
Y\><
r\
X
1
X
/
/
X
XK
*
>
^
X
X
X
X
X
X
^~ Possible leak
area
Packing
Figure 3-1. Simple Packed Seal
1
-------
Gland gasket
co
i
Pump stuffing box
Rotating
seal ring
.Gland ring
.Insert packing
Stationary
/'element
~- Possible
leak
area
Figure 3-2. Basic Single Mechanical Seal
1
-------
3.2.1.2 Pipeline Valves. One of the most common pieces of
equipment in a refinery or an organic chemical production unit is the
valve. The types of valves commonly used are globe, gate, plug, ball,
relief, and check valves. All except the safety/relief valve and
check valve are activated 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," 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
sources of benzene emissions.
Two primary causes of leakage are seal deformation, due to thermal
stress on the adjoining piping or equipment, and opening of the flange
without replacement of the gasket.
3.2.1.5 Compressors. Compressors, like pumps, can be both
centrifugal and positive displacement types. Compressors have a shaft
3-5
-------
HANDWHEEL
STEM
PACKING NUT
DISK
BODY
PACKING
BONNET
SEAT
Figure 3-3. Globe Valve with Packed Seal
3-6
-------
SEAT
SPRPWG
DISK
MOZ.Z.LE
PROCESS SIDE.
1
Figure 3-4. Diagram of a Spring-Loaded Safety/Relief Valve'
3-7
-------
that requires a seal to isolate the compressor interior gas from
atmosphere. 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.
3.2.1.6 Agitators. Agitators are commonly used to stir or
blend chemicals in organic chemical processes. Like pumps and compressors,
agitators may leak organic chemicals at the point where the agitator
shaft penetrates the vessel. Consequently, seals are required to
minimize leakage of process materials from agitators. However, in the
benzene operations that are known to utilize agitators, the agitated
vessels operate at atmospheric pressure, so there would be no leakage
at the seal. For this reason, agitators are not considered to be
a significant source of benzene fugitive emissions.
3.2.2 Other Potential Sources
3.2.2.1 Process Drains. The operation of refinery and organic
chemical process units entails draining condensate water and flushing
water from process equipment. These drains also receive liquid leak-
age, spills, and water used to cool pump glands. Because most of
these drains are open to atmosphere, benzene in the wastewater can be
emitted to atmosphere. However, if leakage and spills are minimized,
benzene emissions from drains are expected to be slight.
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.
3-8
-------
INNER
BUSHING
INTERNAL
GAS
PRESSURE
OIL IN FROM RESERVOIR
OUTER
BUSHING
CONTAMINATED
OIL OUT
TO RESERVOIR
OIL OUT
ATMOSPHERE
Figure 3-5. Liquid-Film Compressor Shaft Seal
3-9
-------
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. Contaminated 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 wastewater undergoes additional
treatment as required. If it is present, benzene will be emitted
wherever wastewater is exposed to atmosphere due to evaporation of
benzene 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
vessels to separate reaction products. Accumulator vessels can be
vented directly to atmosphere or indirectly to atmosphere through a
blowdown 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 be used for
these processes:
Steam ejectors with contact condensers.
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 condensable 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
3-10
-------
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
come 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 atmosphere
from the recirculating water that in turn is used to remove heat from
such process equipment as reactors, condensers, and heat exchangers.
If a leak in the process equipment occurs and if the equipment is
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.
3-11
-------
3.2.2.9 Safety/Relief Valve Discharges. Safety/relief valves
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
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 REFINING AND ORGANIC CHEMICAL
PRODUCTION OPERATIONS
Data are limited on the measurement of benzene fugitive emissions
from sources in the refining and organic chemical industries. However,
recent testing efforts have generated a great deal of information on
VOC emissions from refining operations. Refinery benzene fugitive
emissions are assumed to be similar to the refinery VOC emissions 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 benzene fugitive
emissions in the organic chemical industry originates from equipment
handling benzene and benzene-containing organic streams, the emission
factors developed from the refinery data should apply to organic
chemical industry sources as well. Table 3-1 presents VOC emission
factors for refining and organic chemical industry sources. Benzene
emissions can be related to these VOC emission factors if it is assumed
that the weight percent benzene in the gaseous emissions from a leaking
piece of equipment is equivalent 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 alkylbenzene 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 EMISSION FACTORS FOR NONMETHANE
HYDROCARBONS FROM REFINING AND ORGANIC CHEMICAL INDUSTRY SOURCES
Emission Factor Estimate
Source Type (Kg/hr - sources)
Pumps
Pipeline Valves
a. Gas/Vapor Streams °'°21e
b. Liquid Streams °-010
Safety/Relief Valvesb °-16
cl 0.00026
Flanges
0 44
Compressors
Process Drains c
Sampling Connections °-015 c
Open-Ended Valves °-0032
Wastewater Separators ^
Vacuum-Producing Systems NA
Cooling Towers Negligible
Process Unit Turnarounds NA
Product Accumulator Vessel Vents 1-23
Safety/Relief Valve Discharges NA
NA - No factor available.
aFrom Reference 4 except where otherwise noted.
bGas Service only.
cThis factor was derived by the following equation:
(Purge\ sampling
Loss,]* 8^onnections From
See I 1000 Bbl/hour \KCT. o
Ref. 5/ refinery throughput \ >
+ 0.0032 Kg/hour for one open-ended valve (Seat Leakage, see Ref. 7) = 0.015 Kg/hour
dFrom Reference 8
Emission factor for refinery equipment in light liquid service.
3-13
-------
Table 3-2. ESTIMATED BENZENE EMISSIONS FROM AN AVERAGE PLANT'
Equipment Type9
Number of
Pieces Handling
10 or More
Weight Percent
Benzene
Uncontrolled
Benzenee
Emission
(kg/hr)
Pumps
Pipeline Valves
Gas
Liquid
Safety/Relief Valves
Open-Ended Valves
Gas
Liquid
s+
Sample Connections
Flanges
Drains
Totals
15
91
168
9
9
96
26
600
15
1,003d
1.80
1.91
1.68
1.44
0.22
1.27
0.39
0.16
0.48
9.35
Wastewater separators, vacuum-producing systems, process unit
turnarounds, cooling towers and safety/relief valve discharges
have been excluded due to lack of emission factors.
Uncontrolled benzene emissions include emissions for pipeline valves and
open-ended valves: gas = 0.0242 kg/hr per open-ended valve and
liquid = 0.0132 kg/hr per open-ended valve.
cNumber of sample connections is 25 percent of the number of open-
ended valves.
Total equipment excludes sample connections since they are included
in the total number of open-ended valves.
is 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. II-3, II-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 al. Screening Study for Miscellaneous
Sources of Hyd^carbon Emissions in Petroleum Refineries.
Monsanto Research Corporation. Dayton, OH. For U.b.
Environmental Protection Agency. Research Triangle Park, NC.
Report No. EPA-450/3-76-041. December 1976.
4 VJetherold, 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-600/2-79-044. 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-450/3-76-039. August 1976. 80 p.
6. Powell, D., et al. Development of Petroleum Refinery Plot
Plans Pacific Environmental Services, Incorporated. Santa
Monica, CA. For U.S. Environmental Action Agency.
Research Triangle Park, NC. Report No. EPA-450/3-78-025.
June 1978. 180 p.
7. Wetherold, R.G., et al.. Assessment of Atmospheric Emissions
from Petroleum Refining: Volume 3. Appendix B. Radian
Corooration. Austin, TX. For U.S. Environmental Protection
AgeScy? Industrial Environmental Research Laboratory Research
Triangle Park, North Carolina. (Final) Report No. EPA-600/
2-80-075C. April 1980. p. 266.
8 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-450/3-79-022. October 1978. p. 53
3-15
-------
4.0 EMISSION CONTROL TECHNIQUES
4.1 INTRODUCTION
As identified in Chapter 3, there are several categories of
potential sources of benzene fugitive emissions in refining and organic
chemical operations. These sources include: (1) the cumulative total
of small continuous leaking emission sources caused by seal leakage in
pumps, valves, flanges, safety/relief valves, agitators, and compressors,
(2) continuous emissions from the operation of vacuum-producing systems,
drains, wastewater separators, and cooling towers, and (3) intermittent
emissions from the operation of safety/relief valves, product accumulator
vessel vents, sampling connections, open-ended valves, and process
unit turnarounds.
Three basic control techniques can be applied to reduce benzene
fugitive emissions from these potential sources. These techniques
are as follows:
Leak detection and repair 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.
Process modifications that reduce or eliminate benzene fugitive
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.
Four of the sources described in the previous chapter are not
included in this discussion of emission control techniques wastewater
4-1
-------
separators, cooling towers, process unit turnarounds, and agitators.
No reliable data on emission rates and control techniques are available
for cooling towers and process unit turnarounds, so these sources have
not been included in this chapter. Wastewater separators are also not
included because emission information for controlled and uncontrolled
operations is not available. Agitators are not considered to be a
significant source of benzene fugitive emissions. As stated in
Chapter 3, agitated vessels in benzene operations operate at atmospheric
pressure; consequently, .no leakage is expected at the seal. These
sources may be addressed in the future.
4.2 LEAK DETECTION AND REPAIR PROGRAMS
The types of equipment that have the potential to be benzene
fugitive emission sources (i.e., pumps, valves, etc.) have been identi-
fied and discussed in Chapter 3. When such a piece of equipment
develops a leak, the leak can be detected by using a portable VOC
detector (performance criteria for the instrument and a description of
the leak testing methods are given in Appendix D). When the leak has
been located, it can be repaired through repair 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 leak detection and repair 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), leak detection interval, and allowable repair time
specified.
Recently developed data can be used to predict the potential
number of leaks from the various equipment types. 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
-------
Table 4-1 PERCENTAGE OF SOURCES PREDICTED TO BE LEAKING
IN AN INDIVIDUAL COMPONENT SURVEY1
Equi pment
Typea
Pumps
Pipeline Valves
a. Gasc
b. Liquid
Safety/Relief Valves
Pipeline Flanges
Compressors
Process Drains
PrpHirtpd Percent of Sources Leaking0
>100,000ppmv
6
4
2
1
0
5
0
J -
>50,000ppmv
9
5
4
2
0
10
1
>10,000ppmv
23
10
12
8
0
33
4
>1000ppmv
41
22
25
21
2
68
10
aData are not available for open-ended valves, 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 safety/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
-------
4.2.1 Definition of a Leak
In order to develop a leak detection 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 percent of total mass emissions that can potentially be controlled
by the leak detection and repair 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 repair
leaking components.
The ability to repair leaking equipment from above 10,000 ppm to
2
below 10,000 ppm has been demonstrated in field testing. 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 leak detection 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. This interval
can be related to the type of equipment and service conditions, and
different intervals can be specified for different pieces of equipment.
In the refinery VOC leak Control Technique Guideline (CTG) document ,
the recommended leak detection intervals are: annual pump seals,
4-4
-------
Table 4-2. PERCENT OF TOTAL MASS EMISSIONS
AFFECTED AT VARIOUS ACTION LEVELS1
_ '
1
Source Type
Pumps
i
Pipeline Valves
Gasc
Liquid
Safety/Relief Valves
Compressors
Percent of Mass Emissions Affected
at this Action Level3
100,000 ppmv
56
85
49
19
28
; n
Drains J
50,000 ppmv
68
92
62
33
48
8
10,000 ppmv
87
98
84
69
84
46
1 ,000 ppmv0
97 !
99
96
91
98
82
. 1
aThese figures relate the action level to the percentage of total mass
emissions that can be expected from sources with concentrations at
?he source greater than the action level. If these sources were
rnstantlneously 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
-------
pipeline valves in liquid service, and process drains; quarterly
compressor seals, pipeline valves in gas service, and safety/relief
valves in gas service; weekly visual inspection of pump seals; and
no individual monitoring pipeline flanges and other connections,
and safety/relief valves in liquid service. The choice of the interval
affects the emission reduction achievable, since more frequent inspection
will result 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 the 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 time that leaking sources are allowed to continue to emit benzene.
4.2.4 Visual Inspections
Visual inspections can be performed to detect evidence of liquid
leakage from plant equipment. When such evidence is observed, the
operator can use a portable VOC detector to measure the VOC concentra-
tion of the source. All liquid leaks will not necessarily result in a
3
reading of greater than the action level. 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 signals the need
for seal repair.
4.2.5 Other Leak Detection Techniques
Other leak detection techniques have been proposed to supplement
the individual component survey. These techniques include unit area
surveys (walkthroughs) and fixed-point leak detection systems. In
theory, these techniques allow the operator to reduce the number of
components that must be individually surveyed and hence reduce leak
detection labor requirements.
4.2.5.1 Unit Area Survey. A unit area survey entails measuring
the ambient VOC concentration within a given distance, for example,
4-6
-------
one meter of all equipment located on ground and other accessible
levels within a processing area. These measurements are performed
with a portable VOC detector 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 VOC concentrations are detected. If an elevated
VOC concentration is recorded, the components in that area can be
screened individually to locate the specific leaking equipment.
It is estimated that 50 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 VOC back-
ground concentration sufficiently to allow for reliable detection.
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 VOC emissions from adjacent units can cause false leak
indications; high or intermittent winds (local meteorological conditions)
can increase dispersion of VOC, 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
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 establish
4-7
-------
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
specifically 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 detector 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
5
percent for facilities with a small number of recurring leaks, pro-
vided that major leaking equipment has been repaired, reducing the
benzene background 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 sensing device to signal seal failure. Also, in
cases where rupture disks are installed upstream of safety/relief
valves, leakage through the disk can be monitored if pressure gauges
and/or excess flow valves are installed between the disk and the
valve.
4.2.6 Repair
When leaks are located by the leak detection methods described in
this section, the leaking component can then be repaired or replaced.
Many components can be serviced on-line. This is generally regarded
4-8
-------
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.
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 could be
repaired on-line.
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 benzene service in the refining and organic chemical
industries,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.
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
result from not shutting 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 emissions
from leaking valves until a shutdown is scheduled. Leaks that cannot
be repaired on-line can be repaired by drilling into the valve housing
and injecting a 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.
4-9
-------
For many 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
shutdown, so that repairs can be made. In most cases the shutdown for
repair of the leaking seal and the subsequent startup will involve
flaring the process stream until operations are stabilized. This can
result in the temporary emissions being larger than the continuous
emissions that would occur until the unit was shut down for other
reasons.
4.2.7 Emission Control Effectiveness of Leak Detection and Repair
The control efficiency achieved by a leak detection and repair
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 repair program resulted in repair of all such
sources to 0 ppm, elimination of all sources over the action 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.
4-10
-------
This approach can be expressed mathematically by the following
9
equation:
Reduction efficiency = AxBxCxD
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 Factory
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 leak again before the next inspection (recurrence),
including known leaks that could not be repaired.
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.
n = Imperfecc Repair Correction Factor = correction factor
to account for the fact that some sources which are
repaired are not reduced to zero emission levels, i-or
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 concentrations 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 "D»
correction factors for various possible inspection intervals, allowable
repair times, and action levels.
A.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
4-11
-------
Table 4-3. EMISSION CORRECTION FACTORS FOR VARIOUS INSPECTION
INTERVALS, ALLOWABLE REPAIR TIMES AND ACTION LEVELS (Reference 9)
ro
Leak Occurrence and
Recurrence Correction
Factorb
Inspection Interval
Source
Pumps
Pipeline Valves
Gasf
Liquid
Safety/Relief Valves9
Compressors
Drains
Yearly
0.800
0.800
0.800
0.800
0.800
0.800
Quarterly
0.900
0.900 k
0.900
0.900
0.900
0.900
Monthly
0.950
0.950
0.950
0.950
0.950
0.950
Non-Instantaneous
Repair Correction
FactorC
Allowable Repair
Time (Days)
15
0.979
0.979
0.979
0.979
0.979
0.979
5
0.993
0.993
0.993
0.993
0.993
0.993
1
0.999
0.999
0.999
0.999
0.999
0.999
Imperfect Repair
Correction
Factor^
Action Level
100,000
0.969
0.997
0.984
0.989
0,983
,.--
50,000
0.961
0.996
0.975
0.987
0,984
0.864
10
0.
0.
0.
0.
0,
0,
( ppmv )
,000
923
993
944
976
970
.906
1 ,000e
0.876
0.986
0.898
0.951
0,946
0.868
aNote that these correction factors taken individually do not correspond exactly to the overall emission
reduction obtainable by a leak detection and repair program. The overall effectless of the program is
determined by the product of all correction factors.
Salues are assumed and account for sources that start to leak between inspections (oc^rence) and for
sources that are found to be leaking, are repaired, and start to leak again before the next inspection
(recurrence), including known leaks that could not be repaired.
Accounts for emissions that occur between detection of a leak and subsequent repair.
Accounts for the fact that some sources that are repaired are not reduced to 0 ppm emission levels. The
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.
fValves in gas service carry process fluids in the gaseous state.
Gas service only.
-------
mechanism. This kind of leakage can be reduced to a negligible level
through the installation of improved shaft sealing mechanisms, such as
dual mechanical seals, or it can be eliminated entirely by installing
sealless pumps.
a.a.1.1 Dual Mechanical Seals. By design, dual mechanical
seals (double or tandem) 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 barrier fluid. Double mechanical seal systems have two seals
in a back-to-back arrangement providing an enclosed cavity. The
barrier fluid between the two seals is circulated through the cavity
to lubricate and cool the seals. The barrier fluid typically is
maintained at a pressure greater than the pump stuffing box pressure
(Figure 4-1) so that any leakage between the seals would be from the
barrier fluid to the process fluid. Consequently, no benzene would be
emitted to atmosphere as long as the double mechanical seal system is
operating properly. Field screening has shown that the magnitude of
emissions from new pumps equipped with double mechanical seals is
negligible as long as the integrity of the seals is maintained.
Double seals do fail after extended periods of use, however, and can
develop leaks.
Tandem mechanical seal systems have two mechanical seals in a
front-to-back arrangement. Like double seals, tandem seal systems
utilize a barrier fluid; however, the barrier fluid pressure is maintained
at a pressure lower than the pump stuffing box pressure. In this ar-
rangement, there is the possibility of leakage of the process fluid
into the barrier fluid. Leakage into the barrier fluid can be controlled
by either (1) connecting the barrier fluid degassing system to a
control device (i.e., enclosed combustion or vapor recovery) with a
closed vent system or (2) continuously replacing the fluid with fresh
barrier fluid and properly disposing of the contaminated barrier
fluid. One company, however, has reported mechanical problems associated
with lubrication of the outside seal, which cannot operate in a "dry"
state. When the outside seal fails, it may not be detected until the
inside seal fails, causing release of product. In addition, a bearing
4-13
-------
Possible leak into
sealing fluid
\
\
Sealing-liquid
/ inlet
Fluid_
end
Sealing-liquid
outlet
1
\ YT- ' |i
\ >
\ \. L'
\ u -\
\ , !
\ r-T
\ i '
ri I N.I
JLJ I
~r=- i l^ 1 r ' i /> ^ ^ VTl 1 TT-1
1 .>>, ' n ] f '// ut=t-r
-=n /ICc ^ £=, ^ ^C
,
Inner seal assembly
N0uterseal assembl
Figure 4-1. Double Mechanical Seal
-------
failure, which usually affects the entire pump shaft, can damage or
destroy both tandem seals so that their effectiveness in reducing
emissions is no more than a single mechanical seal.
The barrier fluid system between the dual seals may be a circulating
system, or it may rely on convection to circulate fluid within the
system. While the main function of the barrier fluid is to keep the
pumped fluid away from the environment, it can serve other functions
as well. A barrier flu-d can provide temperature control in the
stuffing box. Furthermore, it can protect the pump seals from the
atmosphere, as in the case of pumping easily oxidizable materials that
form abrasive oxides or polymers upon exposure to air. A wide variety
of fluids can be used as barrier fluids. Some of the more common ones
which have been used are water (or steam), glycols, methanol, oil, and
heat transfer fluid. In cases in which product contamination cannot
be tolerated, clean product, a product additive, or a product diluent
may also be used. ..,.*.
Mechanical seals, single or dual, 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.
In spite of these limitations, dual mechanical seals can be used in
most new pump applications.4 Dual mechanical seals can also be retro-
fitted 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 dual 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 dual
mechanical seal assembly. Data from industrial sources indicate that
this situation may occur for about 10 percent of all pumps in benzene
12
service.
4312 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.
4-15
-------
DISCHARGE
COOLANT CIRCULATING TUBE
SUCTION
STATOR LINER
\
IMPELLER
BEARINGS
Figure 4-2. Seal-less Canned Motor Pump
-------
Standard single-stage canned-motor pumps are available for flows
up to 160 cubic meters per second and heads up to 76 meters. Two-stage
units are also available for heads up to 183 meters. Canned-motor^
pumps are widely used in applications where leakage is a problem.
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, valves 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.
These valves have excellent corrosion resistance characteristics
and are reported to perform well in control valve situations with
minimal maintenance.14 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
low pressure operations.
4-17
-------
HANDWHEEL
PLUNGER
FLEXIBLE
DIAPHRAGM
SADDLE
SHAPED
SEAT
Figure 4-3. Diaphragm Valve
4-18
-------
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 main 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 beyond the limits of diaphragm
valves.
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 rupture 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 between 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
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. Alternately, to
prevent possible overpressure while using a block valve, a parallel
system of relief valves and rupture disks can be installed so that
one rupture disk/relief valve is in operation while the other is
being repaired.
Use of a rupture disk upstream of a safety/relief valve would
eliminate leaks due to improper seating and simmering of the relief
4-19
-------
STEM
YOKE
BELLOWS
Fiaure 4-4. Sealed Bellows Valve
4-20
-------
-Tension-adjustment
thimble
. Spring
To
atmospheric
vent
BLIND FLANGE
CONNECTION FOB
PRESSURE GAUGE
& BLEED VALVE
FROM SYSTEM
Figure 4-5. Rupture Disk Installation Upstream of a Safety/Relief Valve
4-21
-------
valve. Also, the disk can extend the life of a safety/relief valve by
protecting it 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
valve 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 safety/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 destruction 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 flare efficiency.
Calculations and limited test data show efficiencies ranging from
60 percent for low flow to a large flare to 90 to 99 percent for large
flow to a large flare.16"19 The presence of saturated organics or
aromatic compounds may decrease efficiency since such compounds are
not easily oxidized.
4.3.4 0;»on-Ended Valves
Caps, plugs, and double block and bleed valves are devices for
closing off the ends of valves and pipes. When installed downstream
4-22
-------
PILOT
MAIM
HE.ADE.R-S>->
-IXJ 1
V
£L_E-VAT E-D P V_ ARE.
GROUMD
DIVE.R.SIOM
A A A
SEAL OIL-
Figure 4-6. Simplified Closed-Vent System with Dual Flares
4-23
-------
of an open-ended valve, they are effective in preventing leaks through
the seat of the valve from reaching atmosphere. Open-ended valves,
20
about 20 percent of the total valves handling benzene, 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 arid 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 down-
stream 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 refining 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 point of
higher pressure to one of lower pressure in the system and where
sample-line dead space is minimized. Other sampling systems are
available that utilize partially evacuated sampling containers and
21
require no line pressure drop.
4-24
-------
PROCESS. UME.
PR.OCE.SS LIME.
COMTA1MER
0
5AM PL, E.
Figure 4-7. Diagram of Two Closed-Loop Sampling Systems
4-25
-------
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 valves 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
parameters of the combustion device will affect the overall control
efficiency of the closed-vent/combustion device system. Combustion
temperature 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.
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.
4.4 PROCESS MODIFICATIONS
In some instances, benzene fugitive emissions could be eliminated
by process modification. For some chemical processes, feedstocks
4-26
-------
other than benzene can be used, thus eliminating benzene from the
process. The following are some examples of this kind of substitution:
Maleic anhydride can be produced by oxidation of n-butane
rather than by oxidation of benzene.
Cyclohexane can be extracted from refinery products rather
2 5
than by hydrogenation of benzene.
4-27
-------
4.5 REFERENCES
1. Wetherold, R.G. 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 Number EPA-600/2-79-044.
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.
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.
5. 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.
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. Letter from Johnson, J.M., Exxon Co., to Walsh, R.T., EPA. July 28,
1977. Comments on EPA draft document, "Control of Hydrocarbon
from Miscellaneous Refinery Sources."
8. Teller, J.H., Advantages found in On-Line Leak Sealing. The Oil
and Gas Journal. 77J29): 54-59. July 16, 1979.
9. Tichenor, B.A., K.C. Hustvedt, and R.C. Weber. Controlling
Petroleum Refinery Fugitive Emissions Via Leak Detection and
Repair. In Proceedings: Symposium on Atmospheric Emissions
from Petroleum Refineries, November 1979, Austin, TX.
Report Number EPA-600/9-80-013. March 1980. p. 421-440.
10. Letter from Kronenberger, L., Exxon Company, to Goodwin, D.R., EPA.
February 2, 1977. Response to EPA request for information on
miscellaneous hydrocarbon emission sources from refineries.
11. Edwards, J.E. Valves, Pipe and FittingsA Special Staff
Report. Pollution Engineering. 6^:24. December 1974.
4-28
-------
12 Booz Allen and Hamilton, Inc. Benzene Emission Control Costs in
Selected Segments of the Chemical Industry Prepared for the
Manufacturing Chemists Association. June 12, 19/8.
13. Perry, John H. Chemical Engineers Handbook Robert Perry,
Cecil Chilton, Sidney Kirkpatrick, eds. McGraw-Hill Book
Company. New York. 1963. p. 6-7.
14. Telecon. Hanover, K., Rhom & Haas>;^e^
Environmental Services, Incorporated. September b, iy/y.
tee of leakless valves in chloromethyl ether service.
15 Letter from Sienknecht, P.J., Dow Chemical, to Mclnnjs,
Pacific Environmental Services, Incorporated. December
Comments on leak-free technology for control of benzene
emissions.
* *
Hydroscience for flare efficiency.
18- ' so: E«^ 5
flare stack.
19 Straitz, J. Flaring for Gaseous Control in the Petroleum
Industry National Air Oil. Philadelphia, Pennsylvania.
Presented at Air Pollution Control Association. Pittsburgh
June 26-30, 1978.
"Fugitive Emissions Report," February 1979.
calculations for benzene.
TnH,,<;tHal Process Profiles for Environmental Use: Chapter 6.
Ihe IndustrialTrganic Chemicals Industry. Research Triangle
institute Research Triangle Park, NC. Radian Corporation.
Autn TX. ForU?S. Environmental Protection Agency Cincinnati ,
OH Publication Number EPA-600/2-27-023f . February 1977.
p. 6-125, 6-826.
4-29
-------
24. Gunn, T.C. and K. Ring. CEH Marketing Research Report on Benzene.
Chemical Economics Handbook. Stanford Research Institute. Menlo
Park, CA. May 1977.
25. Blackford, J.L. CEH Product Review on Cyclohexane. Chemical
Economics Handbook. Stanford Research Institute. Menlo Park,
CA. February 1977.
4-30
-------
5.0 MODIFICATION AND RECONSTRUCTION
The Environmental Protection Agency plans to propose and promulgate
general provisions for Title 40 of the 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 may 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 refining and organic chemical industry
operations that involve benzene fugitive emissions.
5.1 GENERAL DISCUSSION OF MODIFICATION AND RECONSTRUCTION PROVISIONS
5.1.1 Modification
"Modification" is defined in 40 CFR Part 61, Section 61.02(j), as
any physical or operational change of a stationary source which increases
the emission rate of any hazardous air pollutant or which results in
the emission of any hazardous air pollutant not previously emitted.
Paragraph (j)(l) and (2) list exceptions to the above definition
of physical and operational changes which are not considered modifications
These changes include: (1) routine maintenance, repair, and replacement;
(2) an increase in the production rate not exceeding the operating
design capacity of the source; and (3) an increase in the hours of
operation.
5.1.2 Reconstruction
Under provisions that may be added to 40 CRF Part 61, an existing
source would become a new source upon reconstruction, irrespective of
any change in emission rate. Generally, reconstruction is considered
5-1
-------
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
would be made by the Administrator. The Administrator's final
determinations may be made on the following bases: (1) comparison of
the fixed capital costs of the replacement components and a comparable
entirely new 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 ensure that an owner or
operator does not perpetuate an existing source by replacing all but
vestigial components, support structures, frames, housing, etc.,
rather than totally replacing it in order to avoid being subject to
applicable new source standards.
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 process units, 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 concen-
tration (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.
5-2
-------
In some cases a production unit in the organic chemical industry
can be converted from the production of one chemical 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 during this conversion, the level of benzene
emissions from the unit could be expected to increase. However,
controls could be added in order to reduce or 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 equip-
ment but 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
The replacement of equipment within production units occurs on a
routine basis. In certain cases, these replacements can require the
expenditure of capital that exceeds 50 percent of the fixed capital
cost of the production unit. For example, replacement of existing
pumps with more efficient, new pumps may exceed 50 percent of the
fixed capital cost of a production unit.
5-3
-------
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 pro-
duction 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 and provide varying degrees of emission
reduction. The regulatory alternatives will be applied in the analysis
of both new and existing model units.
6.2 MODEL UNIT PARAMETERS
Production units in the refining and organic chemical industries
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 approximate various levels of process complexity. The
technical parameters of the three model units selected are shown in
Table 6-1.
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 CONTAINING GREATER THAN 10 PERCENT BENZENE
Number of Components per Model Unit
Source Type
Pumps
Pipeline Valves, Gase
Pipeline Valves, Liquid6
Safety/Relief Valves, Gas
Open-Ended Valves, Gas
Open-Ended Valves, Liquid
Drains
Sample Connections
Ab
5
30
56
3
3
23
5
9
Bc
15
91
168
9
7
72
15
26
Cd
25
151
280
16,
12
119
25
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.
°Represents 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 pipeline 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 operations in the refining
and organic chemical industries. Model A represents an average inventory
of such equipment for units involved in the production of benzene from
toluene, ethylbenzene, 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 production 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
characterized by the same set of model units.
It is estimated that 62 percent of existing benzene-related
production units in the refining and organic chemical industries 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 refining and
organic chemical 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
Regulatory Alternative I represents a baseline regulatory
alternative. 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 Alternatives'
Source0
-------
Table 6-2 MONITORING INTERVALS AND EQUIPMENT SPECIFICATIONS
FOR BENZENE FUGITIVE REGULATORY ALTERNATIVES
NOTES:
Regulatory 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 sources in the refining and organic chemical
industries.
Alternative 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 sensing device should be installed between the dual
seals and should be monitored to detect seal failure.
Inspection 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 "as 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.
Quarterly monitoring for leaks would be required for safety/relief
valves, pipeline and open-ended valves in gas service, and compressors.
Annual monitoring for leaks would be required for pumps, drains and
valves. Weekly visual inspections of pump seals would be required;
visual detection of a liquid leak would require 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 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 instal-
lation 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 emissions
reduced than for those included in Regulatory Alternative III. Mechan-
ical seal systems 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.2 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 would 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: (1) to require the use of leakless technology
for all equipment handling benzene-containing streams, (2) to require
the use of substitute feedstocks, thus eliminating the use of benzene,
and (3) to prohibit the production or consumption of benzene.
The use of leakless 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 approach, 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
'
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, 6.E., Pacific Environmental
Services, Inc., to EPA Docket (No. A-79-27). December 1979.
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 designated sources. Total benzene emissions,
incremental benzene emission reductions for the regulatory alternatives,
and projected future benzene emissions are presented. Other environ-
mental 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 units are
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 sources in the refining and
organic chemical industries. As discussed in Chapter 6, neither of
these alternatives is being considered as a viable control option.
