United States .
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
EPA-450/3-79-035a
August 198O
Air
&EPA
Benzene Emissions Draft
from Ethylbenzene/ EIS
Styrene Industry —
Background Information
for Proposed Standards
-------
-------
Errata for ERA-450/3-79-035a
Background Information for Proposed Standards-
Benzene Emissions from Ethy!benzene/Styrene Industry
Pages 7-48.7-49,7-50,7-51, Section 7.3.5
This section presents an estimate of the costs that typical EB/S
facilities have to assume in order to comply with the Occupational
Safety and Health Administration (OSHA) standard reducing the permissible
exposure limit on airborne .concentrations of benzene from the consensus
standard of 10 parts benzene per million parts air (10 ppm) to 1 ppm,
and prohibiting dermal contact with solutions containing benzene. On
pre-enforcement review, however, the United States Court of .Appeals for
the Fifth Circuit held the standard invalid; the Court concluded that
OSHA had exceded.its standard-setting authority because it had not been
shown that the 1 ppm exposure limit was "reasonably necessary or appropriate
to provide safe and healthful employment." The Supreme Court of the.
United States affirmed this judgment on July 2, 1980. As a result of
this action, the costs presented in Section 7.3.5 of this document are
not incurred by typical EB/S facilities. Since the previous 10 ppm OSHA
standard has been a consensus standard, no costs attributable to the 10
ppm standard are available.
-Page 7-16
Replace the word "styrofoam" with the words "polystyrene foams."
-------
-------
EPA-450/3-79-035a
Benzene Emissions from
Ethylbenzene/Styrene Industry
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
August 1980
-------
This report has been reviewed by the Emission Standards and Engineering
Division of the Office of Air Quality Planning and Standards, EPA, and
approved for publication. Mention of trade names or commercial products
is not intended to constitute endorsement or recommendation for use. Copies
of this report are available through the Library Services Office (MD-35) ,
U.S. Environmental Protection Agency, Research Triangle Park, N.C. 27711,,
or from National Technical Information Services, 5285 Port Royal Road,
Springfield, Virginia 22161.
Publication No. EPA-45Q/3-7?-Q35a
11
-------
Background Information
and Draft
Environmental Impact Statement
for Benzene Emissions from the
Ethylbenzene/Styrene Industry
Type of Action: Administrative
Prepared by:
Don R. Goodwin
Director, Emission Standards and Engineering Division
Environmental Protection Agency
Research Triangle Park, NC 27711
Approved by:
(Date)
David G. Hawkins
Assistant Administrator for Air, Noise, and Radiation
Environmental Protection Agency
Washington, DC 20460
Draft Statement Submitted to EPA's
Office of Federal Activities for Review on
This document may be reviewed at:
Central Docket Section
Room 2902, Waterside Mall
Environmental Protection Agency
401 M Street, S.W.
Washington, DC 20460
Additional copies may be obtained at:
Environmental Protection Agency Library (MD-35)
Research Triangle Park, NC 27711
National Technical Information Service
5285 Port Royal Road
Springfield, Virginia 22161
(faate)
(Da'te)
- m -
-------
-------
TABLE OF CONTENTS
Page
List of Tables xiv
List of Figures xviii
Chapter 1. Summary 1-1
1.1 Regulatory Options 1-1
1.2 Environmental Impacts .. 1-2
1.3 Economic Impacts 1-7
Chapter 2. Introduction. 2-1
2.1 Background 2-1
Chapter 3. The Ethyl benzene/Styrene Industry. 3-1
3.1 General 3-1
3.2 Ethylbenzene/Styrene Industry 3-1
3.2.1 Uses and Growth 3-1
3.2.2 Domestic Producers 3-2
3.3 Ethyl benzene/Styrene Production Processes 3-6
3.3.1 Benzene Alkylation and Ethylbenzene
Dehydrogenation 3-6
3.3.1.1 Basic Process 3-7
3.3.1.2 Process Variations ....... 3-12
3.3.2 Ethylbenzene from Mixed Xylenes. 3-13
3.3.2.1 Sources of Xylene Streams 3-14
3.3.2.2 Basic Process 3-17
3.3.3 Styrene from Ethylbenzene Hydro-
peroxidation 3-18
3.3.3.1 Ethylbenzene Hydroperoxidation
Capacity 3-18
3.3.3.2 Basic Process 3-18
-v-
-------
TABLE OF CONTENTS (Continued)
Page
3.4 Emissions 3-21
3.4.1 Benzene Alkylation/Ethyl benzene
Dehydrogenation., ,. 3-21
3.4.1.1 Alkylation Reactor Area Vents 3-22
3.4.1.2 Atmospheric and Pressure Column
Vents 3-24
3.4.1.3 Vacuum Column Vents 3-24
3.4.1.4 Hydrogen Separation 3-25
3.4.2 Emissions from Ethylbenzene via
Mixed Xyl enes ,..,... , 3-25
3.4.3 Emissions from Styrene via Ethylbenzene
Hydroperoxidation. 3-26
3.4.3.1 Ethylbenzene Oxidation Reactor
Vent 3-26
3.4.3.2 Propylene Recycle Purge Vent. 3-27
3.4.3.3 Vacuum Columns 3-27
3.4.4 Baseline Emissions 3-27
3.5 References for Chapter 3 3-29
Chapter 4. Control s , 4-1
4.1 Condensation , 4-1
4.2 Absorption. ,., 4-^2
4.3 Flaring 4-4
4.4 Boilers 4-6
4.5 Manifolding 4-7
4.6 Al ternate Processes 4-9
4.7 References for Chapter 4 4-11
-vi-
-------
TABLE OF CONTENTS (Continued)
Page
Chapter 5. Regulatory Options 5-1
5.1 Model Plants 5-1
5.2 Regulatory Options 5-4
Chapter 6. Environmental and Energy Impacts 6-1
6.1 Air Impacts 6-2
6.1.1 Modeling Results 6-5
6.1.2 Effects of Benzene Controls on Nation-
wide Emissions 6-6
6.2 Water Quality Impact and Consumption 6-10
6.3 Solid Waste Disposal Impact.. 6-10
6.4 Energy Impact 6-11
6.4.1 (Option A) 85 Percent Benzene
Emission Control 6-11
6.4.2 (Option B) 94 Percent Benzene
Emission Control 6-11
6.4.3 (Option C) 99 Percent Benzene
Emission Control 6-12
6.4.4 Summary of Energy Impact 6-12
6.5 Other Environmental Concerns 6-14
6.5.1 Flare Noise 6-14
6.5.2 Flare Thermal Radiation 6-14
6.6 Irreversible and Irretrievable Commitment
of Resources 6-16
6.7 Environmental Impact of Delayed Standards. 6-16
6.8 References for Chapter 6 6-18
-VTI-
-------
TABLE OF CONTENTS (Continued)
Page
Chapter 7. Economic Impact 7-1
7.1 Industry Economic Profile 7-1
7.1.1 Introduction 7-1
7.1.2 Product 7-2
7.1.2.1 Production 7-2
7.1.2.2 Resource Use 7-5
7.1.2.3 Product Use 7-5.
7.1.3 Production Trends 7-8
7.1.4 Prices . 7-11
7.1.5 International Trade 7-13
7.1.6 Market Structure 7-16
7.1.6.1 Firm Characteristics. 7-16
7.1.6.2 Market Concentration 7-20
7.1.7 Supply and Demand 7-22
7.1.7.1 Supply 7-22
7.1.7.2 Demand 7-23
7.1.8 Baseline Projection 7-24
7.1.8.1 Baseline Regulatory Environment 7-24
7.1.8.2 Baseline Growth Rates 7-24
7.1.8.3 Baseline Investment Projections.... 7-26
7.2 Cost Analysis of Alternative Emission Control
Systems 7-27
7.2.1 Introduction 7-27
7.2.2 Equipment Used 7-27
-vn )-
-------
TABLE OF CONTENTS (Continued)
7.2.2.1 85 Percent Control with a
Scrubber and Condensers —
Regulatory Option A 7-27
7.2.2.2 94 Percent Control with a Flare —
Regulatory Option B., 7-28
7.2.2.3 99 Percent Control with a Boiler —'
Regulatory Option C 7-28
7.2.3 Equipment Costs 7-32
7.2.3.1 Introduction 7-32
7.2.3.2 Capital Costs for Control
Equipment 7-32
7.2.3.3 Annualized Costs 7-35
7.2.4 PI ant-By-Plant Costs 7-36
7.2.4.1 Introduction 7-36
7.2.4.2 PI ant-By-Plant Capital Costs 7-39
7.2.4.3 PI ant-By-PI ant Annualized Costs
and Savi ngs 7-39
7.2.5 Cost Comparison and Cost-Effectiveness
of the Regulatory Options 7-42
7.2.5.1 Costs to Achieve 85 Percent
Emissions Reduction 7-42
7.2.5.2 Costs to Achieve 94 Percent Emis-
sions Reduction. 7-42
7.2.5.3 Costs to Achieve 99 Percent Emis-
sions Reduction 7-43
7.2.5.4 Cost-Effectiveness of Regulatory
Options 7-43
-ix-
-------
TABLE OF CONTENTS (Continued)
Page
7.2.6 Costs for Continuous Monitoring 7-44
» «»
7.3 Other Cost Considerations 7-45
7.3.1 Introduction 7-45
7.3.2 Water Pollution Control .. 7-45
7.3.3 Storage and Handling Control 7-47
7.3.4 Fugitive Emission Control 7-48
7.3.5 Compliance with OSHA Regulations 7-48
7.4 Economic Impact Analysis 7-52
7.4.1 Control Costs and Feasibility 7-52
7.4.1.1 Costs of Control by Regulatory
Option 7-52
7.4.1.2 Feasibility of Financing 7-56
7.4.2 Economic Impact Methodology 7-58
7.4.2.1 Pricing Scenarios 7-58
7.4.2.2 Economic Conditions 7-59
7.4.2.3 Estimation Procedure: Simple
Rate of Return Impacts, Under
Ful 1 Cost Absorption 7-60
7.4.2.4 Estimation Procedure: Price
Impacts Under Full Cost Pricing 7-61
7.4.2.5 Other Economic Impacts 7-62
7.4.2.6 Data 7-62
7.4.3 Economic Impacts 7-63
7.4.3.1 Rate of Return Impacts 7-63
7.4.3.2 Price Impacts 7-64
7.4.3.3 Output and Employment Impacts...... 7-70
-x-
-------
TABLE OF CONTENTS (Continued)
Page
7.4.3.4 Investment Impacts 7-71
7.4.3.5 Interindustry Impacts. 7-71
7.4.3.6 Impact Summary 7-72
7.4.4 100 Percent Control (Zero Emissions):
The Closure Option 7-72
7.4.4.1 Direct Impacts 7-73
7.4.4.2 Indirect Impacts 7-77
7.4.4.3 Substitutes for Styrene 7-78
7.4.4.4 Summary 7-81
7.5 Potential Socioeconomic and Inflationary
Impacts 7-83
7.6 References for Chapter 7 7-85
Appendix A: Evolution of the Proposed Standard.. A-l
Appendix B: Environmental Impact Considerations.... B-l
.B.I Cross-indexed Reference System to Highlight
Environmental Impact Portions of the Document.... B-l
B.2 General Modeling Methodology B-4
B.3 References for Appendix B B-12
Appendix C: Emission Source Test Data C-l
C.I Introduction C-l
C.2 Summary.. C-l
C.3 Description of Facilities and Test Procedures.... C-l
C.3.1 Plant A C-2
C.3.2 Plant B C-8
C.3.3 Plant C C-9
-XI-
-------
TABLE OF CONTENTS (Continued)
Paqe
C.4 References for Appendix C C-14
Appendix D: Emission Measurement Methods D-l
D.I General Background D-l
D.2 Field Testing Experience D-3e
D.3 Emission Monitoring D-8
D.4 Emission Test Methods . D-12
D.5 References for Appendix D... D-14
Appendix E: Methodology for Estimating Mortality and
Maximum Risk from Exposure to Benzene
Emissions from Ethylbenzene/Styrene Plants.... E-l
E.I Introduction E-l
E.2 Summary and Overview of Health Effects E-l
E.2.1 Health Effects Associated With Benzene
Exposure E-l
E.2.2 Benzene Exposure Limits E-3
E.2.3 Health Effects at Environmental Exposure
Levels ' E-4
E.3 Population Density Around EB/S Plants E-5
E.4 Population Exposures, Risks, and Mortality E-6
E.4.1 Summary of Methodology E-6
E.4.2 Continuous Emissions E-9
E.4.2.1 Estimates of Leukemia Deaths E-9
E.4.2.2 Example of Leukemia Death
Calculation E-10
E.4.2.3 Estimate of Leukemia Risk E-ll
-xi 1-
-------
TABLE OF CONTENTS (Continued)
E.4.2.4 Example of Leukemia Risk
Calculation... E-12
E.4.3 Excess Emissions E-13
E.4.3.1 Estimates of Leukemia Deaths
from Excess Emissions. E-13
E.4.3.2 Estimates of Leukemia Risk
from Excess Emissions. E-14
E.4.4 Validity of Estimates.. . E-14
E.5 References for Appendix E E-31
AA. ADDENDUM TO BACKGROUND INFORMATION FOR PROPOSED STANDARDS
FOR BENZENE EMISSIONS FROM THE ETHYLBENZENE/STYRENE
INDUSTRY
1. Enforcement Aspects •? AAl-1
2. Reports Impact Analysis AA 2-1
3. Excess Emissions AA 3-1
4. Ethylbenzene Air Oxidation Analysis ...... AA 4-1
-xiii-
-------
LIST OF TABLES
1-1
1-2
3-1
3-2
3-3
3-4
5-1
5-2
6-1
6-2
6-3
6-4
6-5
7-1
7-2
7-3
7-4
7-5
7-6
7=7
Matrix of Environmental and Economic Impacts of
Regulatory Alternatives for Process Vent Streams.
Matrix of Environmental and Cost Impacts of Regulatory
Alternatives for Excess Emissions -.
Styrene Consumption and Growth...' .
U.S. Ethylbenzene/Styrene Capacities
Typical Compositions of Xylene-Containing Stream.
Uncontrolled VOC and Benzene Emissions, EB/S
Production via Alkylation and Dehydrogehation
EB/S Model Plant
Summary of Control s
Maximum Benzene Concentrations From Model Plants.
Predicted Maximum Ambient Benzene Concentrations
for Individual Sources
Estimated Ambient Impacts for Specific EB/S Plants....
Total Net Energy Requirements of Regulatory Options
for 300,000 Mg/Yr Model Plant
Flare Noise Level Relative to OSHA Standards and
EPA Gui del i nes
U.S. Ethylbenzene/Styrene Capacity by Producer
Product Consumption — Percent by Intermediate Use....
Styrene Consumption — Percent by Final Use....
Historical Domestic Styrene Capacity and
Production
Styrene Price History,,-,,-.
Styrene Exports and Imports.
Styrene Unit Sales Values for Domestic and Export
Markets
Page
1-3
1*4
3-2
3-3
3-14
3-23
5-3
5-5
6-7
6-8
6-9
6-13
6-15
7-3
7-6
7-9
7-10
7-12
7-14
7-15
-xiv-
-------
LIST OF TABLES (Contlnued)
Paqe
7-8 U.S. Styrene Producers -- Location, Capacity, and
Integration 7-17
7-9 Financial Characteristics of Selected Styrene
Producers 7-19
7-10 Concentration in the Styrene Industry 7-20
7-11 Control Equipment Needed to Achieve Benzene
Emissions Reduction 7-30
7-12 Assumed Vent Stream Characteristics 300,000 Mg/Yr
Styrene Capacity Model Uncontrolled Plant 7-33
7-13 Annualized Cost Parameters. 7-37
7-14 PI ant-By-PI ant List of Equipment Needed to Achieve
Given Regulatory Options 7-38
7-15 PI ant-By-PI ant Costs 7-40
7-16 Estimated Industry-Wide Cost-Effectiveness for
Each Regulatory Option „
7-41
7-17 Storage Emission Control Costs for Each Tank 7-49
7-18 Cost Basis for Calculations of Operating Costs 7-51
7-19 Cost Summary: 85 Percent Regulatory Option 7-53
7-20 Cost Summary: 94 Percent Regulatory Option 7-54
7-21 Cost Summary: 99 Percent Regulatory Option 7-55
7-22 Representative Capital Expenditures 7-57
7-23 .Rate of Return on Equity Impacts 7-65
7-24 Required Styrene Price Changes: 85 Percent Control... 7-66
7-25 Required Styrene Price Changes: 94 Percent Control... 7-67
7-26 Required Styrene Price Changes: 99 Percent Control... 7-68
7-27 Possible Styrene Substitutes 7-79
-xv-
-------
LIST OF TABLES (Continued)
Page
B-1 Stack Source Characteristics... ..; -. B-5
B-2 Storage Tank and Fugitive Solirce Characteristies•...... B-6
B-3 Ambient Benzene Impacts frb'ril Individual eihd
Group Sources <.. B-ll
C-l Plant A: Benzene Drying Column Vent System C-3
C-2 Plant A: Alkylate Degasser Vent Equipment.. C-5
C-3 Plant A: Catalyst Mix tank Equipment..... . .>. C-6
C-4 Plant A: Benzene-toluene Column Vacuum Equipment..... C-7
C-5 Plant B: Superheater Outlet test Results... >>..... ^. C-lO
C-6 PlantB: Oil Heater'Outlet test Results.... mutnn C-11
C-7 Plarit C: Summary of test Results from Superheater
Offgis ...... i,. ................................ C-l 3
E-l Population Residing Around EB/S Plants*. tn.m* E-16
E-2a Predicted Annual Average Benzene Concentrations in
Parts Per Billion for Model Plant..... >....... E-17
E-2b Predicted Highest Maximum Annual Average Benzene
Concentrations in Parts Per Billion for the Model
Plant *..... E-18
E-3 Scaling Factors E-l9
E-4 Scaled Mean Annual Average Benzene Concentrations in
Parts Per Billion at Distance from Plant in Kilometers:
Process Vent Contribution, Current Control Level E-20
E-5 Scaled Mean Annual Average Benzene Concentrations in
Parts Per Billion at Distance from Plant in Kilometers:
Fugitive and Storage Contribution E-21
E-6 Scaled Mean Annual Average Benzene Concentrations in
Parts Per Billion at Distance from Plant in Kilometers:
Process Fugitive and Storage Emissions E-22
-XVI-
-------
LIST OF TABLES (Continued)
Page
E-7 Scaled Mean Annual Average Benzene Concentrations in
Parts Per Billion at Distance from Plant in Kilometers
Excess Emissions Contribution, Uncontrolled Level
Assumed— „
E-8 Benzene Exposure in Thousands of PPB-Person Years.
E-9 Deaths Per Year, Current Control Levels.... .,
E-10 Scaled Highest Maximum Annual Average Benzene
Concentrations in Parts Per Billion
E-ll Maximum Lifetime Risk.
E-12 Deaths Per Year Expressed as a 95 Percent Confidence
Interval
E-13 Maximum Lifetime Risk Expressed as a 95 Percent
Confidence Interval
E-23
E-24
E-25
E-26
E-28
E-29
E-30
-xvi i-
-------
LIST OF FIGURES
Paqe
3-1 Locations of Plants Manufacturing Ethylbenzene 3-4
3-2 Locations of Plants Manufacturing Styrene 3-5
3-3 Benzene Alkylation Process 3-9
3-4 Ethylbenzene Dehydrogenation Process 3-10
3-5 Process Flow Diagram, Model Plant Uncontrolled,
Styrene from Benzene and Ethylene by Dehydrogen-
ation of Ethylbenzene 3-11
3-6 Ethylbenzene Extraction from Mixed Xylenes 3-15
3-7 BTX Processing Block Flow Diagram 3-16
3-8 Ethylbenzene Hydroperoxidation Process 3-19
3-9 Ethylbenzene Hydroperoxidation Process Block
Di agram 3-20
6-1 Model PI ant Layout 6-3
6-2 Location of Process Emissions for the Four
Regul atory Opti ons 6-4
7-1 Schematic of 94 Percent Regulatory Option 7-29
7-2 Burner System Diagram 7-31
7-3 Control Equipment Costs Without Allowances 7-34
-xvm-
-------
1. SUMMARY
1.1 REGULATORY OPTIONS
The Administrator considered five regulatory options for controlling
continuous process benzene emissions from process vents in the ethyl benzene/
styrene (EB/S) industry: (1) Do not set a standard, assuming benzene
emissions from process vents are controlled adequately to protect public
health; (2) Regulatory Option A -- Require that continuous benzene emissions
be reduced by 85 percent, based on the use of scrubbers and condensers;
(3) Regulatory Option B — Require that continuous benzene emissions be
reduced by 94 percent, based on the use of a flare system coupled to scrubbers
and condensers; (4) Regulatory Option C — Require that continuous benzene
emissions be reduced by 99 percent, based on the use of a boiler*; and
(5) Regulatory Option D — Require that continuous benzene emissions be
reduced by 100 percent, based on plant closure and substitutes for styrene.
Because there are no technological reasons to consider different or alter-
native controls for new or existing sources, these regulatory options apply
to all EB/S plants. These regulatory options are discussed in detail in
Chapter 5 of this document. The control techniques on which these options
are based are discussed in Chapter 4.
To control excess benzene process emissions from EB/S plants, the
Administrator considered four alternatives: (1) Regulatory Option 1 --
Require that excess emissions during startup, shutdown, and malfunctions be
*The term "boiler" includes process heaters and superheaters.
1-1
-------
combusted by one or more smokeless flares at all times during startup,
shutdown, and malfunctions; (2) Regulatory Option 2 — Require that EB/S
owners or operators install backup compressors to control emissions during
malfunctions and to combust the emissions by one or more smokeless flares
during startup and shutdown and until the backup compressor is on line;
(3) Regulatory Option 3 — Require that EB/S owners or operators retrofit
existing boilers to accept all excess emissions at all times; and (4) Regu-
latory Option 4 -- Require 100 percent excess benzene emissions reduction
based on EB/S plant closures and substitutes for styrene. Because there are
no technological reasons to consider different or alternative controls for
new or existing sources, these regulatory options apply to all EB/S plants.
These alternatives are discussed in detail in an addendum to this document.
1.2 ENVIRONMENTAL IMPACTS
The beneficial and adverse economic, environmental, and energy impacts
associated with each regulatory option for both continuous and excess
process emissions are summarized in Table 1-1 for continuous vent streams
and Table 1-2 for excess emissions from EB/S plants. The environmental
impacts are considered in terms of air quality, water quality, solid waste
generation, flare noise, and flare thermal radiation. These are discussed
in detail in Chapter 6 of this document.
In terms of nationwide air quality impacts, estimated total benzene
emissions from continuous process vent streams alone within the EB/S indus-
try are 1,990 megagrams per year (Mg/yr) at current control levels and
production at 100 percent capacity. Regulatory Options A, B, and C would
reduce total benzene continuous process emissions from the industry to
625 Mg/yr, 200 Mg/yr, and 70 Mg/yr, respectively. The estimated total
nationwide excess benzene emissions from process vents within the EB/S
1-2
-------
LU
•p
u
to
Q.
I— I
r—
•
•
I
rO
H-
u
4-
<1)
C
O)
CO
-p
u
rO
Q.
I — i
CD
in
s_
O)
TD
•=t
-p
u
ro
Q.
i — I
0)
-p
u
rO
-p
u
to
a.
-P i— Q.
u
rO
O.
i— i
O
Z
o
•r-
O5
•r-
(— -
OS
O)
E
i— i
i —
P—
rO
.E
to
0)
•P
rO
£_
ai
•a
0
s:
-p
u
rO
a.
E
1 — 1
0)
O)
s_
rO
_J
-p
U
rO
Q.
E.
i — i
E
S-
0)
f—
1
•P
s-
o
-C
to
4J
u
to
a.
i — i
CS
S-
ai
f—
i
05
C
0
-p
u
rO
i — i
CU
•r—
in
s-
O)
CD
s-
S-
1 — 1
c\j ro
* * *
•x *
o
s: LU
i— i a:
o to
o >
o
LU to
to
O LU
z o
-
u
S- tO
O) O.
C E
LU I-H
•P
T3 0) O
•r- -P tO
i — in o.
0 ro E
to 3: 1-1
-P
S- 0
Cl) ro
-P Q.
rO E
rs i — i
•p
u
5- ro
•r- O.
C
C .r- 0
E rO -P
TJ S- U
,
o t^
i
•X
1 —
1
•X
^
rH
•X
•X
rH
O
•X
•X
1
•X
oo
+
>,
O CQ
-p
ro C
i— O
^3 •! —
O5-P
CU Q.
rv Q
•X
•X
1 —
-f-
o
o
•X
•X
^~
-}•
CD
CD
•X
•X
•sf
>.
0 t_5
•P
to c .
r- O
3 •!-
O5-P
cu a.
C£L 0
^
•X
^
^»
1
0
o
o
o
CD
•X
•X
+
>.
o <^^
-p
ra E
r— O
3 *l~
O5+J
cu Q.
o; o
o
o '
0
o
o
CD
•X
•X
^
C\I
1
in
-a
-o s-
O) ro
rO C
•— ro
-------
o
rO
0.-P
E 0
i-i ro
ro i—i
U CU
•r- tO
q- s-
cu cu
c >
CU T3
CQ
en co
O
p i >
s- ro a>
o c s_
CM oo
* -x
CO
O CO
1-1 co
. w
I— M
CO SI
O LU
u
CO
Q CO
Z LU
-
|_ C£
So
CNJ C3
LU
-J
ca
•p o
in ro
o Q.
o E
r— O
CO •!-
OJ
cu u
u> ro
••- a.
C35 U
S- ro
CU Q.
-a C
C -i- O
•I- +->••-
S CO -P
•o s- u
<: -p -p
a co
TD
rO
ro
O -P
•z. co
1-4
-------
industry are 133 Mg/yr under current controls, 21 Mg/yr under Regulatory
Option 1, 10 Mg/yr under Regulatory Option 2, and 1 Mg/yr under Regulatory
Option 3.
The highest estimated maximum annual average benzene concentrations
which occur from continuous process vents within the industry including
emissions from fugitive and storage sources occur at 160 meters (m) for
each option and are 95.6 parts per billion (ppb) under existing conditions,
54.1 ppb under Regulatory Option A, 46.1 ppb under Regulatory Option B, and
46.0 ppb under Regulatory Option C. The highest estimated maximum annual
average emissions within the industry from the process vents alone are
estimated at 63.6 ppb under existing conditions, 18.2 ppb under Regulatory
Option A at a distance of 160 m, 1.4 ppb under Regulatory Option B at a
distance of 500 m, and 0.19 ppb for Regulatory Option C at a distance of
1,000 m.
For excess benzene emissions, the estimated highest maximum annual
average benzene concentrations under uncontrolled conditions occur at 160 m
and are 7.9 ppb. The estimated highest maximum annual average benzene
concentrations occur at 2,000 m for each regulatory option and are 4.3 x 10
ppb under Regulatory Option 1, 1.4 x 10~3 ppb under Regulatory Option 2,
-4
and 2.0 x 10 ppb under Regulatory Option 3.
None of the regulatory options considered result in any significant
increase in wastewater or effluent discharge by EB/S plants nor do they
generate any solid waste. Noise and thermal radiation associated with the
use of flares, which would be required for reducing benzene emissions from
continuous process stream sources if Regulatory Option B were adopted and
under Regulatory Options 1, 2, and 3 for excess emissions, are negligible.
« O
1-5
-------
Under Regulatory Option A, energy is required to operate pumps and
compressors for the condensers and absorbers. Energy also is required for
the cooling system of the condensers. Total energy required for this
control system is less than 0.1 percent of the fuel requirement for the
model plant. Under Regulatory Option B, flare operation consumes both
steam (for smoke control) and natural gas (as a supplemental fuel). In
conjunction with the condensers and absorbers, the total energy demand for
Regulatory Option B also would be less than 0.1 percent of the fuel require-
ments for the model plant. Due to recovered energy, Regulatory Option C
provides a small net energy savings, equivalent to less than 0.1 percent of
the fuel requirements for the model plant.
In terms of nationwide impacts, Regulatory Options A and B for all
existing EB/S plants would result in a total energy demand of less than
0.1 percent of the nationwide EB/S energy demand. Due to recovered energy,
Regulatory Option C would result in a small energy savings, equivalent to
less than 0.1 percent of the current nationwide EB/S fuel requirements.
For excess emissions, energy is used in the production of steam to
ensure smokeless operation of a ten-inch flare for combusting vent streams
during startup, shutdown, and malfunction and for natural gas for pilot and
purge operations. These energy requirements are negligible. Under Regu-
latory Option 1, steam and natural gas requirements are approximately
5.23 x 10 megajoules per year (MJ/yr) based on an average of 12 hours of
operation per year for a 300,000 Mg/yr model plant. Under Regulatory
Option 2, energy requirements for steam and natural gas are approximately
5.14 x 106 MJ/yr. Under Regulatory Option 3, energy requirements for steam
and natural gas are approximately 5.1 x 10 MJ/yr.
1-6
-------
In terms of nationwide energy impacts, assuming no current emissions
control, Regulatory Option 1 would require 71 x 106 MJ/yr, Regulatory
Option 2 would require 70 x 10 MJ/yr, and Regulatory Option 3 would require
69 x 10 MJ/yr of energy, or approximately 11,200 billions of barrels per
year (bbl/yr) (fuel oil equivalent) for all existing plants under each option.
1.3 ECONOMIC IMPACTS
The Administrator considered both capital and annualized costs for
controlling process vent streams under Regulatory Options A, B, and C.
These are discussed in detail in Chapter 7. Nationwide impacts were deter-
mined from these cost estimates.
Regulatory Option A (85 percent control) would result in capital costs
per plant ranging from zero to $268,000, total annualized costs ranging
from $13,500 to a savings of $172,000, a potential maximum styrene price
increase of $0.70/Mg (0.15 percent of product price), and no projected
plant closures. A standard based on Regulatory Option B (94 percent control)
would result in capital costs per plant ranging from zero to $530,000,
total annualized costs ranging from $45,000 to a savings of $172,000, a
potential maximum styrene price increase of $2.14/Mg (0.46 percent of
product price), and no projected plant closures. Regulatory Option C (99
percent control) would result in capital costs per plant ranging from zero
to $555,000, total annualized costs ranging from $26,000 to a savings of
$150,000, a potential maximum styrene price increase of $1.27/Mg (0.27
percent of product price), and no projected plant closures. Regulatory
Option D (100 percent control) would have the most severe economic impact
on the EB/S industry, requiring closure of all EB/S production facilities.
Plants producing styrene derivatives also may close depending upon the
availability of foreign suppliers or substitutes.
1-7
-------
In terms of nationwide impacts, Regulatory Options A, B, and C each
would require 12 plants to install controls. Total nationwide capital
costs to the industry would be approximately $1.4 million, $3.1 million,
and $3.4 million under Regulatory Options A, B, and C, respectively. The
industry's total annualized cost, assuming full utilization of capacity and
including operating and maintenance costs, annualized capital costs, and
fuel and recovered material credits, would be net credits of approximately
$608,000 under Regulatory Option A, approximately $202,000 under Regulatory
Option B, and approximately $460,000 under Regulatory Option C.
Capital costs also were estimated for controlling excess emissions
under Regulatory Options 1, 2, 3, and 4. Regulatory Option 1 Would require
the smallest capital outlay of the regulatory options. Industry-wide
capital costs of Regulatory Option 1 would be approximately $524,000 and
would require four plants to install controls. Regulatory Option 2 would
require capital outlays of approximately $5.5 million and would require six
plants to install controls. Regulatory Option 3 would require the greatest
capital outlays of approximately $20.0 million and would require 13 plants
to install controls. Industry-wide annualized costs for Regulatory Options 1,
2, and 3 would be $171,000, $1.6 million, and $5.2 million, respectively.
As with Regulatory Option D for process vent streams, Regulatory
Option 4 (100 percent control) would have the most severe economic impact
on the EB/S industry, requiring closure of all ethyl benzene and styrene
production facilities. Plants producing derivatives also may close depen-
ding upon the availability of foreign suppliers or substitutes.
1-8
-------
2. INTRODUCTION
2.1 BACKGROUND
The Environmental Protection Agency proposed on October 10, 1979,
"Policies and Procedures for Identifying, Assessing, and Regulating Air-
borne 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
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 implemen-
ting 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, also would be
considered in setting standards.
2-1
-------
(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 regulations will be developed. Although a pollutant may have
been listed because emissions from a particular source category pose a
significant risk, other source categories also may emit the pollutant in
lesser amounts. This may occur, for example, because the sources process
very little of the substance, because the substance is present in only
trace amounts in the sources' raw materials, or because sources have installed
adequate controls on their own initiative or in response to other regulatory
requirements.
The Administrator therefore will 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
2-2
-------
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 stan-
dards 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 regula-
tory 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 indi-
viduals; (3) ease of expeditious standards development and implementation;
and (4) feasibility of significant improvements in controls. In addition,
significant sources of more than one carcinogen may be given priority over
single-pollutant sources, based on the sum of risks from the emitted
substances.
A high priority will be assigned, for example, to a source category
constituting an important problem requiring immediate attention, or where
risks are somewhat lower but an appropriate regulatory solution is both
feasible and readily available. Source categories assigned medium priority
will generally be those that present lower risks and will be scheduled for
standard development as resources become available. Lower risk source
categories for which the extent of feasible control may be limited substan-
tially will be assigned low priority for regulation development. Assignment
2-3
-------
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, technolog-
ically 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 initially will 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 also will
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
2-4
-------
for each option: the number of plant closures predicted and the direct
impact on employment and end product prices; the impact on growth and
expansion of the industry; the resulting changes in profitability; capital
availability for control equipment; the impacts from the availability of
substitute products and foreign imports; the potential increases in national
energy consumption; and the impacts on other environmental media including
increased water pollution and solid waste disposal. On the basis of these
assessments, one of the control options identified will be designated as
the BAT for the control of emissions from the sources in the category.
This level of control will be that technology which, in the judgment of the
Administrator, is the most advanced level of control adequately demonstrated,
considering economic, energy, and environmental impacts.
The control level designated "best available technology" may be dif-
ferent for new and existing facilities in a category. For practical purposes,
this level of control for new sources will, as a minimum, be equivalent to
that which would be selected as the basis for a New Source Performance
Standard (NSPS) under section 111. The requirement of "best available
control te~chnology" 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.
2-5
-------
(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 quantita-
tive 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 controls would not be required. If, however, there is a finding of
unreasonable residual risk, a more stringent alternative would be required.
Among the possible alternatives would be the immediate application of more
restrictive emission standards, including those based on more extensive use
of substitutes, and scheduled or phased reductions in permissible emissions.
The alternative selected would be that necessary, in the Administrator's
judgment, to eliminate the unreasonable residual risks.
Given the differences in the degree of certainty in risk estimates, in
the numbers of people exposed, in benefits, in the distribution of 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. Neverthe-
less, the process followed and the various factors involved can be outlined.
2-6
-------
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 sub-
stance 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.
2-7
-------
-------
3. THE ETHYLBENZENE/STYRENE INDUSTRY
3.1 GENERAL
This chapter describes the ethylbenzene/styrene (EB/S) industry struc-
ture, its production process, and the associated emissions. A more detailed
discussion of the EB/S industry structure is provided in Section 7.1. The
processes described in this chapter are production of ethylbenzene from
(1) benzene alkylation and (2) mixed xylene extraction and production of
styrene from (1) ethylbenzene dehydrogenation and (2) ethylbenzene hydro-
peroxidation.
3.2 ETHYLBENZENE/STYRENE INDUSTRY
For this analysis, production of ethylbenzene and styrene is considered
a single industry since 99 percent of ethylbenzene is used to produce styrene
and styrene is produced only from ethylbenzene. In fact, in many EB/S facili-
ties, ethylbenzene is not separated as a product itself but used only as an
intermediate for styrene production. Furthermore, more than 90 percent of
ethylbenzene is not sold on the market but is used by the producing company.
Thus, growth of ethylbenzene and styrene production are intimately connected.
3.2.1 Uses and Growth
1,2
Polymer manufacture consumes virtually all produced styrene. The produc-
tion of polystyrene, used primarily in packaging and disposable serviceware,
accounts for more than one-half of the styrene produced. Other major styrene
polymers are acrylonitrile-butadiene-styrene (ABS) plastics and styrene-
butadiene rubbers (SBR). The principal uses for ABS are pipe and automotive
parts; the principal use for SBR is pneumatic tires. Table 3-1 shows styrene
34
uses, percentages of consumption, and expected growth rates. '
3-1
-------
TABLE 3-1. STYRENE CONSUMPTION AND GROWTH^
1976 Consumption Expected 1976 to 1980
3
% 10 Mg/yr Growth Rate, Percent
Polystyrene
Styrene copolymer resins*
Styrene-butadiene rubbers
Unsaturated polyester resins
Miscellaneous
Exports
54
17
9
6
1
13
TOO
1,547
467
258
171
28
372
2,843
7.0
7.0
2.5
9.0
12.5
-25.0
^Includes acrylonitrile-butadiene-styrene resins (ABS), eight percent of
styrene consumption; styrene-butadiene resins, six percent; and other
styrene copolymer resins, three percent.
3.2.2 Domestic Producers
The current U.S. styrene capacity is 4,092,000 Mg/yr (4,509,000 tons).
In 1978, production was approximately 78 percent of total capacity. For
ethylbenzene, the current U.S. capacity is 4,742,000 Mg/yr (5,226,000 tons).
In 1978, production was approximately 70 percent of total capacity. By the
end of 1982, based on a projected six percent annual growth in both styrene
production and ethylbenzene consumption, styrene production levels should
nearly equal capacity and ethylbenzene production should be 90 percent of
capacity.
Currently, 14 companies produce ethylbenzene at 15 sites in the U.S.
At ten of these sites, both styrene and ethylbenzene are produced in an
integrated process. At the five remaining sites, only ethylbenzene is
produced. Eleven companies produce styrene at 12 sites in the U.S. At ten
sites both ethylbenzene and styrene are produced, while at two sites styrene
alone is produced. Table 3-2 lists these ethylbenzene or styrene producers,
their locations, capacities, and processes. Figures 3-1 and 3-2 indicate
the site locations.
3-2
-------
TABLE 3-2.
U.S. ETHYLBENZENE/STYRENE CAPACITIES3
Ethyl benzene
Company
American Hoechst Corp.
Baton Rouge, LA
Amoco ,
Texas City, TX
Atlantic Richfield Co. ,
Port Arthur, TX
Atlantic Richfield Co. ,
Kobuta, PA
Charter, Houston, TX
Commonwealth Oil Refining
Co. , Inc. , Penuelas, PR
Cos-Mar, CarvilTe, LA
Dow Chemical U.S.A. ,
Freeport, TX
Dow Chemical U.S.A. ,
Midland, MI
El Paso Products Co. ,
Odessa, TX
Gulf Oil Corp.,
Donaldsville, LA
Monsanto Company,
Alvin, TX
Monsanto Company,
Texas City, TX
Oxirane, Channelview, TX
Sun Oil Co. ,
Corpus Christi, TX
Tenneco, Chalmette, LA
U.S. Steel, Houston, TX
TOTAL
Capacity
(103 Mg/yr)
526
447
227
18
73
689
847
125
313
23
771
545
61
16
61
4,742
Process
c
c
c
d
d
c
c
c
c
d
c,d
c
c,d
d
d
Styrene
Capacity
(103 Mg/yr)
450
380
200
590
680
181
115
272
680
454
36
54
4,092
Process
e
e
e
e
e
e
e
e
e
f
e
e
See References 12 and 13.
b 3
American Hoechst Corp. plans to bring a 408.2 x 10 Mg/yr styrene plant on-stream
at Bayport, TX in 1980.
GBenzene alky!ation.
Mixed xylene stream recovery.
p
Ethyl benzene dehydrogenation.
Ethylbenzene hydroperoxidation.
3-3
-------
FIGURE 3-1
LOCATIONS OF PLANTS MANUFACTURING! ETHYLBENZENE
LEGEND OF PUNT NAMES AND LOCATIONS
1. American Hoechst, Baton Rouge, LA
Z. U.S. Steel, Houstun, TX
3. Atlantic Richfield, Port Arthur, TX
4. Charter, Houston, TX
5. Commonwealth Oil, Penuelas, PR
9. Gulf, Donaldsonvllle, LA
10. Monsanto, Alvin, TX
11. Monsanto, Texas City, TX
12. Oxirane, Channelview, TX
13. Amoco, Texas City, TX
6. Cos-Mar, Carviile, LA
7. Dow, Freeport, TX
8. El Paso Products, Odessa, TX
14. Sun Oil, Corpus Christ!, TX
15. Tenneco, Chalmette, LA
3-4
-------
FIGURE 3-2
LOCATIONS OF PLANTS MANUFACTURING STYRENE
1. American Hoechst, Baton Rouge, LA
2. U.S. Steel, Houston, TX
3. Atlantic Richfield, Kobuta, PA
4. Cos-Mar, Carville, LA
5. Dow, Freepott, TX
6. Dow, Midland, Ml
LEGEND OF PLANT NAMES AND LOCATIONS
7. El Paso Products, Odessa, TX
8. Gulf, Donaldsonville, LA
9. Monsanto, Texas City, TX
10. Oxirane, Channelview, TX
11. Amoco, Texas City, TX
12. Sun Oil, Corpus Christ!, TX
3-5
-------
3.3 ETHYLBENZENE/STYRENE PRODUCTION PROCESSES6 >7'8'9
The EB/S industry produces styrene from ethyl benzene by either (1) ethyl -
benzene dehydrogenation, or (2) ethyl benzene hydroperoxidation. The only
U.S. application of the ethyl benzene hydroperoxidation process is the Oxirane
Corporation unit in Channel view, Texas. This unit represents about ten percent
of the U.S. styrene capacity. The two processes now used in the U.S. to
make ethylbenzene for styrene production are (1) benzene alkylation with
ethylene, and (2) extraction from mixed xylene streams. Benzene alkylation
is the dominant process, accounting for more than 95 percent of production.
There are ten EB/S facilities at which both ethylbenzene and styrene
are produced in an integrated operation. At eight of these facilities,
styrene is produced by this process:
Benzene Alhylatlqn Ethylbenzene Pshydrogenation
Of the two remaining integrated EB/S facilities, one uses the process:
Benzene Alkylation Ethylbenzene Hydroperoxidation styrene;
the other uses the process:
Mixed Xylene Extract1on Ethylbenzene Dehydrogenation styrene
A typical EB/S plant uses benzene alkylation followed by ethylbenzene
dehydrogenation to form styrene. This typical EB/S production process along
with extraction of ethylbenzene from mixed xylenes and the ethylbenzene
hydroperoxidation process is described below.
3.3.1 Benzene Alkylation and Ethylbenzene Dehydrogenation
The primary reactions shown below for the production of styrene are
(1) catalytic alkylation of benzene with ethylene to produce ethylbenzene,
and (2) catalytic dehydrogenation of ethylbenzene to produce styrene.
3-6
-------
(1) Catalytic alkylation of benzene with ethylene to ethylbenzene:
Benzene
Ethylene
Ethyl benzene
(2) Catalytic dehydrogenation of ethylbenzene* to styrene:
Ethylbenzene
3.3.1.1 Basic Process
Styrene
Hydrogen
A step-by-step diagram of the benzene alkylation process and the ethyl -
benzene dehydrogenation process is shown in Figures 3-3 and 3-4. A detailed
process flow diagram of the complete styrene process is shown in Figure 3-5.
(a) Step 1 — Benzene Alkylation to Form Ethylbenzene. First, both feed
and recycled benzene are dried to remove water. The dry benzene (99
percent) (Stream 1, Figure 3-5) and ethylene (95 to 99 percent) are
fed continuously into the reactor operating at essentially atmospheric
pressure. A small amount of ethyl chloride is added to the ethylene
feed as a source of hydrogen chloride which acts as a catalyst promoter.
The catalyst, granular aluminum chloride, is fed to the reactor at a
constant rate.
The alkylation reaction occurs and is maintained at approximately
95 C by cooling water. The aluminum chloride combines with the benzene
and ethylbenzene to form an insoluble complex. Next, the reactor
effluent, which contains benzene, ethylbenzene, and the insoluble com-
plex, goes to a settler, where the crude ethylbenzene is decanted and
the heavy catalyst complex is recycled to the reactor. Approximately
80 percent of the aluminum chloride may be recovered. Reactor vent
gas is routed through a condenser and scrubbers in the alkylation
reaction section to recover any aromatics and to remove hydrogen
chloride.
(b) Step 2 — Treating Liquid Mixture from Alkylation Reactor. The crude
ethylbenzene from the settler (Stream 2, Figure 3-5) is washed with
water and neutralized with a caustic solution to remove traces of
chlorides. Then the crude ethylbenzene is fed to the ethylbenzene
purification section.