The factors used to estimate the emissions from sources controlled
by a leak detection and repair program are calculated for each source
using the methodology presented in Chapter 4. The controlled emission
factors are calculated by multiplying the uncontrolled factors for
each source by a set of correction factors, which account for imperfect
7-1
-------
Table 7-1. CONTROLLED VOC EMISSION FACTORS FOR REGULATORY ALTERNATIVE II
Source
Pumps
Valves1
Gas
Liquid
Gas Service
Safety/Relief
Devices
Drains
Compressors
Inspection
Interval
Yearly
Quarterly
Yearly
Quarterly
Yearly
Quarterly
Uncontrolled
Emission
Factorb
(Kg/hr)
0.12h
0.021
0.010h
0.16
0.032
0.44
I
Correction Factors
Ac
0.87
0.98
0.84
0.69
0.46
0.84
Bd
0.80
0.90
0.80
0.90
0.80
0.90
ce
0.98
0.98
0.98
0.98
0.98
0.98
Df
0.92
0.99
0.94
0.98
0.91
0.97
Control
Efficiency
(AxBxCxD)
0.63
0.86
0.62
0.60
0.33
0.72
Controlled
Emission
FactorS
(Kg/hr)
0.044
0.003
0.004
0.064
0.021
0.123
"From Table 6-2.
bFrom Table 3-1.
cTheoretical maximum control efficiency From Table 4-2.
Leak occurrence and recurrence correction factor Assumed 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)] * 365 = 0.98.'
fImperfect repair correction factor From Table 4-3, calculated as 1 - (f * F), where f = average , 2
emission rate for sources at 1000 ppm and*F = average emission rate for sources greater than 10,000 ppm. '
Controlled emission factor = uncontrolled emission factor x [1 - (A X B X C X D)].
Emission factor for light liquid streams is used. (Reference 1)
Emission factors for in-line and open-ended valves are identical, since emissions from the open end rfould be
essentially eliminated by a cap, plug, blind or second valve.
7-2
-------
Table 7-2. CONTROLLED VOC EMISSION FACTORS FOR REGULATORY ALTERNATIVE III
Pumps
Valves
Gas
Liquid
Gas Service
Safety/Relief
Devices
Drains
Compressors
Inspection
Interval3
Monthly
Monthly
Monthly
None
Monthly
Monthly
Uncontrolled
Emission
Factor"
(Kg/hr)
0.12^
0.021
0.010J
0.15
0.032
0.44
Correction
Ac Bd
0.87 0.95
0.98 0.95
0.84 0.95
NA NA
0.46 0.95
0.84 0.95
Factors
P "f
C D
0.98 0.92
0.98 0.99
0.98 0.94
NA NA
0.98 0.91
0.98 0.97
Control
Efficiency
(AxBxCxD)
0.75
0.90
0.74
--
0.39
0.76
Controlled
Emission
FactorS
(Kg/hr)
0.030
0.002
0.003
0.0
0.020
0.106
aFrom Table 6-2.
bFrom Table 3-1.
cTheoretical maximum control efficiency From Table 4-2.
1,2
dLeak occurrence and recurrence correction factor - Assumed to be 0.80 for yearly inspection, 0.90 for
quarterly inspection, and 0.95 for monthly inspection/
Winstantaneous 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)] * 365 = 0.98.'
fImperfect repair correction factor - From Table 4-3, calculated as 1 - (f t F), where f average
emission rate for sources at 1000 ppm and F = average emission rate for sources greater than 10,000 ppm.
9Controlled emission factor = uncontrolled'emission factor x [1 - (A X 8 X C X D)].
hFor pumps, instrument monitoring should be supplemented v/ith weekly visual inspections for liquid leakage.
^Control equipment for this source is specified in Table 6-2, and about 100 percent control efficiency is
assumed. Therefore, the correction factors are not applicable, and there are essentially no fugitive
emissions from the source.
^Emission factor for light liquid streams is used. (Reference 1)
Emission factors for in-line and open-ended valves are identical, since emissions from the open end would be
essentially eliminated by a cap, blind, plug or second valve.
7-3
-------
Table 7-3. CONTROLLED VOC EMISSION FACTORS FOR REGULATORY ALTERNATIVE IV
Source
Pumps
ValvesJ
Gas
Liquid
Gas Service
Safety/Relief
Devices
Drains
Compressors
Inspection
Interval3
., h
None
Monthly
Monthly
h
None
Monthly
None
Uncontroll ec
Emission
Factorb
(Kg/hr)
0.121
0.021
0.0101
0.16
0.032
0.44
i
Correction Factors
Ac
NA
0.98
0.84
NA
0.46
NA
Bd CS
NA NA
0.95 0.98
0.95 0.98
. NA NA
0.95 0.98
NA NA
Control
f efficiency
DT (AxBxCxD)
NA
0.99 0.90
0.94 0.74
NA
0.91 0.39
NA
Controlled
Emission
Factor9
(Kg/hr)
0.0
0.002
0.003
0.0
0.020
0.0
"From Table 6-2.
From Table 3-1 .
cTheoretical maximum control efficiency From Table 4-2. ^
Leak occurrence and recurrence correction factor Assumed to be 0.80 for yearly inspection, 0.90 for
quarterly inspection, and 0.95 for monthly inspection.'
Q
Non-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)] T 365 = 0.98.2
Imperfect 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
Controlled emission factor = uncontrolled emission factor x [1 - (A X B X C X D)] .
Control equipment for this source is specified in Table 6-2, and about 100 percent control
Sns f^Tt^ou'rcV0^"10" faCtOPS are "Ot a»P1icable> »"< there « essentially no
^mission factor for light liquid streams is used. (Reference 1)
Emission factors for in-line and open-ended valves are identical, since emissions from the open end would be
essentially eliminated by a cap, blind, pluq or second valve.
'
7-4
-------
Table 7-4. CONTROLLED VOC EMISSION FACTORS FOR REGULATORY ALTERNATIVE V
Source
Pumps
Valves j
Gas
Liquid
Gas Service
Safety/Relief
Devices
Drains
Compressors
Inspection
Interval2
None
None
None
'Noneh
M h
None
None
Uncontrolled
Emission
Factorb
(Kg/hr)
0.121
0.021
0.0101
0.16
0.032
0.44
Controlled
Correction Factors Control Emission
Ac
MA
NA
NA
NA
NA
NA
Bd
NA
NA
NA
NA
NA
NA
ce
NA
NA
NA
NA
NA
NA
Df (AxBxCxD) (Kg/hr)
NA -- 0.0
NA -- 0.0
NA 0.0
NA -- 0.0
NA -- 0.0
NA -- 0.0
"From Table 6-2.
bFrom Table 3-1.
cTheoretical maximum control efficiency From Table 4-2.
dLeak occurrence and recurrence correction factor Assumed 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)] * 365 = 0.98.2
^Imperfect 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. '
^Controlled emission factor = uncontrolled emission factor x [1 - (A X B X C X D)3.
hControl eouipment for this source is specified in Table 6-2, and about 100 percent control efficiency is
assumed. Therefore, the correction factors are not applicable, and there are essentially no fugitive
emissions from the source.
Emission factor for light liquid streams is used. (Reference 1)
-'Emission 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
-------
repair, non-instantaneous repair, and the occurrence or recurrence of
leaks between inspections.
For each regulatory alternative and model unit, the number of
components handling greater than 10 percent benzene (Table 6-1) is
multiplied by the controlled VOC emission factor for each source
(Tables 7-1 through 7-4). Summing the VOC emissions from these sources
yields a total VOC emission estimate for equipment in each model unit
handling greater than 10 percent benzene. Since all sources do not
handle pure benzene, it is necessary to convert the VOC emission
estimates to benzene emission estimates. This is accomplished by
first estimating an average percent benzene for equipment in each
process containing greater than 10 percent benzene 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 is then possible to calculate a weighted average percent benzene
for equipment containing greater than 10 percent benzene in each model
unit. From the table, these weighted averages are calculated to be
63, 55, and 75 percent for model units A, B, and C, respectively.
Application of these factors to the VOC emissions yields a benzene
emission estimate for equipment containing greater than 10 percent
benzene in each model unit and for each regulatory alternative. An
example calculation is given below for pumps containing greater than
10 percent benzene in Model Unit A under Regulatory Alternative II.
Benzene emissions from
>10 percent benzene (Bz)
Pumps in Model Unit A
Under Regulatory
Alternative II (kg/hr)
/>10 percent Bz Pumps'
= 5 (in Model Unit A
i(from Table 6-1)
0.63
\
Weighted factor
of percent benzene
in Model Unit A
(from Table 7-5)
x 0.044
7Kg VOC per hourx
per pump for
Regulatory
Alternative II
.(from Table 7-1)
\
= 0.14 kg Benzene per hour from
>10 percent Bz Pumps in Model
Unit A, Regulatory Alternative II
7-6
-------
Table 7-5. CALCULATION OF WEIGHTED
PERCENT BENZENE FOR
EMISSION SOURCES IN MODEL UNITS
Number of
Units,
(1980)°
MODEL UNIT A
Ethylene (1 Unit)
Ethyl benzene
Benzene (Dealkylation)
Styrene
Cumene
Cyclohexane
Maleic Anhydride
Benzenesul fonic Acid
TOTAL
MODEL UNIT B
Benzene (Reformer)
Chlorobenzene
Ethylene (2 or 3 Units)
Al kyl benzenes
TOTAL
MODEL UNIT C
Benzene (Pyrolysis Gas)
Nitrobenzene
Hydroquinone
Ethylene (4 or 5 Units)
TOTAL
41
22
20
19
16
12
' 10
5
145
49 ,
12
7
4
72
12
10
_!_
24
Percent
of Total Average Percent Benzene
Mnrfpl for Equipment in Each
SSS Processa'c
28.3
15.2
13.8
13.1
n.o
8.3
6.9
3.4
100.0
68.0
16.7
9.7
5.6
100.0
50.0
41 .6
a ?
4.2
100.0
33
71
75
30
100
100
100
85
45
100
33
65
.
57
100
75
33
*_
Weighted
Percent
Benzene
9
IV
10
4
11
8
7
3
63
31
17
3
4
55
29
42
3
1
75
7-7
-------
Table 7-5. CALCULATION OF WEIGHTED
PERCENT BENZENE FOR
EMISSION SOURCES IN MODEL UNITS3 (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. Calculations
for all of the other processes are for designated sources handling
greater than ten percent benzene.
blncludes plants under construction expected to be completed in
1980.
cThe average percentage of benzene for designated sources in each product
process containing greater than 10 percent benzene 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
Percent
Benzene
Number of Pieces
of Equipment
Concentration
in Each Stream
Total Pieces of Equipment
Example:
Weighted
Percent
Benzene
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
(5 x 100) + (5 x 50) = 75 Percent
10
7-8
-------
Using a similar technique, it is also possible to estimate
benzene emissions from equipment handling streams with less than
10 percent benzene. VOC emissions from these sources are calculated
using the same VOC emission factors as are used for equipment handling
over 10 percent benzene. The average percent benzene for sources
containing less than 10 percent benzene is estimated by examining flow
diagrams and material balances. For each process, benzene emissions
from these sources are then calculated. A weighted average of the
processes is then applied to give the "residual" benzene emissions
from the equipment handling less than 10 percent benzene for each
model unit. These emissions are calculated to be 0.35, 0.55, and 0.48
kilograms per hour 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.
Tables 7-6 through 7-8 present benzene emission estimates by
source for each of the model units and regulatory alternatives. In
general for each model unit, valves and pumps containing greater than
10 percent benzene contribute about half of the total benzene emissions
from all sources. Table 7-9 presents total national benzene emissions
(megagrams per year) for each alternative based on the number of model
units expected to be in operation in 1980 and the benzene emission
totals per model unit and regulatory alternative from Tables 7-6
through 7-8.
7.2.2 Future Benzene Emissions
In order to assess potential future impacts of the regulatory
alternatives, benzene emissions are projected over a ten-year period
(1981-1990). To estimate future benzene emissions, two assumptions
are made:
1. The average process unit production rate as a percentage of
unit capacity for all benzene products remains the same over
the projected ten-year period.
2. An increase in demand is accounted for by building new
units, expanding the capacity of existing units, or by
renovating existing units.
7-9
-------
Table 7-6. BENZENE EMISSIONS (kg/hr) BY SOURCE FOR THE
REGULATORY ALTERNATIVES - MODEL UNIT A
Regulatory Alternative
Source I II III IV
Pumps
Valves
Gas
Liquid
0.38
0.40
0.36
0.14
0.06
0.14
0.10
0.04
0.11
0.00C
0.04
o.n
o.ooc
o.ooc
o.ooc
Safety/Relief Valves _ _ r
(Gas) 0.30 0.12 0.00^ 0.00L 0.00^
Open-ended Valves
Gas 0.05 0.01 0.00d 0.00d O.QOj:
Liquid 0.19 0.06 0.04 0.04 0.00L
Drains 0.10 0.07 0.06 0.06 0.00C
Sampling Connections 0.09 0.09 0.00C 0.00C 0.00C
Other Equipment5 0.35 0.35 0.35 0.35 0.35
Total 2.22 1.04 0.70 0.60 0.35
Emissions (kg/hr)
aThe regulatory alternatives are summarized in Table 6-2. Alternative I
represents baseline emissions.
bOther equipment includes sources containing less than 10 percent benzene.
cControl equipment for this source is specified in Table 6-2, and about 100
percent control efficiency is assumed. Therefore, there are essentially
no fugitive emissions from the source.
Negligible emissions - less than 0.005 kg/hr.
7-10
-------
Table 7-7. BENZENE EMISSIONS (kg/hr) BY SOURCE FOR THE
REGULATORY ALTERNATIVES - MODEL UNIT B
Regulatory Alternative
Source I " IH IV
Pumps 0.97 °-36 °'24 °-°° °-°°
Valves _
r,, 1.03 0.15 0.10 0.10 0.00
Sjuld 0.9I 0.37 0.27 0.27 0.00C
^(Gas)16' Va1V6S 0.78 0.32 0.00C 0.00C 0.00C
Open-Ended Valves
r,, o.oe o.oi o.oi o.oi o.oo
Squid 0.52 0.16 0.12 0.12 0.00C
0.26 0.17 0.15 0.15 0.00C
Drains
Sampling Connections
Other Equipment5 ' 0.55 0.55 0.55 0.55 0.55
0.21 0.21 0.00C 0.00C 0.00C
ggionsjka/hr^ 5.32 2.30 1 ^_^__Q^^
aThe regulatory alternatives are summarized in Table 6-2. Alternative I
represents baseline emissions.
bOther equipment includes sources containing less than 10 percent benzene.
Control equipment for this source is specified In Table 6-2, and about
100 percent control efficiency is assumed. Therefore, there are
essentially no fugitive emissions from the source.
7-11
-------
Table 7-8. BENZENE EMISSIONS (kg/hr) E5Y SOURCE FOR THE
REGULATORY ALTERNATIVES - MODEL UNIT C
Regulatory Alternative3
Source
Pumps
Valves
Gas
Liquid
I
2.24
2.36
2.09
II
0.82
0.34
0.85
III
0.56
0.23
0.62
IV
0.00C
0.23
0.62
V
0.00C
0.00C
0.00C
Safety/Relief Valves
(Gas) 1.91 0.79 0.00C 0.00C Q.OOC
Open-ended Valves
Liquid 1.17 6.36 6 ".26 6".26 0.66C
Gas 0.22 0.03 0.02 0.02 0.00C
Drains 0.60 0.39 0.35 0.35 0.00°
Sampling Connections 0.49 0.49 0.00C 0.00C 0.00C
Other Equipment *0.48 0.48 0.48 0.48 0.48
Total
Emissions (kg/hr) 11.56 4.55 2.52 1.96 0.48
aThe regulatory alternatives are summarized in Table 6-2. Alternative I
represents baseline emissions.
Other equipment includes benzene sources containing less than .
10 percent benzene.
°Control equipment for this source is specified in Table 6-2, and about
100 percent control efficiency is assumed. Therefore, there are
essentially no fugitive emissions from the source.
7-12
-------
Table 7-9 TOTAL NATIONAL BENZENE EMISSIONS FROM REFINING AND
ORGANIC CHEMICAL PROCESSES IN 1980^
Emission Reduction
From Alternative I
Regulatory
Alternative
I'-
ll
III
IV
V
Emissions
(Mg/yr)b
8250
3570
2230
1860
860d
Mg/yr
4680
6020
6390
7390
_
Percent
57
73
77
90
111
aAnv process that has equipment handling over 10 percent benzene streams is
InclSded in this analysis' The emission totals include the emissions from
pauiOTient over 10 percent and under 10 percent benzene for each of these
Processes. Benzene emissions per year are calculated from an assumed process
operation of 8400 hours per year.
Calculated 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 T^^if^fo^^
unit. The number of existing units operating in 1980 is as follows.
Model Unit A 145
Model Unit B 72
Model Unit C 24
Represents baseline emissions (the industry in the absence of new regu-
lations).
dResidual from equipment handling less than 10 percent benzene.
7-13
-------
In order to calculate benzene emission levels over the 10-year
period, a weighted growth rate for each model unit is 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 7-10 lists the number of plants estimated
in operation in 1980 and the average annual growth rates for each
product. The weighted average annual 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 7-10 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 an example, to calculate the projected number of new Model A units
constructed between 1981 and 1985, the above formula is applied as
follows:
N = 14^ M + n^Tfi'l - 14"S
'Mnoi inoc l*tJ \JL. ^ ,\j-J/uj itj
L9ol-iyob
= 47 new plants.
Assuming an average unit life of 20 years, the number of replacements
(R) is calculated by the following:
R = rf " N'
where N = the number of new units and
rf = a replacement factor as follows:
i T 1 ~ """ ^ */-)V\
7-14
-------
Table 7-10. NUMBERS OF UNITS ESTIMATED TO MEET 1980
DEMAND FOR BENZENE AND BENZENE
DERIVATIVES BY MODEL UNITS
Model A
Ethyl ene (1 unit)
Ethyl benzene
Benzene (Dealkylation)
Styrene
Cumene
Cyclohexane
Maleic Anhydride
Benzenesulfonic Acid
Total
Model B
Benzene (Reformer)
Chlorobenzene
Ethylene (2 or 3 units)
Alky! benzenes
Total
Model C
Benzene (Pyrolysis Gas)
Nitrobenzene
Hydroquinone
Ethylene (4 or 5 units)
Total
Estimated
Number of
Units in
"Year 0"
(1980)
41
22
20
19
16
12
10
5
145"
49
12 *
7
4
72
12
10
1
1
24
Percent
of Total
Units
28.3
15.2
13.8
13.1
n.o
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 Rate4-!*
(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
O2
7.05
2.50
0.00
0.23
9.78
*Does not include replacement of old units.
7-15
-------
The results are presented in Table 7-11, which shows the projected
growth of new and replacement units over a 10-year period (1981-1990).
Using the benzene emission estimates for each regulatory
alternative and model unit from Tables 7-6 through 7-8, the expected
benzene emissions that will be contributed by new units and replace-
ments can be calculated by multiplying the benzene emissions by the
number of new units and replacements estimated to be operating between
1981 and 1990 (Table 7-11). For each alternative, emissions from
Model Units A, B, and C are summed to obtain the total benzene emissions
for the regulatory alternative. Table 7-12 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.
7-16
-------
Table 7-11. CUMULATIVE ANNUAL NUMBER OF PROJECTED
NEW UNITS AND REPLACEMENTS
BETWEEN 1981 AND 1990a
Year
1981
Newb
Replacement
1982
Newb
Replacement
1983
Newb
Replacement
1984
New
Replacement
1985
Newb
Replacement
1990
Newb
Reolacement
A
8
4
17
8
27
13
36
17
47
23
109
53
Model Unit
B
1
0
3
1
4
2
6
3
7
4
15
7
C
2
0
5
1
8
1
11
2
14
3
37
7
7-17
-------
NOTES FOR TABLE 7-11
Projections of new and replacement unit growth over a 10-year
period were based on the following:
Model
Unit
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
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 = Ayear from 1980 (1,2,3... 10)
cProjections of replacements were based on the replacement factor (rf)
rf = 1 -
-TT
assuming an average unit life of 20 years. Number of replacements
(r) is calculated from the following:
R=!L.N,
rf
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.
7-18
-------
Table 7-12. CUMULATIVE ANNUAL ESTIMATED BENZENE
FUGITIVE EMISSIONS FROM
NEW UNITS AND REPLACEMENTS BETWEEN
1981 and 1990 (Mg/year)
Year
1981
New
Replacement
Regulatory AT
ternative
I II III IV
388 165 101
74 35 23
84
20
V
36
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
7-19
-------
Under Regulatory Alternative IV or V, a dual 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" fugitive VOC 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. The 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 Alternative III,
which requires the installation of rupture disks on safety/relief
valves, a closed-loop sampling system, and seals on open-ended valves
and lines would have a negligible impact. For Regulatory Alternative
IV, dual mechanical seals would be retrofitted on pumps. Existing
packing materials and single mechanical seals may not be reusable and
hence would be discarded. As the dual 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
equipment 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
7-20
-------
generated by the replacement of single mechanical seals and packing
material, as in Regulatory Alternative IV, and the disposal of
unrecyclable 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 dual 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
the refining and organic chemical industries is estimated as
8.62 x 106 joules/kg.15 Table 7-13 presents the energy savings over
a five-year period that result from the VOC fugitive emission reductions
associated with Regultory 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 10 joules
per barrel of crude oil, the energy value of the recovered VOC fugitive
emissions 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 oil 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
7-21
-------
Table 7-13. ENERGY IMPACT OF BENZENE EMISSION
REDUCTION FOR REGULATORY ALTERNATIVES
I
r\5
ro
Year
1981
1982
1983
1984
1985
5-year
Total
Reduction from Baseline
Emissions Under Regula-
tory Alternatives3""
(1000 rig)
II III IV
7.93 10.2 10.8
8.59 11.0 11.7
9.16 11.8 12.5
9.82 12.6 13.4
10.5 13.5 14.3
46.0 59.1 62.7
V
12
13
14
15
16
72
5
%
6
5
5
6
7
Energy Value of
Reductions Under
tory
II
68.4
74.0
79.0
84.6
90.5
397
Emission
Requla-
Alternatives0
(109 Btu)
III
87.9
94.8
102
109
116
510
IV
93.1
101
108
116
123
541
V
108
117
125
134
143
627
Crude Oil
Emission
Equivalent of
Reductions'1
(1000 barrels - bbl )
II
11.2
12.1
12.9
13.8
14.8
64.8
III
14.4
15.5
16.7
17.8
19.0
83.4
IV
15.2 ,
16.5
17.6
19.0
20.1
88.4
V
17.6
19.1
20.4
21.9
23.4
102
Alternative I represents baseline emissions.
Emission 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. 15, p. 3-143).
Based on 6.12x10 joules/barrel 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 7-13.
The emission reductions and associated energy savings shown would be
lost at the rates shown for each of the five years.
7-23
-------
7.7 REFERENCES
1. Wetherold, R.G. and L. P. Provost. Emission Factors and Frequency
of Leak Occurrence for Fittings in Refinery 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. In proceedings: Symposium on Atmospheric Emissions from
Petroleum Refineries, Austin, TX. November 6, 1979. Report
Number EPA-600/9-80-013. March 1980. p. 421-440.
3. Calculation of residual benzene emissions from equipment
containing less than 10 percent benzene and calculation of
benzene emissions by source for each model unit and regulatory
alternative. Docket Number A-79-27-II-B.
4. Soder, S.L. CEH Product Review on Styrene. Chemical Economics
Handbook. Stanford Research Institute. Menlo Park, CA.
January 1977.
5. CEH Product Review on Cyclohexane. Chemical Economics Handbook.
Stanford Research Institute. Menlo Park, CA. February 1977.
6. Gunn, T.C., and K. Ring. CEH Marketing Research Report on
Benzene. Chemical Economics Handbook. Stanford Research
Institute. Menlo Park, CA. May 1977.
7.
8.
9.
10.
11.
12.
CEH Product Review on Chlorobenzenes. Chemical Economics
Handbook. Stanford Research Institute. Menlo Park, CA.
July 1977.
CEH Product Review on Ethylene. Chemical Economics Handbook.
Stanford Research Institute. Menlo Park, CA. January 1978.
Cogswell, S.A. CEH Product Review on Resorcinol.
Economics Handbook. Stanford Research Institute.
CA. October 1978.
Chemical
Menlo Park,
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.
7-24
-------
13 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 Greene, R. U.S. Benzene Markets to Face Slower Growth.
Chemical Engineering. 85(3):62-64. January 30, 1978.
15 Perry, J.H. Chemical Engineer's Handbook. Fourth edition.
New York, McGraw-Hill Book Co., 1963. p. 3-143.
7-25
-------
8.0 COST OF CONTROLS
8.1 INTRODUCTION
In order to estimate the economic impact of the regulatory
alternatives on the petroleum refining and organic chemical manu-
facturing 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 unit parameters given in Table 8-1 and the
capital cost data in Table 8-3, capital costs for each model unit
8-1
-------
Table 8-1. MODEL UNIT EQUIPMENT CONTAINING >10 PERCENT BENZENE
Number of Components per
Source Type
Pumps
Pipeline Valves, Gas
Block
Control
Pipeline Valves, Liquidd
Block
Control
Safety/Relief Valves, Gas
Open-Ended Valves, Gas
Open-Ended Valves, Liquid
Drains
Sample Connections
Model Aa
5
27
3
51
5
3
3
23
5
9
Model Bb
15
83
8
153
15
9
7
72
15
26
Model Unit
Model Cc
25
138
13
256
24
16
12
119
25
44
Represents an average inventory of equipment for production of
ethylbenzene, styrene, cumene, cyclohexane, benzene sulforric
acid, resorcinol, maleic anhydride, or 1 ethylene production
unit.
Represents an average inventory of equipment for production of
chlorobenzenes, linear alky!benzenes, or 2 or 3 ethylene produc-
tion units.
Represents an average inventory of equipment for production of
benzene, nitrobenzene, hydroquinone, or 4 or 5 ethylene produc-
tion units.
From Hydroscience, 6 percent of all valves are control valves;
69 percent of all valves are process valves; so, 6 f 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
CO
Regulatory Alternatives*
Source0 'd
1 . Pumps
2. Pipeline Valves
a. Gas Service
b. Liquid Service
3. Safety/Relief
Valves (Gas
Service)
4. Open-Ended Valves
a. Gas Service
b. Liquid Service
5. Drains
6. Sampling
Connections
7. Compressors
8. Product Accumu-
lator Vessel
Vents
Inspection
Interval
Yearly6
Quarterly
Yearly
Quarterly
Quarterly9
Yearly9
Yearly
None
Quarterly
None
,,b
Equipment
Specification
None
None
None
None
Caps, blinds,
or plugs
Caps, blinds,
or plugs
None
None
None
None
Inspection
Interval
Monthly6
Monthly
Monthly
None
Monthly9
Monthly9
Monthly
None
Monthly
None
III
Equipment
Specification
None
None
None
Rupture disks
or tie Into
existing flare
Caps, blinds,
or plugs
Caps, blinds,
or plugs
None
Closed-loop
sampling
None
Tie into
closed control
system
IV
Inspection
Interval
None8
Monthly
Monthly
None
Monthly9
Monthly9
Monthly
None
None
None
Equipment
Specification
Dual Seals
and controlled
degassing vent
None
None
Rupture disks
or tie into
existing flare
Caps, blinds,
or plugs
Caps, blinds,
or plugs
None
Closed-loop
sampling
Mechanical Seals
Tie Into
closed control
system
V
Inspection
Interval
None
None
None
None
None
None
None
None
None
None
Equipment
Specification
Dual Sealsf
and controlled
degassing vent
Diaphragm or
sealed bellows
valves
Diaphragm or
sealed bellows
valves
Rupture disks
or tie into
existing flare
Diaphragm or
sealed bellows
valve plus caps,
blinds, or plugs
Diaphragm or
sealed bellows
valve plus caps,
blinds, or plugs
drain systems
Closed-loop
sampl ing
Mechanical Seals
Tie into
closed control
system
-------
Table 8-2. MONITORING INTERVALS AND EQUIPMENT SPECIFICATIONS
FOR BENZENE FUGITIVE 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 refining and organic chemical industry
sources.
Alternative 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.
A sensing device should be installed between the dual
seals and should be monitored to detect seal failure.
^Inspection applies to the valve.
8-4
-------
Table 8-3. CAPITAL COST DATA
(May 1979 Dollars)
Item
Capital Cost
1. Monitoring Instrument
2. Caps for Open-Ended Valves
3. Double Mechanical Seals
2 x 4,250 = $8,500
$50/cap
$590/pump (new)
$870/pump (retrofit)
4. Barrier Fluid Recirculation $l,530/pump
System for Double Seals
5. Degassing Vents
6. Rupture Disks for Safety/
Relief Valves
7. Closed-Loop Sampling
Connections
8. Sealed Bellows Valves
1. 61 m of 5.1 cm
carbon steel pipe
(installed)
$2,700
2. 3 x 5.1 cm valves
(installed)
$940
3. 1 x 5.1 cm flame
arrestor
(installed)
$450
Total = $4,090/pump
$l,800/relief valve
(new)
$3,230/relief valve
(retrofit)
$480(new or retrofit)
33,700/valve
(retrofit)
$2,500/valve (new)
Reference
1
2, 3, 4C
2, 3, 4L
2, 3, 4L
2, 3, 4'
2, 3, 4C
2, 3,
2, 3, 4,'13, 14
a,e
2, 3, 4'
5
a,d
5-5
-------
Table 8-3. CAPITAL COST DATA (Concluded)
(May 1979 Dollars)
Item
9. Sealed Drain Covers
10. Replacement Pump
Capital Cost
$1 ,000/drain
$3,000/pump
Reference
6f
NOTES:
aPlant cost indices were used.
Pump and compressor cost indices were used.
cPiping, valve and fitting cost indices were used.
No retrofit penalty.
eCost of rupture disks includes rupture disk, block valve, and
replacement of relief valve.
^Consists of sealed drain cover and sealed pump drain line.
gln 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 valves. 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 safety/relief valves that vent to atmosphere and
the installation 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 safety/relief valves explains the difference in
capital costs between existing and new units. Regulatory Alternative IV
specifies that pumps and compressors be equipped with mechanical seals
for which degassing vents on pump seal reservoirs are installed.
Costs are given for double seal systems for pumps, even though Regulatory
Alternatives IV and V specify dual seals (double or tandem). 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). Alternative 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)
Capital Cost Item
Regulatory
I II
Al
III
ternative
IV
V
Model Unit A (Existing)
1. Monitoring Instrument
2. Caps for Open-Ended
Valves
3. Double Mechanical Seals
8.5
1.8
8.5
1.8
8.5
1.8
4.4
1.8
4.4
4. Barrier Fluid Recirculation
System for Double
Mechanical Seals
5. Vents for Barrier Fluid
Degassing Reservoirs
6. Replacement Pumps
7. Rupture Disks for Safety/
Safety/Relief Valves
8. Closed Loop Sampling
Connections
9. Sealed Bellows, Valves
10. Hard Piping and Drain
Covers
9.7
7.7
9.7
7.7
20.5 20.5
3.0
9.7
4.3 4.3 4.3
447.7
50.0
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
Model Unit A (Mew)
1. Monitoring Instrument
2. Caps for Open-Ended
Valves
3. Double Mechanical Seals
4. Barrier Fluid Recirculation
System for Double
Mechanical Seals
5. Vents for Barrier Fluid
Degassing Reservoirs
6. Rupture Disks for Safety/
Safety/Relief Valves
7. Closed Loop Sampling
Connections
8. Sealed Bellows, Valves
9. Hard Piping and Drain
Covers
Tf\T & 1 (
II III IV
8.5 8.5 8.5
1.8 1.8 1.8
3.0
7.7
20.5
5.4 5.4
4.3 4.3
vn 10.3 20.0 51.2
V
1.8
3.0
7.7
20.5
5.4
4.3
302.5
50.0
350.2
8-9
-------
Table 8-4. CAPITAL COST ESTIMATES PER MODEL UNIT
(Thousands of May 1979 Dollars)
Regulatory Alternative
Capital Cost Item
II
III
Model Unit B (Existing)
1. Monitoring Instrument
2. Caps for Open-Ended
Valves
3. Double Mechanical Seals
4. Barrier Fluid Recirculation
System for Double
Mechanical Seals
5. Vents for Barrier Fluid
Degassing Reservoirs
6. Replacement Pumps
7. Rupture Disks for
Safety/Relief Valves
8. Closed Loop Sampling
Connections
9. Sealed Bellows, Valves
10. Hard Piping and Drain
Covers
8.5 8.5
IV
8.5
5.3 5.3 5.3 5.3
13.1 13.1
23.0 23.0
61.4 61.4
6.0
29.1 29.1 29.1
12.5 12.5 12.5
1346.8
15.0
TOTAL
0.0 13.8 55.4 152.9 1512.2
8-10
-------
Table 8-4. CAPITAL COST ESTIMATES PER MODEL UNIT (Continued)
(Thousands of May 1979 Dollars)
Regulatory Alternative
Capital Cost Item I
Model Unit B (New)
1. Monitoring Instrument
2. Caps for Open-Ended
Valves
3. Double Mechanical Seals
4. Barrier Fluid Recirculation
System for Double
Mechanical Seals
5. Vents for Barrier Fluid
Degassing Reservoirs
6. Rupture Disks for
Safety/Relief Valves
7. Closed Loop Sampling
Connections
8. Sealed Bellows, Valves
9. Hard Piping and Drain
Covers
TATAI 1
II III IV
_
8.5 8.5 8.5
5.3 5.3 5.3
8.9
23.0
61.4
16.2 16.2
12.5 12.5
3.0 13.8 42.5 135.8
V
5.3
8.9
23.0
61.4
16.2
12.5
910.0
15.0
1052.3
8-11
-------
Table 8-4. CAPITAL COST ESTIMATES PER MODEL UNIT
(Thousands of May 1979 Dollars)
Regulatory Alternative
Capital Cost Item
II
III
IV
Model Unit C (Existing)
1. Monitoring Instrument
2. Caps for Open-Ended
Valves
3. Double Mechanical Seals
4. Barrier Fluid Recirculation
System for Double
Mechanical Seals
5. Vents for Barrier Fluid
Degassing Reservoirs
6. Replacement Pumps
7. Rupture Disks for
Safety/Relief Valves
8. Closed Loop Sampling
Connections
9. Sealed Bellows, Valves
10. Hard Piping and Drain
Covers
8.5 8.5
8.5
8.8 8.8 8.8 8.8
21.8 21.8
38.3 38.3
102.3 102.3
9.0
51.7 51.7
51.7
21.1 21.1 21.1
2242.2
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
1 """"
Model Unit C (New)
1. Monitoring Instrument
2. Caps for Open-Ended
va 1 ves
3. Double Mechanical Seals
4. Barrier Fluid Recirculation
System for Double
Mechanical Seals
7. Closed Loop Sampling
Connections
8. Sealed Bellows, Valves
9. Hard Piping and Drain
Covers
TOTAL
n
m
8.5 8.5
<:i>i
IV
8.5
8.8
14.8
38.3
8>8
14'8
38>3
21 1
l
0.0 17.3 67.2 222.6 1754.1
8-13
-------
(1) monitoring instrument 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 = Cl + C2
where C = total annualized costs; (^ = annual depreciation and interest
charges; and C2 = property taxes and insurance. Now Cj and C2 are
described as:
C, = C x (CRF) and C2 = Cc x (n)
JL \*
where: C = capital cost of the equipment
\*
. /1+-\n where: i = annual interest rate (10%)
CRF = (l+i)n-l n = lifetime of equipment, years
(n=6 for the monitoring
instrument, and n=10 for
control equipment)
Annualized leak detection and repair labor costs are derived by
means of the formula:
L 4 + 4
where L = total annual leak detection and repair labor costs; L^ -
annual leak detection cost; and l_2 = annual repair cost. Now LI and
L0 are described as:
8-14
-------
I = (AxBxDxE)xN and LZ = FxGxE
where: A = Number of model unit components affected
B = Monitoring time, hours (leak detection)
D = 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 benzene emissions have
been made from refining and organic chemical operations, based^on the
market price of benzene as of May 1979 ($370/Mg or $1.30/gal).
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
8-15
-------
Table 8-5. MONITORING AND MAINTENANCE LABOR-HOUR REQUIREMENTS
FOR REGULATORY ALTERNATIVE II
co
i
Source
Type
Pumps
Valves
Gas
Liquid
Safety/
Relief
Valves
Drains
MONITORING
Number of
Components
Per Model
Unit
A
5
34
87
4
5
B
15
100
264
11
15
C
25
167
439
19
25
Type of
Monitoring
Instrument
Visual
Instrument
Instrument
Instrument
Instrument
Monitoring
Time a
Per Person
(Minutes)
5
0.5
It
1
1
3
1
Times
Monitored
Per Year
1
52
4
1
4
1
Mom" toring
Labor-Hours
Required
A
0.8
2.2
4.5
2.9
4.3
0.2
B
2.5
6.5
13.3
8.8
11.7
0.5
C
4.2
10.8
22.3
14.6
20.3
0.8
MAINTENANCE
Estimated
Number of
Leaksb
A
1
1
2
c
1
B
1
4
6
c
1
C
1
7
11
c
1
Repair
Time
(Hours)
80
1.13
1.13
0
4
Maintenance
Labor-Hours
Required
A
80
1
't
0
4
B
80
5
7
0
4
C
80
8
12
0
4
Total Monitoring ,. g 43 3 73 0 Total Maintanence
Hours = ~ ~ Hnurs =
v $15.50/hour
Total Monitoring 230 670
Dollars =
Hours
x $15.50/hour
Total Maintenance
Dollars =
87
96 104
1350 1490 1610
Instrument monitoring requires a two-person team. Visual monitoring requires one person.
Recurrence factors of 0.6, 0.4, and 0.2 have been applied for monthly, quarterly, and annual
instrument inspections. It is assumed that 5 percent of leaks initially detected recur each
month (0.5 x 12 = 0.6), that 10 percent of leaks initially detected recur each quarter
(.10 x 4 = 0.4), and that 20 percent of leaks initially detected recur annually (.20 x 1 - 0.2)
cThese leaks are repaired by routine maintenance at no incremental increase in manpower requirements.
Safety/relief valves are normally reset during routine maintenance without a leak detection and
repair program. (Reference 8).
-------
Table 8-6 MONITORING AND MAINTENANCE LABOR-HOUR REQUIREMENTS
FOR REGULATORY ALTERNATIVE III
oo
i
Source
Type
Pumps
/alves
Gas
Liquid
.
Safety/
Raltef
Valves
Orains
MONITORING
Number of
Components
Per Model
Unit
A
5
34
87
4
I 5
B
15
100
264
11
15
C
25
167
439
19
25
Type of
Monitoring
Instrument
Visual
Instrument
Instrument
Instrument
Instrument
Monitoring
Time
Per Person
(Minutes)
5
0.5
1
1
8
1
Times
Monitored
Per Year
12
52
12
12
0
12
Total Monitoring
Hours »
x $15.50/hou
Monitoring
Labor-Hours
Required
A
10.0
2.2
13.6
34.8
0.0
2.0
B
30.0
6'.5
40.0
105.6
0.0
6.0
C
50.0
10.8
66.8
175.6
0.0
10.0
62.6 188.1 313.2
r
MAINTENANCE
Estimated
Number of
Leaksb
A
1
2
6
--
1
B
2
6
19
1
C
4
10
32
--
1
Repair
Time
(Hours)
80
1.13
1.13
0
4
Maintenance
La.bor-Hours
Required
A
80
2
7
0
4
B
160
7
21
0
4
C
320
11
36
0
4
Qi 1 Q? 17 1
Total Maintanence j£_ 1jlf_ zli
Hours =
Total Monitoring 970 2920 4850 n iiV:. - _V"- Z^Z ^^~
Dollars = Dollars -
Instrument monitoring requires a two-person team. Visual monitoring requires one person.
Recurrence factors of 0.6, 0.4 and 0.2 have been applied for monthly, quarterly, and annual instrument inspections.
It is assumed that 5 percent of leaks initially detected recur each month (0.5 x 12 = 0.6), that 10 percent of leaks
initially detected recur each quarter (.10 x 4 - 0.4), and that 20 percent of leaks initially detected recur annually
(.20 x 1 * 0.2).
-------
Table 8-7. MONITORING AND MAINTENANCE LABOR-HOUR REQUIREMENTS
FOR REGULATORY ALTERNATIVE IV
00
oo
Source
Typs
_
?'jr:ps
Valves
Gas
Liquid
Safety/
Relief
Valves
drains
MONITORING
Number of
Components
Per Model
Unit
A
34
87
4
5
B
100
264
11
15
C
167
439
19
25
Type of
Monitoring
Instrument
Visual
Instrument
Instrument
Instrument
Instrument
Monitoring
Time a
Per Person
(Minutes)
5
0,5
t
1
8
1
Times
Monitored
Per Year
0
52
12
12
0
12
Monitoring
Labor-Hours
Required
A
0
2.2
13.6
34.8
0
2.0
B
0
C.5
40.0
105.6
0
6.0
C
n
u
10.8
66.8
175.6
0 '
10.0
MAINTENANCE
Estimated
Number of
Leaksb
A
o
2
6
0
1
B
n
6
19
0
1
C
o
10
32
0
1
Repair
Time
(Hours)
80
1.13
1.13
0
4
Maintenance
Labor-Hours
Required
A
0
2
7
0
4
6
.0
7
21
0
4
C
0
11
36
0
4
Total Monitoring 52.6 158.1 263.2
Hours «
x $15.50/hour
Total Monitoring 320 - 2450 4080
Dollars =
Total Maintanence
Hours =
x $15.50/hour,
Total Maintenance
Dollars =
13
32 51
200 500 790
^Instrument monitoring requires a two-person team. Visual monitoring requires one person.
bRecurrence factors of 0.6, 0.4, and 0.2 have been applied for monthly, quarterly, and annual instrument inspections.
It is assumed that 5 percent of leaks initially detected recur each month (0.5 x 12 = 0.6), that 10 percent of leaks
initially detected recur each quarter (.10 x 4 = 0.4), and that 20 percent of leaks initially detected recur annually
(.20 x i = 0.2).
-------
Table 8-8 MONITORING AND MAINTENANCE LABOR-HOUR REQUIREMENTS
FOR REGULATORY ALTERNATIVE V
CO
Number of
Components
Per Model
Unit
Monitoring
Labor-Hours
Required
Total Monitoring 2.2 6.5 10.8 Total Malntanence
,~~ * !l5.50/hour
Total Maintenance
O 100 170 Dollars-. .
T 4. i
.onitorin, requires . two-person t«. Visual Storing requires one p.rson.
-------
Table 8-9. RECOVERED PRODUCT CREDITS
oo
ro
o
I
11
III
IV
V
VI
tory
ative
1
Model Unit A
VOC
Emissions
(Mg/Year)
29.3
13.7
9.2
8.0
d
4.6
0.0
Emission
Reduction
From Recovery
Uncontrolled Credit
(Mg/Year) (Mg/Year)
15.6 12.8
20.} 15.5
21.3 16.8
24.7 19.3
29.3
Recovered
Product
Value
($/Year)
4,740
5,740
6,220
7,140
Model Unit B
VOC
Emissions
(Mg/Year)
82.9
36.0
22.6
18.9
8.8
0.0
Emission
Reduction
From
Uncontrolled
(Mg/Year)
46.9
60.3
64.0
74.1
82.9
Recovery
Credit
(Mg/Year)
38. G
46.5
50.3
57.9
Recovered
Product
Value
($/Year)
14,200
17,200
18,600
21,400
Model Unit C
VOC
Emissions3
(Mg/Year)
129.7
50.8
28.0
21.7
5.0d
0.0
Emission
Reduction
From
Uncontrolled
(Mg/Year)
78.9
101.7
108.00
124.7
129.7
Recovery
Credit0
(Mg/Yef.r)
64.1
77.4
83.7
96.5
Recovered
Product
Value
($/Year)
23,700
28,700
31,000
35,700
aIncludes emissions from equipment handling less than 10 percent benzene.
bDoes not include emissions from safety/relief valves or drains, since these may not be recovered.
cThis value is obtained by multiplying the recovery credit in Mg per year by $37ii per Mg (second quarter, 1979 value for benzene, Ref. 9).
d"Residue" From uncontrolled equipment (less than 10 percent benzene).
-------
Alternative II) to $35,700 (Unit C, Regulatory Alternative V) per
year, the credits being a function of model unit size and the
regulatory alternative.
The anr.ualized 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 by a 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. Annualized control costs for the existing
units range from $8,100 (Model A, Regulatory Alternative II) to
$644,700 (Model C, Regulatory Alternative V). For new units, which do
not incur the high retrofitting costs of existing units, the range is
from $7,700 (Model A, Regulatory Alternative II) to $448,600 (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 18.6 Mg/year (A),
44.7 Mg/year (B), and 97.1 Mg/year (C). Under Regulatory Alternative II,
benzene emissions from Model Unit A are estimated to decrease 53
percent from the baseline emissions (from 18.6 Mg/year to 8.7 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
oo
i
no
Source Type
'Pumps
Valves
Gas
Liquid
Safety/rel ief
Valves
Drains
Percent of
Number of Components Sources
per Model Unit Leaking in
- im uiui
A B C Survey
5 15 25 23
34 100 167 it)
87 264 439 12
4 11 19 8
5 15 25 4
Estimated
Number of Leaks Repair
.... ,_ . . _ TT mp
ABC (Hours)
1 3 6 80
3 10 17 1.13
10 32 53 1.13
a a a o
111 4
Total Hours
X
TOTAL
Repai
A
80
3
11
0
4
98
-
r Labor
B
240
11
36
0
4
291
Hours
C
480
19
60
0
4
563
$15.50/hour
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)
CO
OJ
II
Source Type
. ^^=
Pumps
Valves
Gas
Liquid
Drains
-'i^_ - '" " ^~~~
1
Number of Percent
Components Sources
Per Unit Leaking in
_ ,. Initial
ABC Survey
~
5 15 25 23
34 100 167 * 10
87 ' 264 439 12
5 15 25 4
......
Number of Leaks Repair Repair Labor Hours
Time ~ ~~ ~
ABC (Hours) ABC
1 3 6 80 80 240 480
3 10 17 1.13 3 11 19
10 32 53 1.13 11 36 60
1114 --4 44
~
Total Hours = 98 291 563
x $15.50/hour
TOTAL = 1520 4510 8730
x CRF = 0.25 (10 year,
10 Percent)
Anm.alized5 = 380 1130 2180
-------
Table 8-10. INITIAL SURVEY START-UP COSTS
FOR REGULATORY ALTERNATIVE IV (Continued)
00
I
(X)
Number of Percent
Components . Sources
Per Unit Leaking in
T »*; »- ; -i i
initial
Source Type ABC Survey
Valves
i»
Gas 34 100 167 10
liquid 87 ' 264 439 12
Drains 5 15 25 4
Estimated
Number of Leaks
A B C
3 10 17
10 32 53
1 1 1
Repair
Time
(Hours)
1.13
1.13
4
Total Hours =
X
TOTAL
X
Annual ized =
Repair Labor Hours
A B C
3 11 19
11 36 60
-4 4 4
J18 11 .83
$15.50/hour
280 790 1290
CRF = 0.25 (10 year,
10 Percent)
70 200 320
-------
Table 8-10. INITIAL SURVEY START-UP COSTS (Concluded)
NOTES:
aLeaks are repaired by routine maintenance at no incremental increase
in manpower requirements. Safety/relief valves are normally reset
during routine maintenance without a leak detection and repair program.
Since 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 UNIT3
(Thousands of May 1979 Dollars)
Model Unit; A (Existing)
Cost I ten
Annual ized Capital Charges
1. Control Equipment
b
a. Instrument
b. Caps
c. Double Seals
Seals
Installationd
d. Seal Oil System
e. Vents-Pumps and Compressors
f. Replacement Pumpr
g. Rupture Disks'
« Disks
Installation
h. Closed-Loop Sampling^
i. Sealed-Bellow.5 Valves
j. Hard Piping9'1
2. Initial Leak Repair1
!I III IV
? . 0 2.0 2.0
(i '. 3 0.3 0.3
1 .7
0.2
1.2
T d
J . H
0.3 0.3
15 1.5
0.7 0.7
0.4 0.4 0.1
V
0.0
0.3
1 .7
0.2
1.2
3.4
o'.s
0.3
1 .5
0.7
72.9
0.8
0.0
Operating Costs
1. Maintenance Charges
a. Instrument
b. Caps
c. Double Seals
d. Seal Oil System
e. Vents-Pumps and Compressors
f. Replacement Pumps
g. Rupture Disks
h Closed-Loopi*Sampling
i. Sealed-Bellows Valves
2.
3.
Total
j. Hard Piping
Miscellaneous (Taxes, Insurance,
Administration)
a. Instrument
b. Caps
c. Double Seals
d. Seal Oil System
e. Vents-Pumps and Compressors
f. Replacement Pumps
g. Rupture Disks
h. Closed-Loop Sampling
i. Sealed-Bellows Valves
j. Hard Piping
Labor
a. Leak Detection Labor
b. Repair Labor
c. Administrative and Support
Before Credit
2.7 2.7
0.1 0.1
0.4
0.2
0.3 0.3
01 n i
.1 U.I
0.5
0.2
0.2 1.0
14 1 .4
0.6 1.0
0.0 8.1 13.1
2.7
0.1
0.2
0.3
0.8
0.4
0.2
0.3
0.1
n ?
u . c
n A
U .H
i n
1 U
0.5
0.2
0.8
0.2
0.4
20.2
0.0
0.1
0.2
0.3
0.8
0.1
0.4
0.2
17.9
0.2
0.0
0.1
0.2
0.4
1 0
0.2
o-'.s
0.2
99 a
L £ . H
0.3
0 0
0.0
0.0
129.0
8-26
-------
Table 8-11. ANNUALIZED CONTROL COST
ESTIMATES PER MODEL UNITa (Continued)
(Thousands of May 1979 Dollars)
Model Unit: A (New)
Cost Item
II III IV
Annualized Capital Charges
1. Control Equipment
a. Instrument" 2.0 2.0 2.0 0.0
b. Capst 0.3 0.3 0.3 0.3
c. Double Seals
Seals 1-0 1.0
Installation, 0.2 0.2
d. Seal Oil System 1-2 1.2
e. Vents-Pumps and Compressors 3.4 3.4
f. Rupture Disksf
Disks 0.3 0.3 0.3
Installation 0.8 0.8 0.8
g. Closed-Loop Sampling9, 0.7 0.7 0.7
h. Sealed-Bellows Valves" "9.3
i. Hard Piping9>' °-°
2. Initial Leak RepairJ 0.0 0.0 0.0
Operating Costs
1. Maintenance Charges
2.
a.
b.
c.
d.
e.
f.
g.
h.
i.
Instrument
Caps
Double Seals
Seal Oil System
Vents-Pumps and Compressors
Rupture Disks
Closed-Loop Sampling
Sealed-Bellows Valves
Hard Piping*
2.7 2.
0.1 0.
0
0
,7
,1
.2
.2
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
Miscellaneous (Taxes, Insurance,
Administration)
3.
Total
a.
b.
c.
d.
e.
f.
g.
h.
i .
Instrument
Caps
Double Seals
Seal Oil System
Vents-Pumps and Compressors
Rupture Disks
Closed-Loop Sampling
Sealed-Bellows Valves
Hard Piping
0.3 0
0.1 0
0
0
.3
.1
.3
.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
Labor
a.
b.
c.
Leak Detection Labor
Repair Labor
Administrative and Support
Before Credit
0.2 1
1.4 1
0.6 1
0.0 7.7 11
.0
.4
.0
.6
0.8
0.2
0.4
18.2
0.0
0.0
0.0
89.6
8-27
-------
Table 8-11. ANNUALIZED CONTROL COST
ESTIMATES PER MODEL UNIT9 (Continued)
(Thousands of May 1979 Dollars)
Model Unit- B (Existing)
Cost HPm I 'I I!I
Annual ized Capital Charges
1. Control Equipment
a. instrument5 2-« 2-°
b. Caps' °-8 °'8
c. Double Seals
« Seals
Installation.
d. Seal Oil System
e. Vents-Pumps and Compressors
f. Replacement Pumps6
g. P-upture Disks^
« Disks '''
» Installation *-3
h. Closed-Loop Samplinguh 2-'
i. Sealed-Bel lovts Valves
j. Hard Piping^'1
2. Initial Leak Repair^ 1-1 ] -1
Operating Costs
1. Maintenance Charges
? 7 27
a. "nstrument <-.i .
b. Caps °-2 °'2
c. Double Seals
d. Seal Oil System
e. Vents-Pumps and Compressors
f. Replacement Pumps
g. Rupture Disks '£
h. Closed-Loop Sampling u-b
i. Sealed-Belrtws Valves
j. Hard Piping
2. Miscellaneous (Taxes, Insurance,
Administration)
n i n i
a. [nstrument Jr'^ J:'^
b. Caps °'J
c. Double Seals
d. Seal Oil System
e. Vents-Pumps and Compressors
f. 'Replacement Pumps
g. Rupture Disks '*
h. Closed-Loop Sampling u-°
i. Sealed-Bellows Valves
j. Hard Piping
3. Labor
a. Leak Detection Labor 0-7 2>9
b. Repair Labor 'b ^.u
c. Mministrative and Support °-y '-4
Total Before CredU °-° 10'5 26'9
IV
2.0
0.8
5.0
0. 7
3.7
9.9
i i
i . i
4.3
2.1
0.2
2.7
0.2
0.5
On
. y
0 C
L . J
1 ?
1 . C
0.5
0.3
0^3
07
. /
1 ?
1 . i-
2 -j
1 .4
0.6
2.5
0.5
K2
50.1
V
0.0
0.8
5.0
0.7
3.7
9Q
.y
i n
1 .U
1 . 1
4.3
2.1
219.5
0.0
0.4
0.2
0.5
n Q
u . y
o c
L. , J
n ?
U . L
1 2
0.5
c-i q
jj . y
0.6
0.0
0.3
0 7
1 2
3 1
0. 3
1 .4
0.6
67.3
0.8
0.0
0.0
o!o
386.7
8-28
-------
Table 8-11. ANNUALIZED CONTROL COST
ESTIMATES PER MODEL UNITa (Continued)
(Thousands of May 1979 Dollars)
Model Unit. B (New)
Cost Item I " " IV V
Annualized Capital Charges
1. Control Equipment
. . ..b 2 0 2 0 2.0 0.0
a. Instrument Q_& Q_8 _
b. Laps
c. Double Seals , ^ Q
Seals ,,', n'c
Installation, "? "'°
d. Seal Oil System £' ,
e. Vents-Pumps and Compressors ''
f. Rupture Disksf , , , , ,
Disks I'} \'\ \'\
« Installation <* '* '*
a. Closed-Loop Samplingyh '' f'' '
h. Sealed-Bellows Valves" |q°'^
i. Hard Piping9.' ^-*
2. Initial Leak RepairJ °-° °-° °'° °'°
Operating Costs
1. Maintenance Charges
2.
3.
Total
a. Instrument
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 '
Miscellaneous (Taxes, Insurance,
Administration)
a. Instrument
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
Labor
a. Leak Detection Labor
b. Repair Labor
c. Administrative and Support
Before Credit
2.7 2.7
0.2 0.2
0.6
0.5
0.3 0.3
0.3 0.3
0.8
0.6
0.7 2.9
1.5 3.0
0.9 2.4
0.0 9.4 22.7
2.7
0.2
0.4
0.9
2.5
0.6
0.5
0.3
0.3
n d
U , H
1 9
1 . i.
\ 1
J . 1
00
. O
0.6
2.5
0.5
1.2
44.3
0.0
0.2
0.4
0.9
2.5
0.6
0.5
36.4
0.6
0.0
0.3
0 4
1 2
\ i
j . i
0 8
o'.6
£C C
HO . D
0.8
0.0
0.0
0.0
269.1
8-29
-------
Table 8-11. ANNUALIZED CONTROL COST
ESTIMATES PER MODEL UNIT3 (Continued)
(Thousands of May 1979 Dollars)
Model
Unit: C (Existing)
Cost Item I
II III
IV
V
Annual i zed Capital Charges
1.
2.
Control Equipment
a. Instrument
b. Caps
c. Double Seals
Seals
Installation
d. Seal Oil Systemd
e. Vents-Pumps and Compressors
f. Replacement Pumps6
g. Rupture Disksf
Disks
Installation
h. Closed-Loop Sampling3,
i. Sealed-BelloKS Valves"
j. Hard Piping?-1
Initial Leak Repair-1
2 0 2.0
14 1.4
1.9
7.8
3.4
2.2 2.2
2.0
1 .4
8.3
1.2
6.2
16.7
1 .9
7.8
3.4
0.3
0.0
1.4
8.3
1.2
6.2
16.7
1 .4
1 .9
7.8
3.4
365.5
4.0
0.0
Operating Costs
1.
2.
3.
Total
Maintenance Charges
a. Instrument
b. Caps
c. Double Seals
d. Seal Oil System
e. Vents-Pumps and Compressors
f. Replacement Pumps
g. Rupture Disks,
n. Closed-Loop Sampling
i. Sea ied-Bel lows Valves
j. Hard Piping
Miscellaneous (Taxes, Insurance,
Administration)
a. Instrument
b. Caps
c. Double Seals
d. Seal Oil System
e. Vents-Pumps and Compressors
f. Replacement Pumps
g. Rupture Disks
h. Closed-Loop Sampling
i. Sealed-Bellows Valves
j. Hard Piping
Labor
a. Leak Detection Labor
b. Repair Labor
c. Administrative and Support
Before Credit
2.7 2.7
0.4 0.4
2.0
0.8
0.3 0.3
0.4 0.4
2.5
1 .1
1.1 4.9
1 5 5.8
1 1 4.3
13 2 43.9
2.7
0.4
0.9
1.6
4.1
2.0
0.8
0.3
0.4
1 .1
1.9
5.1
2.5
1 .1
4.1
0.8
2.0
81 .0
0.0
0.4
0.9
1 .6
4.1
0.4
2.0
0.8
89.7
1.0
0.0
0.4
1 .1
1 .9
5.1
0.5-
2. 5
1 .1
112.1
1 .3
0.0
0.0
0.0
644.7
8-30
-------
Table 8-11. ANNUALIZED CONTROL COST
ESTIMATES PER MODEL UNITa (Concluded)
(Thousands of May 1979 Dollars)
Model Unit: C (New)
Cost Item I
Annualized Capital Charges
1. Control Equipment
a. Instrument
b. Caps
c. Double Seals
Seals
Installationd
d. Seal Oil System
e. Vents-Pumps and Compressors
f. Rupture Disks'
2.
Disks
Installation
g. Closed-Loop Samplingsh
h. Sealed-BelloKS Valves
i. Hard Piping^'1
Initial Leak RepairJ
II III
2.0 2.0
1.4 1.4
1 ."
4.
3.
0.0 0.
y
2
,4
0
IV
2.0
1.4
5.0
1.0
6.2
16.7
1 ."
4.
3.
0.
a
2
4
0
V
0.0
1.4
5.0
1.0
6.2
16.7
1 Q
1 .9
4.2
3.4
246.9
4.0
0.0
Operating Costs
1.
2.
3.
Total
Maintenance Charges
a. Instrument
b. Caps
c. Double Seals
d. Seal Oil System
e. Vents-Pumps and Compressors
f. Rupture Disks
g. Closed-Loop Sampling
h. Sealed-Bel Vows Valves
i. Hard Piping
Miscellaneous (Taxes, Insurance,
Administration)
a. Instrument
b. Caps
c. Double Seals
d. Seal Oil System
e. Vents-Punps and Compressors
f. Rupture Disks
g. Closed-Loop Samoling
h. Sealed-Bellows Valves
i. Hard Piping
Labor
a. Leak Detection Labor
b. Repair Labor
c. Administrative and Support
Before Credit
2.7 2.
0.4 0.
1.
0
0.3 0
0.4 0
1
1
1,1 4
1.6 5
1 .1 4
11.0 36
7
,4
.1
.8
.3
.4
.4
.1
.9
.8
.3
.1
2.
0.
0.
1.
4.
1 .
0.
0
0
0
1
s
1
1
4
0
2
70
7
4
6
6
,1
,1
.8
.3
.4
.7
.9
.1
.4
.1
.1
.8
.0
.9
0.0
0.4
0.6
1 .6
4.1
1 .1
0.8
60.6
1 .0
0.0
0.4
0.7
1 .9
5.1
1 .4
1.1
75.8
1 .3
0.0
0.0
0.0
448.6
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
iv. CRF = 0.23
v. Miscellaneous = 0.04
b. Materials and Maintenance
i. Cost = $2,700 Per Year
2. Control Equipment
4
a. Capital Charges
b. Operating Life = 10 Years (2 yrs. for double seal & rupture di:
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
3. Administration and Support = 40 Percent of Leak
Detection and Repair Labor Cost
Cost is for Century Systems Corporation's Organic Vapor Analyzer
(Model OVA-108).
cUsed to seal open-ended valves.
Used as auxilliary for double seal.
"8-32
-------
Table 8-11. ANNUALIZED COST ESTIMATES PER MODEL UNIT (Concluded)
NOTES: (Concluded)
eln retrofitting double seals, it has 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 safety/
relief valve,
9Cost is for new or retrofitted installation.
hUsed for control valves.
Consists of sealed cover on drain with line leading to pump.
ere equipment standards are applied, as in Regulatory Alternatives
IV and V, the amount of .leak detection and repair labor decreases.
kBased on an average price of $370/Mg (Ref. 9).
lumbers in parentheses represent savings (net credit).
8-33
-------
Table 8-12. BENZENE EMISSION REDUCTIONS
oo
i
CO
-Pi.
Regulatory
Alternative
I
11
III
IV
V
VI
Model Unit
VOC Benzene
Emission Emission
(Hg/Year) (Mg/Year)
29.3
13.7
9.2
8.0
4.6
0.0
18.6
8.7
5.9
5.1
2.9
0.0
A
Benzene
Emission
Reduction0
(Mg/Year)
9.9
12.7
13.5
15.7
18.6
Model Unit
VOC Benzene
Emission3 Emission
(Mg/Year) (Mg/Year)
82.9
36.0
22.6
18.9
8.8
0.0
44.7
19.3
12.1
10.1
4.6
0.0
B
Benzene
Emission.
Reduction
{Mg/Year)
25.4
32.6
34.6
40.1
44.7
voc a
Emission
(Mg/Year)
129.7
50.8
28.0
21.7
5.0
0.0
Model Unit
Benzenea
Emission
(Mg/Year)
97.1
38.2
21.2
16.5
4.0
0.0
C
Benzene
Emission b
Reduction
(Mg/Year)
58.9
75.9
80.6
93.1
97.1
Calculations of VOC and benzene emissions for each model unit and regulatory alternative are presented in EPA
Docket Number A-79-27-II-B.