3-7
-------
(c) Step 3 -- Purifying the Ethyl benzene. The crude ethyl benzene consists
of 36 to 42 percent ethyl benzene, 40 to 55 percent benzene, and ten to
20 percent polyethylbenzene (PEB) and high-boiling materials. The
first phase in the purification of the crude ethyl benzene is to sepa-
rate and remove benzene (Stream 3, Figure 3-5) in the benzene recovery
column. After distillation to remove water, the dry benzene is
returned to the alkylation reactor. In the second phase, the product
ethylbenzene (Stream 4, Figure 3-5) is separated from the remaining
high-boiling materials and PEB in the ethylbenzene recovery column.
Finally, the high-boiling materials are distilled in the PEB column to
separate the residue oil from the PEB which is recycled.
(d) Step 4 — Ethylbenzene Dehydrogenation to Produce Styrene. The purified
ethylbenzene is preheated with steam (to 160°C) and by heat exchange
(to 520°C). Superheated steam (710°C) and the ethylbenzene vapors are
mixed continuously and fed into the reactor. The reactor contains a
selective, fixed dehydrogenation catalyst such as zinc, aluminum,
chromium, iron, or magnesium oxide. The dehydrogenation of ethylben-
zene to form styrene occurs.
The reaction product leaves the top of the reactor and is cooled
first by the incoming ethylbenzene and then by steam in heat exchangers.
The mixture then goes through a water-cooled and/or air-cooled con-
denser, where the steam and crude styrene Vapor are condensed, and
then flows to the separators.
(e) Step 5 ~ Separating Crude Styrene from Other Reaction Products. In
this step, the liquid crude styrene, consisting of water, styrene,
ethylbenzene, toluene, benzene, and high-boiling materials, and the
vent gases, hydrogen, carbon monoxide, carbon dioxide, and hydrocarbons,
are separated.
After separation, the condensable hydrogen-rich gas (Stream 9,
Figure 3-5) is sent to the recovery section; the process water con-
densate is decanted (Stream 10, Figure 3-5) and goes to a stripper;
and the remaining crude styrene (Stream 11, Figure 3-5) is sent to
the styrene purification section. The hydrogen-rich gas (Stream 9,
Figure 3-5) is compressed and then cooled in the recovery section to
recover aromatic organics, which are returned to the process. After
removal of the organics, the hydrogen-rich gas (Stream 12, Figure 3-5)
usually is sent to the steam superheater, where it is used as fuel,
The process water condensate (Stream 10, Figure 3-5) is fed to the
process water stripper, where dissolved aromatic organics are removed
and returned to the process. The purified water is sent to the plant
boiler for use as a boiler feedwater.
(f) Step 6 — Purifying the Styrene. In the styrene puri
benzene and toluene (Stream 13, Figure 3-5) are first
the crude styrene in the benzene-toluene column. The
toluene then typically are separated by distillation;
sold and the benzene is either sold or recycled. The
benzene (Stream 6, Figure 3-5) is separated from the
in the ethylbenzene recycle column and reprocessed.
product styrene (Stream 14, Figure 3-5) is separated
the styrene finishing column.
fication section,
separated from
benzene and
the toluene is
recycled ethyl-
styrene and tars
Finally, the
from the tars in
3-8
-------
FIGURE 3-3
BENZENE ALKYLAT10N PROCESS
ALKYLATION
REACTION
SECTION
• Benzene Storage
• Benzene Drying Column
• Catalyst Preparation
• Reactor
• Catalyst Settling Vessel
STREAM 2
Liquid Product Exits Reactor
STEPl
BENZENE ALKYLATION TO
FORM ETHYLBENZENE
a. Benzene is dried
b. Benzene, ethyl chloride, and aluminum chloride
catalyst are fed to reactor
c. Alkylation reaction occur*
d. Catalyst i* recovered
TREATING
SECTION
Water Wash
Caustic Wash
Crude Ethylbenzene
Storage
ETHYLBENZENE
PURIFICATION
SECTION
• Benzene Recovery
Column
* Ethylbenzene
Recovery Column
« Polyethylbenzene
Column
STREAM 4
Ethylbenzene
STEP 2
TREATING THE LIQUID MIXTURE FROM
ALKYLATION REACTOR
a. Crude ethylbenzene is washed with water
and caustics to remove chlorides
STEP 3
PURIFYING ETHYLBENZENE
a. Benzene is separated from crude ethylbenzene
b. Ethylbenzene is separated from 'heavies'
c. Polyethylbenzenes are recovered
ETHYLBENZENE
STORAGE
3-9
-------
FIGURE 3-4
ETHYLBENZENE DEHYDROGENATION PROCESS
DEHYDROGENATION
REACTION
SECTION
• Heat Exchangers .
• Catalytic Dehydrogenation
Reactor
» Condensers
HYDROGEN
SEPARATION
SECTION
• Separator
• Recovery Section
• Process Water Stripper
STYRENE
PURIFICATION
SECTION
* Benzene-Toluene Column
» Ethylbenzene Recycle Column
• Styrene Finishing Column
STEP 4
DEHYDROGENATING ETHYLBENZENE
TO PRODUCE STYRENE
a. Ethylbenzene is vaporized and superheated,
mixed with superheated steam,
and fed to reactor
b.. Dehydrogenation reaction occurs
c.. Reaction mixture is cooled and crude styrene
and water are condensed
STEP 5
SEPARATING CRUDE STYRENE FROM
OTHER REACTION PRODUCTS
a. Non-condensible hydrogen-rich gas is
separated from water-styrene mixture
b. Hydrogen-rich gas is compressed and cooled
to remove aromatic organics
and then burned
c. Water is removed from styrene and purified
d. Styrene is sent to purification section
STEP 6
PURIFYING THE STYRENE
a. Benzene and toluene are separated
from styrene
b. Unreacted ethylbenzene is separated
from styrene
c. Styrene is separated from residue
and sent to storage
3-10
-------
s
3-1 I
-------
3.3.1.2 Process Variations
The benzene alkylation and ethylbenzene dehydrogenation processes
described above are typical of most ethylbenzene or styrene facilities.
However, among the plants, several process variations do exist.
Ethylene used to produce styrene usually is 95 to 99 percent pure.
However, a dilute stream containing as little as ten percent ethylene can
be used.
The pressure in the alkylation reactor can range from near ambient to
2759 kPa and temperatures may range from 80 to 400°C, depending on the
pressure and the catalyst used. Some high-pressure, vapor-phase processes
use solid catalysts, such as the newer Mobil/Badger process and the Alkar
process by UOP. The Mobil/Badger process was used in a demonstration unit
at American Hoechst's Baton Rouge, Louisiana, plant. The Alkar process by
UOP is used by El Paso Products Company at Odessa, Texas, with a boron tri-
fluoride catalyst. El Paso uses a dilute ethylene stream (50 percent) to
feed the process and sends the offgases from the alkylation reaction to its
boiler as fuel.
the catalyst lasts for several years.
Another process variation in the reaction step is operating the alkyl-
ator at pressures greater than atmospheric pressure with high-purity
ethylene. In this variation, natural gas or nitrogen is added periodically
to maintain the desired pressure in the reactor. Few, if any, emissions
are vented from the reactor. Instead, the natural gas or nitrogen fed with
the ethylene and any inert gases produced as by-products in the reactor go
with the alky!ate to a degassing step where they are removed when the pres-
sure on the alky!ate is reduced.
12
This process does not produce by-products or sludge and
13
3-12
-------
The dehydrogenation reactor can be operated either at constant tempera-
ture with heat added in the reactor or under adiabatic conditions with all
the heat supplied by diluting the ethylbenzene feed entering the reactor
with the superheated steam. After recovery of the aromatics, the hydrogen-
rich offgas from the separator following the dehydrogenation reactor can be
either burned as fuel or processed further to recover hydrogen or carbon
dioxide for use in other processes.
A variation of the styrene purification process from that shown in
Figure 3-5 is to first separate a mixture of ethylbenzene, toluene, and
benzene from the styrene and tars. The ethylbenzene, toluene, and benzene
stream then is fed to a column where the benzene and toluene are separated
from the ethylbenzene. The high-boiling materials from the first separation
go to another column, where the styrene is separated from the tars.
14-17
3.3.2 Ethylbenzene from Mixed Xylenes
In addition to typical benzene alkylation to produce ethylbenzene,
ethylbenzene also can be extracted from mixed streams. In fact, of the ten
integrated EB/S operations, one plant extracts ethylbenzene from xylene
streams and then dehydrogenates the ethylbenzene to produce styrene. Three
plants producing only ethylbenzene also recover it from xylene streams. In
contrast to benzene alkylation, which produces one major product and has
definitive start and end points, mixed xylene extraction processes multi-
component streams in many steps for production of benzene, toluene, gaso-
line components, and other products. Since many of these steps do not
involve pure ethylbenzene, the extraction process can be defined as the one
step which distills ethylbenzene from xylene. However, upstream steps must
be altered greatly to change the amount of ethylbenzene available from mixed
3-13
-------
xylene extraction.* Thus, the extraction process will be reviewed starting
with the sources of ethyl benzene-containing xylenes. Figure 3-6 shows a
conceptual flow chart of the mixed xylene extraction process and Figure 3-7
shows the benzene, toluene, and xylene processing flow diagram.
3.3.2.1 Sources of Xylene Streams
The two major sources of such ethyl benzene-containing xylenes are
(1) catalytic reformate from refineries, and (2) pyrolysis gasoline from
the ethylene plant. The refinery reforming process produces aromatics by
catalytically dehydrogenating a naphtha feed containing mainly alkanes in
the 65°C (150°F) to 200°C (400°F) boiling range. The exact composition of
the reformate depends on the catalyst, feed composition, and operating con-
ditions. Table 3-3 shows a typical composition for reformate streams.
TABLE 3-3.18 TYPICAL COMPOSITIONS OF XYLENE-CONTAINING STREAM
All Components
Benzene
Toluenes
Ethyl benzene
Xylenes o/m/p
Cg + Aromatics
N&n- Aromatics
Cg's Only
Ethyl benzene
o-xylene
m-xyl ene
p-xyl ene
Refinery Reformate
(volume percent)
5
24
4
5/9/4
10
39
20
23
40
17
Pyrolysis Gasoline
(volume percent)
32
14
6
1/3/1
13
30
53
12
25
10
In the ethylene process, thermal cracking of various hydrocarbons pro-
duces short chain olefins. With newer ethylene plants moving to higher
molecular weight hydrocarbon feeds, increasing amounts of ethylbenzene and
xylene-containing streams (pyrolysis gasoline) will be available from ethylene
3-14
-------
FIGURE 3-6
ETHYLBENZENE EXTRACTION
FROM MIXED XYLENES
STEP I
Identify sources of
ethylbenzene containing
xylene strains
i
r
Separate
aromatics
from aliphatics
Alkenes
and alkanes
Aromatics
STEP 2
Separate
benzene, toluene,
and xylenes
o-, p-, m-xylene
and ethylbenzenes
• Toluene
• Benzene
STEP 3
Crystallize and
fractionate to
isolate xylenes
and ethylbenzenes
m-xylene
ethylbenzene o-xylene
p-xylene
3-15
-------
cs
Si =
.s I
-
if
if i J
i| S .»
• *o .«
-------
production. Table 3-3 shows typical aromatics yields in pyrolysis
gasoline from ethylene plants.
3.3.2.2 Basic Process
The reformate and pyrolysis gasoline streams are either blended directly
into gasoline or processed further for recovery of specific aromatics.
This processing includes extracting aromatics, separating benzene-toluene-
xylene (BTX), and processing mixed xylenes. Briefly, ethylbenzene is
extracted from the xylene streams in the steps described below.
(a) Step 1 — Aromatics Separation. Aromatics extraction removes the
alkanes and alkenes from the reformate by preferential absorption of
the aromatics in a suitable solvent. The aromatics are recovered from
the solvent and sent to BTX separation; the alkanes and alkenes are
sent to distillate oil blending or to olefins.
(b) Step 2 — BTX Separation. In BTX separation, the aromatics first are
pretreated to remove residual alkenes and then split by straight dis-
tillation to isolate the specific aromatics: benzene, toluene, and
mixed xylenes. In cases in which only one or two aromatics are desired,
the extraction step often is preceded by a prefractionation unit.
This unit isolates a specific cut of the reformate boiling in a narrow
range around the desired aromatic. The aromatic then is recovered by
routing only this specific cut to extraction and sending the remaining
aromatics to gasoline blending. Using such a prefractionation unit
greatly reduces the size of the downstream extraction unit.
(c) Step 3 — Crystallization and Fractionation. In producing xylenes,
mixed xylenes from BTX separation typically are crystallized to produce
p-xylene, fractionated to isolate o-xylene and small amounts of ethyl-
benzene and m-xylene, and isomerized to convert the remaining ethyl-
benzene and m-xylene to o- and p-xylene. Crystallization of mixed
xylenes is required for separation since the close boiling points of
m- and p-xylene make distillation impossible. Isomerization of mixed
xylenes is used to enhance production of o- and p-xylene since demand
for them is larger than their occurrence in mixed xylene streams. The
fractionation of ethylbenzene and m-xylene requires very large distil-
lation units. Since separation of m- and p-xylene by crystallization
is costly, several new processes based on adsorption and extraction
have been developed as alternatives.
At present, very little ethylbenzene is recovered in mixed xylene
processing; production is less than 100,000 Mg/yr (110,000 tons) and
capacity is less than 200,000 Mg/yr (220,000 tons). However, very large
3-17
-------
quantities are potentially available. Based on 1978 reformate and ethylene
production capacities, more than 6 x 106 Mg/yr of ethylbenzene are available
from these streams. Virtually all this ethylbenzene becomes a high octane
component in gasoline or is converted to xylenes.
3.3.3 Styrene from Ethylbenzene Hydroperoxidation
In addition to the typical ethylbenzene dehydrogenation process,
styrene is produced from ethylbenzene by a hydroperoxidation reaction.
3.3.3.1 Ethylbenzene Hydroperoxidation Capacity
The only U.S. application of the ethylbenzene hydroperoxidation process
is the Oxirane Corporation unit in Channelyiew, Texas. This unit represents
about ten percent of U.S. styrene capacity.
3.3.3.2 Basic Process
The basic process shown in Figures 3-8 and 3-9 is described below.
(a) Step 1 ~ Ethylbenzene Oxidation. The initial step in the process is
the oxidation of ethylbenzene with air to produce ethylbenzene hydro-
peroxide and minor amounts of a-methyl-benzyl alcohol and acetophenone.
The exit gas from the reactor columns contains inert gas (principally
nitrogen) and a mixture of organic vapors and is cooled and scrubbed
to recover aromatics before venting.
After the oxidation reaction, the mixed organic stream is fed to
an evaporation system where some of the unreacted ethylbenzene, along
with low-boiling contaminates, is removed. The recovered ethylbenzene
is sent to the ethylbenzene recovery system for treatment prior to
reuse. The concentrated solution of ethylbenzene hydroperoxide and
reaction by-products goes to the epoxidation reaction system.
(b) Step 2 ~ Epoxidation of Propylene with Ethylbenzene Hydroperoxidation.
Next, propylene is mixed with the concentrated ethylbenzene hydroper-
oxide solution and epoxidized over a proprietary catalyst mixture at
high pressures to form propylene oxide. The main by-product of this
reaction is acetophenone. After epoxidation, the pressure is reduced
and most of the residual propylene and other low boilers are separated
from the reaction mixture by a distillation system. The vent stream
from this distillation system containing propane, propylene, and minor
amounts of other gases is fed to the plant fuel gas manifold as fuel.
The recovered propylene is recycled for reuse in the epoxidation
reaction.
The crude epoxidate stream from the epoxidation reactor next is
treated to remove the acidic impurities and residual catalyst by
3-18
-------
FIGURE 3-8
ETHYLBENZENE HYDROPEROXIDAT10N PROCESS
Ethylbenzene
ot Methyl-benzyl alcohol
plus acetophenone
Propylene-
Propylene oxide-
Air
ETHYLBENZENE
OXIDATION
REACTOR
Ethylbenzene
Hydroperoxide
EPOXIDATION
REACTOR
Acetophenone
\
DEHYDRATION
Styrene
STYRENE
PURIFICATION
STEP 1
Ethylbenzene oxidation to ethylbenzene
hydroperoxide, CV-tnethyl-benzyl alcohol,
and acetophenone
STEP 2
Epoxidation of propylene with ethylbenzene
hydroperoxide to propylene oxide
and acetophenone
STEP 3
Dehydration of acetophenone and Q'-methyl-benzyl
alcohol to styrene
STEP 4
Purification of styrene
3-19
-------
UJ
o
cc
a
os
LU
ce
C3
LU a. —
a: o o
=3 a: v-.
S2 So
Ul CO
LU
M
LU
OQ
LU
D
o
3-20
-------
washing^with a caustic water salt. The washed epoxidate stream then
is distilled to separate the propylene oxide product and to purify the
crude propylene oxide from this distillation section. Recovered
propylene is recycled to the epoxidation process; residual water is
returned to the epoxidate washing operation; and residual liquid
impurities are collected in a fuel tank and fed to the process furnace
for disposal. The purified propylene oxide is sent to storage tanks
prior to shipping.
The organic layer remaining after propylene oxide removal goes to
the ethylbenzene, a-methyl-benzyl alcohol recovery operation. In the
recovery section, the organic layer again is washed in caustic and
water and then distilled. The distillation removes any remaining
ethylbenzene and then separates organic waste streams from the a-methyl -
benzyl alcohol and acetophenone. The recovered ethylbenzene, along
with ethylbenzene from other parts of the process and fresh ethylben-
zene, is treated prior to use in the oxidation reactors and the organic
waste liquids are stored and used as fuel.
(c) Step 3 — Dehydration of Acetophenone and A-Methyl-Benzyl Alcohol. The
mixed stream of a-methyl-benzyl alcohol and acetophenone from the ethyl-
benzene recovery section then is dehydrated over a solid catalyst to
produce styrene. The residual catalyst solids and high-boiling impuri-
ties are separated from the organic product stream and collected for
disposal.
(d) Step 4 — Styrene Purification. The crude styrene goes to a series of
vacuum distillation columns where pure styrene monomer is separated
and recovered for sale.
After the pure styrene monomer is removed, the residual organic
stream contains crude acetophenone, along with a mixture of various
impurities. A solid catalyst is dispersed in this crude acetophenone
stream. The mixture then is treated under pressure with hydrogen gas
to convert the acetophenone to a-methyl-benzyl alcohol. The residual
hydrogen stream containing some organic vapors is injected into the
waste vapor fuel manifold for combustion in the process furnace. The
solid catalyst is activated by treating with hydrogen in a nitrogen
medium. The solid catalyst is separated from the crude a-methyl-benzyl
stream and is sent to an intermediate storage area. The crude a-methyl-
benzyl alcohol from the hydrogenation section is returned to the ethyl-
benzene, a-methyl-benzyl alcohol recovery section for separation into
its useful components.
3.4 EMISSIONS
The primary process sources of benzene and volatile organic chemical
(VOC) emissions for the four EB/S processes are described below.
3-4.1 Benzene Alkyl at ion/Ethyl benzene Deh.ydrogenation
These two processes are discussed together because they represent the
typical EB/S production process. Table 3-4 shows the VOC and benzene
3-21
-------
emissions for a typical plant. This plant uses the process described in
Section 3.3.1 and shown in Figures 3-3 and 3-4 and has a capacity of
300,000 Mg/yr (330,000 tons) and an ethylbenzene capacity of 345,000 Mg/yr
(380,000 tons).
The four process emission sources in a typical EB/S plant are (1) the
alkylation reactor area vents, (2) the atmospheric and pressure column
vents, (3) the vacuum column vents, and (4) the hydrogen separation vent.
The emission rates for these sources are shown in Table 3-4 along with a
national emission estimate and typical compositions and flows. The first
three process vent streams are mainly low-flow, high-concentration streams,
whereas the hydrogen separation stream is a high-flow, low-concentration
stream. Fugitive and storage emissions from EB/S production are not dis-
cussed in this document. These emissions will be regulated under separate
standards.
The emissions in Table 3-4 are for an integrated plant producing both
ethylbenzene and styrene. For plants producing ethylbenzene only, the
emissions can be estimated by summing the reactor area and atmospheric and
pressure column vent emissions. For plants producing styrene only, the
emissions are estimated by summing the vacuum column vent and hydrogen
separation vent emissions.
3.4.1.1 Alkylation Reactor Area Vents
Reactor area vents remove various inerts plus entrained aromatics
(benzene) from the benzene alkylation reactor. These inerts include nitro-
gen or methane used in pressure control, unreacted ethylene, reaction by-
products, and impurities in the ethylene feed (ethane, methane, and three-
and four-carbon hydrocarbons). In most plants which use the traditional
3-22
-------
TABLE 3-4. UNCONTROLLED VOC AND BENZENE EMISSIONS^ EB/S PRODUCTION
VIA ALKYLATION AND DEHYDROGENATION
a
300,000 Mg/yr
Estimated Total
Nationwide Benzene
Benzene
Alkylation Reactor
Area Vents
Atmospheri c/Pressure
Column Vents
Vacuum Column Vents
B/T Column01
Other
3.
12.
3.
0.
0
0
0
5
Other
3
6
1
0
VOC
.0
.0
.3
.3
Benzene
11.
45.
11.
1.
2
0
2
8
Other VOCD
H-:
22.
5.
1.
2
5
0
0
Emissions,
200
1,200
500
100
Mg/yrc
Vent Stream Characteristics3
Alkyl Reaction
Area Vents
Atm/Pressure
Column Vents
Vacuum Column Vents
B/T
a
Other
Hydrogen
Separation Vent
Total
kg/hr
liters/hr at STP
112
100,000
100
60,000
25
15,000
11.2
10,000
6,700 6
14 x 10b
Weight %
Benzene
Other Aromatics
Cl
cl/Cc _
kg/hr
Benzene
Total Combustible
VOC
10
32
10
48
11.2
58.0
45
5
12
20
18
45
82
45
5
12
20
18
11.2
20.0
15
5
30
5
45
1.7
7.0
6
30
2
62
The figures are significant to one figure at best. More than one digit is shown in
cases to keep the proper ratios between the various vent streams and between the
kg/10,000 kg and kg/hr typical plant figures.
Other VOC does not include benzene, ethane, and methane.
"Nationwide benzene emissions from the EB/S industry factoring in current levels of
control and current production levels.
Benzene/toluene removal column.
B "
Inerts are all other compounds except combustible VOC and include H0, CO
m M n anH on nr, i
and so on.
3-23
-------
liquid phase aluminum chloride (A1C13) catalyst, high purity ethylene is
used; therefore, inerts other than those from the ethylene feed are impor-
tant in creating the vent streams. Before release, the vent streams
typically are cooled and then scrubbed with polyethylbenzene, water, or
both to recover aromatics.
In contrast to plants using Aid- catalyst, plants using the newer
solid support catalysts of the UOP or Mobil/Badger process generally employ
low-purity ethylene. The resulting reactor vent flows are very large and
process economics require the vent gases to be used as fuel. Thus, although
these vent gases can be considered potential emissions, they are never
vented in practice.
3.4.1.2 Atmospheric and Pressure Column Vents
These column vents mainly remove non-combustibles dissolved in the
column feeds, light straight-chain hydrocarbons not released in the reac-
tion area, and any entrained aromatics. The benzene drying column vent
also removes any low-boiling impurities in the benzene feed. Most of the
emissions occur from the first column in the distillation train, since most
of the compounds generating vent streams would be vented in this column.
3.4.1.3 Vacuum Column Vents
The vacuum column vents remove air that leaks into the column, light
hydrocarbons and hydrogen formed in dehydrogenation, non-combustibles dis-
solved in the column feed, and any entrained aromatics. The vacuum columns
typically are used in the three distillation steps in styrene purification
and in the last distillation step in ethyl benzene purification. As with
atmospheric and pressure columns, most of the emissions from vacuum columns
would come from the first column in the distillation train.
3-24
-------
3.4.1.4 Hydrogen Separation
The hydrogen-rich gas stream from the separator, following the dehydro-
genation reaction, consists of hydrogen, methane, ethane, ethylene, carbon
dioxide, carbon monoxide, and aromatics formed in the dehydrogenation reactor.
In normal operation this stream is cooled and compressed to recover aromatics,
rather than vented, and the remaining hydrogen-rich gas is sent to the steam
superheater as fuel. This stream is vented directly to the atmosphere or
to a flare during plant startup and shutdown and outages of the vent gas
compressor in the recovery section and during startup and shutdown. For
this report, the hydrogen separation stream under normal conditions is con-
sidered an emission stream since, at present, one plant vents this stream
to a flare. In view of the uncertainties in flare control efficiencies
(Section 4.3), the practice potentially results in high benzene emissions.
Although the hydrogen separation stream is classified as an emission
stream, most plants presently burn the stream as fuel because of its high
Btu content. Thus, control method, cost, and impact analyses were not per-
formed since the effect on the industry of regulating the stream would be
minimal.
3.4.2 Emissions from Ethylbenzene via Mixed Xylenes
Present analysis indicates that ethylbenzene from mixed xylenes emits
no benzene because all benzene would likely be removed prior to ethyl benzene
extraction (see Figure 3-6). Furthermore, no new benzene emissions would
be created in reforming or pyrolysis gasoline production since existing
pyrolysis gasoline capacity would provide sufficient ethylbenzene to meet
demand. Thus, expansion of these units, with the potential for new benzene
emissions, is not required.
3-25
-------
The above analysis assumes (1) that producers would not isomerize mixed
xylenes to ethylbenzene, but only extract the ethylbenzene already present
in the mixed xylenes, and (2) that producers would not transalkylate toluene
to mixed xylenes for ethylbenzene. Both of these processes can produce
benzene emissions. The above conditions were assumed since extensive analy-
sis would be required to predict the degree of use of these processes.
Since preliminary study of ethylbenzene via mixed xylenes showed that this
process was not a viable control method (Section 4.6), this extensive analy-
sis was not done.
3.4.3 Emissions from Styrene via Ethylbenzene Hydroperoxidation
This section discusses emissions from the ethylbenzene hydroperoxida-
tion process. The emission data below are based on the capacity of 545 x 10V
Hg/yr (600 x 103 tons/yr) styrene at the one U.S. plant.
There are three main process emissions sources in the hydroperoxidation
process: (1) ethylbenzene oxidation reactor vent, (2) propylene recycle
purge vent, and (3) vacuum column vents. Of these three sources, the vacuum
column vents are the only large benzene emission point. The benzene present
in the vent sources results from benzene impurities in the ethylbenzene
feed and minor side reactions. Each source is described further below.
3.4.3.1 Ethylbenzene Oxidation Reactor Vent
This vent releases nitrogen and excess oxygen brought into the process
with the reaction air; carbon monoxide, carbone dioxide, and light organics
produced in the reaction; and organics entrained in the vent gas. The vent
o o
stream has a reported flow of approximately 16.4 m /sec (580 ft /sec) and
contains approximately 4.4 percent by weight aromatics, or 3,400 kg/hr
(7,480 Ibs/hr).
3-26
-------
At the one U.S. plant using this process, the vent gas is scrubbed
with oil and water for 99 percent removal of organics. The resulting stream
is vented to the air and contains approximately 35 ppm of benzene or
7.2 kg/hr.19 .
3.4.3.2 Propylene Recycle Purge Vent
This vent releases propane, propylene, ethane, and other impurities
that accumulate in recycling propylene to the epoxidation reactors. No
data are available on the flow and composition of this vent. However, based
on process flow diagrams and the use in the process of chemical-grade propylene
(up to eight percent propane), the stream is a high-Btu gas representing about
50 x 10 Btu/hr and likely would be used as fuel. Again, based on process
flow diagrams, no significant benzene is expected in the vent stream.
3.4.3.3 Vacuum Columns
The ethylbenzene hydroperoxidation process contains numerous vacuum
columns and evaporators. The vents from these vacuum devices release air
drawn into the equipment, inerts and light organics dissolved in the column
feeds, nitrogen used for process control, and entrained aromatics. The
combined vacuum device vents have a reported flow of 1.0 x 10 liters/hr
and contain approximately two percent by weight benzene, or about 27 kg/hr
(59 Ib/hr). These small quantities of benzene present in the process accumu-
late at the top of the columns and, due to their high volatility, are drawn
out through the vents.
3.4.4 Baseline Emissions
Benzene and VOC emission rates assigned to the uncontrolled model plant
were derived from test data collected both by EPA and by plant personnel at
existing EB/S facilities. Variations in emissions potential among plants
3-27
-------
were averaged, assuming a constant capacity, to yield a "typical" emission
rate for each vent or group of vents in a plant. Vent stream composition
was measured similarly using test data collected at existing plants.
Total nationwide benzene emissions were estimated using the uncon-
trolled model plant emission rates, which were scaled for both current
production levels and benzene control equipment at each plant. In plants
where only ethyl benzene or styrene alone is produced, the corresponding
process area vents of the model plant were used in estimating the contribu-
tion of that source.
3-28
-------
3.5 REFERENCES FOR CHAPTER 3
1. Paul, S.K. and S.L. Soder. Ethylbenzene. In: Chemical Economics
Handbook. Menlo Park, California, Stanford Research Institute.
May 1977.
2. Soder, S.L. Styrene. In: Chemical Economics Handbook. Menlo Park,
California, Stanford Research Institute. January 1977.
3. Reference 2. '.
4. Manual of Current Indicators. In: Chemical Economics Handbook.
Menlo Park, California, Stanford Research Institute. October 1978.
p. 252.
5. References 2 and 4.
6. Lowenheim, F.A. and M.K. Moran. Faith, Keyes, and Clark's Industrial
Chemicals. New York, John Wiley and Sons. 1975. p. 365-370,
779-785.
7. Key, J.A., Hydroscience. Trip report for Cos-Mar Plant, Cosden Oil
and Chemical Company, Carville, Louisiana. July 28, 1977.
8. Key, J.A., Hydroscience. Trip Report for Dow Chemical U.S.A.,
Freeport, Texas. September 9, 1977.
9. Responses to EPA Section 114 letters, Houdry questionnaires, and State
emission inventory files. (See Addendum on Excess Emissions).
10. Reference 6.
11. Better Path to Ethylbenzene. Chemical Engineering, p. 120-121.
December 5, 1977.
12. Reference 9.
13. Ethylbenzene (Alkar) — UOP Process Division. Hydrocarbon Processing.
p. 52. November 1977.
14. Carlson, E. Xylenes. Chemical Economics Handbook. Menlo Park,
California, Stanford Research Institute. December 1975.
15. Brownstein, A.M. U.S. Petrochemicals, Technologies, Markets, and
Economics. Tulsa, Oklahoma, The Petroleum Publishing Company. 1972.
p. 98-132.
16. Atkins, R.S. Which Process for Xylenes. Hydrocarbon Processing.
p. 127-136. November 1970.
3-29
-------
17. PEDCo Environmental. Evaluation of Emissions from Benzene-Related
Petroleum Processing Operations. Cincinnati, Ohio. October 1978.
Draft.
18. References 14-17.
19. Letter from Fretwell, S., Oxirane, to Farmer, J., EPA.
1979. Concerning ethyl benzene air oxidation process.
November 19,
3-30
-------
4. CONTROLS
This chapter describes controls for benzene and VOC emissions in the
ethylbenzene/styrene (EB/S) industry and analyzes their efficiencies, advan-
tages, and limitations. The controls are condensation, absorption, flaring,
combustion in boilers,* and use of an alternate process. Since new sources
should not differ greatly from existing sources, these controls are appli-
cable to both new and existing EB/S plants.
4.1 CONDENSATION
Refrigerated condensers can be used to remove benzene from the vent
stream. A complete control system includes the condenser, piping for the
vent streams, and refrigerated coolant.
Control of emissions by condensation has several advantages. Condens-
ers are applied easily to the low volume, elevated temperature, saturated
vent streams in EB/S plants. They recover benzene and other aromatics as
usable products. They do not require new refrigeration and purification
equipment since all EB/S plants already have refrigerated brine systems to
provide the required coolant (assuming the existing systems have sufficient
excess capacity to accommodate the load). Finally, the condensed organics
can be routed easily into the production distillation columns for recovery.
The control efficiencies of condensers are determined by the pressure
and temperature at which the condenser operates and by the concentration
and vapor pressure of the organics in the vent streams. Condensers on EB/S
*"Boilers" indicates combustion in any device such as process heaters,
superheaters, and other similar units.
4-1
-------
vent streams generally operate at one to three atmospheres^and 2 to 5°C on
the condenser coils. In addition, they would be applied to streams at 70
to 100 percent saturation in benzene at 40 to 50°C. Condensers under such
conditions would achieve 80 to 90 percent reduction of benzene.
A higher efficiency for condensers traditionally can be achieved by
increasing pressure and lowering temperatures. However, this is not possible
with EB/S plants because of specific pressure and temperature limitations.
Higher pressure generally would not be used since compressor costs rise
rapidly with increased pressure, while benzene removal efficiency would
increase only slowly. Lower temperatures are not possible because the
potential exists for the benzene and water in the vent stream to freeze.
The use of condensation is limited by other factors. A condenser cannot
handle a short duration, large volume stream resulting from the release of
the hydrogen separation vent. Furthermore, although most of the low volume
vent streams are saturated, emission streams like those of the alkylation
reactor vents and the vacuum column vents in the model plant are not close
to saturation with benzene even at 2 to 5°C. Therefore, a system operating
at 2 to 5°C and one to three atmospheres would achieve very little control
of benzene. Condensation also achieves little control of the two- to
five-carbon straight-chain hydrocarbons in the vent streams. At 2 to 5°C,
none of the vent streams approaches saturation for any of these compounds.
4.2 ABSORPTION
Another technique for reducing benzene emissions is absorption. In an
EB/S plant, an absorption system includes a packed tower and pumps to circulate
the scrubbing liquid, a heat exchanger to cool the scrubbing liquid, piping
for the vent gases, a brine cooling system, and the scrubbing liquid itself.
4-2
-------
Absorption of benzene from an EB/S plant has several advantages. The
system could be applied easily. For example, polyethylbenzene (PEB), a
mixture of mono-, di-, and triethyl benzene produced during benzene alkyla-
tion, is a good benzene solvent. If benzene is absorbed in PEB, the resulting
solution can be recycled back to the process and the benzene in solution
can be recovered and used as part of the feedstock in the alkylation reactor.
Since the PEB is recycled in any case, it would be inexpensive to use as
the absorbent and would alleviate the need for special scrubbing liquid or
absorbate recovery. An additional advantage for benzene absorption at EB/S
plants is that no new refrigeration equipment would be required since all
EB/S plants already have refrigerated brine systems providing coolant
(assuming the existing systems have sufficient excess capacity to accommo-
date the load).
The main disadvantages of an absorption system are that (1) it is
unsuitable for handling high volume, intermittent releases of gases, and
(2) absorbers also produce a pressure drop in the gas stream as it passes
through the towers. The use of additional ejectors or the use of packed
towers instead of plate or tray towers, however, can compensate for the
loss of pressure.
Absorption systems are capable of maintaining good benzene removal
efficiencies (between 80 and 99 percent) with vent streams both saturated
and unsaturated with benzene. The removal efficiency achieved for a par-
ticular absorber and a given solvent is a function of gas composition, sol-
vent temperature, flow rate, tower height and diameter, and the nature of
the packing material; removal efficiency increases with tower height and
solvent flow rate.
4-3
-------
4.3 FLARING
Various vent streams can be manifolded to a flare for thermal destruc-
tion. The flare system includes the manifold piping from the vents, an
oxygen monitoring system to warn of possible explosive concentrations in
the manifold piping, and a flare. The system also includes a knockout
drum to remove liquid from the vent gases, a water seal, a flare tip flame
retention ring, and the necessary controls to provide for automatic ignition
and steam application.
The main advantage of the flare is its ability to control streams and
compounds for which condensation or absorption are not applicable. For
example, flares can control upset releases and streams such as the alkyla-
tion vents which are not saturated in aromatics. In addition, flares can
control the light straight-chain hydrocarbons (VOC) in EB/S vent gases not
controllable by condensation or absorption.
The major technical difficulties with flares occur in manifolding. As
noted in Section 4.5, designs and equipment exist to overcome these
difficulties.
Various factors affect flare combustion efficiency, but the chief one
is the thoroughness with which flare exit gases are mixed with air. The
key design variables affecting mixing are as follows:
• Flare tip diameter — As flare diameter increases, the turbulence required
to mix air thoroughly with all exit gases becomes harder to achieve.
• Exit velocity — The kinetic energy of the exit gases is important to
mixing; that is, the greater the velocity, the more complete the mixing,
until a maximum is reached when the flame blows out or the noise becomes
unacceptable.
• Steam injection — Injecting steam also adds kinetic energy to improve
mixing .and entrains large amounts of air. Increased steam improves
mixing until a point is reached where the quenching by the steam and
the excess air begins to decrease combustion efficiency.
4-4
-------
Furthermore, the presence of saturated orgam'cs or aromatic compounds may
decrease efficiency since such compounds tend to be thermally stable. Also,
if nitrogen is used for pressure control or continuous purging, or if vent
streams contain large volumes of compounds with low heats of combustion or
non-combustibles, flare combustion may be affected with a decrease in combus-
tion efficiency.
At present, no conclusive data are available on flare efficiency.
Calculations and limited test data show benzene destruction efficiencies
ranging from 60 to 99 percent. These include engineering evaluations by
Hydroscience, Inc., ' calculations based on residence time,3 limited and
unverified test data on a flare burning methane from a flare manufacturer,4
and EPA in-house estimates.
For the purposes of this study, a 60 percent destruction efficiency of
benzene in flares will be used. This efficiency was chosen because many
existing flares in the EB/S industry are not of optimum design. There are
large diameter flares designed to handle emergency releases being used to
control the low volume, continuous EB/S vent gases. With such designs,
optimum mixing is not achieved since the vent gas exit velocity is low and
large flares generally cannot inject steam into low volume streams properly.
Second, benzene is a stable ring compound requiring longer residence time
for combustion than straight-chain organics. Flares which are not optimally
designed do not provide extended residence times and thus are judged incap-
able of achieving high benzene destruction. Until better flare efficiency
data are available, a conservative estimate of flare efficiency is necessary.
4-5
-------
4.4 BOILERS
Vent streams can be ducted to the burner of a boiler. This control
system would use existing EB/S boilers retrofitted appropriately to accept
the vent streams.
The boiler control system can be retrofitted in two ways. In the first,
lower pressure vents first are compressed and then piped along with higher
pressure vents into the boiler fuel header and burned with the regular fuel.
In the second, one or several burners in the boiler are replaced by forced
draft burners capable of accepting low pressure streams. Streams from the
lower pressure vents then are piped to the replacement burners for destruc-
tion and from the higher pressure vents to the fuel header or replacement
burners. Forced draft burners applicable to this second method are avail-
able as standard items for fuel pressures down to one inch of water gauge. '
A chief advantage of boiler combustion over conventional incineration
is that mixing of the waste stream with fuel prior to introduction to the
combustion device results in the passage of that gas mixture directly through
the burner ports and along the entire flame zone. In this way, optimum
gas/air mixing in a controlled environment exposes the benzene to tempera-
tures of 3,000°F. Literature sources indicate that a marked shift occurs
789
both in the mechanism and speed of combustion under these conditions. ' '
Since combustion occurs in only fractions of a second and essentially is
complete by the time the gas exits the flame zone, conventional combustion
efficiency parameters, such as firebox temperature and residence time, are
not critical to this design.
Two other advantages of boilers are energy savings and ease of instal-
lation. Energy is saved by recovering the Btu content of the vent streams
4-6
-------
as steam or process heat. Installation is simple because EB/S plants already
require boilers as part of the process and, since the vent streams are com-
bustible, only minor retrofit is required for the boiler to accept these
streams.
The major technical problems with the boilers are the same as with
flares, i.e., manifolding.
4.5 MANIFOLDING
This section discusses the difficulties of flare and boiler manifold-
ing and describes the designs and equipment available to overcome them.
Two problems may be present in manifolding vent streams to boilers and flares:
(1) pressure differences in the vent streams, and (2) potential explosion
hazards.
In the first area, certain vent or process streams may be at higher
pressure than others; thus, if these streams are ducted together, backflow
in the vent system may occur. Such backflow can result in off-specification
product by introducing impurities in the vent streams into the process or
by upsetting process conditions. For example, during outages of the hydro-
gen offgas compressor, the temporary high pressure in the flare header may
place too high a back pressure on steam ejectors tied to the header and .
upset the vacuum columns.
In the second area, potential explosion hazards may result from ducting
vacuum sources to either a flare or boiler. If a leak occurs in the equip-
ment under vacuum, air could be drawn into the system venting a potentially
explosive mixture to a combustion source.
Designs and equipment to overcome both problems have been demonstrated
and are straightforward to apply. For backflow, two solutions are present:
4-7
-------
(1) segregating the low pressure sources, or (2) installing equipment to
increase the pressure. The first solution avoids pressure differentials by
routing high pressure and low pressure sources differently. At one plant,
a separate header for emergency releases is used to prevent any temporary
back pressure on other process vessels. At two other plants, a separate
header and flare are used specifically for the low pressure vacuum column
vents.11'12 The second solution is to install equipment to increase the
pressure of vent streams, either a compressor or an extra ejector stage.
Two plants have installed a compressor ' and two others are installing
or have installed ejectors driven by natural gas to increase vent stream
pressure. '
For explosion hazards, a number of designs are available. One design
uses oxygen analyzers which divert the vent streams to a flare or add inerts
17 18 19
if higher oxygen levels are detected. Three plants have such a system. ' '
Other designs include welded pipe to prevent air infiltration, orifices to
20 21
restrict flow from vacuum equipment in case of leaks, and flame arrestors. '
In addition, some minor process changes may be required to reduce the explosion
potential. For example, several plants use steel-covered concrete hotwells
for the collection of condensate from steam ejectors. Since these hotwells
are prone to air leaks, they would need to be replaced by steel drums or
other equipment when ducting the hotwell vents to a combustion source.
Alternatively, the steam ejectors could be replaced by a vacuum pump which
would do away with the need for a hotwell.
The ability to overcome explosion hazards is demonstrated further by
the number of plants that have tied vacuum sources to either a boiler or
4-8
-------
flare. Three plants have ducted vacuum columns to a flare22'23'24 and three
25 26 27
others to a boiler. ' '*- In addition, the basic design of the process
at several plants ties a vacuum condensate stripper into the hydrogen-rich
pQ
offgas used as fuel from styrene production. Finally, a number of plants
can operate their dehydrogenators under vacuum; the hydrogen offgas from
this step is used as a fuel.29'30'31
There should be no manifolding problems associated with an add-on absorp-
tion system as a control device.
4,6 ALTERNATE PROCESSES
One final control technique in reducing EB/S benzene emissions is to
produce ethylbenzene from mixed xylenes instead of benzene. If this process
were used more widely, benzene now extracted from reformate and purified
for alkylation to ethylbenzene would be sent directly to gasoline blending.
Mixed xylenes now blended into gasoline as part of reformate streams would
be extracted from reformate and processed for ethylbenzene.
Technological problems limit the application of this technique in both
existing and new plants. Existing EB/S units are of radically different
designs and capacities than those required for mixed xylene extraction. If
the substitute control technique were applied to existing plants, substi-
tuting units of mixed xylenes for units of benzene would be required. How-
ever, design and capacity problems limit this. First, most existing benzene,
ethylbenzene, and styrene processing equipment is of improper type, con-
figuration, and sequence for mixed xylenes. Mixed xylene extraction is a
distillation process, whereas benzene alkylation is a chemical reaction
process. The equipment would have to be changed and large parts of the
present units scrapped. In most cases, building new mixed xylene units
would be simpler. Second, in addition to design limitations, capacity
4-9
-------
problems exist. Major expansion of aromatics processing capacity would be
required to apply this technique to existing plants since the production of
one unit of ethyl benzene would require more mixed xylene than it would benzene.
The same capacity and design limitations do not apply to new plants; however,
for new plants, the special feedstocks and low-cost energy required for
producing ethylbenzene from mixed xylenes generally would not be available.
In addition to the design and capacity problems which exist with ethyl-
benzene extraction from mixed xylenes, benzene emission concerns also exist.
As noted in Section 3.4.2, ethylbenzene from mixed xylenes is not being
examined as a benzene emission source since the actual separation of ethyl-
benzene has no benzene emissions. However, as a result of switching ethyl-
benzene production to extraction from mixed xylenes, benzene emissions may
occur.
The analysis of these "global" emissions is complex. Switching pro-
cesses involves more than substituting one discrete process for another.
The switching involves many stages of aromatics processing; each one has
benzene emissions which are affected differently.
Switching ethylbenzene processes would likely result in offsetting
changes in benzene emissions. First, switching processes would remove large
quantities of benzene from benzene alkylation to ethylbenzene. Thus, the
process, storage, and fugitive emissions from this process would be elimi-
nated. However, switching processes also would remove large quantities of
high octane ethylbenzene from gasoline and the most likely replacement would
be the benzene previously alkylated to ethylbenzene. The resulting increase
in benzene emissions from gasoline distribution would offset the decrease
from the ethylbenzene process. Therefore, switching ethylbenzene production
to mixed xylenes is estimated to have no effect on benzene emissions.