Benzene emission
reductions are calculated for each model unit as the difference between the benzene emissions for
D C 11 i. C I IC CIIIIOOIIMI I V. «J U \* w I w I I -» *>. w. . ^ / T T \
the baseline alternative (I) and each of the other alternatives (II-VI).
-------
Table 8-13. COST-EFFECTIVENESS FOR EXISTING MODEL UNITS
co
Regulatory Alternative
i ii'
Total Capital Cost ($1.000)"
Total Annualized Cost ($1,000)"
,c,d
Total Annual Credit ($1.000)
Net Annualized Cost ($1.000) '
Total Benzene Reduction
(Kg/Year)
Cost-Effectiveness9(Net Annualized
Sl,000/Mg Benzene)
aFrom Table 8-4.
I II III IV V
.
0.0 10.3 24.3 56.9 504.1
0.0
8.1 13.1 20.2 129.0
0.0 (4.7) (5.7) (6.2) (7.1)
3.4 7.4 14.0 121.9
9.9 12.7 13.5 15.7
0.34 0.58 1.0 7.8
each model unit and regulatory alternative.
0.0 13.8 55.4 152.9 1512.2
0.0 10.5 26.9 50.1 386.7
, '
0.0 (14.2) (17.2) (18.6) (21.4)
0.0 (3.7) 9.7 31.5 366.3
_
25.4 32.6 34.6 40.1
0.0 0.30 0.91 9.1
I 11 111 IV V
0.0 17.3 90.1 252.5 2520.5
0.0 13.2 43.9 81.0 644.7
-
0.0 (23.7) (28.7) (31.0) (35.7)
(10.5) 15.2 50.0 609.0
*-
58.9 75.9 80.6
0.0 0.20 0.62 6.5
f°r
-------
Table 8-14. COST-EFFECTIVENESS FOR NEW MODEL UNITS
o =~
Regulatory Alternative
Total Capital Cost ($l,000)a
Total Annualized Cost ($l,000)b
Total Annual Credit (Sl,000)c'd
d,e
Net Annualized Cost ($1,000)
Total Benzene Reduction
(Mg/year)
Cost-Effectiveness9(Net Am.ualized
Sl.OOO/Mg Benzene)
Model Umt
I II
0.0 10.3
0.0 7.7
0.0 (4.7)
0.0 3.0
9.9
0.30
III
20.0
11.6
(5.7)
5.9
12.7
0.46
A
IV
51.2
18.2
(6.2)
12.0
13.5
0.89
V
350.2
89.6
(7.1),
82.5
15.7
5.3
Model Unit
I II III
0.0 13.8 42.5
0.0 9.4 22.7
0.0 (14.2) (17.2)
0.0 (4.8) 5.5
25.4 32.6
0.0 0.17
B
IV
135.8
44.3
(18.6)
25.7
34.6
0.74
V
1052,3
269.1
(21.4)
247.7
40.1
6.2
Model Unit
I II
0.0 17.3
0.0 11.0
0.0 (23.7)
0.0 (12.7)
58.9
0.0
III
67.2
36.1
(28.7)
7.4
75.9
0.10
C
IV
222.6
70.9
(31.0)
39.9
80.6
0.50
V
1754.1
448.6
(35.7)
412. y
93.1
4.4
co
i
CO
CTi
dFrom Table 8-4.
bhrom lable 8-11.
cFrom Table 8-9 - Recovered product credit is based on VOC emission reductions from the baseline Alternative I.
lumbers in parentheses represent savings. n,Mo n Ql fnr
"Net annual,zed cost is calculated as the difference between the total armualized cost (Table 8-11) and the recovered product credit (Table 8-9) for
each model unit and regulatory alternative.
by'dividing the net annuitized cost ($1000) by the total benzene reduction for each model unit and regulatory alternative
-------
Units B and C, respectively, are estimated. For Regulatory
Alternative III, emissions are expected to decrease 68, 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 B and C, Regulatory
Alternative II) to $9,100/Mg of benzene (Model B, Regulatory
Alternative V). For new units, the cost-effectiveness ranges from no
cost (Models B and C, Regulatory Alternative II) to $6,200/Mg benzene
(Model B, Regulatory Alternative V).
8.4 TOTAL INDUSTRY IMPACTS
The total nationwide estimated costs, recovered product credits,
and benzene fugitive emissions for the refining and organic chemical
industries are presented in Table 8-15 for existing units and in
Table 8-16 for new units.
8.4.1 Existing Units
For existing sources, which are estimated at 241 production
units, economic impacts range from a capital cost of $2.9 million for
Regulatory Alternative II to a capital cost of $242 million for Regulatory
Alternative V, based on second quarter 1979 dollars. These costs
result in annualized costs ranging from a net savings of $25,000 for
Regulatory Alternative II to an annual cost of $58.6 million for
Regulatory Alternative V. The alternatives would result in nationwide
benzene fugitive emissions ranging from 3,600 Mg/yr for Regulatory
Alternative II to 900 Mg/yr for Regulatory Alternative V. These rates
compare with an uncontrolled rate of 8,300 Mg/yr for Regulatory
Alternative I. The resulting cost-effectiveness of the regulatory
alternatives ranges from a savings of $10/Mg for Regulatory Alternative II
to a cost of $7,900/Mg for Regulatory Alternative V.
8.4.2 New Units
The costs and estimates in Table 8-16 represent fifth-year estimates
based on the number of new model units predicted to be operating in
1985. Fifth-year impacts for an estimated 68 production units ranged
8-37
-------
Table 8-15. NATIONWIDE COSTS FOR THE
EXISTING INDUSTRY
(May 1979 Dollars)
REGULATORY ALTERNATIVE
II III IV
Total Capital Cost 0.0 2.9 9.7 25.3 242
($ Million)
Total Annualized Cost 0.0 2.2 4.9 8.5 62.0
($ Million)
Total Annual Credit 0.0 (2.3)b (2.8)b (3.0)b (3.4)b
($ Million)
Net Annualized Cost 0.0 (0.03)b 2.1 5.5 58.6
($ Million)
Total Benzene 8.3 3.6 2.2 1.9 0.9
Fugitive Emissions
(1000 Mg/yr)
Total Benzene Fugitive - 4.7 6.0 6.4 7.4
Emission Reduction
(1000 Mg/yr)
Cost-Effectiveness 0.0 0.4 0.9 7.9
Net Annual Cost, $1000/
Mg Benzene Reduced
^Calculated by multiplying cost and emission estimates for each model unit
and regulatory alternative (from Table 8-13) by the following numbers of
existing model units:
Model Number of Existing
Unit Units in 1980
A 145
B 72
C 24
Numbers in parentheses represent savings.
cFrom Table 8-12, calculated by multiplying emissions for each model unit
and regulatory alternative by the number of existing model units in foot-
note a.
8-38
-------
Table 8-16. NATIONWIDE COSTS FOR
NEW UNITS (FIFTH YEAR IMPACT)3
Total Capital Cost
($ Million)
Total Annual! zed Cost
($ Million)
Total Annual Credit
($ Million)
Net Annuall zed . Cost
($ Million)
Trt-hal RpnTPnP
REGULATORY
I II
0.0 0.8
0.0 0.6
0.0 (0.7)b
0.0 (0.07)b
2.5 1.1
ALTERNATIVE
III
2.2
1.2
(0.8)b
0.4
0.7
IV V ',
6.5 48.4
2.2 12.4,
(0.9)b (1.0)*
3
1.3 11.4,
0.5 0.2 .
I \j t*CL I wCIl«»wll%» A
Fugitive Emissions
(1000 Mg/yr) *
Total Benzene Fugitive 1-5 ]'9 2>0 ' J
Emission Reduction T^
(1000 Mg/yr) *.
Cost-Effectiveness °-° °'2 °'7 5> ;:
(Net Annual Cost, $1000/
Mg Benzene Reduced
Calculated by multiplying cost and emission estimates for each model unit
and regulatory alternat^e (from Table 8-14) by the following numbers of
new model units predicted to be in operation in 1985:
,.,««=, Number °f Ne*
Unit Units in 1985
7~ 47
B 7
C
b
'Numbers in parentheses represent savings.
cFrom Table 8-12, calculated by multiplying emissions for each model unit
and regulatory alternative by the number of new model units in footnote
8-39
-------
from a capital cost of $820 thousand for Regulatory Alternative II
to a capital cost of $48.4 million for Regulatory Alternative V, based
on second quarter 1979 dollars. These costs would result in annualized
costs ranging from a net savings of $70,000 for Regulatory Alternative II
to an annual cost of $11.4 million for Regulatory Alternative V. The
alternatives would result in nationwide benzene fugitive emissions
ranging from 1,100 Mg/yr for Regulatory Alternative II to 200 Mg/yr
for Regulatory Alternative V. These rates compare with an uncontrolled
rate of 2,500 Mg/yr for Regulatory Alternative I. The resulting
cost-effectiveness for new sources ranges from a savings of $50/Mg for
Regulatory Alternative II to a cost of $5,000/Mg for Regulatory Alternative V.
8.5 COST COMPARISON
The costs of controlling benzene fugitive emissions from the
refining and organic chemical industries 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. " Table 8-17 summarizes capital and
annual control cost data for benzene fugitive emissions from the refining
and organic chemical industries, the maleic anhydride and ethylbenzene-
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 (refining and
organic chemical industries) 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 8-18 and 8-19 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 ethylbenzene-styrene
8-40
-------
Table 8-17. RANGE OF CONTROL COSTS FOR THE BENZENE SOURCE
CATEGORIES FOR EXISTING AND NEW UNITS
(Thousands of Dollars)
Total Capital Cost Total Annual Cost3
Source Category Existing New Existing New
Refinery-organic chemical xo-2502 10-1754 (11)-609 (13)-413
(Benzene Fugitives)
Maleic Anhydride 1160-1440 354-600
(Process Emissions)
Ethylbenzene-Styrene 268-555 (150)-45
(Process Emissions)
12
Benzene Storage 7-2BQ 6-290 0.2-68 0-69
(Producers)
ip u
Benzene Storage 16-249 15-249 2-68 (0.8)-6r
(Consumers) ^
alncludes recovered credits.
bNumbers in parentheses represent savings.
8-41
-------
Table 8-18. COSTS FOR THE CONTROL OF TOTAL BENZENE EMISSIONS
FROM THE MAELIC ANHYDRIDE INDUSTRY1U '^ »a
(Thousands of Dollars)
Cost Item Existing New
Capital Cost
Annual Costb
1193-3018
359-725
31-1241
4-54
alncludes control costs for benzene fugitive emissions
from refineries and chemical plants.
Includes recovered credits.
Table 8-19. COSTS FOR THE CONTROL OF TOTAL BENZENE EMISSIONS
FROM THE ETHYLBENZENE-STYRENE INDUSTRY11'^'a
(Thousands of Dollars)
Cost Item
Capital Cost
Annual Costb
Existing
301-2133
New
31-1241
4-54
alncludes control costs for benzene fugitive emissions
from refineries and chemical plants.
Includes recovered credits.
cNumber in parentheses represents savings.
8-42
-------
industries, respectively. Total control cost data are given in
Table 8-20 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 8-21.
8-43
-------
Table 8-20. TOTAL COSTS FOR THE CONTROL OF BENZENE EMISSIONS.
STORAGE TANKS AND BENZI
(Thousands of Dollars)
FROM PRODUCER BENZENE STORAGE TANKS AND BENZENE FUGITIVE SOURCES12'3
Cost Item
Capital Cost
Annual Costb
Existing
17-2792
00-677
New
16-2044
(13)-482
alncludes control costs for benzene fugitive emissions
from refineries and chemical plants.
Includes recovered credits.
cNumbers in parentheses represent savings.
Table 8-21. TOTAL COSTS FOR THE CONTROL OF BENZENE EMISSIONS FROM
CONSUMER BENZENE STORAGE TANKS AND BENZENE FUGITIVE SOURCES12'9
4
(Thousands of Dollars)
Cost Item Existing New
Capital Cost 26-2751 25-2003
Annual Cost5 (9)-677 (14)-475
alncludes control costs for benzene fugitive emissions
from refineries and chemical plants.
Includes recovered credits.
cNumbersin parentheses represent savings.
8-44
-------
8.6 REFERENCES
1. Letter from Amey, G.C., Century Systems Corporation, to Seme,
J.C., 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). Chemical Engineering.
86(9):7. April 23, 1979.
4. Economic Indicators (for April and May 1979). Chemical Engineering.
86(16):7. July 30, 1979.
5. Telecon. Mclnnis, J.R., PES, Incorporated, with Hetrick, C.,
Crane Chempum Division, Warrington, PA. August 24, 1979.
6. PES estimate.
7. Letter from Crutchfield, B., Duriron Company, Incorporated, to
Mclnnis, J.R., PES, Incorporated. August 31, 1979. Cost data
for replacement pump.
8. Letter from Johnson, J., Exxon Company, to Walsh, R.T., 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. Chemical
Marketing Reporter. 215J21):43. May 21, 1979.
10. 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.
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.
12. Letter and attached cost tables from Ailor, D.C., Energy Systems
Group of TRW, Incorporated, to Markwordt, D.W., EPA, CPB. July 24,
1979.
8-45
-------
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-Q36. June 1978. p. 4-6.
8-46
-------
9.0 ECONOMIC IMPACT
9.1 INDUSTRY CHARACTERIZATION
9.1.1 General Profile
Throughout the United States there are 74 petroleum refining and
organic chemical companies operating 131 plant sites that manufacture
iic 32-33
benzene and derivatives of benzene. " ' " 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 of 1 or more units, 14
existing 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., Sulfolane, UDEX). Benzene is
also produced as an impure by-product in the manufacture of ethylene,
which is produced by 28 companies at 46 sites and 58 units. There are
50 companies that manufacture benzene derivatives at 76 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 refining and
organic chemical production units was estimated to be 5260 gigagrams
(Gg).16'26 Total ethylene production was higher at 10,600 Gg, while
total production 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 ORGANIC
CHEMICAL MANUFACTURING SITES , 1t- - ,,
WITH BENZENE FUGITIVE EMISSION POTENTIAL1""15'^""'3'3
Benzene-Related
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
Plant
Allied Chemical
Allied Chemical
American Cyanamid
American Cyanamid
Amerada Hess
American Hoechst
American Hoechst
American Petrofina
(of Texas)
American Petrofina
(Cosden Oil)
American Petrofina
(Cosden Oil/Petrogas)
American Petrofina/
Union Oil of CA
Ashland Oil
Ashland Oil
Ashland Oil
Atlantic Richfield
Atlantic Richfield
City/State
Geismar, LA
Moundsville, WV
Bound Brook, NJ
Willow Island, WV
St. Croix, VI
Baton Rouge, LA
Bayport, TX
Port Arthur, TX
Big Spring, TX
Groves, TX
Beaumont, TX
Ashland, KY
Neal, WV
North Tonawanda, NY
Beaver Valley, PA
(Kobuta)
Channel view, TX
Products
At Site
Et
NiBz
NiBz
NiBzc
Bz
EtBz
St
EtBzd
std
Bz
Bz
Cyx
EtBz6
St
Et
Bz
Cyx
Bz
Cu
Cyx
MAN
Bz
St
BzC
Et (2 units)
K
Capacity
(Gg/yr)
340
25
48
34
217
526
ND9
469
409
67
194
35
20
41
9
73
88
214
181
ND9
27
77
200
107
1179
9-2
-------
Table 9-1. REFINERIES AND ORGANIC
CHEMICAL MANUFACTURING SITES
WITH BENZENE FUGITIVE EMISSION POTENTIAL (CONTINUED)
Benzene-Related h
Plant
17. Atlantic Richfield
18. Atlantic Richfield
(ARCO/Polymers)
19. Atlantic Richfield
(ARCO/Polymers)
20. Charter
International
21. Chemetics International
22. Chemplex
23. Cities Service
24. Clark Oil
25. Coastal States Gas
26. Commonwealth Oil
27. Continental Oil
28. Continental Oil
29. Core-Lube
30. Corpus Christi
Petrochemicals
31. Cos-Mar, Inc.
32. Crown Central
33. Denka (Petrotex)
34. Dow Chemical
City/State
Wilmington, CA
Houston, TX
Port Arthur, TX
Houston, TX
Geismar, LA
Clinton, 10
Lake Charles, LA
Blue Island, IL
Corpus Christi , TX
Penuelas, PR
Baltimore, MD
Lake Charles, LA
Danville, IL
Corpus Christi , TX
Carrville, LA
Pasadena, TX
Houston, TX
Bay City, MI
Products
At Site
Bz
Et
Bzc
Et
EtBz
St
EtBz
Bz
EtBz
NiBz
Et
Bz
Et (2 units)
Cu
Bz
Cue
Bz
Cyx
EtBz6
LAB
Et
BSA
Bzd
j
Etd
EtBz
St
Bz
MAN
Bz
Et
Capacity
(Gq/yr)
40
45
140
227
61
54
114
17
i ^
16
173
227
83
400
50
234
64
618
73
122
302
ND9
100
544
690
r~ f\ f\
590
77
23
100
86
9-3
-------
Table 9-1. REFINERIES AND ORGANIC
CHEMICAL MANUFACTURING SITES
WITH BENZENE FUGITIVE EMISSION POTENTIAL (CONTINUED)
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
Plant
Dow Chemical
Dow Chemical
Dow Chemical
Dow Chemical
Dupont
Dupont
Dupont
Eastman Kodak
El Paso Natural Gas
El Paso Products/
Rexene Polyolefins
Exxon
Exxon
First Chemical
Georgia-Pacific
Getty Oil
City/State
Freeport, TX
Midland, MI
Orange, TX
Plaquemine, LA
Beaumont, TX
Gibbstown, NJ
Orange, TX
Longview, TX
Odessa, TX
Odessa, TX
Baton Rouge, LA
Bay town, TX
Pascagoula, MS
Houston, TX
Delaware City, DE
Benzene-Related
Products
At Site3
Bz
Et (5 units)
EtBz
St
ClBz
EtBz6
St
Et
Bzd
Et (2 units)
NiBz
NiBz
Et
Et
Et
EtBz
St
Et
stc
Bz
Et
EtBz
St
Bz
Cyx
Etc
NiBz
Cu
Bz
h
Capacity
(Gg/yr)
167
1136
794
658
129
249
181
375
200
545
159
110
374
580
NDg
125
68
236
47
234
816
NDg
NDg
200
147
36
152
340
37
9-4
-------
Table 9-1. REFINERIES AND ORGANIC
CHEMICAL MANUFACTURING SITES
WITH BENZENE FUGITIVE EMISSION POTENTIAL (CONTINUED)
__
~ Benzene-Related
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63,
64,
65
66
67
68
/-r\
Plant
Getty Oil
B.F. Goodrich
Goodyear Tire & Rubber
Gulf Coast Olefins
Gulf Oil
Gulf Oil
Gulf Oil
Gulf Oil Chemicals
Gulf Oil Chemicals
Hercules
Howe! 1
ICC Industries
, Independent Refining
Corp.
. Jim Walter Resources
. Kerr-McGee Corp.
. Koppers
. Koppers
. Koppers
. Marathon Oil
M n W -i t < r*Ui /^m T r* a 1
Ci tv/State
El Dorado, KA
Calvert City, KY
Bayport, TX
Taft, LA
Alliance, LA
Donaldsonville, LA
Philadelphia, PA
Cedar Bayou, TX
Port Arthur, TX
McGregor, TX
San Antonio, TX
Niagara Falls, NY
Winnie, TX
Birmingham, AL
Corpus Christi, TX
Bridgeville, PA
Cicero, IL
Petrol ia, PA
Texas City, TX
New Martinsville, WV
Products
At Site
Bz
Cu
Et
Hqn
r
Et
Bz
EtBz
St
Bz
Cu
Et (2 units)
Bzc
Cu
Cyx
Et (2 units)
ClBzf
Bz
ClBz
Bz
BSA
Bz
MAN
MAN
Rcnol
Bz
Cue
NiBz
. b
Capacity
(Gq/yr)
43
61
136
5
O 1 O
218
224
313
272
124
209
719
134
204
106
558
0.05
NO9
11
10
Q
NDy
53
15
1 r
16
O ">
23
95
61
9-5
-------
Table 9-1. REFINERIES AND ORGANIC
CHEMICAL MANUFACTURING SITES
WITH BENZENE FUGITIVE EMISSION POTENTIAL (CONTINUED)
70.
71.
72.
73.
74.
75.
76.
77.
78.
79.
80.
81.
82.
83.
84.
Plant
Mobil Oil
Monsanto
Monsanto
Monsanto
Monsanto
Montrose Chemical
National Distillers
(U.S.I.)
Nease Chemical
Northern Petrochemical
01 in Corporation
Oxirane
Pennzoil (Atlas)
Phillips Petroleum
Phillips Petroleum
Phillips Petroleum
City/State
Beaumont, TX
Alvin, TX
(Chocolate Bayou)
Sauget, IL
St. Louis, MO
Texas City, TX
Henderson, NV
Tuscola, IL
State College, PA
Morris, IL
Brandenburg, KY
Channelview, TX
Shreveport, LA
Borger, TX
Pasadena, TX
Sweeny, TX
Benzene-Related
Products
At Site
Bz
Et
Cur
Etc
EtBz
LAB
ClBz
NiBz
MAN
Bz
Et
EtBz
St
ClBz
Et
BSAe
Et
Et
EtBz
St
Bzc
Cyx
EtBz
Et
Bz
Cyx
Et (3 units)
Capacity
(Gq/yr)
200
410
340
285
27
102
80
5
48
284
45
744
680
32
181
NDg
400
50
525
454
49
104
ND9
13
33
250
973
9-6
-------
Table 9-1. REFINERIES AND ORGANIC
CHEMICAL MANUFACTURING SITES
WITH BENZENE FUGITIVE EMISSION POTENTIAL (CONTINUED)
Benzene-Related K
Plant
85. Phillips Puerto Rico
86. Puerto Rico Olefins
87. PPG
88. PPG
89. Quintana-Howell
90. Reichhold Chemicals
91. Reichhold Chemicals
92. Reichhold Chemicals
93. Rubicon
94. Shell Chemical
95. Shell Oil
96. Shell Chemical
97. Shell Oil
98. Shell Oil
99. Specialty Organics
100. Standard Chlorine
101. Standard Chlorine
102. Standard Oil (CA)/
Chevron Chemical
103. Standard Oil (CA)
Chevron
104. Standard Oil (CA)
Chevron
105. Standard Oil (IN)/
Amoco
City/State
Guayama, PR
Penuelas, PR
Natrium, WV
New Marti nsvi lie, WV
Corpus Christi , TX
Elizabeth, NJ
Morris, IL
Tuscaloosa, AL
Geismar, LA
Houston, TX
Deer Park, TX
Norco, LA
Odessa, TX
Wood River, IL
Irwindale, CA
Delaware City, DE
Kearny, NJ
El Segundo, CA
Pascagoula, MS
Richmond, CA
Alvin, TX
Products
At Site
Bz
Cyxc
Et
ClBz
ClBz
Bzc
MAN
MAN
BSA
NiBz
Et
Bzc
Cu
Et
Bzd
Etd
Bz
Bz
ClBz
ClBz
ClBz
Bz
Cu
Bz
Bz
Et (2 units)
Capacity
(Gq/yr)
367
212
454
NDg
64
23
14
20
ND9
170
590
301
326
681
133
681
40
150
2
125
7
77
45
ND9
ND9
907
9-7
-------
Table 9-1. REFINERIES AND ORGANIC
CHEMICAL MANUFACTURING SITES
WITH BENZENE FUGITIVE EMISSION POTENTIAL (CONTINUED)
Benzene-Related
106.
107.
108.
109.
110.
111.
112.
113.
114.
115.
116.
117.
118.
119.
Plant
Standard Oil (IN)/
Amoco
Standard Oil (OH)/
BP Oil
Stauffer Chemical
Sun Oil
Sun Oil
Sun Oil
Sun Oil
Sun-Olin
Tenneco
Tenneco
Texaco
Texaco
Texaco/Jefferson
Chemical
Texaco/Jefferson
Chemical
City/ State
Texas City, TX
Marcus Hook, PA
Henderson, NV
Corpus Christi, TX
Marcus Hook, PA
Toledo, OH
Tulsa, OK
Claymont, DE
Chalmette, LA
Fords, NJ
Port Arthur, TX
Westville, NJ
Bellaire, TX
Port Neches, TX
Products
At Site
Bz
Cu
EtBz
St
Bz
BSA
Bz
Cu
Et
"EtBz
St
Bz
BzC
Bz
Cyxc
Et
Bz
EtBz
MAN
Bz
Cyxc
Et
Bz
Cu
Et
Et
i
Capacity
(Gq/yr)
284
14
386
381
27
4
127
113
9
61
54
97
164
80
83
109
33
16
12
150
117
454
117
64
240
238
9-8
-------
Table 9-1. REFINERIES AND ORGANIC
CHEMICAL MANUFACTURING SITES
WITH BENZENE FUGITIVE EMISSION POTENTIAL (CONTINUED)
Benzene-Related h
Plant
120. Union Carbide
121. Union Carbide
122. Union Carbide
123. Union Carbide
124. Union Carbide
125. Union Carbide
126. Union Oil of CA
127. Union Pacific/
Champ! in
128. U.S. Steel
129. USS Chemicals
130. Vertac/Transvaal
131. Witco Chemical
aBSA = Benzenesulfonic
Bz = Benzene
ClBz = Chlorobenzene
Cu = Cumene
Cyx = Cyclohexane
Et = Ethyl ene
City/State
Institute, WV
Penuelas, PR
Seadrift, TX
Taft, LA
Texas City, TX
Torrance, CA
Lemont, IL
Corpus Christi , TX
Neville Island, PA
Houston, TX
Jacksonville, AR
Carson, CA
Acid
Products
At Site
EtBz
LAB
St
Bz
Cu
Et
Et
EtBz
St
Bzc
Et
Et
Et
Bz
Bz
j
Cud
Cyx
MAN
Et
ClBz
LAB
Hqn
LAB
MAN
NiBz
Rcnol
St
Capacity
(Gq/yr)
N°9
64
ND9
ND9
290
454
546
154
136
234
500
546
73
57
33
uu
65
38
227
nu
20
= Hydroquinone
= Linear Alkyl benzene
= Maleic Anhydride
= Nitrobenzene
= Resorcinol
= Styrene
EtBz = Ethyl benzene
9-9
-------
Table 9-1. REFINERIES AND SYNTHETIC ORGANIC
CHEMICAL MANUFACTURING SITES
WITH BENZENE FUGITIVE EMISSION POTENTIAL (CONCLUDED)
Annual 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
ClBz - 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, 13, 14, 33
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
Rcnol - Ref. 6
St - Refs. 1 (1977 year-end), 14
cProduct unit under expansion
Product unit under construction
eProduct unit on standby or not currently in use
Product unit in engineering phase
9No data available
9-10
-------
Table 9-2. NUMBER OF COMPANIES AND PLANT SITES
THAT MANUFACTURE BENZENE DERIVATIVES
Number of
Companies
8
8
15
10
4
9
12
8
1
4
1
Number of Sites
(and Units)
10
10
18
14
4
11
13
10
1
4
1
Benzene Derivative
Chlorobenzene
Nitrobenzene
Ethyl benzene
Styrene
Linear Alkylbenzene
Cyclohexane
Cumene
Maleic Anhydride
Resorcinol
Benzenesulfonic Acid
Hydroquinone
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
disproportionate, 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
-------
Table 9-3. SUMMARY OF PRODUCTION AND CAPACITY FOR
BENZENE, ETHYLENE, AND BENZENE DERIVATIVES1"10
Total U.S.
Capacity
Product (Gg 1977)
Benzene
Ethyl ene
Chlorobenzenes
Nitrobenzene
Ethyl benzene
Styrene
Linear
Alky! benzenes
Cyclohexane
Cumene
Maleic Anhydride
Resorcinol
Benzenesulfonic
Acid
Hydroquinone
7,008
15,100
440
441
5,070
3,741
308
1,395
1,653
236
16b
61 b
3b
Total U.S.
Production
(Gg 1977)
5,256a
10,600
234
252
2,829
2,694a
239
983
1,281
132
14a
48b
N.D.d
Capacity
Utilization
75 Percent
70 Percent
53 Percent
61 Percent
56 Percent
72 Percent
78 Percent
70 Percent
77 Percent
56 Percent
86 Percent0
79 Percent
M.D.d
Estimated from the percent capacity utilization for the product,
Represents 1977 capacity for one company.
cBased on benzene consumption estimated for 1976.
N.D. designates no data available.
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 streams also produce ethylene by pyrolysis.
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.
g.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-
18
dichlorobenzene being co-produced.
g.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.
g.1.3.5 Ethylbenzene/Styrene. More than 95 percent of domestic
ethylbenzene 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 ethylbenzene-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
21
production is by dehydrogenation of ethyl benzene.
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 linear alpha olefins 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
22
detergent with a slightly different balance of detergent properties.
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-
23
stitutes the remaining 15 percent.
9.1.3.8 Maleic Anhydride. The two major processes used to
manufacture maleic anhydride (MAN) in the United States are benzene
oxidation and butane oxidation. Most major U.S. producers are employ-
ing the first 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
24
France and Japan.
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-
20
tion, oligomerization, and other side reactions. Essentially all
g
cumene produced is consumed in the manufacture of phenol and acetone.
9-14
-------
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
not been proved commercially successful.
9.1.3.11 Benzenesulfonic Acid (BSA). 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 Hvdroguinone. 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 method is used by the only U.S. producer
of hydroquinone.
9-15
-------
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
20
crystallizes.
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
?fi
54 percent of production in 1977. 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, benzene 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 consumption. Household goods furniture,
appliances, carpeting use nylon fibers and resins, ABS, polystyrene,
phenolics, and epoxies, among 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 benzene
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
-------
Figure 9-1.
Percentages of Total Benzene Production Consumed by
Intermediate and Final Products (1976)*3
I CUMENE/PHENOL (17%)-
I
^-J
BENZENE
-ETHYLBENZENE/STYRENE (51%)-
-CYCLOHEXANE (15%)-
ANILINE (4%)-
-POLYSTYRENE (28%)
- molded plastic
- packaging
-STYRENE COPOLYMER RESINS (9%)
- construction
- automobiles
- appliances
SBR elastomers (5%)
- tires
-PHENOLIC RESINS (9%)
- plywood adhesives
-CAPROLACTAM & BISPHENOL (5%)
- epoxy
- polycarbonate
- nylon 6
-NYLON FIBERS
- nylon 66
- nyIon 6
& RESINS (14%)
-RIGID POLYURETHANE (2%)
- construction
- insulation
- refrigerators
- transport equipment
- marine products
- packaging
RUBBER CHEMICALS (2%)
*Percentages are from 1976 data for benzene apparent consumption.