4-10
-------
4.7 REFERENCES FOR CHAPTER 4
1. Memorandum from Seeman, W. R., Hydroscience, to White, R., Hydroscience
May 19, 1978. Flare Efficiency.
2. Memorandum from Seeman, W.R., Hydroscience, to Kalcevic, V., Hydro-
science. September 29, 1978. Flare Efficiency.
3. Letter from Bergman, H. , EPA Region 6, to Strader, W.C., Ethyl Corpora-
tion. August 30, 1977.
4. Straitz, J. Flaring for Gaseous Control in the Petroleum Industry.
National Air Oil. Philadelphia, Pennsylvania. (Presented at Air Pol-
lution Control Association. Pittsburgh. June 26-30, 1978.)
5. Telecon. Reed, R., North American Manufacturers, with Mascone, D.C
EPA. August 7, 1978.
6. Telecon. Miller, F., Bloom Engineering, with Mascone, D.C., EPA
October 23, 1978.
7. Rolke, R.W. et al. Afterburner System Study, U.S. EPA. S-14121.
Shell Development Company. 1971.
8. Barnes, R.H., M.J. Saxton, R.E. Barrett, and A. Levy. IERL Report
U.S. EPA. EPA-600/7-79-096. Battelle Columbus Laboratories
April 1979,
9. Lee, K., J.L. Hansen, and D.C. Macauley. Predictive Model of the Time
Temperature Requirements for Thermal Destruction of Dilute Organic
Vapors.
10. Telecon. FretweTl, S.W., Oxirane, with Mascone, D.C., EPA.
January 4, 1979.
11. Letter from Hughes, M.P., El Paso Products, to Miles, A., EEA.
November 5, 1979.
12. Key, J.A. Hydroscience. Trip Report for Dow Chemical U.S.A., Freeport
Texas. July 28-29, 1977.
13. Letter from Brennan, H.M., Amoco, to Mascone, D.C., EPA. November 30,
1978.
14. Mascone, D.C., EPA. Trip Report for Monsanto Chemicals, Texas City
Texas. June 27, 1979.
15. Letter from Bufkin, L.T., American Hoechst, to Walsh, R.T. EPA
November 20, 1978.
4-11
-------
16. Letter from'Crist, J.G., U.S. Steel, to Walsh, R., EPA. January 25,
1979. Response to Section 114 letter.
17. Reference 14.
18. Telecon. Mayfield, G., Tennessee Eastman, with Mascone, D.C., EPA.
October 20, 1978.
19. Mascone, D.C., EPA. Trip notes for Oxirane, Houston, Texas. October 18,
1979.
20. Reference 10.
21. Reference 15.
22. Reference 11.
23. Reference 12.
24. Reference 10.
25. Reference 16.
26. Reference 13.
27. Reference 14.
28. Key, J.A., Hydroscience. Trip Report for Cos-Mar Plant, Cosden Oil
and Chemical Company, Carville, Louisiana. July 28, 1977.
29. Telecon. Berry, F., Gulf Oil Corporation, with Mascone, D.C., EPA.
July 28, 1978.
30. Telecon. Bufkin, L.T., American Hoechst, with Mascone, D.C., EPA.
February 4, 1980.
31. Telecon. Kuykendall, C., El Paso Products, with Mascone, D.C. , EPA.
February 4, 1980.
4-12
-------
5. REGULATORY OPTIONS
This chapter discusses the regulatory options considered for con-
trolling continuous benzene emissions resulting from the ethyl benzene/
styrene (EB/S) production process and describes the model plant used in
assessing the impact of each option.
5.1 MODEL PLANTS
A model plant for the EB/S industry was designed to represent the
emissions and control device requirements of a typical EB/S plant. This
model plant was used, along with available data on existing plants, to
study the pi ant-by-plant and industry-wide impacts of the regulatory
options. Table 5-1 summarizes the characteristics of the model plant used
in this study. The criteria for designing the model plant for EB/S are
described below.
The preliminary criterion in selecting the EB/S model plant was to
consider the number of model plants needed to represent properly the EB/S
industry. Several model plants would be needed if the variations existing
in EB/S processes, plant size, emissions, etc., lead to widely different
impacts on different plants for each regulatory option.
Three major process variations exist: the ethylbenzene hydroperoxida-
tion process, the vapor phase alkylation reaction, and the hydroperoxida-
tion reactor. Though the ethylbenzene hydroperoxidation process differs
from the ethylbenzene dehydrogenation process, its main benzene emission
point, the vacuum column, is the same. Thus, the control methods and costs
for the ethylbenzene dehydrogenation process are applicable to hydroperoxidation.
Vapor phase alkylation can use dilute ethylene which can result in a much
5-1
-------
greater vent flow than the model plant. However, due to the high flow and
Btu content of the offgas from vapor phase alkylation using dilute ethylene,
plants currently using this process burn the stream; therefore, this variation
would not affect the regulatory analysis. Emissions from the hydroperoxidation
reactor differ widely from emissions from the ethylbenzene dehydrogenatiori
process. However, only one reactor is used domestically and its emissions,
controls, and control costs can be studied without using a model plant.
Therefore, rather than designing a second model plant solely for analyzing
the potentially different impacts from the one hydroperoxidation reactor
vent, a separate study of this emission point was completed and is presented
in the Addendum to this document. Because none of these variations would
affect the regulatory analysis, all impacts were studied by scaling or
modifying one model plant. Variations in plant size and levels of control
were scaled according to the model plant parameters. Variations in emissions
resulted in only small changes in cost and no change in applicable control
technology.
Because there are no technological reasons to consider different or
alternative models for new or existing plants, one model plant was con-
sidered representative of both.
The hydrogen separation stream typically is classified as an emission
stream. However, control method, cost, and impact analyses were not per-
formed for this stream since most plants currently burn it as fuel because
of its high Btu content. This variation has minimal impact on the regulatory
analysis.
The second criterion in selecting the model plant was to determine the
specific process, emissions, and other characteristics of the model plant.
This was done by conducting EPA field surveys, tests, and literature searches.
5-2
-------
TABLE 5-1. EB/S MODEL PLANT
1. Process/Capacity
The model plant is an integrated ethyl benzene/
styrene unit using the benzene alkylation and
ethyl benzene dehydrogenation processes outlined
in Section 3.3.1. The capacity is 345,000 Mg/yr
for ethyl benzene and 300,000 Mg/yr for styrene
based on 8,000 hrs/yr of operation.
2. Vent Streams
The vent points for the model plant are described
in Section 3.4.1 and the vent flows and compo-
sition are listed in Table 3-4.
3. Energy Waste-
water Load
The energy use for steam, process heat, elec-
tricity, etc. at the model plant is approximately
5,000 kJ/kg styrene. The wastewater load is 50
to 3,000 gallons per 1,000 Ibs and the TOC content
is 0.2 to 40 Ibs per 1,000 Ibs.
4. Baseline
Controls
The model plant assumes that the hydrogen separa-
tion vent is processed for aromatics recovery and
then burned in the styrene superheater. In the
same model plant, vents are uncontrolled.
However, in the pi ant-by-pi ant cost analysis,
the remaining vents are controlled to the degree
present at each existing plant.
5. Location/Size
The model plant is located in an integrated
chemical complex sharing utilities, storage, and
similar services with other units in the com-
plex. The dimensions of the EB/S unit and
entire complex are shown in Figure 6-1.
5-3
-------
The values of the model plant then were determined by interpreting and
averaging these data in light of basic engineering principles and EPA expe-
rience with the EB/S industry.
5.2 REGULATORY OPTIONS
The regulatory options considered for controlling EB/S benzene emis-
sions are based on application of benzene emission controls, or combination
of controls, discussed in detail in Chapter 4 and summarized, along with
respective control efficiencies, in Table 5-2. Some of these control
techniques, however, were not considered in developing the regulatory
options because of specific limitations in their use or application.
The Administrator has been reluctant to allow the use of flares as a
continuous process emission control device because emissions cannot be
measured practically to determine if the flares are in compliance. Fur-
thermore, due to their relatively low or variable efficiency, the Admin-
istrator does npt consider flares as a single, effective control technique
for continuous process benzene emissions.
Condensation alone cannot control all vent streams, such as short
duration, large volume streams or low volume, unsaturated streams. Absorp-
tion alone cannot handle high volume, intermittent releases of gases.
Because of these limitations, these control devices were rejected as single,
effective benzene emissions controls.
However, a combination of condensers (achieving 80 to 90 percent effi-
ciency) on the vents saturated with benzene and absorption towers (achieving
80 to 99 percent efficiency) on the vents not saturated with benzene can
provide an overall benzene control efficiency of 85 percent. In conjunction
with condensation and absorption, flaring can control the large volume,
intermittent releases for which condensation or absorption is not applicable,
5-4
-------
co
o
Q£
1—
-2£
O
0
LU
o
^— . •
Cti
s-
s:
co
CM
LO
LU
CQ
- c
Qi O
LU •<-
> -P
o re
c_> m
LU c:
C£. CU
•a
1— c:
o o
Z3 C_S
Q
o
o; /-*.
Q- i—
in
E
re
cu
•p
•p
cu
>
in
in
cu
u
o
i-
a.
^^
en
cn
o
00
•
in
E
re
cu
s-
•p
in
-P >>
c u
0) C
> Q)
•r~ .
in u
in «i—
CU 4-
O 4-
o cu
O-r—
o
CU J-
•p -p
3 C
i— O
•r- U
T3
S- Ol
O T-
4- .C
^^
-p •—
•r- re
3 -r-
in -P
C
-P CU
in -P
cu o
CQ a.
e
o
•r—
-p
a.
o
in
JQ
^
O
c
cu
•r~
u
4-
4-
CU
. C
o'
•p
u
3
S-
•p
in
cu
-a
• ^
03
£.
re
s-
0
"u
c •
re
3s
• o
— 1
O5
•z. c
O T-
H-l 5-
(— re
CO i—
Z3 U_
CQ
O s~~^
C_3 r—
^^
en
cn
in
in
cu
o
o
a.
^~
r—
re
j-,
0
4-
^^
U
C
cu
•r—
U
<£
4-
CU
"o
•P
C .
O in
0. E
re
-C CU
O) S-
••I- .p
3= in
c.
o
c
• r—
O
1 — 1
CM
^3
cn
en
m
in
cu
u
o
s-
a.
i—
P—
re
S-
o
4-
5>^
O
CU
(J
4-
4-
01
O
•P
C .
O in
U E
^= CU
O) S-
•r- H->
3: in
S-
d) ai
-P r-
in -i—
re o
3: ca
4- O
0 -P
Cn in
C E
•r- re
•P CU
3 S-
O -P
Q£ CO
CO
, re
^^
f-*i
- O
^•*^
in
cu in
c. cu
cu u
N O
C S-
CU Q.
r— C
>>.°
•P 4J
cu re
i —
4- i—
O -r-
-P
CU in
S- T-
3'O
u re
re ••—;
4-
3 in
c cu
re c
E cu
21?
•r-
3 -a
cr cu
^ *r—
E
T3
's o
O S-
3: 4-
i
r—
^j "O
x: cu
ZZ -P X
O LU T-
n s:
1— 4-
in c
cu re
4- C i —
o cu p.
cu >j in
s- -p cu
3 in E
-P 3
u 4- .in
re o in
<3C<
c o
re T- .
E +-> m
re cu
cu -P u
s- s- s-
•^- 0 3
3 a. o
cr E m
, C
•P t? 0
X CU T-
CU C -P
•i- re
O 4- C
•P CU
s- c
C •!—
o ~o
•r- 0) C
-P -P O
u re •!-
3 -r- -p
"O (J O
O O 3
s- in -a
O. in cu
re j-
a>
c: in o
cu re c
N
C -£= >,
CU O i—
J3 3 qj
r— in ^
>, re
-C " E
-P in •!—
cu cu -P
O r—
O) i- 3
c: 3
•r- O -C
-E to 4->
O T-
-P s- 3:
•r- \
cu c: re
S- O E
re in «*
in c .
in *r— o in
re E T- cu
CU -P O
S- CU J3 3
CU C •!- 0
> O) i- in
CU N +J
3: c in r—
O CU T- t—
re
5-5
-------
achieving an overall benzene destruction efficiency of 94 percent. These
combinations of equipment were considered in developing the regulatory
option.
Although EB/S plants could modify their processes and produce ethyl-
benzene from mixed xylenes instead of benzene, the Administrator does not
consider this process modification as a viable regulatory option. The
actual separation of ethyl benzene from mixed xylenes has no emissions.
However, as a result of switching ethylbenzene production to extraction
from mixed xylenes, benzene emissions from other sources, such as associated
refinery operations and gasoline distribution, may increase with ultimately
no reduction in nationwide benzene emissions from all sources.
A retrofitted boiler, process heater, or superheater can achieve a
high, but varying, degree of control depending on the inlet level of benzene.
Based on the range of inlet at existing EB/S plants, these control devices
can achieve approximately 99 percent destruction efficiency and therefore
were considered as single, effective control devices in developing the
regulatory options. A combination of condensers, absorption towers, and
boilers,* however, cannot achieve any greater reduction in emissions than
boilers alone because the condenser and absorption tower reduce the inlet
concentrations and, as a result, the boiler operates at a lower destruction
efficiency than if it were receiving concentrated, uncontrolled inlet.
This combination of equipment was not considered in developing the regula-
tory options because it is no more effective in reducing benzene emissions
than a single boiler.
*Boilers also include process heaters, superheaters, and other similar
units.
5-6
-------
Based on these considerations, the Administrator therefore is con-
sidering five regulatory options for controlling benzene emissions from
EB/S plants:
• Do not set a standard.
• Regulatory Option A: Require that EB/S plants achieve 85 percent
benzene emissions reduction in the main process vents. This alterna-
tive is based on condensers for some main process vents and absorbers
for others.
• Regulatory Option B: Require that EB/S plants achieve a 94 percent
benzene emissions reduction in the main process vents. This alterna-
tive is based on using condensers for some main process vents and
absorbers for others followed by a flare. ' ' .
9 Regulatory Option C: Require that EB/S plants achieve 99 percent
benzene emissions reduction in the main process vents. This alterna-
tive is based on the use of a boiler for controlling the main process
vents.
« Regulatory Option D: Require that EB/S plants achieve 100 percent
reduction in benzene emissions. This alternative is based on the use
of substitutes for styrene and assumes that all EB/S plants will be
forced to close.
Under all options, all continuous releases from the hydrogen separation
vent would be required to achieve 99 percent control. We believe this to
be possible based on the current industry practice of sending the hydrogen-
rich gas stream to the steam superheater.
5-7
-------
-------
6. ENVIRONMENTAL AND ENERGY IMPACTS
This chapter discusses the environmental and energy impacts of each
regulatory option presented in Chapter 5. Impacts are discussed with regard
to air quality, water quality, solid waste, and energy requirements. Both
beneficial and adverse effects are presented, the major emphasis being on
the incremental impact of the regulatory options.
Five regulatory options are considered in the ethylbenzene/styrene
(EB/S) analysis, including not setting a standard. Four of these options
involve limiting benzene emissions from the process vents, consisting of
the alkylation reaction vents, the atmospheric pressure column vents, and
the vacuum column vents. Process vent emissions also include the continu-
ous release of the hydrogen separation section.
With regard to these source categories, five basic regulatory options
were considered and three of these were analyzed:
• Do not set a Standard -- This option assumes that current levels of
benzene emissions from process vents are controlled adequately to
protect public health. EB/S benzene emissions are controlled to
some extent via State Implementation Plans (SIP) hydrocarbon emission
regulations.
8 Regulatory Option A (85 percent control) — A benzene emission control
efficiency of 85 percent is achieved by routing process streams to a
scrubber or condenser system. The scrubber is sized to control con-
tinuous venting from the alkylation reactor and all vacuum columns in
the EB/S facility with the exception of the benzene/toluene columns.
The condenser system controls releases from the benzene/toluene hotwell
vents and from all columns operating at atmospheric pressure and above.
The continuous release of the hydrogen separation vent is routed to a
boiler* after primary product recovery (60 percent benzene recovery)
to achieve an overall benzene emission control of 99 percent.
• Regulatory Option B (94 percent control) — A benzene emission control
efficiency of 94 percent is achieved by coupling a flare system to the
Boilers include process heaters, superheaters, and other similar units.
6-1
-------
scrubber/condenser system of Regulatory Option A. All process streams
are first handled as in Regulatory Option A; the vent streams from the
scrubber and condenser subsequently are manifolded into a flare system
sized to handle the continuous vent releases. Emissions from the
hydrogen separation section also are treated as in Regulatory Option A.
• Regulatory Option C (99 percent control) — Under this regulatory option,
all process vent streams are manifolded together and subsequently routed
to an existing boiler to achieve a benzene control efficiency of 99
percent. Emissions from the separation section are treated as in
Regulatory Option A.
• Regulatory Option D (100 percent control) -- This regulatory option
entails 100 percent benzene emission control via plant closure and
substitutes for styrene.
Each level of control was analyzed with respect to a model plant of
300,000 Mg/yr styrene capacity; where applicable, pi ant-by-pi ant impacts at
existing facilities were analyzed and presented. (The fifth regulatory
option of 100 percent control involves closure of the industry and therefore
is discussed on an industry-wide rather than a model plant basis. The
discussion of this option is presented in Chapter 7.)
6.1 AIR IMPACTS
Air quality impacts were analyzed through the use of the Industrial
Source Complex (ISC) dispersion model. ISC predicts the ambient concen-
trations that would result from air pollutant sources based on meteorologi-
cal data and the characteristics of the emitting sources (such as emission
rate, stack height, and stack gas temperature). The three regulatory options
(A, B, and C) and one uncontrolled scenario were examined. Appendix B contains
a discussion of the overall modeling methodology.
A 300,000 Mg/yr styrene production capacity model plant was used as
the basis for this analysis. Figure 6-1 shows the layout of the model plant
and the location of all sources, including fugitive emissions from the EB/S
«K>
unit, three storage tanks, and the hydrogen separation vent. Figure 6-2
The hydrogen separation vent emissions are discussed in the Addendum on
Excess Emissions.
6-2
-------
FIGURE 6-1
MODEL PLANT LAYOUT
260
T
20
40
20
90
60
100
40
Small
Benzene
Tank
H2 Vent-Uncontrolled
EB/S Unit*
0 0
Benzene/
+ ) Toluene
Tank
Large
+ 1 Benzene
Tank
500
+ H2 Section-Controlled (Flare)
KEY:
1 cm = 20 Meters, All Numbers are Distances in Meters.
® = Center Point for Dispersion Model Receptor Grid.
* For Fugitive Emissions the EB/S Unit was Modeled as Two Square Area Sources.
6-3
-------
FIGURE 6-2
LOCATION OF PROCESS EMISSIONS FOR THE FOUR REGULATORY OPTIONS*
Uncontrolled:
'
Dm
r
L
1m
-t- +
Atmospheric
Pressure Columns
-f- +
Alkylation
Reaction
Boiler Stack
_j.
+
.f.
Vacuum Columns
Condenser/
Absorber:
Condenser/
Absorber/Flare:
Boiler:
-t-
Boiler Stack
Boiler Stack
Boiler Stack
Process Flare
* Large Rectangle Represents the EB-'S Unit in Figure 6-1
6-4
-------
shows the specific locations of the process emission sources for each of
the four regulatory options.
6.1.1 Modeling Results
The results of the dispersion analysis are summarized in Tables 6-1
*
and 6-2. Table 6-1 shows the maximum ambient benzene concentration for
each regulatory option and for each of the averaging periods considered.
These concentrations include uncontrolled fugitive and current control
levels for storage emissions. The table also shows the contribution of the
process emissions to the total plant annual average impacts. Table 6-2
shows the impacts of the individual sources or groups of sources that com-
3%)k
prise the model plant.
It should be noted that the "uncontrolled" model plant is included in
this analysis solely as a basis for comparison and does not reflect current
levels of control practiced by the industry. Actual emission levels at
existing facilities average about 70 percent less than those assumed for
the uncontrolled model plant (based on an overall 70 percent benzene control
efficiency). In this respect, current control levels fall somewhere between
the uncontrolled case and Regulatory Option A.
Table 6-1 shows high benzene concentrations from the uncontrolled plant,
due mostly to process vent emissions. Impacts are considerably lower under
Regulatory Option A (condenser/absorber), with process vent emissions
accounting for approximately half of total plant impact. For Regulatory
Options B (condenser/absorber/flare) and C (boiler), process vent emissions
*Table B-3 in Appendix B gives more detailed modeling results and shows
concentrations at various distances from the plant.
**Ambient concentrations in all tables may be converted to ppb by the fol-
lowing relationship: 3.19 yg/m = 1 ppb.
6-5
-------
no longer exert the predominant impact. Instead, storage and fugitive
emissions account for virtually all of the impact recorded for these latter
two regulatory options.
Combined impacts from all sources under each regulatory option are
located near the assumed boundaries of the plant property (160 to 300 meters
(m) from the source). For individual emission sources or source groups,
the highest estimated maximum annual average concentrations, including emissions,
usually occurred at 160 m from the model plant, except in the case of
Regulatory Options B and C. Maximum concentrations for process emission
points occurred at 500 m and 1,000 m, respectively, from the source. The
greater distance is due to the greater buoyancy of the hot plume associated
with the flare and boiler source compared to the source characteristics of
the other regulatory options.
Table 6-3 depicts estimated maximum annual average benzene concentra-
tions associated with individual existing plants. These estimates are based
on the emission rates and pollution control relationships between existing
plants and the model plant under each regulatory option. These estimates
are useful for comparative purposes only and should not be taken to represent
actual conditions since aspects such as individual source characteristics
and source locations are not considered.
6.1.2 Effects of Benzene Controls on Nationwide Emissions
Utilization of the condenser/absorption tower system (Regulatory
Option A) at all existing facilities would reduce nationwide benzene emis-
*
sions by 1,365 Mg/yr. The integrated condenser or absorption tower/flare
^Estimated total benzene emissions from process vents alone in the EB/S
industry are approximately 1,990 Mg/yr. Current levels of control and
100 percent capacity are.factored in this estimate.
6-6
-------
1—
3
Q.
1
LU
O
O
^"•"1
•y
-^-
o
Q;
u_
z.
o
I — 1
H-
*5
O£
r—
2;
LU
CJ
z
o
o
LU
LU
M
2
LU
CO
_.
=;
s:
i— i
X
.
i
LU
~«1
CO
«c
p~"
E
in. 3
in E
O) -i—
U X
o ro
t- s:
0. C
-a o
Ol 0) •!-
-C C .p /-%
•P *r~ ftfO
XJ S- E
M- E -P \
o o c a
c_> o) a.
C US-*
0 O> C
•r- -C O
-P -P C_)
-Q O
•r- .p
s-
•P 10
C -P
o c
U 01
""*
r— Ol
ro O)
3 ro
c s-
C 01
c
o
•r—
•P
ro
S-
c
0)
U
c
0
o
3 E
•r- CD
X 3.
re v_/
s:
o»
c
•r-
E
O
0
ro D>
3 ro
C S-
C 01
«c >
CD CD 00
CD «^- «j-
CO VD to
O «tf- ^J-
"-^ 00 00
CO i-^ r-T
^ CO CM
*^^ ^*^i /""N
Q1^ ^? ^^
00 CD CD
^ ^^^^ ^"^"^
O
O O>
in i —
ai -P
> c
•r- Ol
-P S-
*05 3
3 U
I4t-.
•P
£ C
0 0)
s- in
if- 0)
t.
in o.
c 01
O t-
•r-
in jc
in o
.
/— ^
Ol
OS
ro
S-
ai
ro
j_
13
O
^r
ai
c
o
ft
,
o>
•r—
0)
ai
01 ai
"O OJ
3 ro
r— S-
U 0
C -P
•i— in
in E
C 0
0 S-
•i- q-
•p
ro m
s- c
•P 0
C. •!-
Sin
in
C T-
0 E
U 01
i — "O
r— C.
-------
V)
111
o
r— CO
re a
3 re
c co
co
00
co
CO
ir>
o r>» oo
CM CO
Q>
U
X
0)
(U
O O
« •
o oo
in r—
co
CM
i— O O
• • •
o <— oo
CM r--
CM
0)
re
•P
cn
CM CO
o
i— CO
co
i— in
VD r-
r— CT»
O CM
CO
uu
LU
•i-re
IA -P IA
•i- re-
E OS
LU
CM
co
CM r-
• *
i— O
in
o
CO
Ul
o:
o.
CM
«£>
LU
CO
U
CO
U
(A
c
O>
U) S-
•i- (0
in S-
ui 0) in co
o o u O
01—
(A JO
O o o o
CO
co s e e
I— .£= 3 3 3
re -p s £ s
H-i -r- -r- «r-
O S-
re
re re
3 u
6-8
-------
03
-^
1
Q_
co
""^x
CQ
LU
O
I 1
LL.
r 1
o
1 i 1
1 !|
Q.
co
o
ij-
.
Cn
3.
C
•r—
in
O
•P
ro
S-
E
CU
u
o
o
CU
D)
S-
>
>^^
5— cn
o cn
•P ^— ^
re
p~-
3
cn
cu
Cc
CO
E
O
*r""
-p
CX
o
2>>s^
O %•
re >— '
3
O3
CD
QC
, "^^
•P OJt/}1
U CQ
nro LU
CXO >— •
re i —
o
-p
E
re
51
r— CO CM «3"
CD r*** r— ^*
P~~
i — LO CM «*•
CD r^* P— *sj"
p—
oo r*» i — to
LO r^*» CM r**»
LO O f^- «3-
o cn LO r-
CO CM
o o o
LO 00 O XI
•* CO CM O
\ ^. \ "^
CM >^- CD CM
LO >^- CM
_p
in
XT
O
O)
o in
2Z E re
E X
E fl) CU
re a. i—
CJ
S- O O O
cu o u u
E e s- s-
-P in
CU E 3
i- re o
CU CU -P T-
5 x: cu 3 s-
+* m r— re cu
tn in o i — > x:
E i- E .E O -P
O O O -P Q. CU
•r- -P x: cn
re re cu -P E o ui
S_ lf~. t/) Q) ^) g ^_>
+-> re m T- E re o
E cnx2 1 — x) o re
cu E cu E E in ex
E r— T3 CU U •!-.
o re cu r— +J ^-^ E
u u > u -P cu -P
in -r- r— S '. T- 0
•p- re /•> •!— cu -p re
in i — in -P in v-' re
in o_ c re s- P- cn
•r- O P- O in (D E
E P— T- T- E E E x: Q.
cu -p- E t- W -o
o E cn cu E O--P in
s- T- re T- ro
Q. S- -P S- ^ >>
S_ O T- CU U i — "O
WO-PI — "oroEcu
ro u_ in «r- T- +j o s-
u in m cu
p~* ^3 re E p~ T3
p- . E 4- O "O 3 •!-
cu in ro u E 4— tn
S -P cu re i —
i — E re s- ro 3
E -a E cu o -P x:
rooos-cninEin
E -P E CU
cu -a in •<- E u -p
> CU E r— .p- E 3
•p- xz ro in re o <"**
+J -P in u in u
•p- tn cu in cu
O5 E E U U CU in
•a in Q_I— s_ x: re
E CU in cu -P r—
o in *r- 4- 4- Q_
s- ro E o .4- " -
a. E -5 s- 3
in •> in «r- re o o
UP- cumcxrecus-
ro s- o E s- re
o re o o r— r~ cu z>
E cu i- T- cu ox:
•p- E CX-P -a 3i—4-
•r- re o in o
in p— S- p— E
a) o cu •> . m
•OT34- S- CU-P in-P
3 cu x: o -P o
r— p-cucu-preEre
o re E N o re ex
EOO-p-4-Ep-E
i— i in >— ' in o •!— p ••—
re j
-p0
E
re
cu
cu
s_
4->
in
£2
O
ex
cu
i-
r"
A
a_
*i
re
+i
3
0
o
S-
<:
o
-p
-^
cu
s_
cu
in
E .
•P S-
3
E O
re ro
4-
X 3
I — E
fft
*u
O
o cu
S- S.
,
•P -P
LU tn
D
6-9
-------
system (Regulatory Option B) would reduce nationwide benzene emissions by
1,790 Mg/yr. Finally, a special burner and controls (Regulatory Option C),
retrofitted in an existing boiler, will reduce benzene emissions by 1,920 Mg/yr
for the model plant. In this latter option, neither benzene nor other
organics are recycled back to the process; however, energy savings are
realized by recovering the thermal energy content of the vent streams in
the form of steam.
6.2 WATER QUALITY IMPACT AND CONSUMPTION
None of the regulatory options considered results in any significant
increase in wastewater discharge by EB/S facilities. No water effluents
are discharged from either flares or boilers themselves (Regulatory Options B
and C, respectively). A small amount of water is collected from the vapor
streams running through the scrubbers and condensers (Regulatory Options A
and B). Most of this condensed water can be recycled back to the process,
increasing total plant wastewater flows by less than 50 gallons/hour, an
amount which would not affect current plant waste treatment capacity.
The net water consumption for the boiler and condenser/absorption sys-
tems is zero. The condenser/absorption tower utilizes polyethylbenzene as
an absorbent medium which is chilled via a heat exchanger coupled to the
plant's brine refrigeration system.
The flares use water in the form of steam which is injected into the
combustion zone of the stack. Water consumption for the 300,000 Mg/yr model
plant incorporating a three-inch flare is approximately 380 m /yr. Water use
for the water seal in the base of the flare stack is negligible.
6.3 SOLID WASTE DISPOSAL IMPACT
None of the regulatory options being considered generates solid waste.
6-10
-------
6.4 ENERGY IMPACT
Table 6-4 presents the total energy requirements for each control
strategy as applied to a 300,000 Mg/yr model plant. For Regulatory
Options A and B, energy requirements amount to less than 0.1 percent each
of total plant energy requirements; conversely, Regulatory Option C results
in an energy credit of less than 0.1 percent of total plant consumption.
All energy impacts are essentially linear with respect to plant capacity.
6.4.1 (Option A) 85 Percent Benzene Emission Control
In the condenser and absorber emission control (or product recovery)
system, electrical energy is required due to a slight demand on the brine
system's cooling load. Electricity is required to operate pumps and com-
pressors. Pumps are required to circulate the refrigerant and absorbent;
compressors are required for vacuum vents and other waste streams under
negative or extremely low pressure. Total electrical requirements for pumps
and compressors are on the order of 60 x 103 MJ/yr for the model plant.
Energy demand on the cooling system for condenser operation is approximately
750 x 103 MJ/yr for the model plant.
6-4.2 (Option B) 94 Percent Benzene Emission Control
Flare operation consumes both steam (for smoke control) and natural
gas (for pilot and purge requirements). A flare system placed downstream
of a condenser/absorption tower system on a 300,000 Mg/yr model plant would
consist of a three-inch flare for process vent emissions. Steam consumption
for the three-inch flare is approximately 9.4 x 105 MJ/yr. Natural gas con-
sumption for the flare is approximately 1.2 x 106 MJ/yr (natural gas is
necessary for both pilot operation and purging the flare of organics). The
total energy demand of this regulatory option, including the condenser/
absorber system, would be approximately 2.9 x 106 MJ/yr for the model
6-11
-------
plant. (ElectricaFenergy needs for the oxygen monitoring equipment is
negligible.)
6.4.3 (Option C) 99 Percent Benzene Emission Control
In the event a compressor is needed to direct low-pressure vent streams
to the burner assembly, electrical energy consumption would be approximately
30 x 103 MJ/yr. Electrical energy used for the oxygen monitoring equipment
would be negligible. Assuming a conservative 75 percent heat recovery in
the form of steam from combustion of the waste gas streams, an energy credit
of approximately 53 x 10 MJ/yr could be attained.
6.4.4 Summary of Energy Impact
Regulatory Option C, at 99 percent benzene control, provides an energy
credit of 53 x 106 Mg/yr. This is equivalent to 8,500 barrels of oil. The
energy credit, in the form of steam, is derived by assuming 75 percent heat
recovery of the combusted waste streams.
The condenser/absorption tower option (A) and the condenser/absorption
tower/flare option (B), at 85 and 94 percent benzene control, respectively,
3
provide substantial product recovery capability. Each recovers 613 x 10 Kg
of benzene per year (300,000 Mg/yr plant) which is recycled continuously to
the alkylation process.
It is difficult to present an accurate estimate of possible national
energy requirements due to a standard based on one of the regulatory options.
The actual total national energy impact under each regulatory option depends
upon both the control technique chosen and the existing level of control
for a particular plant. Insufficient data exist to determine the current
energy demand of "the control equipment present on existing facilities.
However, a "worst-case" situation, i.e., completely uncontrolled plants,
can be used to estimate the national energy impact for each level of control,
Therefore, assuming no energy credits for controls already in place, the
6-12
-------
1—
s- s-
•i— x^ x^
O 0 r—
01 Jg- O
-Q
c
o
p_
o
/••"•s 4-5 .p
CQ c c:
CU 0
coo
0 i.
•i— CU CU
•P a. c
O- CU
O "3- N
cu
CQ
^3
QQ
S- CU
OS- .— «,
m re s- s-
-a •- >, >,
C 0
CU H-
•a
c
o
o
_a
o
S-
N
X-/CO C
cu
CQ
s-
cu
S- O
CU 1— /-N
in s- s-
c c >, >,
cu o •\\
T3 -r- --3 i—
C 4-> S _Q
0 Q. ^3
O S- > — '
O
f^
<:
in
c
cu
CU -P
in S- c
in -i- re
cu 3 r- re
o O~Q_ <— •«.
0 CU S- S-
cf* "cu,-a o i—
i — cn o s; .a
re &- s: ja
•P CU ^^
O C <4-
1— LU O
|Q
O ^"^
i — O
o
X ir>
CO 00
LO x^1
1
CO
CD <*^*»
r— O
X ^
o
cn
C\J
co
O /"^
co
X r—
o
00
f*^
^o
cn o
O r—
j
X
x
cn "~I
VD •*-*
,
f~—
re
O)
"^"X
^
4-^
CQ
^^
O
O
w»
O
^t"
r— ,
O)
c
•r—
£
13
(/>
U)
n3
*l
-p
C
CU
r— —
re
^>
•r"
3
o-
S-
&_
re
CQ
re
tl
^"
x^
cn
*NX*
0
r>».
^i
r—
cu
-p
re
E
•n-
X
O
i-
a.
o.
re
M-
o
•P
•r~
r*"
•i^
^"}
fO
a.
re
o
*^
S-
01
o
u
cu
S-
cu
<~
cu
N
c
cu
n
re
in
re
r-
i-
ai
3:
0
•p
c
o
• I—
-p
Q.
s-
o
in
o
re
*x^
S-
01
in
c
cu
T3
C
O
o
cu
f~
f—
JD
CU
D5 O -P
CU -P S-
C C O
CU CU 4—
E
C O) r—
re s- T-
•r- O
in 3
CU CTH—
0 CU 0
3 S-
T3 in
S- D5 CU
Q- S- S-
CU S-
« c re
S- Ol -Q
CU
r— C l£>
•i- re co
O LO
c oo"
cn re
c s-
'•p >> °
in s- s-
•r- CU >,
X > \
CU 0 O
o s
c cu
re s*o
o
C -P r—
•r- re
cu x
i. -C
a> co
c -p tn
s- c
3 CU O
.a o -P
S-
CU CU -P
s- a. c
3 3
in LO o
in jx, E
cu re
s- ra
a. c in
t T- -P
3t £-•-
O 3 T3
i — t/5 CU
in s_
re >
•r- . CD
> e s-
re cu
- 0) C 4-5
in 4-5 cu c
re in re
D5 i— i—
M- re Q.
CU O 3
4J C r—
in E c cu
re s- re T3
So o
t- - E
CU J-
^= cu >> s-
-p -c<; >,
-P r) 00
1/5 4-> i — O
3-r- O
E CU o"
0 S- 0 0
O O CO CO
o
6-13
-------
total national energy requirements would be 11.0 x 10 MJ/yr or 1,800 bbl/yr
of energy under Regulatory Option A; 39.6 x 10 MJ/yr or 6,000 bbl/yr under
Regulatory Option B; and a savings of 723.0 x 106 MJ/yr or 117,000 bbl/yr
under Regulatory Option C.
6.5 OTHER ENVIRONMENTAL CONCERNS
6.5.1 Flare Noise
The flaring process and the noise it generates may have an impact on
the surrounding environment. Estimated sound pressure levels decibels (dB)
of the three-inch flare designed to burn continuous process emissions is
*
approximately 79 dB at the base of the flare stack. This estimate assumes
no steam injection. While the introduction of steam into the combustion
zone of the stack increases its destruction efficiency, it may increase
overall noise levels of the flare. However, it is believed that the flare.
2
can be designed properly to minimize noise. Table 6-5 compares flare noise
levels with the Occupational Health and Safety Administration (OSHA) standards'
and EPA recommendations for noise levels requisite to protect public health
A
and welfare in the industrial environment. Sound pressure level is reduced
as a function of distance. At 47 meters, the sound pressure level for the
three-inch diameter flare falls to 70 dB.**
6.5.2 Flare Thermal Radiation
Flares also emit thermal radiation. Stack height must be sufficient
to ensure the safety of personnel working in the vicinity of the flare.
Calculations of thermal radiation intensity indicate that the three-inch
diameter flare employed for burning continuous process emissions emits 750
^Assumes noise originates at center of flame.
**EPA recommendation for maximum dB level to protect against hearing loss
(24-hour exposure).
6-14
-------
LU
z
1 — 1
_J
LU
a
t— )
T^
C3
1— I
1—
•^
1
LU
Q£
— 1
LU
>
LU
— 1
LU
to
I— I
O
z
LU
a:
CD
—I
CD
in
•1™
O
z
cl)
>.
•r-
•P
i-
re
o.
E
0
CJ
in
•r- S-
e. 3
S- 6
a> 3:
O- 1
oo
«^ ^^
T
to
o
o
CTt
re
CD
s-
•O 05 (!)
C C _J
HH O
o i-
S- 3
o w o
M— w nr
O 1
CD —J •*
C Csl
•r- O5
i— C J-
a> «r- o .
TJ S- M-
•r- re
3 CD
v»^
in
•r—
O
z
CD
S-
re
^— •
Ll_
•a
O)
re
m
'-p
10
LU
LU
M
I™H
CO =
CO
LU
a:
— i
u_
o
CT>
0)
•p
(0
3
cr
ai
•a
to
c
IB
-P
•I—
ID
re
a>
to
re
a)
.e
u
3
a.
ai
-p
u
c .
•S^
••- Q»
10 re
i— in
CD
a> o
c
a>
in
O
Q.
ai
g
QQ
•o
CD
CD
C O)
3 S-
o re
re^£
•i- T3
3
O
6-15
-------
Kcal/hr sq m,* which," when added to the maximum ambient solar radiation,
amounts to approximately 1,550 Kcal/hr sq m.** The acceptable heat intensity
for short-term exposures is 4,000 Kcal/hr sq m (this will harm personnel if
skin is exposed for more than a few seconds).
6.6 IRREVERSIBLE AND IRRETRIEVABLE COMMITMENT OF RESOURCES
Regulatory Options A (85 percent control) and B (94 percent control)
recover a significant quantity of benzene. Benzene is considered a valuable
raw material and is used in many industrial applications. Although Regu-
latory Option C (99 percent) does not provide recovered benzene as do the
other regulatory options, an energy credit from the combusted waste stream
may be realized. Both the benzene credit from Regulatory Options A and B
and the energy credit from Regulatory Option C can be expressed in terms of
the barrels of oil required to produce the benzene or energy. The benzene
recovered in Regulatory Options A and B, assuming no current controls, would
require 80,000 bbl/yr of oil to produce and the energy recovered under
Regulatory Option C would require 117,000 bbl/yr of oil. These figures are
small in comparison to the total nationwide oil consumption for all uses.
In this respect, the energy credit for each option would be considered
approximately the same as would their savings of non-renewable resources.
6.7 ENVIRONMENTAL IMPACT OF DELAYED STANDARDS
In the event that the standard is delayed, significant negative impacts
on ambient air quality could result. Projected benzene emissions from plants
*As measured on the ground adjacent to the flare stack -- assuming heat
source originates from center of flame.
**
The maximum ambient solar radiation ranges up to 800 Kcal/hr sq m at
30 degrees north latitude.
6-16
-------
at current control levels should remain at least, an order of magnitude greater
than those from plants equipped with the controls under Regulatory Option C.
EB/S plants now are operating at less than full capacity but are expected
to reach full capacity within five years. Based on current levels of con-
tinuous process emissions in relation to those under Regulatory Option C
(1,990 Mg/yr and 70 Mg/yr, respectively), an annual growth rate of six
percent should increase emission levels linearly.
As stated in Sections 6.2 and 6.3, there are no significant water quality
or solid waste impacts (positive or negative) associated with the control
equipment. Therefore, there are no anticipated impacts in these areas in
the event of a delay in the standard.
Although Regulatory Options A and B require energy for their
operation and Regulatory Option C results in a small energy savings, each
accounts for less than 0.1 percent of total plant energy requirements.
Energy impacts, therefore, would be considered negligible in the event of
a delay in proposal or promulgation of the standard.
6-17
-------
6.8 REFERENCES FOR CHAPTER 6
1.
2.
3.
4.
5.
6.
7.
J. F. Bowers, et a!., H. E. Cramer, Inc. Draft Industrial Source
Complex (ISC) Dispersion Model User's Guide -- Volumes I and II. Pre-
pared for EPA, Source Receptor Analysis Branch. Research Triangle
Park, North Carolina. January 1979.
Straitz, John F. III. Solving Flare-Noise Problems. National Air Oil
Burner Co., Inc. Philadelphia, PA. p. 2.
Bureau of National Affairs. Noise Regulation Reporter. OSHA Stan-
dardized Procedure for Noise Measurements Chapter VI, p. 41-3242.
October 10, 1977.
U.S. EPA, Office of Noise Abatement and Control. Information on Levels
of Environmental Noise Requisite to Protect Public Health and Welfare
with an Adequate Margin of Safety. March 1974. p. 29.
Swithenbank, J. Ecological Aspects of Combustion Devices With Refer-
ence to Hydrocarbon Flaring. Department of Chemical Engineering and
Fuel Technology, University of Sheffield. Sheffield, England. May
1972. p. 553.
Flaregas, Inc.
Issue 2.
Ground Pollution From Elevated Flare Effluents.
Lowenheim, F.A. and Moran, M.K. Industrial Chemicals, 4th Edition.
John Wiley and Sons. New York, 1975.
6-18
-------
7. ECONOMIC IMPACT
This chapter discusses the potential economic impacts of each regulatory
option presented in Chapter 5. Impacts on both individual firms and the
general economy are considered.
7.1 INDUSTRY ECONOMIC PROFILE
7.1.1 Introduction
Styrene is produced by 11 companies at 12 locations in the United
States (Figure 3-2). Ten of these 12 plants are located in Texas and Loui-
siana. The remaining two plants are the Atlantic Richfield facility at
Kobuta, Pennsylvania, and the Dow Chemical facility at Midland, Michigan.
Approximately 98 percent of the styrene produced in these two facilities is
captively consumed. Consequently, virtually all shipments of styrene origi-
nate in the same basic geographic area.
Ethylbenzene (EB) is produced by 14 companies at 15 locations (Figure 3-1).
Only one of these plants is located outside Texas or Louisiana: Commonwealth
Oil in Penuelas, Puerto Rico. The 11 companies that produce styrene also
produce EB. These companies account for more than 97 percent of the U.S. EB
capacity.
For the purposes of this analysis, ethylbenzene/styrene (EB/S) is
considered a single industry. It has been estimated that 99 percent of the
EB produced in the U.S. is used in domestic styrene manufacture. (EB also
can be used as a solvent.) In fact, in many integrated EB/S operations, EB
o
is not separated as such. Most EB (90 percent or more) is not sold on the
market but is consumed by the producing company.
7-1
-------
7.1.2 Product
7.1.2.1 Production
Styrene is a fragrant liquid unsaturated hydrocarbon that is produced
from EB. It may be produced as a prime product or as a coproduct with
propylene oxide. EB may be produced as a prime product or recovered from
mixed xylene streams in refineries. Each of these processes will be dis-
cussed separately.
(a) Ethyl benzene: benzene alkylation. The predominant production
process for EB is the alkylation of benzene with ethylene. In 1978,
97.8 percent of total EB capacity was based on benzene alkylation. From 1965
to 1975, the percentage of EB produced through alkylation increased from
2 3
88.6 percent to 97.4 percent. ' It is anticipated that this percentage will be
maintained or even increased in the future (see mixed xylene extraction,
below). As shown in Table 7-1, 11 of the 14 companies that produce ethyl-
benzene use this process.
(b) Ethylbenzene: mixed xylene extraction. Mixed xylene extraction
is a refinery recovery process requiring feedstocks that yield xylenes with
high EB content. It is economically competitive with benzene alkylation
only when large quantities of fuel are available at low cost. Capacities
based on xylenes are small compared to capacities based on benzene alkylation.