-------
Table 9-4. ETHYLENE USAGE17
Percent of
Consumption
End Use (1976)
Low-Density Polyethylene 27.4
High-Density Polyethylene 14.7
Ethylene Oxide 18.6
Vinyl Chloride 12.0
Ethylebenzene, Styrene 9.1
Ethyl Alcohol 3.5
Aliphatic Alcohols 2.5
Acetaldehyde 2.6
Vinyl Acetate ' 2.2
Ethyl Chloride 1.4
Alpha Olefins 1.5
Other 4.5
100.0
9-18
-------
18
Table 9-5. MONOCHLOROBENZENE USAGE10
Percent of
Consumption
End Use (1977)
Solvents 30
Nitrochlorobenzene
(Agricultural Products) 35
DDT, Silicones, etc. 15
Diphenyl Oxide 10
Rubber Intermediates 10
I O
Table 9-6. DICHLOROBENZENES USAGE10
Percent of
Consumption
End Use (1976)
o-Dichlorobenzene
3,4,Dichloroaniline, etc. 65
TDI* Process Solvent 15
Solvents 10
Dye Manufacture 5
Pesticides, etc. 5
£-Dichlorobenzene
Space Deodorant 90
Intermediate for Pesticides 10
*Toluene Diisocyanate
9-19
-------
Table 9-7. NITROBENZENE USAGE7
Percent of
Consumption
End Use (1978)
Aniline 98
Solvent, Dichloroaniline 2
Table 9-8. ANILINE USAGE7
Percent of
Consumption
End Use (1978)
MDI* 52
Rubber Chemicals 29
Dyes 4
Hydroquinone 3
Drugs, Pesticides 12
*p,p'-Methylene Diphenyldi isocyanate
9-20
-------
Table 9-9. ETHYLBENZENE USAGE
20
End Use
Percent of
Consumption
(1976)
Styrene
Solvent
99
Table 9-10. STYRENE
20
""
End Use
Percent of
Consumption
(1976)
Polystyrene ^
Styrene Copolymer Resins
Styrene-Butadiene Elastomers
Unsaturated Polyester Resins
Miscellaneous
Exports
54
17
9
6
1
9-21
-------
Table 9-11. LINEAR ALKYLBENZENE USAGE8
Percent of
Consumption
End Use (1977)
Linear Alkylbenzene
Sulfonates* 90
Export 10
*Detergent Surfactant.
Table 9-12. CYCLOHEXANE USAGE22
Percent of
Consumption
End Use , (1977)
Adi pic Acid 53
Exports 18
Caprolactam 23
1,6,-Hexamethylenedi ami ne
(HMDA) 3
Miscellaneous 3
9-22
-------
Table 9-13. CUMENE USAGE9
Percent of
Consumption
End Use (1976)
Phenol and Acetone 99
a-Methylstyrene, Solvent 1
?4
Table 9-14. MALEIC ANHYDRIDE USAGE^
Percent of
Consumption
End Use (1975)
Unsaturated Polyester Resins 51.1
Fumaric Acid 6-4
Agricultural Chemicals 10.0
Alkyd Resins I-3
Lubricating Additives 7.8
Copolymers 5.3
Reactive plasticizers 3.6
Maleic Acid 3.8
Chlorendic Anhydride and
Acid !!
Surface-Active Agents 2.9
Other 5-7
9-23
-------
Table 9-15. RESORCINOL USAGE6
Percent of
Consumption
End Use (1977)
Rubber Products 59.6
Wood Adhesive Resins 25.5
Miscellaneous 14.9
Table 9-16. BENZENESULFONIC ACID USAGE3
Percent of
End Use Consumption
Phenol ' N.D.*
Dyes N.D.*
*N.D. designates no data available.
Table 9-17. HYDROQUINONE USAGE3
Percent of
End Use Consumption
Rubber Antioxidant N.D.*
Photographic Developer N.D.*
Dye Intermediates N.D.*
*N.D. designates no data available.
9-24
-------
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. 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
Product
Benzene
Ethyl ene
Chlorobenzenes
Nitrobenzene
Aniline
Ethylbenzene0
Styrene
Linear Alkyl benzene
Cyclohexane ,
Cumene
Maleic Anhydride
Benzenesulfonic Acid/Phenol6
Resorcinol
Hydroquinone
May
1979
0.
0.
0.
N.
0.
N.
0.
0.
0.
0.
0.
0.
3.
3.
37
31
30
D.a
79
D.a
63
75
46
40
88
68
09
39
Price History
1978
0
0
0
0
0
N
0
0
0
0
0
0
2
3
.22
.29
.27
.51
.75
.D.a
.43
.65
.27
.31
.68
.46
.87
.39
1977
0
0
0
0
0
0
0
0
0
0
0
0
2
3
.24
.27
.27
.51
.75
.15
.46
.58
.29
.31
.82
.59
.72
.30
(Dollars/Kilogram)
1976
0
0
N
0
0
N
N
0
0
0
0
N
N
N
.23
.25
.D.a
.51
.71
.D.a
.D.a
.56
.27
.31
.82
.D.a
.D.a
.D.a
1975
0
0
0
0
0
N
N
0
0
0
0
N
N
N
.23
.19
.29
.42
.71
.D.a
.D.a
.25
.27
.35
.82
.D.a
.D.a
.D.a
1974
0.23
0.17
0.25
0.21
0.24
0.13
N.D.a
0.25
0.27
N.D.a
0.35
N.D.a
N.D.a
N.D.a
aN.D. designates no data available.
bSince 97 to 98 percent of nitrobenzene production goes into aniline
manufacture, aniline prices have been provided for comparison.
cPrices are not generally available for ethylbenzene because greater
than 99 percent of production is captively consumed.
dSince most ethylbenzene production 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
Reference
Benzene Derivative Alternative Processes Number
Cumene Separation from petroleum liquids
Cyclohexane Separation from petroleum liquids 2
Ethyl benzene Extraction from mixed xylene streams 20
Maleic Anhydride Oxidation of n-butane; by-product of 3,20
phthalic anhydride production (xylene
derivative)
Styrene Propylene oxide coproduct; extraction 1
from pyrolysis gasoline; production
from toluene and ethylene via stilbene
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 disproportionate,
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
?7 2Q 30
about 20 percent by the end of this period. '' 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
27
the largest new source of the aromatic.
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 exoected to be satisfactory through 1982-83, with only minimal
27
needs for additional capacity beyond this point until 1986-87.
9-27
-------
Table 9-20. PROJECTED ANNUAL GROWTH RATES FOR DEMAND OF
:ENE, ETHYLENE. AND BENZENE DERIVATIVES1-'0' 17-T9, 21-24, 27
Product
Projected
Average Annual
Percent Growth
(1977 - 1983)
Benzene
Ethylene
Chlorobenzenes
Nitrobenzenes
Ethylebenzene/Styrene
Linear Alky!benzene
Cyclohexane
Cunene
Maleic Anhydride
Resorcinol
Benzenesulfonic Acid
Hydroauinone
5.5a
5.5
1.5
6.0
6.0
2.0
5.0
7.5
11.0
N.D.1
N.D.
N.D.1
Growth rate for benzene is estimated at 5 to 5.5 percent per
year through 1985.
N.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 refining and organic chemical
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 units are .07, .20, .59, and 4.47 for Regulatory
Alternative 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 units 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
-------
play a 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
derivatives tends to be dominated by a relatively small number of firms. The
degree of dominance in the various product areas is illustrated by the con-
centration ratios presented in Table 9-21. 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 producer 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 line'ar alkylbenzene (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 as economies of scale. Important economies can be
9-31
-------
Table 9-21. CONCENTRATION RATIOS3 FOR
BENZENE, ETHYLENE, AND BENZENE DERIVATIVES
Chemical
Benzene
Ethyl enec
Benzenesulfonic Acid
Chlorobenzene
Cumene
Cyclohexane
Ethyl benzene
Hydroquinone
Linear Alkyl benzene
Maleic Anhydride
Nitrobenzene
Resorcinol
Styrene
# of
f i rms
32
28
4
8
12
9
15
1
4
8
8
1
10
Concentration
2 firm
18
23
d
56
34
54
35
100
73
41
47
100
45
ratio (%)
4 firm
34
43
100
87
63
72
59
-
100
67
82
-
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
Ratio for top 8 firms is 58 percent.
GRatio for top 8 firms is 68 percent.
Productive capacity data not available.
9-32
-------
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 derivatives may be charac-
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 importance tends to be reduced by three
9-33
-------
factors: 1) joint product cost accounting; 2) low price elasticity of
demand; and 3) the consumer's interest in an uninterrupted supply.37 jn
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
both 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
supply, 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 ethyl benzene/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
-------
by more than 30 percent, and while 1976 witnessed a substantial recovery,
consumption was still below the peak levels attained in 1973 and 1974.
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-
portionate, 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_Acld. 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 Chlorobenzene. 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
DDT. 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
-------
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 production 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.10 It is estimated that demand for ethylben-
zene will grow at an average annual rate of 6 percent through 1983 (refer
to Section 9.1.8).
9.2.3.7 Hydroquinone. 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 Lin^ar_AJkylbenzene. The dominant use for linear alkylbenzenes
(LAB) is as a ravTmaterial in the production of linear alkylbenzene 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 Maleic Anhydride. 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, going from 83 Gg to 128 Gg. Most of this growth was due to
increased use for the manufacture of unsaturated polyester resins. In 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,p'-methylene diphenyldi-
isocyanate (MDI) (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 R_eso_rc_i_nol_. Resorcinol is produced in the U.S. by only one
company, and is consumed mainly in the production of resorcino!-formaldehyde
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 growth 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 that 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 unit 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:
Historical and projected growth in demand - High rates of growth,
especially during periods of price increases, are indicative of a
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.
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 elasticity of demand is determined to be low.
9.2.4 Supply Characteristics
At present it appears the existing benzene and benzene derivative pro-
ductive capacity, augmented by projected additions to capacity (Table 7-9)
9-39
-------
Table 9-22. QUALITATIVE EVALUATION
OF PRICE ELASTICITY OF DEMAND
Product
Benzene
.
Benzenesulfonic Acid
Chlorobenzene
Evaluative factors ~ "
Demand growth8
1950-1974. 9% per year.
Decline of more than 30%
1n 1975. Recovery followed.
Projected 5-5.5% per year
over period 1977-1985.
Continuing upward move-
ment of prices.
No data available on
historical or projec-
ted demand growth.
Small amount of growth
anticipated.
Principal chlorobenzene
product is monochloro-
benzene. From 1960-
1976, demand for mono-
chlorobenzene declined
from 174 Gg to 159-163
Gg. For chlorobenzenes
as whole, growth pro-
jected at 1.5% per year
over period 1977-1985.
Prices have been rela-
tively stable in recent
years.
Captive consumption
Substantial. Estimated
to be more than 50% of
production.
No data available, but
captive consumption
appears to be important.
Much of monochloro-
benzene production
consumed captively.
Ortho-8 para-dichloro-
benzenes mainly sold.
Potential for substitution
Main uses are ethylbenzene/
styrene, cumene/phenol , &
cyclohexane. No good sub-
stitutes for benzene as
raw material for these.
Some indirect substitution
possible-e.g. substitution
of polyester fibers for
nylon fibers.
Main use is an intermediate
in production of phenol.
Only one domestic producer
by this route. Dominant
route to phenol is via
cumene.
No direct substitutes
available in current
applications. Some
indirect substitution
at end-product level
may be possible in
future.
Foreign
competition
Imports small.
None
Imports
negl i gible.
P«-
-------
Table 9-22. QUALITATIVE EVALUATION
OF PRICE ELASTICITY OF DEMAND
(Continued)
Product
Cumene
i
-P.
Cyclohexane
Ethyl benzene
Evaluative factors
Demand growth
During 1960's, about 25*
per year. With matura-
tion of phenol markets,
growth has slowed some-
what. Growth projected
at 7.5% per year over
period 1977-1983. Up-
ward movement 1n prices.
1971-1974, 5.9* per year.
Decline in consumption
1n 1975, followed by
reversal 1n 1976. Growth
projected at 5% per year
over period 1977-1983.
Prices continuing to
move upward.
1960-1972, 10.8% per
year. Consumption con-
stant from 1972 to 1974.
Big drop 1n 1975, fol-
lowed by reversal In
1976. Future growth
projected at 6% per
year through 1983.
Price not a factor
since most ethyl benzene
consumed captlvely.
Captive consumption
About 50% consumed
captlvely.
Mainly sold on
merchant market.
Virtually all pro-
duction captlvely
consumed for styrene.
Potential for substitution
Other routes to phenol
exist, but cumene 1s
preferred.
Limited substitution
possible. Adipic acid
can be made from phenol,
but this is not an
economic substitute.
Several other routes are
available in production
of HMD. Acrylic and
polyester fibers can be
substituted for nylon.
No substitute for ethyl-
benzene in production of
styrene.
Foreign
competition
Substantial amounts
of cumene are
imported.
Imports negligible.
Imports negligible.
Price elasticity
of demand
Low
Low
Low
-------
Table 9-22. QUALITATIVE EVALUATION
OF PRICE ELASTICITY OF DEMAND
(Continued)
Product
Hydroquinone
Linear Alkylbenzene
Maleic Anhydride
Evaluative factors
Demand growth
No historical data
available. Projec-
tions also unavail-
able. May have growth
prospects. Only one
domestic producer at
present.
1975-1977. slightly
more than 3% per year.
Estimated growth of 25!
per year over period
1977-1983. Price has
risen somewhat in
recent years.
1968-1973, over 9% per
year. In 1974, growth
slowed to 7%. In 1975,
consumption dropped
2&%. Recovery since
then. Growth projec-
ted at 11% per year
over period 1977-1983.
Price has increased
greatly since 1974,
and is continuing upward.
Captive consumption
Both captive use
& sale on merchant
market. Breakdown
unknown.
Both captive use
and sale on merchant
market. Breakdown
unknown.
Produced mainly for
merchant market.
Potential for substitution
No substitutes.
Main use is 1n
production of linear
alkylbenzene sulfon-
ates. No substitu-
tion possible in this
use.
No direct substitutes
in production of un-
saturated polyester
resins, fumarlc add,
and agricultural
chemicals.
Foreign
competition
Not a factor
Imports
negligible.
Imports
negligible.
Price elasticity
of demand
Low
Low
Low
I
-F*
ro
-------
-p.
GO
Table 9-22. QUALITATIVE EVALUATION
OF PRICE ELASTICITY OF DEMAND
(Concluded)
Product
Nitrobenzene
Resorcinol
Styrene
Evaluative factors r
1975-1978, 5% per year.
Projected growth of 6%
per year over period
1977-1983. Price has
been gradually increasing.
1964-1974, about 7% per
year. In 1975, sharp
drop in consumption,
which was not overcome
until 1977. Projection
of future demand not
available. Price has
been advancing.
1960-1972, 10-11% per
year. Drop in growth
rates in 1973 & 1974.
In 1975, consumption
declined 18% from pre-
vious year. Reversal
of downward trend in
1976. Growth projected
at 6% per year through
1983.
Captive consumption
Most consumed captively
in manufacture of
aniline.
Degree of captive use
not known. Appears
that sale on the mer-
chant market 1s impor-
tant.
Captive use estimated
to be around 60%.
Potential for substitution
Aniline can be produced
from ammolysis of chloro-
benzene, but this source
dependent on surplus of
chlorine & sales of by-
product ammonium chloride.
No substitutes.
No good substitutes for
styrene available at
present time.
competition
Imports
negliaible.
Imports
negligible.
Imports
negligible.
Price elasticity
of demand
Low
Low
Low
-
aSee discussion in Sections 9.2.3.1 to 9.2.3.12 for sources of these data.
-------
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 Units and Economic Impacts. In the following sections
the economic implications of applying the regulatory alternatives upon the
industry, as represented by the model units 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 unit
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
annual!zed 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
9-44
-------
full pass through of control costs, to the consumers of benzene and its
derivatives, has been assumed for all regulatory alternatives. 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 unit have been
estimated by expressing the net annualized costs of control for each model
unit and regulatory alternative (Table 8-11) as a percentage of the annual
total revenue of each model unit. These percentages are therefore indicators
of the percentage increase in model unit revenues (and thus product prices)
required if the net earnings (post control) of each plant are to remain
unaffected. While the net annualtzed cost estimates are presented in Section
8.2.3, the total annual revenues of each model unit 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 unit 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 between 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 units 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
-------
Table 9-23. MODEL UNIT
ANNUAL REVENUES - MODEL UNIT A
(May 1979 Dollars)
Chemical Capacity
(Gg/yr}_
Benzene
(toluene dealkylation)
Ethyl benzene/Styrene
Cumene
Cyclohexane
Benzenesulfonic Acid
Resorcinol
Maleic Anhydride
Ethyl ene
(1 unit)
153a
282d
160
120
4
16
23
323
Capacity Output
Utilization (Kg/yr)
.64e
.72
.77
.70
.79
.86
.56
.70
97,920,000
203,000,000
123,200,000
84,000,000
3,160,000
13,760,000
12,880,000
226,100,000
Price0
($/Kg)
.37
.63
.40
.46
.68
3.09
.88
.31
Total
Revenue
$ 36,230,0
$127,890,0
$ 49,280,0
$ 38,640,0
$ 2,149,0
$ 42,518,0
$ 11,334,0
$ 70,091,0
Reference 43, pp. 2-6. *
""Derived from Table 9-3.
:As presented in Table 9-18.
^Capacity presented is for styrene.
Reference 27, p. 64.
9-46
-------
Table 9-24. MODEL UNIT
ANNUAL REVENUES - MODEL UNIT B
(May 1979 Dollars)
Chemical
Benzene
(extraction)
Chlorobenzene
Capacity
(Gg/yr)
80a
46
Linear Alky! benzene 77
Ethylene
(2-3 units)
Reference 43,
Derived from
cAs presented
Reference 27,
755
pp. 2-6.
Table 9-3.
in Table 9-18.
p. 64.
Capacity
Utilization
.83d
.53
.78
.70
Output
(Kq/yr)
66,400,000
24,400,000
60,000,000
528,500,000
Price0
($/Kq)
.37
.30
.75
.31
-
Total
Revenue
$ 24,600,000
$ 7,300,000
$ 45,000,000
$163,800,000
-
Table 9-25. MODEL UNIT
ANNUAL REVENUES - MODEL UNIT C
(May 1979 Dollars)
Chemical
Benzene
(pyrolysis gas
Nitrobenzene
Hydroquinone
Ethylene
(4-5 units)
Capacity
(Gq/yr)
) 107a
63d
5
1136
Capacity
Utilization
.76e
.61
.69
.70
Output
(Kq/yr)
81,320,000
38,430,000
3,450,000
795,200,000
Price0
($/Kq)
.37
.79
3.39
.31
Total
Revenue
$ 30,100,000
$ 30,400,000
$ 11,700,000
$246,500,000
Reference 3.
bDerived from Table 9-3, except for hydroquinone which is based upon the mean
capacity utilization of all plants in Table 9-3.
cAs presented in Table 9-18.
Capacity presented is for Aniline.
Reference 27, p. 64.
9-47
-------
The total annual revenue of each model unit 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 Unit 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 unit have been compared to the total investment represented by
each model unit.
Estimates of total plant investments (including process units, construc-
tion and start-up costs, and working capital) were obtained through several
sources including: trade journal summaries of construction activity; industry
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 unit 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 A-,'7
Cost of B Capacity BJ
In addition, all investment totals have been expressed in May 1979 dollars,
through adjustment according to the Chemical Engineering "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 Model Unit Impact Analysis
9.2.6.1 Price Impacts.. With the exception of Regulatory Alternative V
(i.e., leakless emission control equipment) the full cost pricing policies
9-48
-------
Table 9-26. TOTAL CAPITAL
INVESTMENT REQUIRED - NEW MODEL UNITS
(Millions of May 1979 Dollars)
(Continued)
Chemical Capacity
(Gq/vr)
Model Unit A
Benzene
(toluene dealkylation)
Ethyl benzene/Sty rene
Cumene
Cyclohexane
Benzenesulfonic Acid
Resorcinol
Maleic Anhydride
Ethylene
(1 unit)
153
282 '
160
120
4
16
23
323 >
Investment
In Plant0
$ 14.6
56.7
21.2
9.9
3.8
9.7
20.8
384.5
Investment In Total Capital
Working Capital Investment
$ 2.2
8.5
3.2
1.5
.6
1.5
3.1
57.7
$ 16.8
65.2
24.4
11.4
4.4
11.2
23.9
442.2
Model Unit B
Benzene
(extraction)
Chlorobenzene
Linear Alkylbenzene
Ethylene
(2-3 units)
80
46
77
775
31.4
11.7
13.8
709.6
4.7
1.8
2.0
106.4
36.1
13.5
15.8
816.0
9-49
-------
Table 9-26. TOTAL CAPITAL
INVESTMENT REQUIRED - NEW MODEL UNITS
(Millions of May 1979 Dollars)
(Concluded)
Chemical
Model Unit C
Benzene
(pyrolysis gasoline)
Ni trobenzene/Ani 1 ine
Hydroquinone
Ethylene
(4-5 units)
Capaci ty
(Gg/yr)
107
63
5
1136
Investment
In Plantc
$ 6.8
32.5
22.0
946.8
Investment In
Working Capital
$ 1.0
4.9
3.3
142.0
Total Capital
Investment
$ 7.8
37.4
25.3
1,088.8
Capacity 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.
Capacity in terms of aniline*, investment totals are for an integrated nitro-
benzene/aniline facility in which all nitrobenzene is consumed in the manu-
facture of aniline.
Original 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
Ethyl benzene/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 both the impacts of the alterna-
tives upon individual chemicals as well as those resulting from
the pass-through of benzene and ethylene price increases to
derivatives.
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 unit 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 outweigh 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
manufacturers 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)
I
cn
rv>
Existing
Chemical
Model Unit A
Benzene (toluene dealkylation)
Ethyl benzene/Styrene3
Cumene
Cyclohexane ^
Benzenesulfonic Acid/Phenol
Resorcinol
Maleic Anhydride
Ethylene (1 unit)
Model Unit B
Benzene (extraction)
Chlorobenzene
Linear Alkyl benzene
Ethylene (2-3 units)
Model Unit C
Benzene (pyrolysis gasoline)
Ni trobenzene/Ani 1 i nec
Hydroquinone
Ethylene (4-5 units)
Plants
Regulatory Alternative
II
.01
!oi
.01
.16
.01
.03
(.02)
(.05)
(.01)
(.03)
(.03)
(.09)
III
.02
.01
.02
.02*
.34
.02
.07
.01
.04
.13
.02
.01
.05
.05
.13
.01
IV
.04
.01
.03
.04
.65
.03
.12
.02
.13
.43
.07
.02
.17
.16
.43
.02
V
.34
.10
.25
.32
5.67
.29
1.08
.17
1.48
5.00
.81
.22
2.02
2.00
5.21
.25
New Plants
Requlatory Alternative
II
.01
!oi
.01
.14
.01
.03
(.02)
(.07.)
(.01)
(.04)
(.04)
(.11)
(.01)
III
.02
!oi
.02
.27
.01
.05
.01
.02
.08
.01
.02
.02
.06
IV
.03
.01
.02
.03
.56
.03
.11
.02
.10
.35
.06
.02
.13
.13
.34
.02
V
.23
.06
.17
.21
3.84
.19
.73
.12
1.01
3.39
.55
.15
1.37
1.36
3.53
.17
*Price increase for Styrene
DPrice increase for Phenol
"Price increase for Aniline
-------
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 Unit C, Alternative III
(.05%) and the increase in cumene price from Model Unit 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 Unit C) were used in the
calculation of cumulative price increases. In reality, the manufacture
of derivatives through the u^e of benzene from Model Units 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
9-53
-------
Table 9-28. CUMULATIVE PERCENTAGE PRICE INCREASES
(May 1979 Prices)
Chemical
Ethylbenzene/Styrene
Cumene
Cyclohexane
Benzenesulfonic Acid/Phenol
Resorcinol
Maleic Anhydride
Chlorobenzene
Linear Alkylbenzene
Nitrobenzene/Aniline0
Hydroquinone
Regulatory Alternative
II
.03
.02
.02
.17
.02
.04
(-04)
.00
(.02)
(-08)
III
.07
.07
.07
.39
.07
.12
.18
.07
.10
.18
IV
.20
.20
.21
.82
.20
.29
.60
.24
.33
.60
V
2.37
2.27
2.34
7.69
2.31 -
3.10
7.02
2.83
4.02
7.23
Price increase for Styrene
Price increase for Phenol
CPrice increase for Aniline
9-54
-------
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 unit. Estimates of model unit 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 unit and regulatory alter-
native are presented in Table 8-4.
Table 9-29 summarizes the additional percentages of invested capital
required of new model units. With regard to the increased capital require-
ments of Regulatory Alternative II, III and IV, these increases are generally
low and cannot be considered potential obstructions to new plant construction.
Regulatory Alternative V (i.e., leakless 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.
9.3 MACROECONOMIC IMPACT
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 price increases (Regulatory Alternatives II,
III, and IV) for benzene and its derivatives are low (Table 9-28), price
9-55
-------
Table 9-29. PERCENTAGE INCREASE IN
NEW PLANT CAPITAL INVESTMENT REQUIRED
Chemical Regulatory Alternative
ii m iv v
Model Unit A
Benzene (toluene dealkylation) .06 .12 .30 2.08
Ethylbenzene/Styrene .02 .04 .09 .54
Cumene .04 .08 .21 1.44
Cyclohexane .09 .18 .45 3.07
Benzenesulfonic 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 Unit B
Benzene (extraction) .04 .12 .38 2.91
Chlorobenzene -10 .31 1.01 7.79
Linear Alky!benzene -09 .27 .86 6.66
Ethylene (2-3 units) <-01 .01 .02 .13
Model Unit C
i .. -i.....111...-..-... ^
Benzene (pyrolysis gasoline) .22 .86 2.85 22.49
Nitrobenzene/Aniline .05 .18 .60 4.69
Hydroquinone .07 .27 .88 6.93
Ethylene (4-5 units) <-01 .01 .02 .16
9-56
-------
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 Section 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, I" 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) applicable4 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.
9-57
-------
9.4 REFERENCES
1. Soder, S.L. CEH Product 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. Chemical
Economics Handbook. Stanford Research Institute. Menlo Park,
CA. February 1977. 29 p.
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. July 1977. 10 p.
5. Soder, S.L. et al. CEH Product Review on Ethylene. Chemical
Economics Handbook. Stanford Research Institute. Menlo Park,
CA. January 1978. 108 p.
6. Cogswell, S.A. CEH Product Review on Resorcinol. Chemical
Economics Handbook. Stanford Research Institute. Menlo
Park, CA. October 1978. 11 p.
7. Klapproth, E.M. CEH Product Review on Aniline and Nitrobenzene.
Chemical Economics Handbook. Stanford Research Institute. Menlo
Park, CA. January 1979. 10 p.
8. Bradley, R.F. CEH Product Review on Linear and Branched
Alkylbenzenes. Chemical Economics Handbook. Stanford
Research Institute. Menlo Park, CA. January 1979. 16 p.
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.
11. Hydrocarbon Processing, Section 2: World-Wide HPI Con-
struction Boxscore. June 1978. p. 4-17.
12. Hydrocarbon Processing, Section 2: World-Wide HPI Con-
struction Boxscore. October 1978. p. 3-18.
13. Hydrocarbon Processing, Section 2: World-Wide HPI Con-
struction Boxscore. February 1979. p. 3-16.
14. Hydrocarbon Processing, Section 2: World-Wide HPI Con-
struction Boxscore. June 1979. p. 3-13.
9-58
-------
15 Wett T Ethylene Report - Ethylene Capacity Outruns Demand
Growth! OH and Gas Journal. 77(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 (Draft). Hydroscience, Inc., Knoxville, TN For
U S. Environmental Protection Agency, Emission Standards and
Engineering Division. Research Triangle Park, NC. July
1978. 200 p.
18. Dylewski, S.W. Emission Control Options for the Synthetic
Organic Chemicals Manufacturing Industry, Chlorobenzenes
Product Report (Draft). Hydroscience, Inc., Knoxville, TN
For U.S. Environmental Protection Agency, Emission Standards
and Engineering Division. Research Triangle Park, NC.
August 1978. 90 p.
19. Emission Control Options for the Synthetic Organic Chemical
Manufacturing Industry, Nitrobenzene Product Report (Draft).
Hydroscience, 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-600/2-27-023f. February
1977. 1014 p.
21 Hobbs, F.D., and J.A. Key. Emission Control Options for the
Synthetic Organic Chemicals Manufacturing Industry, Ethylbenzene
and Styrene Product Report (Draft). 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 Alkylbenzene
Product Report (Draft). Hydroscience, Inc., Knoxville, TN.
For U.S. Environmental Protection Agency, Emission Standards
and Engineering Division. Research Triangle Park, NL.
September 1978. 137 p.
23 Blackburn, J.W. Emission Control Options for the Synthetic
Organic Chemicals Manufacturing Industry, Cyclohexane Product
Report (Draft). Hydroscience, Inc., Knoxville, TN. For
U S. Environmental Protection Agency, Emission Standards and
Engineering Division. Research Triangle Park, NC. May
1977. 78 p.
20
9-59
-------
24. Lawson, J. F. Emission Control Options for the Synthetic Organic
Chemicals Manufacturing Industry, Maleic Anhydride Product Report (Draft).
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 Production and Sales, 19-77.
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. Chemical Engineering.
85(3):62-64. January 30, 1978.
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. Chemical Week. 124(16):31.
April 18, 1979,
30. Ethylene Oversupply Could Last Until 1980. Chemical and Engineering
News. 55(14): 10. April 4, 1977.
31. Aromatics Seen Entering Slow-Growth Era as Energy, Government Strictures
Hobble Trade. Chemical Marketing Reporter. 213.(24):11. June 12, 1978.
32. CE Construction Alert. Chemical Engineering. 86_(23):136. October 22,
1979.
33. Hydrocarbon Processing, Section 2: World-Wide HPI Construction Boxscore.
October 1979. p. 13.
34. Lurie, M. Oil and Chemicals: Era of Peaceful Coexistence? Chemical
Week. _125_(16): 70-92. October 1979.
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 Protection Agency.
Washington, DC. Final Report SW-158c. 1978. p. 85.
36. Reference 35, p. 87.
37. Reference 35, p. 87-88.
38. Reference 3, p. 35.
39. "Refer to Section-9.1.8.1.
9-60
-------
40. Blackford, J. L. CEH Product Review on Maleic Anhydride. Chemical
Economics Handbook. Stanford Research Institute. Menlo Park, CA.
July, 1976. 35 p.
41. As Polyester Goes, So Goes Maleic. Chemical Weeks. 123(22) : 21-22.
November, 1978.
42. Is Benzene Losing to Gas Tanks? Chemical Week. 123(4) :26. July,
1978.
43. Evaluation of Emissions from Benzene-Related ,
Operations, PEDCo Environmental, Inc., EPA Report No. 450/3-79-OZZ.
October 1978, p. 7.
44. Part 1, Research-Project Evaluations. Hydrocarbon Processing. Decem-
ber, 1976. p. 137.
45. CE Cost Indexes Maintain 13-Year Ascent. Chemical Engineering. 85(11):
189. May 8, 1978.
46. Economics of the Petrochemical Industry. Hyplan Consulting Group. June
1978. p. 32.
47. CE Construction Alert. Chemical Engineering. 86J23) :136. October 22,
1979.
48. Oil and Gas Journal. 64(44) :46. October 31, 1966.
49. How Synthetic Phenol Processes Compare. Oil and Gas Journal. 64(1):
83-88. January 1966.
50. Hydrocarbon Processing, Section 2: World-Wide HPI Construction Box-
score. 56(2) :18. February 1977.
51. Ethylene: Makke It or Buy It? Chemical Engineering Progress. 74(12):
17. December 1978.
52. Worldwide Construction. Oil and Gas Journal. 75_(41):127. October
1977.
53. Chemical Week. 95J5):23. August 1964.
54. Worldwide Construction. Oil and Gas Journal. 65(3):79-81. January
1967.
55. Hydrocarbon Processing. 1977 Petrochemical Handbook Issue. November
1977. p. 132.
56. CE Construction Alert. Chemical Engineering. 81(20) :105. September
1974.