Firms that distill EB from mixed xylenes and do not produce styrene are not
included in this analysis.
Between 1965 and 1975, the percentage of total EB production based on
xylenes declined from 11.4 percent to 2.6 percent. After increasing in
response to demand pressure in the late 1960s, the absolute quantity produced
from xylenes also declined.
7-2
-------
TABLE 7-1. U.S. ETHYLBENZENE/STYRENE CAPACITY BY PRODUCER'
Ethyl benzene
Company
Dow Chemical
Monsanto
Cos-Mar, Inc.
Oxi rane
American Hoechst
Amoco
Gulf
Atlantic Richfield Co.
El Paso Products
Commonwealth Oila
United States Steel
Sun Oil Co. , Inc.
Charter Co.a
Tenneco9
TOTAL
Capacity
(103 Mg/yr)
847
794
689
545
526
447
313
227
125
73
61
61
18
16
4,742
Cumulative
% of total
17.9
34.6
49.1
60.6
71.7
81.1
87.7
92.5
95.2
96.7
98.0
99.3
99.7
100.0
100.0
Styrene
Capacity
(103 Mg/yr)
861
680
590
454
450
380
272
200
115
54
36
4,092
Cumulative
% of total
21.0
37.7
52.1
63.2
74.2
83.5
90.1
95.0
97.8
99.1
100.0
100.0
These companies distill ethyl benzene from mixed xylene streams for merchant
sale. They do not produce styrene and will not be regulated under the EB/S
standard. EPA plans to regulate benzene emissions from these operations
with a petroleum refinery standard.
Although overall production in the EB/S industry is strongly sensitive
to fluctuations in aggregate demand, the choice of production process is
determined predominantly by supply considerations. As in the recent past,
future production from xylenes may be limited by increasing production costs
and limited availability of feedstocks. In the past, large amounts of cheap
natural gas have been available to most producers. This will not continue
7-3
-------
to be true. Feedstocks that provide xylenes high in EB content may not
always be available in the future. This combination of circumstances renders
mixed xylene extraction relatively unattractive on economic grounds and is
not expected to provide significant quantities of EB in the future.
(c) Styrene: ethyl benzene dehydrogenation. Over 88 percent of U.S.
styrene capacity is based on catalytic dehydrogenation of EB. This process
has been improved over time and production efficiencies vary with the age of
A
the producing unit. Prior to 1977, all U.S. styrene capacity was based on
catalytic dehydrogenation. It is expected to continue to be the predominant
process in the near future. Only one alternate process, propylene oxide
coproduct, has had commercial application. Although it is competitive with
dehydrogenation, it is not expected to replace dehydrogenation because it is
a coproduct process.
(d) Styrene: propylene oxide coproduct. EB can be oxidized to its
hyperoxide and reacted with propylene to produce propylene oxide and methyl
phenyl carbinol, which is dehydrated to styrene. Although this process has
been used in other parts of the world, only Oxirane Corp. uses it in the
U.S. (the technology was developed by Halcon International, a joint owner
c
of Oxirane).
The economic efficiency of direct oxidation is determined by the mar-
kets for both styrene and propylene oxide. Although growth prospects for
propylene oxide are good, some excess capacity is expected to persist in the
near term. Moreover, the market for propylene oxide is highly concentrated,
with only five producers. Oxirane already was marketing propylene oxide
from a different production facility when it constructed its coproduct
plant. Other producers are reticent to adopt a process whose economic
viability depends upon the firm entering a new, highly concentrated market.
Similarly, propylene oxide producers currently are not interested in entering
7-4
-------
the styrene market, which also suffers from excess capacity. Consequently,
major inroads by this process are not expected in the near term. However,
it is one of the most efficient ways to produce propylene oxide7 and
provides styrene at competitive costs, assuming the producer has an outlet
for propylene oxide. Consequently, it can coexist with dehydrogenation.
7.1.2.2 Resource Use
Estimates of base employment and energy use in the EB/S industry are
necessary to assess the potential impact of controls on such resource use.
A rough estimate of direct employment in the EB/S industry is 650 to 950
8
workers. It is difficult to be more specific because of variations between
plants and because companies sometimes are reluctant to provide this infor-
mation. Energy use in the EB/S industry is estimated to be 25 megajoules
per kilogram (MJ/kg) of production. This corresponds to approximately
78 million gigajoules (GJ) in T9787 In that year, EB/S production accounted
for 47 percent of all benzene consumed in the United States, or 2.67
million megagrams (Mg).9 Benzene consumption is of interest because benzene
availability has been a problem for the industry in the recent past and
could be again.
7.1.2.3 Product Use
Styrene is used in the manufacture of plastics and resin materials.
These materials are high molecular weight polymers that can be shaped or
otherwise processed through the application of heat and/or pressure.
Plastics are solid forms that are molded, cast, or extruded; resins are
solutions, pastes, or emulsions that can be used as coatings. There are two
basic types of plastics and resin materials: thermosetting materials, which
can be softened or shaped only once by heat or solvents, and thermoplastic
materials, which can be repeatedly softened and shaped by heat.10 Styrene
materials are generally thermoplastics.
7-5
-------
The predominant thermoplastic material produced from styrene is poly-
styrene. As shown in Table 7-2, 1976 production of polystyrene accounted
for 62 percent of total styrene consumption. Polystyrene is a medium
strength, rigid, easily dyeable material of relatively low cost. It is used
in packaging, food containers, insulation, furniture parts, and "styrofoam."
Polystyrene demand grew by more than 13 percent per year in the 1960's
because of its low cost vis-a-vis competing materials. As its price
increased during the 1970's, it became less competitive. As a result, some
TABLE 7-2. PRODUCT CONSUMPTION — PERCENT BY INTERMEDIATE USE5
Intermediate Uses (1976)
Ethyl benzene uses:
Styrene
Other
Styrene uses:
Polystyrene
Other styrene copolymers
Styrene-butadiene elastomers (SBR)
Unsaturated polyester resins
Miscellaneous
Net exports
TOTAL
Percent of
Domestic
Production
98a
2a
100
54
17
9
6
1
13
100
Percent of
Domestic
Consumption
99
1
100
62
20
11
7
1
—
101
'Estimates for 1975.
7-6
-------
substitution of polypropylene, wood, and paper products for polystyrene has
taken place. Substitution may continue, but it is not expected to prevent
all growth in the polystyrene market. Projected growth through 1985 of six
percent per year is lower than it has been in the past because of product
competition and mature markets for the final goods produced. But growth
will continue because of the low capital cost requirements in processing
polystyrene and the high costs of switching input materials. Also, in some
situations polystyrene is so much more suitable technically than other
materials that substitution is unlikely in the near term.
Other styrene copolymers include acrylonitrile-butadiene-styrene (ABS)
resins, styrene-acrylonitrile (SAN) resins, and styrene-butadiene copolymer
latexes. ABS resins are used in pipe and in automotive and appliance appli-
cations. SAN resins are used in houseware and automobile transparent panels
and windows. Copolymer latexes are used in paper coatings and carpet
backing. These uses accounted for 20 percent of U.S. styrene consumption in
1976 (Table 7-2). Like polystyrene, these products face mature markets but
still some growth is expected.
In 1976, styrene-butadiene elastomers (SBR) accounted for 11 percent of
domestic consumption of styrene, its primary use being the manufacture of
synthetic rubber. Final goods produced include tires, hoses, belting, and
adhesives. This market currently is depressed because of the weak tire
4
market. Increased tire life has created a substantial inventory buildup
which will persist for some time. Little, if any, growth is expected in
this market nor is substitution of other materials.
Unsaturated polyester resins are used in transportation-related and
construction materials. These applications accounted for seven percent of
domestic consumption in 1976. This market should continue to grow.
7-7
-------
Substitutes for styrene materials include more basic products like
wood, glass, and paper. As noted, there has been some substitution of these
goods for styrene products due to styrene price increases. However, this
type of substitution is limited because, in many cases, plastic materials
originally replaced these products as a result of their technical
superiority. A return to basic products will be avoided unless the price of
plastics becomes very high indeed.
Styrene plastics may be supplanted by other plastics to some degree if
prices continue to increase. As noted, polypropylene has replaced poly-
styrene in some applications. There also may be cost advantages to substi-
tuting polyvinyl chloride or polyethylene for some styrene-based plastics.
Thus far, the fear that other polymers, especially polypropylene, would
12
substantially replace styrene materials has proved unjustified, indicating
that either the technical difficulties or the costs of substitution are
greater than initially indicated.
Table 7-3 presents final uses for styrene materials ranked according to
percentage of domestic styrene production and consumption. Packaging is the
predominant use and this is not expected to change in the near future. Con-
struction-related uses are second in importance, followed by electrical
goods, household goods, and transportation-related applications.
7.1.3 Production Trends
From 1961 to 1976, styrene production increased by an average of over
eight percent per year while the Gross National Product (GNP) increased by an
average of less than four percent per year. However, this average is not
representative of activity during the entire period. From 1960 to 1972,
annual growth averaged ten to 11 percent.
7-8
-------
TABLE 7-3. STYRENE CONSUMPTION — PERCENT BY FINAL USE13
Final Uses (1974)
Packaging
Construction-related
Electrical goods (incl. appliances)
Household goods (excl. appliances)
Transportation- related
Recreational goods (incl. toys,
sporting gds.)
Disposable serviceware
Miscellaneous
Exports
TOTAL
Percent of
Domestic
Production
22
16
12
12
10
8
4
n
5
100
Percent of
Domestic
Consumption
23
17
13
12
10
9
5
11
__
100
As shown in Table 7-4, production increased most dramatically in 1969
and again in 1972. However, year-to-year variations during this period
resulted predominantly from the normal vicissitudes of business. Beginning
in 1973 this was not the case. Production remained approximately constant
from 1973 to 1974. Demand was not static in this period but supply was.
The Mideast oil crisis affected the EB/S industry as it did most
petroleum-related production. EB/S production was constrained by overall
benzene shortages. Note that while production remained constant, capacity
increased by over nine percent. Capacity utilization declined from 91 percent
to 83 percent.
In 1975, production declined from 2,701,000 Mg to 2,119,000 Mg, or by
over 21 percent. Capacity increased by over nine percent and utilization
7-9
-------
TABLE 7-4. HISTORICAL DOMESTIC STYRENE CAPACITY AND PRODUCTION1'14'15
Year
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
Capacity (10s Mg)
844
998
1,068
1,109
1,268
1,417
1,460
1,574
1,846
2,086
2,086
2,696
2,812
2,984
3,256
3,392
3,712
4,200
4,092
Production (103 Mg)
791
799
881
977
1,166
1,299
1,447
1 ,487
1,677
2,108"
1,966
2,123
2,694
2,710
2,701
2,119
2,858
3,114
3,119a
Provisional Estimate, U.S. International Trade Commission.
7-10
-------
decreased to 62 percent. Just as benzene shortages eased, demand decreased
substantially as a result of general worldwide.recession. In 1976, produc-
tion increased by almost 35 percent to levels above those characteristic of
the early 1970's. However, capacity again increased by over nine percent.
Capacity utilization in 1976 was approximately 77 percent. Production
increased by nine percent in 1977 and remained approximately static in 1978.
Finally, capacity increased by 13 percent in 1977 and decreased by over three
percent in 1978.
The phenomenon of increasing capacity in the presence of a production
slack during this period requires some explanation. The 1960 to 1978 period
was characterized by increasingly larger individual production facilities.
The replacement of small capacity facilities with "large capacity facilities
was dictated by the economics of production over the long term. In partic-
ular, the potential for economies of scale was recognized. This process
exacerbated excess capacity problems at times.
7.1.4 Prices
Table 7-5 presents a price history for styrene which includes published
list prices as well as unit sales values. Unit sales values represent
average prices actually paid and are generally indicative of contract prices
for styrene. Since large amounts of styrene are traded in this manner, unit
sales values frequently are more useful than list prices.
Both list prices and unit sales values decreased from 1960 to 1965. In
fact, list price and market sales values had been declining since 1952 when
the list price was $463/Mg and the unit sales value was $353/Mg. Increased
technical efficiency and scale of operation are the most probable explana-
tions of this decline. From 1965 to 1972, list price fluctuated at about
7-11
-------
TABLE 7-5. STYRENE PRICE HISTORY15'16'17
Year
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
Cents
List Price
27.6
24.3
23.2
23.2
20.9
17.6
19.8
18.7
16.5
18.1
17.6
17.6
17.6
15.9
44.1
48.5
46.3
46.3
46.3
per Kilogram
Unit Sales Value
25.1
23.5
22.1
20.4
17.9
16.8
17.7
17.0
14.8
14.0
14.3
13.3
12.6
15.5
38.0
41.3
43.4
41.0
—
Dollars per
List Price
275.58
242.51
231.48
231.48
209.44
176.37
198.41
187.39
165.34
180.78
176,37
176.37
176.37
158.73
440.92
485.01
462.97
462.97
462.97
Metric Ton (Mg)
Unit Sales Value
251.32
235.45
221.12
204.15
178.57
168.21
177.03
169.97
147.49
139.99
142.64
132.94
125.88
154.54
380.07
412.48
433.87
409.84
—
the 1965 level but unit sales value continued to decline. In 1972, list
price was almost 40 percent above unit sales value. Then, in 1973, list
price declined and unit sales value rose, bringing the two values back into
line with each other.
Late in 1973, styrene prices began to be forced up by benzene shortages.
In 1974, list price almost tripled and unit sales value more than doubled.
This increase was the result of more than just benzene shortages. Other
factors influencing price included: (1) the removal of 1973 price controls;
7-12
-------
(2) the fourfold increase in crude oil prices, which increased both raw
material and energy costs; and (3) an increase both in investment values,
reflecting higher capital costs, and in returns on investment, generating
18
higher profitability levels. List price rose again in 1975 and then
declined slightly in 1976 to $462.97/Mg, where it stayed for several years.
In early 1979, list price again began to move upward. In May 1979, announced
prices reached $683.43/Mg. These increases reflect rapidly escalating
benzene prices and strong European demand for many petrochemicals. It is
not yet clear where the price will stabilize.
Both list prices and published unit sales values must be interpreted
with some caution. An announced list price may or may not have associated
with it a discount, sometimes called a "temporary allowance."19 Even though
list price may remain unchanged for a substantial period, it is not unlikely
that discounts offered vary over that period. Similarly, producing company
policy with regard to the payment of transportation costs can introduce a
differential between either contract or spot price and the effective price
paid. Producers sometimes absorb transportation costs, but not always.
Absorption of such costs will vary with customer, delivery location, and
20
market conditions. Consequently, both list price and unit sales value
should be treated as average values only.
7.1.5 International Trade
In 1976, approximately 40 percent of world styrene capacity was located
in the United States and Puerto Rico. At that time, Western Europe had
34 percent of world capacity and Asia had 20 percent. These proportions
should change some in the near future because European styrene capacity is
increasing vis-a-vis U.S. capacity. These increases are expected to alter
the current U.S. export position during the 1980's.
7-13
-------
Traditionally/ the U.S. has been a net exporter of styrene (see Table 7-6).
In many years, exports have represented a substantial portion (ten to 15 percent)
of U.S. production. In 1975, 67 percent of exports went to Western Europe
and 27 percent went to Latin America. At maximum, imports have represented
one to two percent of total consumption. Imports generally have come from
Canada.
TABLE 7-6. STYRENE EXPORTS AND IMPORTS (103 Mg)21
Year
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
Exports
71.7
62.6
55.7
54.6
155.0
166.2
163.1
197.8
243.8
377.1
257.4
167.4
299.8
260.6
282.3
260.4
431 . 6b
445. 9C
357. 2C
Imports
N.A.a
N.A.
N.A.
N.A.
N.A.
N.A.
N.A.
N.A.
N.A.
N.A.
4.6
0.2
11.2
14.0
27.6
3.0
11. 2b
3.0d
14.0d
Net Exports
—
—
—
—
—
—
—
—
—
—
252.8
167.2
288. 6
246.6
254.7
257.4
420.4
442. 9
343.2
aAccurate estimates of imports in this period are difficult to obtain
because of the manner in which data are grouped in Department of Commerce
publications:
bPersonal communication. Bureau of the Census, Foreign Trade Division,
Chemicals Section.
'Chemical Marketing Reporter.
Chemical Marketing Reporter.
March 26, 1979. p. 11.
March 19, 1979. p. 11.
7-14
-------
Exports are expected to decline as capacity increases in Western Europe.
This decline may be proceeding more slowly than anticipated because of
problems with naphtha availability in Europe in 1979. Small European capac-
ities, together with lower freight rates to Europe from the U.S. than from
Japan, have given U.S. producers an advantage in the past. However, this
and the other U.S. advantage, lower raw material costs, are disappearing.
In addition, the styrene industry has been increasingly characterized by
U.S. companies with European affiliates and European companies with U.S.
affiliates. As a result, experts anticipate smaller future styrene price
differentials between the U.S. and Europe.22 Table 7-7 presents unit sales
values for both domestic and export sales. Note that differences between
TABLE 7-7. STYRENE UNIT SALES VALUES FOR DOMESTIC AND EXPORT MARKETS
(Dollars per Metric Ton or Mg)23'24
Year
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
Domestic
Unit Sales Value
168.21
177.03
169.97
147.49
139.99
142.64
132.94
125.88
154.54
380.07
412.48
433.87
409.84
Export
Unit Sales Value
180.78
154.30
145.50
138.89
138.89
141.09
123.46
127.87
299.83
529.10
392.42
— ,
399.76
7-15
-------
these values vary considerably from year to year in both magnitude and
direction.
7.1.6 Market Structure
7.1.6.1 Firm Characteristics
Table 7-8 shows U.S. styrene producers, their plant locations and
capacities, and some indication of the degree of vertical integration in the
EB/S industry. Note that all styrene manufacturers produce ethyl benzene
also. In some integrated operations EB is not separated as such; however,
in most cases, EB and styrene are produced in the same basic location. When
they are not, styrene production coincides locationally with the production
of styrene derivatives, as opposed to feedstocks (EB). Eight of the 11 EB/S
firms also produce styrene derivatives. Of those three that do not, two
(El Paso Products and Sun Oil Co., Inc.) produce relatively small amounts of
styrene.
Something akin to horizontal integration is also characteristic of the
EB/S industry in the 1970's. Two producers, Cos-Mar and Oxirane, are sep-
arate operating companies, each of which is owned jointly by other companies.
Cos-Mar is owned jointly by Borg-Warner (30.8 percent) and American Petrofina
(69.2 percent). Each of these companies previously has owned and operated a
styrene plant of its own. Oxirane is owned by Atlantic Richfield (50 per-
cent) and Halcon International (50 percent). Atlantic Richfield owns other
existing styrene capacity.
These operating companies were not "taken over" by existing companies,
as would be typical of horizontal integration. Rather, they were created by
the existing companies to meet their needs for EB/S capacity in the most effi-
cient manner. A jointly owned operating capacity allows each owner to achieve
the economies of large-scale production without building a much larger plant
7-16
-------
TABLE 7-8. U.S STYRENE PRODUCERS — LOCATION, CAPACITY. AND INTEGRATION1'25
Company
American Hoechst
Corp.
Amoco
Atlantic Richfield
Company (ARCO)
Cos-Mar, Inc.
Dow Chemical USA
El Paso Products
Gulf Oil Corp.
Monsanto Co.
Oxirane Corp.
Sun Oil Co. , Inc.
United States Steel
Company
r -t . Produces
Plant capacity Company styrene
Location (10 Mg) Produces EB? Derivatives?
Baton Rouge, LA
Texas City, TX
Port Arthur, TX
Kobuta, PA
Carville, LA
Bay City /Midi and,
MI
Freeport, TX
Odessa, TX
Dona! dsonvi lie, LA
Texas City, TX
Channel view, TX
Corpus Christi, TX
Houston, TX
450
380
227 (EB)
200 (S)
590
181
680
115
272
680
545
36
54
yes
yes
yes
no
yes
no
yes
yes
yes
yes
yes
yes
yes
yes
yes
no
yes
yesa
yes
yes
no
no
yes
yesa
no
yes
a .....
Cos-Mar is a joint venture involving Borg-Warner Corp. and American
Petrofina, Inc. Each produces a styrene derivative. Oxirane is a joint
venture of Halcon International and Atlantic Richfield which produces
styrene derivatives.
7-17
-------
than it needs or wants. Oxirane, in particular, functions as a reliable
source of additional styrene for several existing companies. Its existing
capacity has been sold out "long term" to several producers: Atlantic
oc
Richfield (an owner), American Hoechst, Borg-Warner, and Gulf.
The economies of large-scale operation are not process-related. They
do not involve lower costs in terms of raw material, energy, or labor usage.*
These economies do involve spreading overhead and possibly achieving lower
capital requirements per unit of capacity. This analysis is supported by an
examination of styrene plants over time. In 1960, total styrene capacity
was 844,000 Mg. There were nine plants with an average size of 94,000 Mg,
the largest of which produced 236,000 Mg. By 1965, all of those plants had
expanded and several had been added. By 1970, more plant expansion had
occurred and Cos-Mar came on stream. In 1970, average plant size was
152,000 Mg and the largest had a capacity of 363,000 Mg. By 1978, average
size was 337,000 Mg and the largest had a capacity of 680,000 Mg. The
average size increased because new large plants came on stream, some existing
plants were expanded, and some older, small plants were phased out.
It is important to note that not all large plants represent single
large production units. Some, like Oxirane, represent a single production
train. Others, like the Dow plant in Texas, represent multiple units.
Moreover, multiple units are not necessarily of the same age, so that a
given plant may not have "an" age, but rather, multiple ages.
Table 7-9 presents some further information on selected owners of
styrene capacity. Rough estimates of styrene sales by company were made to
*Existing smaller plants are less efficient in these terms, tjut this is not
because they are smaller so much as because they are older.
7-18
-------
TABLE 7-9. FINANCIAL CHARACTERISTICS OF SELECTED STYRENE PRODUCERS28
Company
American Petrofina
(joint owner, Cos-Mar)
Amoco
Atlantic Richfield
(ARCO + joint owner, Oxirane)
Borg-Warner
(joint owner, COS-MAR)
Dow Chemical
Gulf Oil
Monsanto
Sun Oil Co. , Inc.
Year
1977
1977
1977
1977
1977
1977
1977
1977
Debt-
Equity
Ratio
0.638
0.910
1.245
0.664
1.462
0.939
0.812
0.211
After-Tax
Rate of
Return on
Equity
10.3
15.0
14.2
11.9
17,8
10.2
11.5
13.1
Styrene
Sales as
a % of .