57. CE Construction Alert. Chemical Engineering. 77_(7):123. April 1970.
9-61
-------
APPENDIX A
EVOLUTION OF THE BACKGROUND
INFORMATION DOCUMENT
A-l
-------
APPENDIX A - EVOLUTION OF THE
BACKGROUND INFORMATION DOCUMENT
Date
August 9-12, 1976
November 3-4, 1976
November 8-10, 1976
November 16-17, 1976
February 8-14, 1977
April 19-20, 1977
May 1977
June 8, 1977
April 1978
April 26-28, 1978
June 1978
Nature of Action
Plant visit to Los Angeles Air Pollution
Control District and four Los Angeles area
petroleum refineries (Fletcher Oil and Re-
fining Company, Atlantic Richfield Watson
Petroleum Refinery, Shell Oil Company Wilmington,
Champlin Wilmington Refinery) to obtain
background information on miscellaneous
sources of hydrocarbon emissions from petroleum
refineries.
Meetings with Exxon Company, USA and Shell
Oil Company to discuss EPA request for information
on hydrocarbon emission sources and controls.
Plant visits to four New Orleans, Louisiana,
petroleum refineries (Murphy, Gulf, Tenneco,
and Shell) to obtain background information
on miscellaneous sources of hydrocarbon
emissions in petroleum refineries.
Meetings with Standard Oil of California and
Union Oil of California to discuss EPA requests
for information on hydrocarbon emission
sources and controls.
Emission source testing at Atlantic Richfield
Watson Petroleum Refinery, Carson, California,
and Newhall Refining Company, Newhall, California.
Plant vist to "Refinery A," Corpus Christi,
Texas, to gather information for Control
Techniques Guideline (CTG) documents.
First draft CTG, "Control of Hydrocarbons
from Miscellaneous Refinery Sources."
Benzene listed as hazardous air pollutant in
Federal Register (42 FR 29332).
Second draft CTG, "Control of VOC leaks from
Petroleum Refining Equipment."
Radian/IERL Symposium on refinery emissions,
Jekyll Island, Georgia.
Publication of final CTG, "Control of Volatile
Organic Compound Leaks from Petroleum Refinery
Equipment."
A-2
-------
June 29, 1978
June 30, 1978
July 6, 1978
July 13, 1978
July 14, 1978
November 13-17, 1978
December 1978 -
January 1979
March 5-8, 1979
March 7, 1979
May 31, 1979
June-July 1979
June 20, 1979
Plant visit to Phillips Petroleum Company,
Sweeny, Texas, to collect information on
emissions from benzene-related petroleum
refinery operations.
Plant visit to Exxon Chemical Company, Baytown,
Texas, to collect information on emissions
from benzene-related petroleum refinery
operations.
Plant visit to Sun Petroleum Products Company,
Toledo, Ohio, to observe and discuss BTX and
THD units.
Plant visit to Gulf Oil Refinery, Philadelphia,
Pennsylvania, to collect information on
emissions from benzene-related petroleum
refinery operations (UDEX and toluene dealkylation
unit).
Plant visit to Sun Petroleum Products Company,
Marcus Hook, Pennsylvania, to collect infor-
mation on emissions from benzene-related
petroleum refinery operations.
Plant visit and emission source testing at
Sun Petroleum Products Company, Toledo, Ohio,
of BTX and HDA units.
Telephone survey of refineries to obtain
information on accumulator vessel vent emissions.
Plant visit and emission source testing at
Phillips Petroleum Company, Sweeny, Texas,
refinery.
Plant visit to Phillips Petroleum Company,
Sweeny, Texas, refinery and NGL Processing
Center.
Completion of technical portion of
preliminary draft background document
and distribution to industry, environmental
groups, and other interested persons.
Public comments on draft background document.
Visit to Chevron Company, U.S.A., El Segundo,
California, refinery to discuss fugitive VOC
emissions.
A-3
-------
June 21, 1979
July 19, 1979
October 10, 1979
March 1980
April 16, 1980
July 14, 1980
Visit to Atlantic Richfield Company, Carson,
California, refinery to discuss fugitive VOC
emissions.
Chemical Manufacturers Association Fugitive
Emission Seminar, Washington, D.C.
Carcinogen Policy proposed in Federal Register
(44 FR 58642).
Completion of Benzene Fugitives preliminary
draft background document and distribution to
NAPCTAC, industry, environmental groups,
and other interested persons.
Meeting of the National Air Pollution
Control Techniques Advisory Committee to
review the Benzene Fugitive Emissions
Standard, Raleigh, N.C.
Meeting between EPA and the American
Petroleum Institute to discuss proposed
performance and equipment specifications
of Benzene Fugitive Emissions Standard,
Durham, N.C.
A-4
-------
APPENDIX B
INDEX TO ENVIRONMENTAL CONSIDERATIONS
This appendix consists of a reference system which is cross
indexed with the October 21. 1974, Federal RecQster (39 FR 37419)
containing the Agency guidelines for the preparation of Environmental
Impact Statements. This index can be used to identify sections of
the document which contain data and information germane to any portion
of the Federal Register guidelines.
B-l
-------
APPENDIX B
INDEX TO ENVIRONMENTAL IMPACT CONSIDERATIONS
Agency Guidelines for Preparing
Regulatory Action Environmental
Impact Statements (39 FR 37419)
Location Within the Background
Information Document (BID)
1. Background and Description
Summary of the Regulatory
Alternatives
Statutory Authority
Industry Affected
Sources Affected
Availability of Control
Technology
2. Regulatory Alternatives
Regulatory Alternative I
No Action (Baseline)
Environmental Impacts
Costs
The regulatory alternatives are
summarized in Chapter 1,
Section 1.2.
Statutory authority is given in
Chapter 1, Section 1.1, and
Chapter 2.
A description of the industry
to be affected is given in
Chapter 9, Section 9.1.
Descriptions of the various
sources to be affected are
given in Chapter 3, Section 3.2.
Information on the availability
of control technology is given
in Chapter 4.
Environmental effects of Regulatory
Alternative I are considered in
Chapter 7.
Costs associated with Regulatory
Alternative I are considered in
Chapter 8.
(Continued)
B-2
-------
INDEX TO ENVIRONMENTAL IMPACT CONSIDERATIONS (Continued)
Agency Guidelines for Preparing
Regulatory Action Environmental
Impact Statements (39 FR 37419)
Location Within the Background
Information Document (BID)
Regulatory Alternative II
Environmental Impacts
Costs
Regulatory Alternative III
Environmental Impacts
Costs
Regulatory Alternative IV
Environmental Impacts
Costs
Environmental effects associated
with Regulatory Alternative II
emission control systems are
considered in Chapter 7.
The cost impact of Regulatory
Alternative II emission control
systems is considered in
Chapter 8.
The environmental effects associated
with Regulatory Alternative III
emission control systems are
considered in Chapter 7.
The cost impact of Regulatory
Alternative III emission control
systems is considered in
Chapter 8.
The environmental effects associated
with Regulatory Alternative IV
emission control systems are
considered in Chapter 7.
The cost impact of Regulatory
Alternative IV emission control
systems is considered in
Chapter 8.
(Continued)
B-3
-------
INDEX TO ENVIRONMENTAL IMPACT CONSIDERATIONS (Concluded)
Agency Guidelines for Preparing
Regulatory Action Environmental
Impact Statements (39 FR 37419)
Location Within the Background
Information Document (BID)
Regulatory Alternative V
Environmental Impacts
Costs
Regulatory Alternative VI
Environmental Impacts
Costs
The environmental effects associated
with Regulatory Alternative V
emission control systems are
considered in Chapter 7.
The cost impact of Regulatory
Alternative V emission control
systems is considered in
Chapter 8.
The implementation of this alter-
native would require the elimination
of benzene use and production.
Benzene emissions would be zero.
This alternative could not be
implemented without closing all
plants in several industries.
This would not be feasible, and
hence costs were not evaluated.
B-4
-------
APPENDIX C
EMISSION SOURCE TEST DATA
C-l
-------
APPENDIX C - EMISSION SOURCE TEST DATA
C.I INTRODUCTION
A survey was conducted by EPA to collect data from refineries
throughout the United States on leak detection and repairs of equipment
emitting volatile organic compounds (VOC) in excess of a defined action
level. This appendix summarizes the results of data from seven sources:
(1) one reference summarizing valve repair data from four U.S. refineries;
(2) Phillips Petroleum Company, Sweeny, Texas; (3) Shell Oil Company,
Martinez, California; (4) Union Oil Company of California, San Francisco
refinery, Rodeo, California; (5) benzene-producing Units E and F from a
U.S. petroleum refinery; (6) ethylene and cumene units at organic chemical
manufacturing plants; (7) cyclohexane unit at Exxon Chemical Company,
Baytown, Texas; (8) benzene-producing unit at refinery "E," Gulf Coast,
U.S.; and (9) benzene-producing unit at AMOCO refinery. 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.
Data collected by Monsanto Research Corporation (MRC) concerning
fugitive emissions from synthetic organic chemical plant monochlorobenzene
units are not summarized in this appendix. This data were never published
in final form because of the following: (1) the data exhibited very
wide confidence intervals; (2) the experimental design, which was developed
for testing several units, was not appropriate for application to the
one unit tested; (3) the statistical treatment of the data was questionable;
and (4) the quality assurance and control procedures were poorly documented.
C.2 DATA SUMMARIES
2
C.2.1 Refinery Valve Maintenance Data
Table C-l presents the results of maintenance data for gas service
and light liquid valves from four U.S. petroleum refineries. The number
C-2
-------
Table C-l. REFINERY VALVE MAINTENANCE 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 Valves Percent of
for which Main- Number of Valves Emission
tenance was After Maintenance Reduction
Source Type Attempted (<10,000 ppm) (Leak Rate)
Gas Service Valves
Undirected
Maintenance 8 7 53
Directed .
Maintenance 7 6 92
Light Liquid Valves
Undirected
Maintenance3 4 2 79
Directed .
Maintenance 4 2 92
aAs a result of undirected maintenance,valves were tightened, but
no concentration data were recorded during maintenance.
bAs a result of directed maintenance,valves were tightened until
emissions were below 10,000 ppm.
C-3
-------
Table C-l. REFINERY VALVE MAINTENANCE (Concluded)
Action Level: >10,000 ppm
Distance from Source: 0 cm
Instrument: Bacharach "TLV Sniffer"
II. Hydrocarbon Leaks Less Than 10,000 ppmv
Source Type
Number of Valves Percent of
for which Main- Number of Valves Emission
tenance was With Increased Rate
Attempted Emissions Rate Reduction
Gas Service Valves
Undirected
Maintenance
Directed b
Maintenance
1
0
66
54
Light Liquid Valves
Undirected
Maintenance 7
Directed ,
Maintenance 10
1
2
56
76
aValves repaired after undirected maintenance were tightened, but
no concentration data were recorded during repair.
bValves repaired after directed maintenance were tightened until
emissions were below 10,000 ppm.
Source: Reference 2
C-4
-------
of valves on which maintenance was attempted and the number of valves
for which screening values after maintenance were below 10,000 ppmv are
shown as well as the percentage reduction in leak rates (Ib/hr). Main-
tenance data for leaks less than 10,000 ppmv are also given in Table
C-l, showing the number of valves on which maintenance was attempted and
the'number of valves with increased emissions after maintenance.
Maintenance for the valves was classified as directed or undirected.
Directed maintenance valves were those that were tightened and screened
until no further reduction in hydrocarbon emissions could be detected by
the screening instrument. Undirected maintenance valves were tightened,
but no concentration data were recorded during valve maintenance.
C 2 2 Phillips Petroleum Company Data
Equipment leak testing was performed 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
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. g
C 2 4 Union Oil Company, San Francisco Refinery Data
" ' 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 from the
valves, using a VOC detector (Model OVA - 108). Valves leaking at or
above 3,000 or 10,000 ppm by volume hydrocarbon were Identified, and
repairs were attempted. Table C-6 summarizes the number of valves
C-5
-------
Table C-2. LEAK DATA FOR THE PHILLIPS PETROLEUM COMPANY, SWEENY
REFINERY AND NATURAL GAS LIQUIDS PROCESSING COMPLEX, SWEENY, TEXAS
Action Level: >10,000 ppm
Instrument: "OVA-108" VOC detector
Distance from Source: Maximum concentration at seal interface
I. Equipment Type
Valves
Pumps
Compressor Seals
Drains
Control Valves
Open-Ended Lines
TOTALS
Number of
Equipment
Checked
2,564
190
33
150
68
420
3,425
Number of
Leaking
Equipment
222
41
1
9
13
39
325
Percent of
Equipment
Type
Leaking
8.7
21.6
3.0
6.0
19.1
9.3
Continued ..
C-6
-------
Table C-2 LEAK DATA FOR THE PHILLIPS PETROLEUM COMPANY, SWEENY
REFINERY AND NATURAL GAS LIQUIDS PROCESSING COMPLEX, SWEENY, TEXAS (Concluded)
Action Level: >10,000 ppm
Instrument: "OVA-108" VOC detector .
Distance from Source: Maximum concentration at seal interface
II. Units Tested
Number of Number of Percent
Unit
Number
, -;. i. ~..i. -i i *
4
9
10B
11
12
15
Unit
Name
-_ __ _ ' '"" '
FCCU Gas
Concentration
Crude
Distillation
NGL Manufacturing
High End Point *
Reformer
Ethyl ene
Manufacturing
Hexane
Equipment
Checked
-
297
443
178
847
1,096
564
Leaking
Equipment
18
8
13
61
170
55
i_eaKin<
Unit Te
=====
6.1
1.8
7.3
7.2
15.5
9.8
TOTALS
3,425
325
Source: Reference 3
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
Tag
Number
32
o 28
oo
16
10
7
4
367
366
364
362
Initial
Reading
>1 0,000
>10,000
>1 0,000
>1 0,000
>1 0,000
>10,000
>1 0,000
>1 0,000
>1 0,000
>1 0,000
>10,000
>1 0,000
>10,000
Date
Screened
__.. , . ., _-_.,....
03/06/79
03/06/79
03/06/79
03/06/79
03/06/79
03/06/79
03/06/79
03/06/79
03/06/79
03/05/79
03/05/79
03/05/79
03/05/79
Maintenance
Attempted
No
No
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
Undirected Directed Maintenance Readings
Maintenance
Reading 12 3
>10,000 1,100
>10,000 >10,000 >10,000
>10,000 >10,000 >10,000
2,000 100
2,000
>10,000 >10,000 >10,000 700
100
>10,000 >10,000
200
500
NCb
>10,000 >1 0,000
Comments
Only checked one valve
with tag meter lines
Only checked one valve
with tag
Only checked one valve
with tag
Repaired when valve was
backseated
Bolts all the way down
Bolts need replacing
Leak at gland, not stem -
*-i/-ivn^/-\r*T/"\m nva\/£in"!~~inn nr^nr
seating of gland
-------
Table C-3. PHILLIPS SWEENY REFINERY ETHYLENE UNIT BLOCK VALVE REPAIRS (Continued)
o
Undirected Directed Maintenance Readings
Tag Initial Date Maintenance Maintenance
Number Reading9 Screened Attempted Reading 1 2 3
360 >10,000 03/05/79
359 >10,000 03/05/79
None >10,000 03/05/79
358 >10,000 03/05/79
361 >10,000 03/05/79
None >10,000 03/05/79
356 >10,000 03/05/79
Yes
Yes
No
Yes
Yes
No
Yes
2,000
4,000
»10,000
354
352
65
64
/ _1 _
>1 0,000
>1 0,000
>1 0,000
>10,000
« J \
03/05/79
03/05/79
03/06/79
03/06/79
Yes
Yes
Yes
Yes
900
NCb
3,000
1,000
XL 0,000
»1 0,000
NC
>10,000 >10,000
>10,000 >10,000
NIT
>10,000 7,000
Comments
Mistagged originally so no
initial repair attempted
tightened bolts needs
new packing
Leak reduced but needs new
packing
Near No. 361 needs new
packing
Was not leaking before
maintenance (mistagged)
Leak detected by soap
solution missed by
instrument operator
-------
Table C-3. PHILLIPS SWEENY REFINERY ETHYLENE UNIT BLOCK VALVE REPAIRS (Concluded)
o
I
o
Tag
Number
--'_..;_.--- -"- ' :
315
311
316
313
312
314
Initial
Reading
>10,000
NCb
>1 0,000
>10,000
>10,000
XL 0,000
Date Mi
Screened >
03/06/79
03/06/79
03/06/79
03/06/79
03/06/79
03/06/79
aintena
Attempt
Yes
Yes
Yes
Yes
Yes
No
-
Undirected Directed Maintenance Readings
Maintenance
Reading 123
._.
3,000
NCb
>10,000 2,000
>10,000 >10,000
1ft
>10,000 >10,000 5,000
>1 0,000
_ =
T =
Comments
Drain still >10,000
All the
packing
All the
packing
Bad bol
replaci
: .==
way down on
way down on
ts need
ng
*A11 readings are in parts per million by volume calibrated to hexane using OVA-108 detector,
3NC = No change detected in reading above ambient level.
Source: Reference 3
-------
Table C-4. SUMMARY OF PHILLIPS SWEENY BLOCK VALVE LEAK AND REPAIR DATA
Action Level: >10,000 ppm
Instrument: "OVA-108" VOC detector 1ntprfare
Distance from Source: Maximum concentration at seal interface
1. Total number of valves with VOC >10,000 ppm 121
from unit survey
2. Total number of valves tested for 46
maintenance effectiveness
% Tested 38%
UNDIRECTED MAINTENANCE9
3. Total number subjected to repair attempts 37
4. Successful repairs (VOC <10,000 ppm) 22
% Repaired
Follow-up h
DIRECTED MAINTENANCE"
5. Number of valves unrepaired by undirected c 14
maintenance subjected to directed maintenance
6. Number repaired by follow-up directed 5
maintenance
% of unsuccessful repaired 36%
by directed maintenance
7. Total number repaired based on undirected 27
maintenance subset (3) above
% repaired 73%
8. Total number of repairs including leaks 29
not found before initial maintenance
Total % repaired 63%
Total % not repaired 37%
aValves repaired through undirected maintenance were tightened, but no
concentration data were recorded during repair.
bValves repaired through directed maintenance were tightened until VOC
emissions were below 10,000 ppmv.
cValves not repaired are valve stems found emitting volatile organic
compounds (VOC) at or above 10,000 ppmv.
Source: Reference 2
C-ll
-------
Table C-5. LEAK AND REPAIR DATA FOR REFINERY VALVES FROM THE SHELL
OIL COMPANY, MARTINEZ MANUFACTURING COMPLEX, MARTINEZ, CALIFORNIA3
Equipment: Valves
Action Level: > 10,000 ppm
Instrument: "OVA-108" VOC detector
Distance from Source: 1 cm
Valves checked 9,277
Valves leaking 293
(Percent of valves checked)
In-service valve repairs
attempted 230
Successful repairs 199
(Percent of repairs attempted) (87%)
successful repairs0 31
(Percent of repairs attempted) (13%)
Two sets of emission measurements taken before and after
repair during March and April 1979.
Successful repairs resulted in valves leaking below 10,000 ppm.
cUnsuccessful repairs are valves still leaking at or above
10,000 ppm after repair.
Source: Reference 4
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: "OVA-108" VOC detector
Distance from Source: 1 cm
Action Level (ppm)
Leaking valves
In-service repairs attempted
Qn/roccf ill rpnairs*
>3,000
300
158
107
>10,000
215
125
74
(Percent of repairs attempted
that were successful)
Valves with increased emissions
after repair
(Percent of repairs attempted
with increased emissions)
(68%)
17
(11%)
(59%)
7
(6%)
*Successful repairs resulted in valves leaking below 10,000 ppm.
Source: Reference 5
C-13
-------
"successfully repaired" that resulted in emissions below each action
level and the number of valves with increased emissions after repair.
In cases where emissions 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 (Ib/hr) from valves after repair increased 5 percent.
C.2.5 Benzene-Producing Units E and F
Units E and F are part of an intermediate size integrated petroleum
refinery located in the North Central United States. Testing was conducted
during November 1978 as part of an EPA test program to gather data on
leaking sources (defined by a VOC concentration at the leak interface
of >10,000 ppmv).
An attempt was made to screen all potential leak sources (generally
excluding flanges) on an individual component basis with a portable
organic vapor analyzer. Normally all pumps were examined, and approxi-
mately 33 to 85 percent of valves carrying VOC were screened. All tests
were performed with a Century Systems Corporation Organic Vapor Analyzer,
Model 108, with the probe placed as close to the source as possible.
Unit E is an aromatics extraction unit that produces benzene,
toluene, and xylene by extraction from refined petroleum feedstocks.
Unit E is a new unit, and special attention was given during the design
and start-up to minimize equipment leaks. All valves were repacked
before start-up, adding 2 to 3 times the original packing. All pumps in
benzene service had double mechanical seals with a barrier fluid.
Unit F produces benzene by hydrodealkylation of toluene. Unit F
was originally designed to produce a different chemical, but it was
redesigned to produce benzene.
Table C-9 presents the results of the screening study for Units E
and F. Equipment tested included valves, open-ended lines, pump seals,
and control valves. Unit E had fewer leaks at or above 10,000 ppmv than
Unit F.
C-14
-------
Table C-7. ATTEMPTED REPAIR DATA FOR VALVES
FROM THE UNION-SAN FRANCISCO REFINERY
Instrument: "OVA-108" VOC detector
Distance from Source: 1 cm
Action Level
(ppm)
0 - 999
1,000 - 9,999
>10,000
Source: Reference
Total Number
of Valves
Before Repair
0
33
125
5
Total Number
of Valves
With Increased
Emissions
After Repair
0
10
7
Percent of
Valves With
Increased
Emissions
30
6
C-15
-------
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)
Instrument: "OVA-108" VOC detector
Distance from Source: 1 cm
BEFORE REPAIR
ppmv Ib/hr
5,000
3,000
5,000
5,000
4,000
8,000
4,000
5,000
4,000
1,000
7,000
9,000
5,000
4,000
3,000
2,000
5,000
9,000
8,000
3,000
4,000
5,000
0.04
0.03
0.04
0.04
0.03
0.05
0.03
0.04
0.03
0.01
0.05
0.06
0.04
0.03
0.03
0.02
0.04
0.06
0.05
0.03
0.03
0.04
AFTER REPAIR
ppmv Ib/hr
10,000
300
100
1,000
7,000
1,000
400
100,000
1,500
1,000
4,000
4,500
400
2,000
2,000
400
1,000
700
10,000
30,000
10,000
10,000
0.06
0.01
c
0.01
0.05
0.01
0.01
0.30
0.02
0.01
0.03
0.04
0.01
0.02
0.02
0.01
0.01
0.01
0.06
0.13
0.06
0.06
C-16
-------
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
ppmv
b
lb/hr ppmv 1b/hr
b
5,000
7,000
8,000
4,000
6,000
4,000
4,000
3,000
^==^=====
0.04
0.05
0.05
0.03
0.04
0.03
0.03
0.03
8,000
1,500
2,000
1,500
10,000
9,000
1,500
100
=========
0.05
0.02
0.02
0.02
0.06
0.06
0.02
c
TOTALS
!.12 1.19 (+5%)
aSource: Reference 5
Emission rate calculations are derived from a correlation of
ppmv and lb/hr and not from actual emission rate testing.
cRate is less than 0.005 Ib/hr.
C-17
-------
TABLE C-9. FREQUENCY OF LEAKS FROM FUGITIVE
EMISSION SOURCES IN SYNTHETIC ORGANIC
CHEMICAL UNITS E AND F
Equipment Type
Uni
BTX
Number of
Sources
Tested
t Ea'b
Recovery
% with
Screening
Values
£10,000 ppmv
Unit Fa
Toluene
,b
HDA
% with
Number of Screening
Sources Values
Tested >10,000 ppmv
Valves
Open-ended lines
Pump Seals
Control Valves
715
33
33C
53
1.1
0.0
3.0
4.0
427
28
30
44
7.0
11.0
10.0
11.0
aSource: Reference 6
bNo data were available for compressor seals, safety/relief valves,
or flanges.
**
Pump seals in benzene service have double mechanical seals.
C-18
-------
C.2.6 Ethvlene and Cumene Units at Organic Chemical Manufacturing
Plants7
Three ethylene units and two cumene units were screened for fugitive
emissions as part of a program to develop data on the frequency of
occurrence of fugitive emissions from various sources in the organic
chemical manufacturing industry. Each source was screened by measuring
the maximum repeatable concentration of total hydrocarbons (expressed in
parts per million by volume) detected at the source with a portable
hydrocarbon detector (i.e., Century Systems Models OVA-108 or OVA-128).
Sources were screened by a two-person team at an average rate of
1.7 minutes per source (including instrument calibration and repair).
Tables C-10 and C-ll present the screening data by source for the cumene
and ethylene units. 8
C.2.7 Cvclohexane Unit at Exxon Chemical Company. Bavtown. Texas
Data were supplied by Exxon Chemical Company concerning fugitive
emissions at its cyclohexane unit. Sources were "bagged" to measure
emissions. The clean air apparatus was combined into a single unit to
expedite sampling by reducing set-up time. Calibration of the rotameter
for measuring gas flow was performed using a positive displacement wet
gas meter.
The total number of valves, pumps and compressor seals, and safety/
relief valves were sampled. For valves, however, a soap solution was
used to determine leaking components.
Calculations were made to translate the flow rate and concentration
data into leak rates for each component sampled from each source type.
Confidence limits were calculated at 99.8 percent, higher than normal
for engineering applications. No action level was specified.
Table C-12 presents numbers of screened sources, percentage of
sources leaking, average emission factors, and 99.8 percent confidence
intervals. g
C.2.8 Benzene-producing Unit at Refinery "E." Gulf Coast, U.S.
Leaks were measured from seals, valves, control valves, and drains
of the aromatics extraction (BTX) unit at refinery "E." A portable
hydrocarbon analyzer was used to determine the localized VOC concentration
near individual sources and the ambient VOC levels in the unit processing
areas. Ambient VOC concentrations along the refinery perimeter were
C-19
-------
Table C-10. SCREENING DATA FOR CUMENE UNITS
Action Level: slO.OOO ppmv
Distance from Source: Unspecified
Instrument: OVA-108 or OVA-128
Source
Flanges
Process
Drains
i
o Open-Ended
Valves
Safety/Relief
Valves
Pipeline
Valves
Pumps
Service9
Gas
Liquid
Gas
Liquid
Gas
Liquid
'Gas
Gas
Liquid
Liquid
Number of
Screened Sources
367
568
6
31
6
15
1
448
799
25
Number of
Sources > 10,000
ppmv
19
9
0
1
0
2
1
63
84
4
Percentage of
Sources Leaking
> 10, 000 ppmv
5.2
1.6
0.0
3.2
0.0
13.3
100.0
14.1
10.5
16.0
95% Confidence
Interval
3.2-8.1
0.7-3.0
0.0-45.9
0.1-16.7
0.0-45.9
1.7-40.5
2.5-100
11.1-17.8
8.5-12.8
4.5-36.1
aLight liquid service only.
Data were collected on two cumene units.
is a weighted average of the two units.
Source: Reference 7
The percentage of sources with total hydrocarbon concentrations >10,000 ppmv
-------
Table C-ll. SCREENING DATA FOR ETHYLENE UNITS*
Action Level: >10,000 ppmv
Distance from Source: Unspecified
Instrument: OVA-108 or OVA-128
Source
Flanges
Process
Drains
o
^ Open- Ended
1-1 Valves
Safety/Relief
Valves
Pipeline
Valves
Pumps
Compressors
a
i
Service
Gas
Liquid
Gas
Liquid
Gas
Liquid
Gas
Liquid
Gas
Liquid
Liquid
Gas
~\ n f^± t*is4 SXM 4* ti i
Number of
Sources Screened
627
407
8
407
305
214
51
11
6294
4176
78
17
hA/tA a+hwlano imi+C Poi
Number of
Sources > 10, 000
ppmv
39
25
1
4
37
41
2
1
932
969
20
1
-rontano nf ^niirrp^ Wlt.h t.i
Percentage of
Sources Leaking
>1 0,000 ppmv
6.2
6.1
12.5
1.0
12.1
19.2
3.9
9.1
14.8
23.2
25.6
5.9
nt.al hvdrnrarhnn r.nncenl
95% Confidence
Interval
4.5-8.5
4.0-8.9
0.3-52.6
0.3-2.5
8.7-16.4
11.8-21.1
0.5-13.5
0.2-41.3
13.9-15.8
21.8-24.7
16.4-36.8
0.2-28.7
trations > 10. 000
ppmv is a weighted average of the three units.
Light liquid service only.
Source: Reference 7
-------
Table C-12. SCREENING DATA FOR CYCLOHEXANE UNIT AT EXXON CHEMICAL COMPANY, BAYTOWN, TEXAS
Action Level: Unspecified
o
1
ro
Source
Valves
Gas
Liquid
Safety/Relief
Valves
Pumps
Liquid
Compressors
Number of
Screened Sources
136
100
15
8
NA
Percentage of
Sources Leaking
32
15
87
83
100
Emission Factor
(kg/hr)
0.017.
0.008°
0.064
0.255
0.264
99.8% Confidence
Interval (kg/hr)
0.008-0.035.
0.003-0.007
0.013-0.500
0.082-0.818
0.068-1.045
aSource handles light liquid only, cyclohexane unit does not handle heavy streams. Total vapor pressure is 3 psia or
higher.
bNote that the mean emission factor does not fall within the 99.8 percent confidence interval. This discrepancy existed
in the Exxon report (reference 8).
Source: Reference 8
-------
also measured and recorded. Walkthrough surveys were conducted, but due
to high background readings, the results for the BTX unit could not
easily be evaluated. Individual component surveys were conducted at
5 cm from the potential leak source. Previous studies have indicated
that a concentration of 1,000 ppmv at 5 cm is approximately equivalent
to a concentration of 10,000 ppmv at 0 cm. For the BTX unit, 6 percent
of the 122 components surveyed were found leaking at greater than 1,000 ppm
VOC emission concentration.
C.2.9 Benzene-producing Unit at the Amoco Texas Refining Company,
Texas City, Texas
Leaks were measured form seals, valves, control valves, and drains
of the aromatics extraction (BTX) unit at the Amoco Refining Company,
Texas City, Texas. A portable hydrocarbon analyzer was used to determine
the localized VOC concentration near individual sources and the ambient
VOC levels in the unit processing areas. Individual component surveys
were conducted at 5 cm from the potential leak source. Previous studies
have indicated that a concentration of 1,000 ppm at 5 cm is approxi-
mately equivalent to a concentration of 10,000 ppm at 0 cm. Of all
the equipment tested in the unit, 4.2 percent of the total valves and
15 percent of the pump seals were found to have concentrations greater
than 1,000 ppm at 5 cm.
C-23
-------
C.3 REFERENCES
1. Memorandum from B.A. Tichenor, CPB/ESED/OAQPS/EPA, to K.C. Hustvedt,
CPB/ESED/OAQPS/EPA, dated October 27, 1980. SOCMI Fugitive Emission
Sampling by Monsanto Research Corporation - Final Report.
2. U. S. Environmental Protection Agency. Air Pollution Emission
Test, Petroleum Refinery Fugitive Emissions, Phillips Petroleum
Company, Sweeny, Texas, December, 1979. EMB Report 78-OCM-12E.
Office of Air and Waste Management, Office of Air Quality Planning
and Standards, Research Triangle Park, North Carolina.
3. Equipment Summary from Phillips Petroleum Company, Sweeny, Texas.
March 14, 1979.
4. Valve Repair Summary and Memo from R.M. Thompson, Shell Oil Company,
Maritnez Manufacturing Complex, Martinez, California, to Milton
Feldstein, Bay Area Quality Management District. April 26, 1979.
5. Valve Repair Summary and Memo from F.R. Bottom!ey, Union Oil Company,
Rodeo, California, to Milton Feldstein, Bay Area Quality Management
District. April 10, 1979.
6. Hustvedt, K.C. Trip report to J.F. Durham, Chief, Petroleum Section,
U.S. Environmental Protection Agency. January S, 1979 (Plants E
and F).
7. Blacksmith, J.R., Harris, G.E., and Langley, G.J. Frequency of
Leak Occurrence for Fittings in Synthetic Organic Chemical Plant
Process Units. Final Report. Radian Corporation, Austin, Texas.
For U.S. Environmental Protection Agency, Industrial and Environmental
Research Laboratory, Research Triangle Park, North Carolina.
September 1980.
8. Letter from Cox, J.B., Exxon Chemical Company, to Walsh, R.T., EPA,
CPB. March 21, 1979. Fugitive emissions from cyclohexane unit.
9. Kelly, W. and K.C. Hustvedt. Emission Test Report, Miscellaneous
Refinery Equipment VOC Sources at Refinery "E", Gulf Coast, United
States. December, 1979. EMB Report 78-OCM-12F. Emission Standards
and Engineering Division, Office of Air Quality Planning and Standards,
U.S. Environmental Protection Agency, Research Triangle Park, North
Carolina.
10. Hustvedt, K.C., R.A. Quaney, and VI.E. Kelly. Control of Volatile
Organic Compound Leaks from Petroleum Refinery Equipment. U.S.
Environmental Protection Agency. Research Triangle Park, North
Carolina. Report Number EPA 450/2-78-036, June 1978. 72 p.
11. Memorandum from K.C. Hustvedt, EPA/CPB, to File 1.2.3.7.
November 12, 1980. Screening Data Summaries for Testing at Amoco,
Texas City, Texas (October 1977).
C-24
-------
APPENDIX D
EMISSION MEASUREMENT
AND CONTINUOUS MONITORING
D-l
-------
APPENDIX D ~ EMISSION MEASUREMENT AND CONTINUOUS MONITORING
D.I EMISSION MEASUREMENT METHODS
To develop data in support of standards for the control of fugitive
emissions, EPA conducted leak surveys at six petroleum refineries and
three organic chemical manufacturing plants. The resulting leak
determination procedures contained in Reference Method 21 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. To 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
IP
procedure has been used in several studies, * 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 with 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.
D-2
-------
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. This
information was used to formulate the test plan for initial testing.
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 the presence of a leak was a
measurement at 5 cm, with a leak definition concentration of 100 or
1000 ppmv. The localized meteorological conditions affected dispersion
significantly at greater distances. Also, it was more difficult to
define a leak at greater distances because of the small changes from
ambient concentrations observed. Surveys were conducted at 5 cm from
the source during the next three facility tests.
The procedure was distributed for comment in a draft control
techniques guideline document.5 Many commentors felt that a measurement
distance of 5 cm could not be accurately repeated during screening
tests. Since the concentration profile is rapidly changing between 0
and about 10 cm from the source, a small variance from 5 cm could
significantly affect 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 change
required that the leak definition level be increased. Additional
testing at two refineries and three chemical plants was performed by
measuring volatile organic concentrations at the interface surface.
A complication that this change introduces is that a small mass
emission rate leak ("pin-hole leak") can be totally captured by the
instrument and a high concentration result will be obtained. This
has occurred occasionally in EPA tests, and a solution to this problem
has 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. Since all analyzer types do not
D-3
-------
respond equally to different compounds, it was necessary to establish a
reference calibration material. Based on the expected compounds and the
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 10,000 ppmv levels were required. Commentors pointed
out that hexane standards at this concentration were not readily avail-
able commercially. Consequently, modifications were incorporated to
allow alternate standard preparation procedures or alternate calibration
gases in the test method recommended in the Control Techniques Guideline
Document for Petroleum Refinery Fugitive Emissions. Since that time,
additional studies have begun to develop response factor data for two
instrument types. Based on preliminary results, it appears that methane
is a more representative reference calibration material at 10,000 ppmv
levels. Based on this conclusion, and the fact that methane standards
are readily available at the necessary calibration concentrations, the
recommended calibration material for this regulation was changed to
methane.
The alternative of specifying a different calibration material for
each type stream and normalization factors for each instrument 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 potential stream compositions possible. The amount of prior
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 two or three would change the decision as to
whether or not a leak exists is small in comparison to the total number
of potential leak soirees.
An alternative approach to leak detection was evaluated by EPA
during field testing. The approach used was an area survey, or walkthrough,
using a portable analyzer. The unit area was surveyed by walking through
D-4
-------
the unit positioning the instrument probe within 1 meter of all valves
and pumps. The concentration readings were recorded on a portable strip
chart recorder. After completion of the walkthrough, the local wind
conditions were used with the chart data to locate the approximate
source of any increased ambient concentrations. This procedure was
found to yield mixed results. In some cases, the majority of leaks
located by individual component testing could be located by walkthrough
surveys. In other tests, prevailing dispersion conditions and local
elevated ambient concentrations complicated or prevented the interpre-
tation of the results. Additionally, it was not possible to develop a
general criteria specifying how much of an ambient increase at a distance
of 1 meter is indicative of a 10,000 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.
D.2 CONTINUOUS MONITORING SYSTEMS AND DEVICES
Since the leak determination procedure is not a typical emission
measurement technique, there are no continuous monitoring approaches
that are directly applicable. Continual surveillance is achieved by
repeated monitoring or screening of all affected potential leak sources.
A continuous monitoring system or device could serve as an indicator
that a leak has developed between inspection intervals. EPA performed a
limited evaluation of fixed-point monitoring systems for their effective-
ness in leak detection. The systems consisted of both remote sensing
devices with a central readout and a central analyzer system (gas
chromatograph) with remotely collected samples. The results of these
tests indicated that fixed point systems were not capable of sensing all
leaks that were found by individual component testing. This is to be
expected since these systems are significantly affected by local dispersion
conditions and would require either many individual point locations, or
very low detection sensitivities in order to achieve similar results to
those obtained using an individual component survey.
D-5
-------
It is recommended that fixed-point monitoring systems not be
required since general specifications cannot be formulated to assure
equivalent results, and each installation would have to be evaluated
individually.
D.3 PERFORMANCE TEST METHOD
The recommended benzene fugitive emission detection procedure is
Reference Method 21. This method incorporates the use of a portable
analyzer to detect the presence of volatile organic vapors at the
surface of the interface where direct leakage to atmosphere could occur.
The approach of this technique assumes that if an organic leak exists,
there will be an increased vapor concentration in the vicinity of the
leak, and that the measured concentration is generally proportional to
the mass emission rate of the organic compound.
An additional procedure provided in Reference Method 21 is for the
determination of "no detectable emissions." The portable VOC analyzer
is used to determine the local ambient VOC concentration in the vicinity
of the source to be evaluated, and then a measurement is made at the
surface of the potential leak interface. If a concentration change of
less than 2 percent of the leak definition is observed, then a "no
detectable emissions" condition exists. The definition of 2 percent of
the leak definition was selected based on the readability of a meter
scale graduated in 2 percent increments from 0 to 100 percent of scale,
and not necessarily on the performance of emission sources. "No
detectable emissions" would exist when the observed concentration change
between local ambient and leak interface surface measurements is less
than 200 ppmv.
The test procedure does not detect benzene specifically; instead,
the volatile organic compound concentration is measured. There is
commercially available one type of portable analyzer that has the capa-
bility 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
D-6
-------
benzene would not yield additional information since the affected
facilities are those in which benzene is transported and a measure of
organic vapor leakage is indicative of a benzene leak.
Reference Method 21 does not include a specification of the
instrument calibration basis or a definition of a leak in terms of
concentration. Based on the results of EPA field tests and laboratory
studies, methane is recommended as the reference calibration basis for
benzene fugitive emission sources in the refining and organic chemical
manufacturing industries.
There are at least four types of detection principles currently
available in commercial portable instruments. These are flame ionization,
catalytic oxidation, infrared absorption (NDIR), and photoionization.
Two types (flame ionization and catalytic oxidation) are known to be
available in factory mutual certified versions for use in hazardous
atmospheres.
The recommended test procedure includes a set of design and
operating specifications and evaluation procedures by which an analyzer's
performance can be evaluated. These parameters were selected based on
the allowable tolerances for data collection, and not on EPA evaluations
of the performance of individual instruments. Based&on manufacturers'
literature specifications and reported test results, commercially
available analyzers can meet these requirements.
The estimated purchase cost for an analyzer ranges from about
$1,000 to $5,000 depending on the type and optional equipment. The cost
of an annual monitoring program per unit, including semiannual instrument
tests and reporting is estimated to be from $3,000 to $4,500. This
estimate is based on EPA contractor costs experienced during previous
test programs. Performance of monitoring by plant personnel may result
in lower costs. The above estimates do not include any costs associated
with leak repair after detection.
D-7
-------
D.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 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 Number EPA-600/2-79-044.
February 1979.
3. Telecon. Harrison, P., Meteorology Research, Inc. with Hustvedt,
K.C., EPA, CPB. December 22, 1977.
4 Miscellaneous Refinery Equipment VOC Sources at ARCO, Watson
Refinery, and Newhall Refining Company. U.S. Environmental
Protection Agency, Emission Standards and Engineering Division.
Research Triangle Park, NC. EMB Report Number 77-CAT-6.
December 1979.
5. 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.
QAQPS Guideline Series. Report Number EPA-450/2-78-036. June 1978.
6 Letter from McClure, H.H., Texas Chemical Council, to Barber, VI.,
EPA, OAQPS. June 30, 1980.
D-8
-------
APPENDIX E
METHODOLOGY FOR ESTIMATING LEUKEMIA MORTALITY AND MAXIMUM LIFETIME
METHODOLOGY hOK "*^ FUGmvE BENZENE EMISSIONS FROM
PETROLEUM REFINERIES AND SYNTHETIC ORGANIC CHEMICAL
MANUFACTURING PLANTS
-------
-------
APPENDIX E
METHODOLOGY FOR ESTIMATING LEUKEMIA MORTALITY AND MAXIMUM LIFETIME
RISK FROM EXPOSURE TO FUGITIVE BENZENE EMISSIONS FROM
PETROLEUM REFINERIES AND SYNTHETIC ORGANIC CHEMICAL
MANUFACTURING PLANTS
E.I INTRODUCTION
The purpose of this appendix is to describe the methodology used in
estimating leukemia mortality and maximum lifetime risk attributable to
population exposure to fugitive benzene emissions from petroleum refineries
and synthetic organic chemical manufacturing plants. The appendix is
presented in three parts:
Part E.2, Summary and Overview of Health Effects, summarizes and
references reported health effects from benzene exposure. The
major reported health effect is leukemia. Mortalities cited in
the BID include only the -estimated leukemia deaths attributable
to exposure to benzene emissions from existing petroleum refiner-
ies and synthetic organic chemical plants although other, some-
times fatal, effects are known to result from benzene exposure.
Part E.3, Population Density Around Petroleum Refineries and
Synthetic Organic Chemical Plants, describes -the method used to
estimate the population at risk; i.e., persons residing within
20 km of existing petroleum refineries and synthetic organic
chemical plants.
Part E.4, Population Exposures, Mortalities, and Risks, describes
the methodology for estimating fugitive benzene emissions from
model plants, calculating expected population exposures, and
estimating number of leukemia deaths and maximum risk of leukemia
attributable to fugitive benzene emissions from 133 existing U.S.
petroleum refineries and synthetic organic chemical plants.
i
E.2 SUMMARY AND OVERVIEW OF HEALTH EFFECTS
E.2.1 Health Effects Associated with Benzene Exposure
A large number of occupational studies over the past 50 years have
provided evidence of severe health effects in humans from prolonged inhala-
E-l
-------
tion exposure to benzene. Some 300 studies of the health effects of
benzene have recently been reviewed and analyzed in terms of application to
low-level ambient benzene exposures that might occur in a population resid-
ing near a source of benzene emissions.
The reviewers concluded that benzene exposure by inhalation is strongly
implicated in three pathological conditions that may be of public health
concern at environmental exposure levels:
Leukemia (a cancer of the blood-forming system),
Cytopenia (decreased levels of one or more of the formed elements
in the circulating blood), and
Chromosomal aberrations.
Leukemia is a neoplastic proliferation and accumulation of white blood
cells in blood and bone marrow. The four main types are acute and chronic
myelogenous leukemia and acute and chronic lymphocytic leukemia. The
causal relationship between benzene exposure and acute myelogenous leukemia
and its variants in humans appears established beyond reasonable doubt.
The term "pancytopenia" refers to diminution of all formed elements of
the blood and includes the individual cytopenias: anemia, leukopenia,
thrombocytopenia, and aplastic anemia. In mild cases, symptoms of pancyto-
penia are such nonspecific complaints as lassitude, dizziness, malaise, and
shortness of breath. In severe cases, hemorrhage may be observed, and
death may occasionally occur because of hemorrhage or massive infection.
Patients with pancytopenia may subsequently develop fatal, acute leukemia.
Chromosomal aberrations include chromosome breakage and rearrangement
and the presence of abnormal cells. These aberrations may continue for
long periods in hematopoietic and lymphoid cells. Ample evidence exists
that benzene causes chromosomal aberrations in somatic cells of animals and
p
humans exposed to benzene. The health significance of these aberrations
is not fully understood. However, aberrant cells have been observed in
individuals exposed to benzene who have later developed leukemia. Some
types of chromosomal aberrations may be heritable. Quantitative estimates
of heritable genetic damage due to benzene cannot be made from data on the
frequency of somatic mutations, although this damage may be occurring at
concentrations as low as 1 ppm in air.
E-2
-------
The review1 concluded that man may be the only species yet observed to
be susceptible to benzene-induced leukemia. Evidence for production of
leukemia in animals by benzene injection was considered nonconclusive.
Moreover, benzene exposure by oral dosing, skin painting, or inhalation has
not been shown to produce leukemia or any other type of neoplastic diseases
in test animals, although other effects, including pancytopenia, have been
widely observed.
E.2.2 Benzene Exposure Limits
It should be noted that where the health effects described above have
been associated with benzene exposure, the exposure has been at occupa-
tional levels. That is, the benzene exposure levels associated with the
effects have been high (10 ppm up to hundreds of parts per million of
benzene, except in a few cases of exposures to 2 to 3 ppm benzene) or they
have been unknown.
Benzene exposure was first associated with health effects in occupa-
tional settings, so initial attempts to limit benzene exposures were aimed
at occupational exposures. With recognition of the toxic effects of ben-
zene and its greatly expanded use after 1920, several occupational exposure
limits were established in the United States.3 These limits, originally in
the range of 75 to 100 ppm, were successively lowered as more information
on benzene toxicity became kgown.
For example, the American Conference of Governmental Industrial Hygien-
ists (ACGIH) recommended a benzene threshold limit value of 100 ppm in
1946, 50 ppm in 1947, 35 ppm in 1948, 25 ppm in 1949, and 10 ppm in 1977. '
The National Institute for Occupational Safety and Health (NIOSH) recommended^
an exposure limit of 10 ppm in 1974 and revised it downward to 1 ppm in 1976.
The current Occupational Safety and Health Administration (OSHA) permissible
exposure limit is 10 ppm6 (a lower limit of 1 ppm is currently in litigation*)
Occupational exposure limits were initially established to protect
workers from adverse changes in the blood and blood-forming tissues. The
*A benzene standard with a limit of 1 ppm was proposed by OSHA May 27,
1977 (42 FR 27452) and promulgated February 10, 1978 (43 FR 5918). This
standard was struck down October 5, 1978, by the U.S. Fifth Circuit Court
of Appeals. The U.S. Department of Labor appealed the decision, and the
Supreme Court agreed to hear arguments on the case^during its fall, 19/9,
term. No decision has been announced at this writing.
E-3
-------
most recently recommended or pending limits of 1 ppm and 1.0 ppm are based
5U
,4
c 7
on the conclusion that benzene is leukemogenic in man (NIOSH and OSHA ) or
a suspected carcinogen in man (ACGIH ).
E.2.3 Health Effects at Environmental Exposure Levels
Little information is available on health effects of nonoccupational
exposures of the general populace to benzene. Virtually all of the studies
1 2
cited ' were on the working population (mostly males) exposed to higher
than ambient benzene levels on a work cycle. Applying these studies to
chronic (24 hours per day) low-level exposure to the general population
(including infants, the ill, and the elderly) requires extrapolation.
The recent analysis of benzene health effects concluded that the
evidence of increased risk of leukemia in humans on exposure to benzene for
various time periods and concentrations was overwhelming but that data were
not adequate for deriving a dose-response curve.
However, EPA's Carcinogen Assessment Group (CAG), acknowledging the
absence of a clear dose-response relationship, has estimated the risk of
2
leukemia in the general population from low-level benzene exposure. Data
from three epidemiological studies of leukemia in workers (mostly adult
white males) were used to estimate the risk of developing leukemia. The
annual risk factor derived for benzene-induced leukemia was 0.34 deaths per
year per 10 ppb-person years of exposure.
A no-threshold linear model was used to extrapolate this estimated
risk to the low levels (below 5 to 10 ppb) to which some populations may be
exposed. For example, if 3 million persons are chronically exposed to I
ppb benzene, the model predicts there will be 1.02 leukemia deaths (3 x
0.34) per year in that population. Use of a "linear" model means that the
model would predict the same number of leukemia deaths among 3 million
people exposed to 1 ppb benzene as among 1 million people exposed to 3 ppb.
The risks factor (0.34 deaths per year per 10 ppb-person years) was
used in estimating the number of leukemia deaths attributable to benzene
emissions from petroleum refineries and synthetic organic chemical plants.
Other effects of benzene exposure (including deaths from causes other than
leukemia) were not included in the estimated number of deaths. The risk
factor equated one leukemia case to one death (that is, each case was
presumed fatal).
E-4
-------
Several sources of uncertainty exist in applying the risk factor.
First, the retrospective occupational exposure estimates may be inaccurate.
CAG calculated the 95-percent confidence intervals for this risk factor to
be 0.17 to 0.66 deaths per 106 ppb-person years if exposure estimates in
the three studies extrapolated are precisely correct, and 0.13 to 0.90 if
exposure estimates are off by a factor of 2. Second, the composition of
the exposed populations around petroleum refineries and synthetic organic
chemical plants may vary from that of the populations used as a basis for
the CAG estimate; the risk factor assumes that the susceptibility to leuke-
mia associated with a cohort of white male workers is the same as that
associated with the general population, which includes women, children, the
aged, nonwhites, and the ill. Third, the true dose-response relationship
for benzene exposure may not be a linear no-threshold relationship at the
low concentrations to which the general population may be exposed.. Fourth,
the risk factor includes only leukemia deaths and not other health risks.
No quantitative estimate of the uncertainty in the risk factor due to the
latter three factors has been attempted.
E.3 POPULATION DENSITY AROUND PETROLEUM REFINERIES AND SYNTHETIC ORGANIC
CHEMICAL PLANTS
The population "at risk" to benzene exposure was considered to be
persons residing within 20 km of petroleum refineries and synthetic organic
chemical plants. Populations residing within radial distances of 1, 5, 10,
and 20 km from each plant were estimated from an existing population file.
This file consists of a grid of 1-km2 cells covering the continental United
States, each with an assigned population. The population assigned to each
cell was the 1975 estimated population, extrapolated from the 1960 and 1970
populations of the census enumeration district in which each cell occurs,
assuming that the population is uniformly distributed within each of the
256,000 census enumeration districts. The population around each plant was
determined by summing the populations of all cells occurring in annular
areas at radial distances from the plant center of 0.5 to 1 km, 1 to 5 km,
5 to 10 km, and 10 to 20 km. The estimated total populations exposed as a
function of distance from the plant site are reported in Reference 8, Table
A-5.
E-5
-------
There are some uncertainties in the above method. First, the assump-
tion of uniform population distribution, both within enumeration districts
and annular areas, may not be precisely correct. For urban areas the
assumption is probably reasonably valid, but there is some uncertainty for
rural areas 10 to 20 km from the site. Another area of uncertainty is the
use of 1960 and 1970 population data. However, these are the latest avail-
able in the form required. No attempt, was made to quantify the range of
variability in the population figures.
E.4 POPULATION EXPOSURES, MORTALITIES, AND RISKS
E.4.1 Summary of Methodology for Calculating Deaths
The locations and descriptions of all 133 known U.S. benzene-using and
Q
benzene-producing plants were compiled. From these data, a basic "model"
plant was developed. The model plant was assumed to be an intermediate-size
processing complex along the Texas-Louisiana Gulf Coast, with 20 benzene-
emitting units placed at various plant locations. Fugitive benzene
emission rates from equipment leaks at these sources were estimated from
data available at the time this analysis was performed.
The omnidirectional and maximum annual average benzene concentrations
in ambient air resulting from each benzene-emitting process were determined
to a distance of 20 km outsi.de the model plant boundary, according to the
Industrial Source Complex (ISC) dispersion model, urban mode I.10 Omni-
directional annual average benzene concentrations are averages for all
directions from the source. They take into account changes in wind direc-
tion throughout the year and so are lower than maximum annual averages,
which are the concentrations downwind from the plant in the prevailing wind
direction.
Houston meteorological data for the 1973-1975 period were used in the
dispersion model. This period was considered representative of poor disper-
sion conditions in the area, in order to develop a potential worst-case
situation.
Ambient air benzene concentrations were estimated originally based on
available data for benzene emission rates from petroleum refineries. These
concentrations were proportioned with recently developed data on benzene
Q
emission rates from refineries and synthetic organic chemical plants.
E-6
-------
Several variations of the original model plant were developed on the
basis of different combinations of units currently operating in the 133 exist-
ing plants. Ambient air benzene concentrations at several distances from
each model plant were determined by summing the benzene concentrations (at
each distance) contributed by all the processes incorporated into each
particular model variation.8'9 The result was a range of 56 models, each
of which closely resembled an existing plant. Each model was then matched
with an appropriate plant, on the basis of unit operations.
The population around each actual plant location was then correlated
with its modeled benzene concentrations to yield a benzene dose to that
population in ppb-person years. The methods for determining populations
are described in Part E.3 of this appendix.
From health effects data, the EPA Carcinogen Assessment Group derived
a leukemia risk estimate of 0.34 deaths per year per 106 ppb-person years
from exposure to benzene. The methodology for estimating the leukemia risk
factor is described in Part E.2.3 of this appendix.
The leukemia deaths per year attributable to exposure to fugitive ben-
zene emissions from petroleum refineries and synthetic organic chemical
manufacturing plants were estimated by multiplying 0.34 x 10 deaths per
year per ppb-person year exposure times the exposure in "ppb-person years,"
as described in Part E.4.2. ,The leukemia deaths so calculated are summar-
ized in Table E-l for each plant, with a total for all plants of 0.42
deaths per year.
E.4.2 Estimates of Leukemia Deaths
The general equation for estimating the number of leukemia deaths
attributable to fugitive benzene emissions from a particular plant (e.g.,
Plant X) is:8
D = b+2 - b+2
x
i = 0.5-1.0
(RX1/3.2) (2)(n)(p,)(a)(D.b+2 - D.b+2)/(b+2) , (1)
'21
in which
D = estimated number of leukemia deaths per year from fugi-
x tive benzene emissions from the plant (e.g., Plant X).
E-7
-------
R = the risk factor (0.34 deaths per year per 10 ppb-person
years).
p. = density of population at risk, in area (i) around Plant
1 X.
D. and D. = distances from plant to outer edge (D. ) and inner
i2 il i2
edge (D. ) of area i (e.g., for the area 5-10 km from
h
the plant, D. = 10 km and D. = 5 km).
i2 T!
(1/3.2) = factor converting ug/m3 to ppb, the units in which R
is expressed.
a and b = values describing the dispersion pattern of benzene in.
air around Plant X, according to the equation Bj = a D-,
in which B. is the benzene concentration at distance D^
from the plant. Values of a and b are unique to each
annular area i around each model plant.
i = the particular area in which p. occurs (i progresses
from the area 0.5 to 1.0 km from the plant to the area
10 to 20 km from the plant).
I = summation of deaths per year from all areas (i).
This equation is a mathematically rigorous method for estimating the
exposure to the population within any area between i-, and ip km from the
plant, taking into account that with constant population density (p.) more
people reside near the outer edge of the area than near the inner edge, and
that the benzene concentration (B^) decreases with distance from the plant.
The equation is derived in Reference 8.
Values of a and b were calculated for each annular area for each model
plant as follows:
ln(B. /B. )
- 2 nl . (2)
ln(Di /Di )
a = B, /(D, )b , (3)
in which Bi-, is the benzene concentration at the inner edge of area i
n B
(i.e., at distance i-i), and i'2 is the benzene concentration at the outer
E-8
-------
edge of area i (i.e., at distance Di2). Bi values for each distance (D.)
from each model plant are listed in Reference 8.
Population density (p.) for a particular annular area around a partic-
ular plant is obtained by dividing the total population in that area (P^
by the area in square kilometers; i.e.:
2 - D . (4)
In summary, Pi values for each plant and annular area are listed in
Reference 8. For each annular space around a particular plant, 12 and ^
values are taken from Reference 8. BI values at all distances (D^ are
also taken from Reference 8. Values of b, a, and p. are calculated from
Equations 2, 3, and 4 for each annular area. Then, with Equation 1, expo-
sures in ug/m3-person years are calculated for each annular area, divided
by 3.2 to convert pg/m3 to ppb, the units in which R is expressed, and
multiplied by R to yield the number of deaths in each annular area. These
deaths are summed to give Dv, the annual leukemia deaths attributable to
/\
benzene emissions from Plant X.
The total estimated number of leukemia deaths per year attributable to
benzene emissions from all plants was determined by the equation:
Total estimated number _
of leukemia deaths/yr (Dt) = DI + D£ + . . . + D133 . W
The total numbers of estimated leukemia deaths attributable to fugitive
benzene emissions from petroleum refineries and synthetic organic chemical
manufacturing plants are given in the last column of Table E-l on a plant-
by-plant basis, in deaths per year, assuming current control conditions for
benzene emissions. The number of deaths expected under each of the control
alternatives can also be derived with the same methodology.
E.4.3 Example of Leukemia Death Calculation
Plant no. 4 was chosen for an example calculation of the number of
leukemia deaths attributable to plant fugitive benzene emissions. For a
determination of the number of deaths according to Equation 1, numerical
values are needed for R, a, p., Di2, Di1} and b. In turn, for a determina-
tion of p. from Equation 4, the numerical value of PI must be known for
E-9
-------
each annular area. For a determination of b and a from Equation 2 and
B B
Equation 3, numerical values of ip and i'2 must be known for each dis-
tance.
Calculations are shown in Table E-2. The values in the first three
lines of Table E-2 are common to all plants. They show the distances at
which concentrations and populations were measured and the risk factor (R).
Line 4 shows the population (Pp in each annular ring, obtained from
Table A-5, Reference 8, for Plant 4.
Lines 5 and 6 show the benzene concentrations at various distances
from the plant for the applicable model. Table A-l, Reference 8, indicates
that Model 2 is used for Plant 4. From Table A-2, Reference 8, the benzene
concentrations at each distance are found for Model 2.
Note that concentrations for 1, 5, and 10 km apply to the outer edge
( i2) of one ring and the inner edge (Bi;L) of the adjacent ring. Note also
that the concentrations at 0.1 km are not used in these calculations. The
population within 1 km of a plant is assumed to reside in the area 0.5 to
1 km from the plant.
Lines 7 through 11 show the calculations. These are shown below for
the outer ring. From Equation 4:
P,- = PjE/TK0^ - °i2)] ~ 4,515,175/n(202 - 102) = 4,791 .
From Equation 2:
b = ln(Bi2/Bi1)/ln(Di2/Di1) = ln(0.0251/0.0695)/ln(20/10) = -1.469 .
From Equation 3:
a = Bi2/°i2b = 0.0251/C20)"1-469 = 2.048 .
All the values needed for using Equation 1 are now available, so:
Deaths for 10- to 20-km ring only = (R/3.2)2npia(Di^+2 - DiJ+2)/(b+2) ,
D10_20 = (0.34 x 10~5/3.2)2n(4,791)(2.048X200'531 - 10°'531)/0.531 , and
E-10
-------
D = 0.0182 (0.018623 if decimals are retained during calculations) .
The same calculations are made for ranges 0.5 to 1 km, l^to 5 km, and
5 to 10 km. The total annual leukemia deaths for the plant (Dx) are the
sum of the deaths for each ring; i.e.:
Deaths for Plant 4 (U4) = K^^ + ^.5 + &5-10 + U10-20) '
D. = 0.001391 + 0.010064 + 0.011609 + 0.018623 .
D4 = 0.041687 deaths/yr .
Deaths attributable to fugitive benzene emissions from any of the 133
plants may be calculated in the same manner.
E.4.4 Estimate of Leukemia Risk
The estimated leukemia deaths shown in Table E-l are based on estimates
of mean annual average benzene concentrations around petroleum refineries
and synthetic organic chemical manufacturing plants. Because atmospheric
dispersion patterns are not uniform, some population groups will receive
above-average benzene exposures and will therefore incur a higher risk (or
probability) of developing leukemia.
Maximum annual risk is the estimated probability to a person, who is
constantly exposed to the highest maximum annual average benzene concentra-
tion in the ambient air around a particular source for a year, of develop-
ing leukemia because of exposure to benzene emissions from that source.
Maximum lifetime risk is estimated by multiplying the maximum annual risk
by 70 years.
The maximum lifetime risk of leukemia was calculated for this person,
who is assumed to reside at the point of highest maximum annual average
benzene concentration outside the model plant with the greatest benzene
emissions. This was found to be Model 14 (Table A-2, Reference 8), repre-
senting a plant manufacturing nitrobenzene and chlorobenzene. Model 14 was
matched to one existing plant-the Monsanto facility at Sauget, Illinois.
The maximum risk of leukemia associated with these emissions is calculated
as follows.
First, the fugitive benzene emissions from both of these units are
determined'(Reference 9, Table 1-1): 15.79 kg/hr benzene from the nitro-
E-ll
-------
benzene unit and 9.49 kg/hr from the chlorobenzene unit. These units are
then assumed to be located in the positions occupied by the highest and
second highest benzene-emitting units in the model plant, in order to be
consistent with the original model and so ensure a corresponding maximum
benzene concentration and location. These are Groups 18-1, 2, and 3, and
17-1, 2, and 3 (Table 1, Reference 10), with 0.124 g/sec total fugitive
benzene emissions from Group 18 and 0.0858 g/sec from Group 17.
The maximum annual average benzene concentrations associated with
fugitive emissions from Groups 18 and 17 are then taken from Table 3,
Reference 10. For both groups, the highest maximum concentrations occur at
_1 O
0.1 km outside the plant boundary: 8.74 x 10 ug/m benzene due to emis-
-1 3
sions from Group 18, and 9.46 x 10 ug/m from Group 17. These emissions,
however, are based on the basic model plant emission rates stated in the
preceding paragraph (0.124 and 0.0858 g/sec) and must be scaled up, propor-
tionately to reflect the benzene emission rates actually estimated for the
nitrobenzene and chlorobenzene units (15.79 and 9.49 kg/hr). This process
is carried out as follows.
FOR NITROBENZENE
Fugitive benzene emissions estimated
from nitrobenzene unit (Reference 9,
Table 1-1)
Fugitive benzene emissions from
Group 18 (from Table 1, Reference 10)
Maximum annual average benzene
concentration due to Group 18
(from Table 3, Reference 10)
Maximum annual average benzene
concentration from nitrobenzene
unit emissions in Location 18
FOR CHLOROBENZENE
Fugitive benzene emissions estimated
from chlorobenzene unit from (from
Table 1-1, Reference 9)
= 15.79 kg/hr (4.39 g/sec)
= 0.124 g/sec
8.74 x lo'1 ug/m3
= (8.74 x 10"1)(4.39/0.124)
= 30.94 ug/m3
= 9.49 kg/hr (2.64 g/sec)
E-12
-------
Fugitive benzene emissions from
Group 17 (from Table 1, Reference 10) = 0.0858 a/sec
Maximum annual average benzene
concentration due to Group 17 -1 3
(from Table 3, Reference 10) = 9-46 x 10 Mg/m
Maximum annual average benzene
= 29.11 ug/m3
Summing maximum concentrations from both processes gives:
Maximum annual average benzene _ ,3
concentration from Model plant 14 = 30.94+29.11 - 60.05 ug/m .
This figure is converted from pg/m3 to ppb by dividing by 3.2, so
Maximum annual average benzene
concentration from Model plant 14 = 60.05/3.2 - 18.7 ppb.
This figure, 18.7 ppb, indicates that the person most exposed to
benzene from any of the 133 plants resides 0.1 km from the boundary of
Model plant 14 and receives an exposure of 18.7 ppb continuously, or for
1 person year annually. By applying the risk factor of 0.34 x 10 deaths
per year per ppb-person year to this exposure, the annual risk can be
calculated, viz: *
Maximum annual = (Q 34 x 10"6 deaths per year/ppb-person year) x
risk of leukemia ^7 ppb-person years), or
Maximum annual = 6 36
risk of leukemia
Because lifetime risk is expressed as a probability to one person of
dying of leukemia, the units have been deleted for convenience. Technically,
the number represents deaths per year for one person. The lifetime risk of
leukemia, assuming a 70-year lifespan, is simply 70 times the annual risk,
or:
Maximum lifetime = (6 36 x 10"5)(70) = 4.45 x 10"4.
risk of leukemia
E-13
-------
The risk associated with emissions from any specific plant or model plant
may be calculated in the same manner.
E.4.5 Validity of Estimates
Several uncertainties exist in the estimated number of leukemia deaths
and the maximum leukemia risk. Primary sources of uncertainty are in the:
Risk factor (R),
Populations at risk,
Estimated benzene concentrations around plants, and
Benzene exposure calculations.
Uncertainties in the risk factor (R) are discussed in Part E.2.3, and
uncertainties in populations "at risk" (P.) are discussed in Part E.3. The
other factors are discussed below.
E.4.5.1 Estimated Benzene Concentrations. The estimated benzene
concentrations are derived from several factors, as follows:
Configuration of the model plant,
Emission rates from the model plant, and
Dispersion patterns of the emissions.
Uncertainties associated with these factors could not be quantified,
but their qualitative effects on the estimated number of leukemia deaths
are discussed below.
The configuration of the model plant assumes a petroleum refinery of
2
given area (2.5 km ), with equipment at specific locations and an effective
emission height for all fugitive emissions of 2.5 m. Current fugitive
Q
emissions from existing plant processes were estimated and uniform emission
rates assumed. When the basic model plant was applied to existing plant
emissions, it was assumed that benzene concentrations in air varied in
direct proportion to plant emissions.
Several sources of uncertainty occur in this model. First, it is
unlikely that any plant duplicates the model plant precisely, so variation
in locations of units within plant boundaries may be expected. Second, the
model used 1973-1975 weather data from Houston (a coastal city) to project
dispersion patterns for all existing plants, and did not take into account
the effects of terrain. Thus, when applied to hilly, inland areas, the
model may introduce inaccuracies. Third, the model assumes there is no
E-14
-------
loss of benzene from atmospheric reactions or ground level absorption. If
such losses occur, the actual concentration of benzene will be less than
the estimated values. Fourth, the model was originally developed for
refineries and has been applied to synthetic organic chemical manufacturing
facilities as well.
A final source of uncertainty is that the model measures benzene
dispersion only to 20 km. If the linear risk model2 is accurate, exposures
at distances greater than 20 km, however small, may be important. If such
exposures occur, the estimated number of deaths would be higher than esti-
mated here.
It is estimated that benzene concentrations predicted by the disper-
sion model are accurate to within a factor of 2,10 barring large inaccur-
acies in estimated benzene emission rates.
E.4.5.2 Benzene Exposure Calculations. Benzene exposure calculations
assume that persons at specific locations are exposed 100 percent of the
time to the benzene concentrations estimated to occur at each location.
The assumption of continuous exposure to residents introduces some uncer-
tainty, both in estimated number of leukemia deaths and in maximum leukemia
risk. No numerical estimates of potential variation are available. Further-
more, the maximum lifetime risk assumes that a particular plant operates at
full capacity for 70 years. >
There is a discrepancy between the methods used to measure distance
from the plant for benzene concentrations and for populations. Benzene
concentrations are measured from the plant boundary, and geographic rings
of population are measured from the plant center. This discrepancy intro-
duces some imprecision in the "ppb-person years" benzene exposure calculations
used to estimate the number of leukemia deaths. The maximum lifetime risk
estimate is not affected.
E-15
-------
TABLE E-l ESTIMATED LEUKEMIA DEATHS FROM FUGITIVE BENZENE EMISSIONS FROM
PETROLEUM REFINERIES AND SYNTHETIC ORGANIC CHEMICAL MANUFACTURING PLANTS
UNDER CURRENT CONTROL CONDITIONS8
I
t'
cr>
Plant
code
no.
Region
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
Region
15.
16.
17.
18.
Company (by region)
II
American Cyanamid
DuPont
Reichhold
Standard Chlorine
Tenneco
Texaco
Allied Chemical9
Ashland Oil
ICC Industries
Commonwealth Oil
Phillips Puerto Rico
Puerto Rico Olefins
Union Carbide
Amerada Hess
III
Getty
Standard Chlorine
Sun-Olin
Continental Oil
State
NJ
NJ
NJ
NJ
NJ
NJ
NY
NY
NY
PR
PR
PR
PR
VI
DE
DE
DE
MD
Location
City
Boundbrook
Gibbstown
Elizabeth
Kearny
Fords
Westville
Solvay
North Tonawanda
Niagara Falls
Penuelas
Guyana
Penuelas
Penuelas
St. Croix
Delaware City
Delaware City
Claymont
Baltimore
County
Somerset
Glouchester
Union
Hudson
Middlesex
Glouchester
Onondaga
Niagara
Niagara
New Castle
New Castle
New Castle
Baltimore
Model
number
10
10
9
2
9
30
2
21
2
45
46
4
25
24
23
2
4
8
Leukemia
deaths per
year from
benzene,
exposure
0.009642
0.018525
0.004384
0.041687
0.001759
0.014268
0.002042
0.000130
0.004375
0.004001
0.004001
0.000575
0.004516
0.001568
0.001063
0.001999
0.001184
0.008464
-------
TABLE E-l (continued)
Plant
code
no.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
Region
34.
35.
36.
37.
38.
Company (by region)
Atlantic Richfield
Gulf Oil
Koppers
Koppers
Nease Chemical
Standard Oil
(Ohio)/BP Oil
Sun Oil
U.S. Steel
Allied Chemical
American Cyanamid
Ashland Oil
Mobay Chemical
PPG
PPG
Union Carbide
IV
Jim Walter Resources
Reichhold Chemicals
Ashland Oil
B. F. Goodrich
01 in Corporation
State
PA
PA
PA
PA
PA
PA
1
PA
PA
WV
WV
WV
WV
WV
WV
WV
AL
AL
KY
KY
KY
Location
City
Beaver Valley
Philadelphia
Bridgeville
Petrol i a
State College
Marcus Hook
»
Marcus Hook
Neville Island
Moundsville
Willow Island
Neal
New Marti nsvi lie
Natrium
New Martinsville
Institute
Birmingham
Tuscaloosa
Ashland
Calvert City
Brandenburg
County
Beaver
Philadelphia
Allegheny
Butler
Centre
Delaware
Delaware
Allegheny
Marshall
Pleasants
Wayne
Wetzel
Marshall
Wetzel
Kanawha
Jefferson
Tuscaloosa
Boyd
Marshall
Mead
Model
number
12
52
9
11
1
48
56
9
10
10
9
10
2
2
19
1
1
49
4
4
Leukemia
deaths per
year from
benzene.
exposure
0.000258
0.029612
0.001924
0.000050
0.000281
0.005566
0.006708
0.001681
0.002003
0.000916
0.000166
0.000671
0.000153
0.000375
0.001677
0.001945
0.000288
0.002214
0.000040
0.000049
See footnotes at end of table.
(continued)
-------
TABLE E-l (continued)
Plant
code
no.
39.
40.
Region
41.
42.
43.
44.
7 45.
i1
00
46.
47.
48.
49.
50.
51.
52.
53.
Region
54.
55.
Company (by region)
Chevron
First Chemical
V
Clark Oil
Core- Lube
Koppers
Monsanto
National Distiller
(U.S.I.)
Northern Petrochemicals
Reichhold Chemicals
Shell Oil
Standard Oil ,
(Indiana)/Amoco
Union Oil (California)
Dow Chemical
Dow Chemical
Sun Oil
VI
Vertac/Transvaal
Allied Chemical
State
MS
MS
IL
IL
IL
IL
IL
IL
IL
IL
IL
IL
MI
MI
OH
AR
LA
Location
City
Pascagoula
Pascagoula
Blue Island
Danville
Cicero
Sauget
Tuscola
Morris
Morris
Wood River
Joliet
Lemont
Bay City
Midland
Toledo
Jacksonville
Geismar
County
Jackson
Jackson
Cook
Vermilion
Cook
St. Clair
Douglas
Grundy
Grundy
Madison
Will
Cook
Bay
Midland
Lucas
Pulaski
Ascension
Model
number
21
10
3
1
9
14
4
4
9
21
9
23
26
13
56
2
4
Leukemia
deaths per
year from
benzene.
exposure
0.000690
0.002044
0.001235
0.000262
0.007317
0.037257
0.000033
0.000081
0.000076
0.001555
0.000367
0.002669
0.003289
0.002391
0.006263
0.000045
0.000090
See footnotes at end of table.
(continued)
-------
TABLE E-l (continued)
Plant
code
no. Company (by region)
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
See
American Hoechst
Cities Service
Continental Oil
Cos-Mar, Inc.
Dow Chemical
Exxon
Gulf Coast Olefins
Gulf Oil
Gulf Oil
Pennzoil United (Atlas
Processing)
Rubicon
Shell Oil
Tenneco
Union Carbide
Sun Oil
American Hoechst
American Petrofina of
Texas
American Petrofina
(Cosden Oil)
footnotes at end of table.
State
LA
LA
LA
LA
LA
LA
LA
LA
LA
LA
LA
LA
LA
LA
OK
TX
TX
TX
Location
City
Baton Rouge
Lake Charles
Lake Charles
Carrville
Plaquemine
»
Baton Rouge
Taft
Alliance
Donaldsonville
Shreveport
Geismar
Norco
Chalmette
Taft
Tulsa
Bayport
Port Arthur
Big Spring
County
East Baton
Rouge
Calcasieu
Calcasieu
Iberville
Iberville
East Baton
Rouge
St. Charles
Plaquemines
Ascension
Caddo
Ascension
St. Charles
St. Bernard
St. Charles
Tulsa
Harris
Jefferson
Howard
Model
number
20
42
4
20
5
40
4
48
20
23
10
4
33
26
54
c
22
53
Leukemia
deaths per
year from
benzene,
exposure
0.001632
0.000784
0.000166
0.000166
0.000008
0.006369
0.000088
0.000084
0.000179
0.003908
0.000772
0.000103
0.004497
0.000538
0.005978
0.000000
0.000616
0.001444
(continued)
-------
TABLE E-l (continued)
I
ro
o
Plant
code
no.
74.
75.
76.
77.
78.
79.
80.
81.
82.
83.
84.
85.
86.
87.
88.
89.
90.
91.
Comp. ny (by region)
American Petrofina/
Union Oil of California
Amoco Chemicals1
Atlantic Richfield
Atlantic Richfield
(ARCO/Polymers)
Atlantic Richfield
(ARCO/Polymers)
Charter International
Coastal States Gas
Corp.;? Ihristi
Petrochemicals
Cosden Oil
Crnwn Central
DENKA (Petrotex)
Oow Chemical
Dow Chemical
DuPont
DuPont
Eastman Kodak
El Paso Natural Gas
El Paso Products/
Rexene Polyolefins
State
TX
TX
TX
TX
TX
TX
TX
TX
TX
TX
TX
TX
TX
TA
TX
TX
TX
TX
Location
City
Beaumont
Chocolate Bayou
Channel view
Houston
Port Arthur
Houston
Corpus Christi
Corpus Christi
Groves
Pasadena
Houston
Freeport
Orange
Beaumont
Orange
Longview
Odessa
Odessa
County
Jefferson
Brazoria
Harris
Harris
Jefferson
Harris
Nueces
Nueces
Jefferson
Harris
Harris
Brazoria
Orange
Jefferson
Orange
- Gregg
Ector
Ector
Modela
number
39
4
27
47
6
43
44
c
4
56
9
29
4
10
4
4
18
17
Leukemia
deaths per
year from
benzene.
exposure
0.002130
0.000026
0.003569
0.011300
0.000359
0.008574
0.000800
0.000000
0.000002
0.007757
0.001650
0.001087
0.000055
0.002980
0.000148
0.000168
0.001083
0.000587
See footnotes at end of table.
(continued)
-------
TABLE E-l (continued)
Nant
dl I o
code
no.
92.
93.
94.
95.
96.
97.
98.
99.
100.
101.
102.
103.
104.
105.
106.
107.
108.
109.
110.
111.
See fo
Company (by region)
Exxon
Georgia-Pacific Corp.
Goodyear Tire and Rubber
Gulf Oil Chemicals
Gulf Oil Chemicals
Hercules
Howe 11
Independent Refining
Corp.
Kerr-McGee Corp.
(Southwestern)
Marathon Oil
Mobil Oil
Monsanto
Monsanto
Oxirane
Phillips Petroleum
Phillips Petroleum
Quintana-Howell
Shell Chemical
Shell Oil
Shell Oil
otnotes at end of table.
State
TX
TX
TX
TX
TX
TX
TX
TX
TX
TX
TX
TX
TX
TX
TX
TX
TX
TX
TX
TX
Location
City
Baytown
Houston
Bayport
Cedar Bayou
Port Arthur
McGregor
San Antonio
Winnie
Corpus Christi
Texas City
Beaumont
Alvin (Chocolate
Bayou)
Texas City
Channel view
Borger
Sweeny
Corpus Christi
Houston
Deer Park
Odessa
-
County
Harris
Harris
Harris
Brazoria
Jefferson
McLennan
Bexar
Chambers
Nueces
Galveston
Jefferson
Brazoria
Galveston
Harris
Hutchinson
Brazoria
Nueces
Harris
Harris
Ector
Model
i Q
number
32
3
7
5
36
c
23
23
21
35
55
15
28
6
16
38
48
4
31
56
Leukemia
deaths per
year from
benzene.
exposure
0.002130
0.000419
0.003867
0.000011
0.002375
0.000000
0.004469
0.000116
0.002053
0.002177
0.003353
0.000192
0.002497
0.000926
0.000868
0.000241
0.001794
0.000365
0.004345
0.001633
(continued)
-------
TABLE E-l (continued)
I
ro
ro
Plant
code
no.
112.
113.
114.
115.
116.
117.
118.
119.
120.
121.
Region
122.
123.
124.
Region
125.
126.
127.
128.
129.
Company (by region)
Standard Oil (Indiana)1
Standard Oil
(Indiana)/Amoco
Sun Oil
Texaco
Texaco/Jefferson Chemical
Texaco/Jefferson Chemical
Union Carbide
Union Carbide
Union Pacif ic/Champl in
USS Chemicals
VII
Chemplex
Getty Oil
Monsanto
IX
Atlantic Richfield
Chevron
Specialty Organics
Standard Oil of
California (Chevron
Chemical)
Union Carbide
State
TX
TX
TX
TX
TX
TX
TX
TX
TX
TX
10
KA
MO
CA
CA
CA
CA
CA
Location
City
Alvin
Texas City
Corpus Christi
Port Arthur
Bellaire
Port Neches
Seadrift
Texas City
Corpus Christi
Houston
Clinton
El Dorado
St. Louis
Wilmington
Richmond
Irwindale
El Segundo
Torrance
County
Brazoria
Galveston
Nueces
Jefferson
Harris
Jefferson
Calhoun
Galveston
Nueces
Harris
Clinton
Butler
St. Louis
Los Angeles
Contra Costa
Los Angeles
Los Angeles
Los Angeles
Model
number3
5
34
51
37
4
4
18
4
50
4
4
35
9
41
23
2
35
4
Leukemia
deaths per
year from
benzene,
exposure
0.000036
0.003654
0.002155
0.003038
0.000165
0.000004
0.000046
0.000322
0.005486
0.000308
0.000068
0.000562
0.003600
0.015754
0.003862
0.011570
0.015163
0.002377
-------
TABLE E-l (continued)
CO
-.. - '-' ' ' -^^
Plant
code
no. Company (by region)
130. Witco Chemical
131. Montrose Chemical
132. Stauffer Chemical
133. Phillips
Location
State city
CA Carson
NV Henderson
NV Henderson
TX Pasadena
Total deaths
County
Los Angeles
Clark
Clark
Harris
Leukemia
deaths per
year from
Model benzeneb
number exposure
8 0.003025
2 0.001527
1 0.000445
4 0.000070
j
0.420656
aModel used to estimate benzene concentrations around plant.
br f-H nr. limit, of 95 oercent are obtained by multiplying and dividing figures shown by 2.64. This
resuH in a 95-percent confiSence interval that assumes the estimated concentrations to which the
workers were exposed, as discussed in the CAG report,2 are within a factor of 2.
cModels were not assigned to three plants still under construction.
al places due to rounding. Total deaths for some plants differ
8, due to correction of a minor mathematical error.
organic chemical manufacturing plants are not quantified.
-------
TABLE E-l (concluded)
fThis total has been revised since three of the plants in the table should be deleted (see footnotes
g-i). This revision represents a decrease of 0.6 percent from the previous total of 0.4^09! deaths
per year.
9Industry sources indicate that Allied Chemical's (Solvay, N.Y.) chlorobenzene unit is no longer in
operation. Therefore, the plant should be deleted from the table.
Standard Oil of Indiana's maleic anhydride unit at Joliet, Illinois, should be deleted since it is
using butane as a feedstock instead of benzene.
i*
^he Amoco Chemical plant at Chocolate Bayou, Texas, and Standard Oil of Indiana at Alvin, Texas
(Plant No. 112) are the same plant (2 ethylene units). Therefore, the Chocolate Bayou plant should
be deleted fr^n the table.
-------
TABLE E-2. EXAMPLE CALCULATION OF LEUKEMIA DEATHS, PLANT 4
Source of data
GIVEN DATA
D,
1 Distance to outer ring edge, i2(km)
1 5 10 20 Refs. 8, 10
0.5 i 5 10 Refs. 8, 10
£_l_/I^V*v^li%rf^-»"- t~r -r - ^ /"
3 R, Hs. facto, (deaths/ppb-person »r> 0.34 x ID"6 0.34 x if6 0.34 x 10' IK34 x M Ref 2
4 PopuUtion in .nnul.r ring (persons) 7,963 271,100 1,025,093 4,515,175 Ref. 8
2 Distance to inner ring edge, i-^km)
5 Benzene cone, at outer ring edge,
Bi2(ug/m3)
6 Benzene cone, at inner ring edge,
24
45
0.189
1.24
0.0695
0.189
ro
en
CALCULATIONS
2
1 Population density, pi (persons/km )
Pi = F
8 b: b = ln(Bi2/b11)/ln(ui2/lli1)
B. //D. Ab
9 a: a = !/>/( i2J
10 Deaths per year in each ring =
3,380
3,596
4,351
0.0251 Ref. 8, Table A-2
for Model 2a
0.0695 Ref. 8, Table A-2
for Model 2a
4,791 Equation 4
BBDD
-0.982 -1-169 -1-443 -1.469 Equation 2
1.240 1-240 1.929 2.048 Equation 3
0.001393 0.010064 0.011609 0.018623 Equation 1
11 Total deaths in all rings -
D
0.041687"
lEquation 1
n Tab!e
bThis total appears in Tab!e E-l of this appendix.
8, and in Table E-l of this
It a!so appears in Tab.e A-4, Reference 8, as 41,686.54
-------
E.5 REFERENCES
1. U.S. Environmental Protection Agency. Assessment of Health Effects of
Benzene Germane to Low Level Exposure. EPA-600/1-78-061. September
1978.
2. U.S. Environmental Protection Agency. Carcinogen Assessment Group
(R. Albert, Chairman). Population Risk to Ambient Benzene Exposures.
January 1980.
3. National Institute for Occupational Safety and Health. Criteria for a
Recommended StandardOccupational Exposure to Benzene. HEW Publica-
tion Number (NIOSH)74-137. 1974.
4. American Conference of Governmental Industrial Hygienists. Threshold
Limit Values for Chemical Substances and Physical Agents in the Work-
room Environment with Intended Changes for 1977. 1977.
5. National Institute for Occupational Safety and Health. Revised Recom-
mendation for an Occupational Exposure Standard for Benzene. August
1976.
6. Occupational Safety and Health Administration. Occupational Safety
and Health Standards, 29 CFR 1910.1000, Table 1-2. Publication 2206.
1976.
7. 42 FR 27452. May 27, 1977.
8. Suta, B. E. Assessment of Human Exposures to Atmospheric Benzene from
Petroleum Refineries and Synthetic Organic Chemical Manufacturing
Plants. SRI Internatiopal, Center for Resource and Environmental
Systems Studies Report No. 118. February 1980.
9. Memorandum from Mclnnis, J. R., Pacific Environmental Services, Inc.,
to Warren, John L., Research Triangle Institute, April 1, 1980,
with "Summary of Methodology for Determining Ambient Benzene Concen-
trations due to Benzene Fugitive Emissions."
10. H. E. Cramer Co., Inc. Dispersion-Model Analysis of the Air Quality
Impact of Benzene Emissions from a Petroleum Refinery. EPA Contract
Number 68-02-2507. October 1978.
E-26
-------
APPENDIX F
F-l
-------
APPENDIX F - ESTIMATES OF BENZENE EMISSIONS AND
CONTROL COSTS OF PRODUCT ACCUMULATOR VESSELS
F.I INTRODUCTION
Product accumulator vessels were discussed in Chapter 3 as a
potential source of benzene fugitive emissions. Controlling accumulator
vessels by routing the emissions to a closed-vent system was discussed
in Chapter 4. No estimates of benzene emissions from or control costs
for accumulator vessels have been included in Chapters 7 and 8 because
of uncertainties in the number of uncontrolled vents and lack of data
on benzene emission rates from uncontrolled vents.
The purposes of this analysis are (1) to estimate the additional
benzene emission and control costs of accumulator vessels and (2) to
describe the effect of adding these estimates to the ones from all
other sources.
F.2 ESTIMATE OF BENZENE EMISSIONS
F.2.1 Uncontrolled Estimates
Benzene emissions from accumulator vessels can be estimated by a
method similar to the one described in Chapter 7. The number of
accumulator vessels per model unit (containing greater t^in 10 percent
benzene) was estimated by examining flow diagrams and ,na .eriil balances
for various benzene processes.
The uncontrolled VOC emission factor estimated for accumulator
vessels is multiplied by the number of vessels in each model unit and
the weighted factor of percent benzene in each model unit to obtain
the uncontrolled benzene emission rate. Table F-l shows the calculation
of the uncontrolled emissions from accumulator vessels by model unit.
F.2.2. Controlled Estimates
Assuming that 95 percent of accumulator vessels are already
controlled by a closed-vent system, the controlled emission rates for
,he baseline alternative (I) are estimated by multiplying the uncon-
trolled rates by 0.05 (1-0.95). It is not necessary to estimate
benzene emissions from accumulator vessels for Regulatory Alternatives III,
IV, and V, since 100 percent control efficiencies are already specified
for each. The controlled emissions for each model unit under Regulatory
F-2
-------
Table F-l. CALCULATION OF UNCONTROLLED BENZENE
EMISSIONS FROM ACCUMULATOR VESSELS BY MODEL UNIT
Model
Unit
A '
B
C
Number of
Vessels Per
Unit
1
2
2
Uncontrolled
VOC Emission
x Factor9
(Kg VOC/hr)
1.23
1.23
1.23
Weighted
x Percent
Benzene15
0.63
0.55
* 0.75
Conversion
Factor
8.4
8.4
8.4
Uncontrolled
* Benzene Emission Rate
(Mg/yr)
6.5
11.4
15.5
'Source- Brings, T. and V.P. Patel. Evaluation of Emissions from Benzene-Related Petroleum Processing
Source. Briggs PEDCo Environmental, Inc. Cincinnati OH For U.S-Environmental Protection
Agency. Research Triangle Park, N.C. Report No. EPA-450/3-79-022. October 1978.
Weighted average benzene concentration for the processes represented by each model unit.
-------
Alternative I are calculated to be 0.33 Mg/yr for Model Unit A, 0.57 Mg/yr
for Model Unit B, and 0.78 Mg/yr for Model Unit C.
F.2.3 Nationwide Estimates
Nationwide benzene emissions from accumulator vessels are estimated
by multiplying the controlled emissions from each model unit by the
number of existing units operating in 1980. The resulting nationwide
total under Regulatory Alternative I is presented in Table F-2.
The national accumulator vessel emissions are then added to the
national baseline emissions from all other sources as shown in Table 7-9
to yield a combined emission estimate of 8,360 Mg/yr. Adding the
accumulator vessel emissions increases the nationwide total by less
than 2 percent. Furthermore, the increase in excess incidence of
lukemia deaths for the baseline alternative would change only slightly
by adding accumulator vessel emissions. The baseline incidence would
increase from 0.6300 deaths per year to 0.6384, a change of 1.3 percent
per year, as shown in Table F-3.
F.3 CONTROL COST ESTIMATES
Capital and annual costs were calculated for accumulator vent
systems for each model unit as presented in Table F-4. Nationwide costs
for controlling existing accumulator vessels, as shown in Table F-5, are
based on the assumption that 5 percent of existing accumulator vessels
are uncontrolled and are tied into existing control devices.
When costs for controlling accumulator vessels are added to the
nationwide costs for controlling all other sources (Chapter 8, Table 8-15),
the total costs increase only slightly. Table F-6 presents the total
capital and (net) annualized costs for Regulatory Alternatives III
through V after accumulator vessel costs are added.
F-4
-------
Table F-2. NATIONAL BENZENE EMISSIONS FROM
ACCUMULATOR VESSELS IN 1980 FOR
REGULATORY ALTERNATIVE I
Model
Units
A
B
C
Benzene
Emissions
Per Unit
(Mg/yr)
0.33
0.57
0.78
Number of Units
Expected in Operation
(1980)
145
72
24
Total Benzene
Emissions
Per Unit
(Mg/yr)
47.9
41.0
.18.7
Nationwide
Total = 107.6
F-5
-------
Table F-3. EFFECT OF ADDING ACCUMULATOR
VESSEL EMISSIONS ON BASELINE RISK
Nationwide Benzene
Emission Total
(Mg/yr)
Excess Incidence of
Leukemia Deaths
(deaths/yr)
Other Sources (exluding
Accumulator Vessels)
Other Sources
Vessels
-i- Accumulator
8250
8360
0.6300
0.6384
aRisk for "Other Sources" represents a mid-range of excess incidence
of leukemia for the baseline regulatory alternative.
Calculated by multiplying 8360 by 0.63.
8250
F-6
-------
Table F-4. CAPITAL AND ANNUALIZED
COSTS FOR VENT SYSTEMS FOR ACCUMULATOR VESSELS
BY MODEL UNIT
Model Unit3 Capital Cost, $
A 2700
B 5400
C 5400
Annual i zed Cost, $
675
1350
1350
aModel A has 1 accumulator vessel per unit, and Model Units B and C
have 2 vessels each.
bCapital cost is based on 61 meters of 5.1 cm carbon steel pipe,
$2700 installed per vent.
t
cAnnualized cost based on a 10-year expected life and 10 percent
interest; CRF = 0.16, Maintenance = 0.05, Miscellaneous = 0.04,
Total = 0.25.
F-7
-------
Table F-5. NATIONAL CAPITAL AND
ANNUAL COSTS FOR CONTROLLING
EXISTING ACCUMULATOR VESSELS
Capital Cost
Model Unit ($1000)
A 19.6
B 19.5
r 6.5
Total 45.6
Annuali zed Cost
($1000)
4.9
4.9
1.7 -
11.5
aAssumes that 5 percent of existing accumulator vessels are
uncontrolled and are tied into existing control devices.
Thus, the national capital costs for Model Unit A would be:
145 (number of existing, units) x $2700 (control cost per
unit) x 0.05 (5 percent uncontrolled) = $19,575.
bAnnualized cost based on a 10-year expected life and 10
percent interest; CRF = 0.16, Maintenance = 0.05, Miscellaneous
0.04, Total = 0.25.
F-8
-------
Table F-6. NATIONWIDE COSTS FOR THE EXISTING INDUSTRY FOR
REGULATORY ALTERNATIVES III, IV, AND V
Regulatory Alternative
III
IV
V
-==========^^
Capital Cost
($ Million)
Accumulator
Vessels
0.05
0.05
0.05
Other
Sources*
9.7
*
25.3
242
New
Total
.
9.8
25.4
242.1
Annual!
($ Mi
Accumulator
Vessels
0.01
O.O1
0 01
zed Cost
llion)
Other
Sources*
2.1
5.5
58.6
.
New
Total
2.11
5.51
58.61
-- -
- '
*0ther sources are ones for which control costs «ere .stl.tod in Chapter 8 (Table 8-15).
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-450/3-30-032a
4. TITLE AND SUBTITLE
Benzene Fugitive Emissions
2.
- Background Information for
Proposed Standard
7. AUTHOR(S)
9. PERFORMING ORGANIZATION NAME Af>
U.S. Environmental Protectii
Office of Air Quality Plann
Emission Standards and Engii
Research Triangle Park, Nor
ID ADDRESS
Dn Agency
ing and Standards
leering Division
bh Carolina 27711
12. SPONSORING AGENCY NAME AND ADDRESS
3. RECIPIENT'S ACCESSION-NO.
5. REPORT DATE
November 1980
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-02-3060 77-5J
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA 200/04
15. SUPPLEMENTARY NOTES
16. ABSTRACT
A national emission standard for petroleum refining and organic chemical manu-
facturing industries is being proposed under authority of Section 112 of the Clean
Air Act (42 U.S.C. 7412, as amended). The purpose of the proposed standard is to
minimize benzene fugitive emissions in these industries. The document provides
background information for the proposed standard. Control technologies and five
regulatory alternatives are evaluated in terms of environmental and economic impacts
on both new and existing emission sources. Included in the evaluation of environ-
mental impacts are estimates of air quality, water, noise, and solid waste impacts.
Included in the evaluation of economic impacts are estimates of total capital and
annualized costs, including recovery credits.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
Air Pollution
Petroleum Refining
Organic Chemical Industry
Performance Standards
18. DISTRIBUTION STATEMENT
Unlimited - Available to the public free of
charge from: U.S. EPA Library (MD-35),
Research Triangle Park, NC 27711
b.lDENTIFIERS/OPEN ENDED TERMS
Air Pollution Control
19. SECURITY CLASS (This Report)
Unclassified
20. SECURITY CLASS (This page)
Unclassified
c. COSATI Field/Group
13 B
21. NO. OF PAGES
292
22. PRICE
EPA Form 2220-1 (9-73)
------- |