Revenues
11.4
0.8
1.3
5.3
4.1
0.4
4.5
2.6
a. _.
own assessments of after-tax income. The size and diversification of the
companies and the types of markets in which they operate make their tax
structures very complex. '
Styrene sales were estimated by applying a capacity utilization rate of
74 percent (1977 average) to each company's capacity to estimate produc-
tion and multiplying this by the 1977 unit sales value ($409.84/Mg).
determine their degree of importance in overall company revenues. Styrene
sales represent between 0.4 percent (Gulf) and 11.4 percent (American-Petrofina)
of total company revenues. All of the companies examined are highly diversi-
fied. In fact, the only styrene producers that are not highly diversified
are the operating companies, Cos-Mar and Oxirane, and the owners of those
companies are themselves highly diversified.
After-tax rates of return on equity vary across the selected companies.
The lowest value is 10.2 percent for Gulf Oil and the highest value is
7-19
-------
17.8 percent for Dow Chemical. Debt-equity ratios also vary, with the
lowest value being 0.211 for Sun Oil Co., Inc. and the highest being
1.462 for Dow Chemical. Even the highest ratio is sufficiently low so that
debt financing should not present a problem for any of these companies.
7.1.6.2 Market Concentration
Table 7-10 presents market concentration ratios for the styrene industry
for 1960, 1965, 1970, 1975, and 1978. Dow Chemical is the largest single
producer in each of those years. However, its share of the U.S. market has
declined from 33.5 percent to 21.3 percent. It is interesting to note that
in 1976, Dow's worldwide holdings represented 17 percent of total world
styrene capacity.30 In 1960, the top two and top four firms represented
61.5 and 81.3 percent, respectively, of total U.S. capacity. Comparable
figures for 1978 are 38.1 percent and 63.9 percent. Concentration ratios
for SIC 2865 (Cyclic Crudes and Intermediates), of which styrene is a part,
show the top four and top eight companies accounting for 34 percent and
52 percent of shipments, respectively. Thus, styrene manufacture is a
1 ?9
TABLE 7-10. CONCENTRATION IN THE STYRENE INDUSTRY1'"
Percent of Total Production Capacity Accounted for by
Largest 1, 2, 4, and 8 Styrene Manufacturers
Year
1960
1965
1970
1975
1978
Largest
1
33.5
26.2
19.2
24.7
21.3
Largest
2
61.5
46.9
35.1
44.8
38.1
Largest
4
81.3
68.8
62.8
69.3
63.9
Largest
8
100.0
95.3
90.4
95.7
96.1
Total
Number of
Manufacturers
8
11
12
11
11
7-20
-------
highly concentrated segment of an overall industry group that is itself
rather concentrated.
Concentration alone is not sufficient to guarantee market power. In
the case of the EB/S industry, several factors weaken the potential power
created by concentration. The fact that the product generated is an inter-
mediate good, which is not differentiated by producer, increases the likeli-
hood of competition.* Currently there is excess capacity in the EB/S indus-
try, which also makes it difficult for producers to manipulate product
price. Of course, as capacity is approached, this constraint is weakened.
The existence of an international market also weakens domestic market power.
The U.S. traditionally has been an exporter and not an importer of styrene,
but existing trends indicate that this will not continue. Price differen-
tials between the U.S. and Europe are expected to decrease, increasing the
competitive potential of foreign producers. Finally, the existence of some
substitute materials in the form of other plastics increases the likelihood
of competition.
Examination of the pricing process in the EB/S industry supports the
theory that market power currently is not as great as concentration ratios
might indicate. There does not appear to be any established price leader in
the industry. Price increases sometimes have been announced by individual
firms and then retracted when other firms did not increase prices also.
Concern about whether price increases will "hold" appears to be not uncommon
in the industry. Thus, while there appears to be some market power in the
industry, it is not unrestrained.
*Chemical products are more likely to be differentiated by buyer than by
producer. That is, the producer sometimes tailors the product to buyer
specifications. J
7-21
-------
7.1.7 Supply and Demand
7.1.7.1 Supply
No attempt was made to estimate econometrically a supply function for
the styrene industry. Nevertheless, it is clear that the quantity of styrene
supplied is affected materially by the price of styrene and the prices of
the inputs used to produce styrene. Also, the quantity supplied reflects
any input availability problems that are not captured fully by the prices of
those inputs. For example, in 1974, benzene shortages substantially con-
strained the production of styrene.
Important input prices which affect styrene supply include prices of
feedstocks, labor, machinery and equipment, and fuels. Feedstocks and fuels
have had particular impact in the recent past. Primary feedstocks in the
EB/S industry are benzene and ethylene. The market for ethylene has been
comparatively stable. This is not expected to change. However, the market
for benzene presents more of a problem. Benzene is a petrochemical and its
market reflects any upheaval in the petroleum market. Benzene prices have
increased considerably in the recent past. Recent price increases for
styrene substantially reflect these increases in feedstock costs.
Energy costs also have increased substantially. Frequently, fuels are
purchased on relatively long-term contracts. When contracts expire and are
renegotiated, fuel prices can increase considerably. In 1977, it was esti-
mated that plants using "old gas," i.e., gas contracted for at "old" prices,
had fuel costs of approximately 1.1$ per kilogram of styrene. It was antici-
pated that at contract renegotiation those prices could increase to between
oc
4.4$ and 5.5
-------
Styrene supply also could be influenced by the market for propylene
oxide as a result of the coproduct process. The profitability of the co-
product process depends upon the markets for both styrene and propylene
oxide. Should the demand for propylene oxide alter substantially, it could
induce changes in the quantity of styrene supplied.
7.1.7.2 Demand
A derived demand function for styrene was estimated using time series
data for 1958 to 1973. Although data for 1974 to 1977 were available, they
were excluded because observations for those years were not consistent with
previous data. Input availability problems in 1974 could have been handled
through the use of dummy variables (an econometric technique) had there been
no further complications. But upheaval in the petroleum market and consequent
price increases created further difficulties. For example, the price index
used to adjust styrene prices probably is not appropriate for the period
after 1973 and cannot be used to generate true relative price changes for
the plastics industry. Please note that any application of estimates gene-
rated for the period 1958 to 1973 to later periods requires the assumption
that there has been no major structural change in the demand relationship.
A shift of some kind certainly occurred after 1973, but a simple upward
shift need not invalidate econometric estimates.
The equation estimated through ordinary least squares is:
(1)
In QSTY = 2.74 - 0.49* In RSP + 1.36* In RING; R2 = 0.986
(1.67) (0.24) (0.34)
where
QSTY = quantity of styrene produced
*These coefficients are significant at the 90 percent confidence level
7-23
-------
RSP = relative price of styrene
= (styrene price x 100)/(styrene price, 1958 x PI)
RING = real income = (GNP/PI) x 100
GNP = Gross National Product
In = natural logarithm
PI = price index.
Numbers in parentheses present the standard errors of the coefficient
estimates. This equation represents a long-run demand function. Note that
the relative price of styrene (RSP) and real income (RINC) both generate
significant coefficients with the expected signs, e.g., the quantity of
styrene demanded varies inversely with the price of styrene and directly
with real income.
Also note that the coefficient of real income is much larger than the
coefficient of styrene price. This supports the statement that styrene
demand is tied closely to the performance of the overall economy. As the
equation is estimated in natural log form, the coefficient of the styrene
price variable is the negative of the price elasticity of demand for styrene.
7.1.8 Baseline Projection
7.1.8.1 Baseline Regulatory Environment
The industry is assumed to be in compliance with existing regulations
prior to enforcement of standards considered here. Accordingly, estimated
costs of control (Section 7.2) are incremental costs associated with increased
pollution control.
7.1.8.2 Baseline Growth Rates
The demand for styrene in the future will be tied closely to the per-
formance of the overall economy. Significant market penetration in existing
7-24
-------
styrene uses indicates that future demand for styrene will depend upon final
demands for the goods produced. While these goods satisfy a variety of
consumer needs, the demands for these goods tend to be sensitive to overall
income trends.
Several other factors have the potential to influence styrene demand,
but are not expected to be of major importance. New use development is
always a possibility and could cause an increase in the demand for styrene.
But significant new use creation is not foreseen in the near future. Compe-
tition among various polymers could alter styrene demands. If styrene were
to replace other polymers, the demand for styrene would increase. During
the 1960's, considerable demand growth occurred in this manner, but this is
not expected to continue. If styrene were replaced by other polymers,
demand for styrene would be affected negatively. Some such replacement has
been feared for several years, but has not been forthcoming. Some experts
now feel that competition among polymers will have only small net effects on
the various markets involved.31 They reason that new-use development is
more profitable than the capture of markets from other polymers. Thus,
producers of competing materials may not be actively seeking to capture
styrene markets. ;
Several sources estimate growth rates for styrene consumption based on
projected demands for derivative products. SRI International predicts
annual growth rates of four to six percent for styrene demand through 1980.32
One industry source predicts an annual growth rate of five percent through
33
1985. The National Petroleum Refiners Association predicts a growth rate
of five percent per year through 1987.34 Thus, five percent is taken to be
a best'estimate for the period 1979 to 1985. Assuming a five percent annual
growth rate, styrene demand in 1985 would be 4,389,000 Mg.
7-25
-------
7.1.8.3 Baseline Investment Projections
Existing capacity is 344,000 Mg short of projected demand for 1985.
Furthermore, a portion of existing capacity certainly will be retired by
1985 and require replacement. Without more specific information on the age
structure of existing plants, it is difficult to estimate replacement needs,
but it is reasonably certain that most of the smaller units still in exis-
tence will be retired over this period.
One industry expert estimates that approximately 30 percent of 1976
capacity was made up of smaller, older, less efficient units that could be
yc
expected to shut down by 1982. Feedstock, fuel, and labor utilization are
all relatively poor for these units. In fact, several such plants have
closed already, representing approximately five percent of 1976 capacity.
If the prediction that 30 percent will close is correct, closures amounting
to 928,000 Mg are still to come. If such closures are forthcoming, total
new capacity of 1,200,000 to 1,300,000 Mg will be required by 1985. Given
typical sizes of units constructed in the recent past, three to four new
units will be needed to meet this demand, assuming full or nearly full
utilization of all units. Note that new units do not necessarily mean new
producers or new plant locations, since additions, expansions, or replace-
ments at existing sites are common in the EB/S industry.
7-26
-------
7.2 COST ANALYSIS OF ALTERNATIVE EMISSION CONTROL SYSTEMS
7.2.1 Introduction
This section presents capital and annualized control" cos'ts=forr three
of the four benzene emission regulatory options under consideration. The
costs are presented on a pi ant-by-plant basis for each of the facilities
producing EB/S identified in Chapter 3. No data are presented for those
facilities producing ethylbenzene from mixed xylenes.
The three regulatory options for which cost data are presented are:
• Regulatory Option A (85 percent control) — achievable by routing
emission streams to either a scrubber or condenser system.
a Regulatory Option B (94 percent control) — achievable by coupling a
flare system to the scrubber/condenser system of Regulatory Option A.
• Regulatory Option C (99 percent control) -- achievable by routing
emission streams to an existing boiler.*
Under each option the continuous release of the hydrogen separation vent is
routed to a process heater (99 percent control).
7.2.2 Equipment Used
7.2.2.1 85 Percent Control with a Scrubber and Condensers — Regulatory
Option A ~
This analysis assumed that 85 percent control can be achieved using a
scrubber and condenser system. The scrubber (absorption tower) operates
with lean** polyethylbenzene (PEB) oil which is recycled back to the process,
The scrubber is designed to control continuous venting from the alkylation
reactor and all vacuum columns in the EB/S facility with the exception of
the benzene/toluene column(s). The condenser system for each facility
consists of two condensers, one of which controls the benzene/toluene
hotwell vent(s) and the other which controls the combined vent emissions
*Boiler includes process heaters, superheaters, and other similar units.
**
Lean indicates an absence of benzene.
7-27
-------
from all of the columns operating at atmospheric pressure and above. The
vent streams from the condenser and scrubber system are then released to
the atmosphere.
7.2.2.2 94 Percent Control with a Flare — Regulatory Option B
This analysis assumed that 94 percent control can be achieved with a
flare system when it is used in conjunction with a scrubber and condenser
system. The 94 percent control is based on an 85 percent reduction of
benzene with a scrubber and condenser system identical to the one mentioned
above in addition to a 60 percent reduction with a flare. The vent streams
from the condenser and scrubber system are then manifolded into a flare
header which is attached to a flare which has been sized to handle the
continuous vent releases. Figure 7-1 presents a schematic of the option
described here. Table 7-11 presents a list of equipment assumed necessary
to achieve this level of control.
7.2.2.3 99 Percent Control with a Boiler — Regulatory Option C
Under this regulatory option all of the vents are manifolded together
and the combined vent stream then is routed to an existing boiler. In
order to combust the vent stream, which is at a relatively low pressure
when compared to normal fuel line pressures, the boiler either must be
retrofitted with a burner which uses high pressure air to aspirate the vent
stream into the burner, or the vent stream must be compressed to 30 psi or
greater and introduced into the boiler fuel system where it can be combusted.
In either case, compressors are necessary when vent streams are to be
compressed up to 300 psig and introduced into the main fuel system. Figure 7-2
presents a schematic of the option described here. Table 7-11 presents a
list of the equipment needed to achieve this level of control.
7-28
-------
LU
a.
o
o:
o
o
o
o
•a;
LU
3Z
O
exs
i i
OQ <^>
7-29
-------
TABLE 7-11. CONTROL EQUIPMENT NEEDED TO ACHIEVE BENZENE EMISSIONS REDUCTION
(Regulatory Option A -- 85 Percent Benzene Emissions Reduction)
Vent condenser Type AEM carbon steel shell 3/4" diameter tubes.
Tube surface areas 68 sq. ft atmospheric and pressure
columns, 32 sq. ft benzene/toluene columns
Scrubber 9" diameter steel shell 6 ft of 1" diameter Raschig ring
packing material, PEB chiller, PEB pumps and controls
Scrubber and condenser Brine piping, vent piping, PEB piping
piping
Compressor
(Regulatory Option B ~ 94 Percent Benzene Emissions Reduction)
Flare — 3" diameter Flare tip, pilot, steam injectors, stack, water seal,
knockout drum, controls and instruments, flare base
Flare piping Vent header, drain and sewer, air and water, natural gas
Vent condenser Type AEM carbon steel shell 3/4" diameter tubes. Tube
surface areas 68 sq ft atmospheric and pressure columns,
32 sq ft benzene/toluene columns
Scrubber 9" diameter steel shell 6 ft of 1" diameter Raschig ring
packing material, PEB chiller, PEB pumps and controls
Scrubber and condenser Brine piping, vent piping, PEB piping
piping
Compressor
Oxygen monitor
(Regulatory Option C ~ 99 Percent Benzene Emissions Reduction)
Piping Vent header and manifolding of vents to a common system
Oxygen monitor
Compressor
7-30
-------
FIGURE 7-2
BURNER SYSTEM DIAGRAM
Vacuum Co lumn .Vents . -—:—
PEB Column - EB
Benzene/Toluene Column-S
Ethylbenzene Co lumn-S
Styrene Column -S
Atmospheric/Pressure Column Vents
Benzene Drying Column- EB
Benzene Recycle Co lumn-EB
Benzene/Toluene Spfitter-S
Ethylbenzene Coiumn-EB
Alkylation
Reaction
Area Vents
Existing
Boiler
Natural
Gas
1
I
H2
Separation
Vent
7-31
-------
7.2.3 Equipment Costs
7.2.3.1 Introduction
The capital, operating,, and maintenance costs for all of the control
equipment considered are based on the stream characteristics shown in
Table 7-12.
The quantities presented in Table 7-12 represent the emissions from a
model facility with a capacity of 300,000 Mg/yr (330,000 tons/yr) of styrene
and 345,000 Mg/yr (380,000 tons/yr) of ethylbenzene. The cost analysis
assumes that vent flows show a linear relation with plant capacity.
7.2.3.2 Capital Costs for Control Equipment
Figure 7-3 presents capital costs curves for flares, flare headers,
burners,* scrubbers, and condensers for the various plant sizes under
consideration. All capital costs presented are in terms of fourth quarter
1978 dollars. The costs presented include installation costs but do not
include allowances for fees, contingencies, taxes, and other indirect
costs. All of the cost curves with the exception of those representing the
total installed costs of scrubbers and condensers exhibit a marked plateau
below 150,000 Mg/yr (165,000 tons/yr) plant capacity. It was assumed that
there would be very little change in total installed cost of a piece of
equipment below this size. This is because installation and piping con-
tribute the most to total installed cost for small pieces of equipment. In
addition, a minimum size was specified for some pieces of equipment. Also,
the analysis assumed that no pipes smaller than two inches in diameter
would be used and that the total length of pipe needed would not change
^Burners refer to the equipment needed to introduce the vent streams into
the combustion device and include a compressor piping and aspirating
burner if necessary.
7-32
-------
UJ
^y
UJ
H-
J
^""^
O5
O
O
0
O
O
CO
f—
re z:
CO
-
Z 1—
UJ HH
> O
c
O5 O
O T-
t \_£
~a re
3; re
Q.
O)
CO
in
•P
c
O) i-
> 01
-C
C -P
E 0
3
'o
CJ
EQ
3 1—
3 \
U CD
re
in
i- -P
3 C
to a>
0) >
s- c
a. E
Xs. 3
•p "o
> t-
"^
co
o
O r—
O CO O CM CM
rx. x co co
CO ^
CM
r— o m in o ITS in
i — o i — ro «3-
o
o
in o in m CM o oo
CM O ^" r— CM r—
f.
in
r—
O O Lf) LO CM O 00
O O <3- i— CM r—
r— O
0
ir>
CM O O 1 CM O 00
r-0 - COr-^J-
O
o
in
•r-
re
s- o
-CO- S-
"^x (^ OJ ^C O
in co S^ c m
i- S- Ol S- IDP
JTOI-P -P NO! OS-
re Q)-r- en 01 -P i— CMC
•p -*: i— -r- cQOOoi— i
o o>
H- 3=
in
Ol "O
v> c
re re
u
O)
*~ o
c o
3 o
o •>
J= O
in i—
Ul O5
•P Ot
O5-P
CM
O
CM
CM
O
O
f^s
I*™
CM
i—
^~
Lrt
•N^
CM
•
r™™
^~
ai
c
O)
N
C
Ol
ca
o
•
r^»
o
•
o
CM
CM
CO
0
•
00
LO
r
OJ
r^
^3
•^
-P
W
3
Q
S
o
0
r™* CJ
re o
•p ^>
o
h-
•r—
T5 C
Ol
Ol Ol
c. 5
0 -P
u
C f}
re
-C T3
-P C
re
Ol
S- Ifl
0 E
"Z. re
Ol
S-
. -p
•P in
in
ai -P
Ol
-P >
re
in
Ol 3
S- O
3 •!—
O) s_
•r- re
M- >
Ol 0)
C *""
O -P
0 C.
•P Ol
01
•P S
c -p •
re 01 m
1 1 o ni
•r- 1-
t- 0) 3
•r- 0 O5
C -r- -r-
O) -P i^*
•r- re
U) S- 4->
c
01 s- re
S- Ol i —
re a. a.
o
tn s- i—
01 a. co
s- -a
3 0) 0
O5-C E
•r- -P
M- S-
o.^:
O) Ol \
•c. oi co
1— -^ ^
c
E
^—
0
(J
i—
re
o
E
Ol
i.
Ol
c
Ol
3
r—
O
-p
0)
c
O)
N
C
Ol
ca
re ^3
CM
1C
Ol
T3
3
U
e
TJ
C
re
o
o
Ol
fn
-Q
^J3
in
3
JD
E
O
«J
•P
Q.
Ol
0
X,
O)
in
•a
c
3
O
a.
E
O
U
S-
Ol
[ *
o
,•_
^_
re
Ol •
S- C
re o
Ul O
•p m
Ol "O
c c
M re
u
7-33
-------
FIGURE 7-3
CONTROL EQUIPMENT COSTS WITHOUT ALLOWANCES
FOURTH QUARTER COST (1000's)
INSTALLED CAPITAL COST
300
200-
100-
50-
10
10
I - Complete Header
11 -Complete Burner
—r~
50
II!-Partial Header
IV-Partial Burner
I' I' I'l
100
V-Scrubber
VI- Condensor
500 1000
• PLANT SIZE (Gg/yr)
VII-Small Flare
Based upon scaling of model plant emissions present in Table 7-12.
• Complete header represents the capital cost necessary to pipe all of the vents to the specified control system. This cost
will be incurred at facilities practicing little or no control.
. Partial header represents the capital cost necessary to pipe some of the vents to the specified control system. This cost
will be incurred at plants currently practicing control of some of the vents.
. Complete burner represents the costs of retrofitting the boiler to accept all of the vents and can include multiple burners
and/or compressor/piping systems.
• Partial burner represents the incremental costs of retrofitting the boiler to accept additional uncontrolled vents.
. Scrubber includes the costs of a packed tower, heat exchanger pumps, and piping for control of specified vents.
• Condenser includes two condensers plus piping allowances.
• Compressors and oxygen monitors were considered to cost $30,000 and $23,000, respectively, without allowances,
regardless of plant size.
7-34
-------
significantly at the small plant sizes. The cost curves for condensers and
scrubbers do not show as marked a plateau as do the other pieces of equip-
ment because no minimum size was specified and pilot-sized pieces of
equipment can be used.
The capital costs for the flare, burner, and header systems were
derived from data presented in the Hydroscience Ethylbenzene/Styrene Product
35
Report. Capital costs for scrubbers and condensers were obtained from
oc oy
vendor data. ' These costs were adjusted to account for installation-
costs by using installation factors from Reference 38.
7.2.3.3 Annualized Costs
Table 7-13 presents a list of factors that were used to derive the
annualized costs for each regulatory option. The total annualized costs of
each piece of equipment are made up of operating and maintenance costs,
annualized capital costs, and fuel or recovered material credits, if any.
Maintenance costs were estimated as a percentage of total installed
capital costs. Supplies (misc. capital) were estimated similarly. Operat-
ing costs are made up of electrical power requirements for pumps, compres-
sors, steam, refrigeration, and natural gas. Each of these was assigned a
unit value (see Table 7-13). The utility consumption of each device was
calculated and summed to yield equipment operating costs. It was assumed
that more of the control equipment would require additional operating
labor. Flare utility requirements for purge pilot and steam were derived
from vendor data.
Benzene credits were calculated assuming a value of $1.00 per gallon
($0.13/lb) for any recovered benzene from the condensers and scrubbers.
Each control device was credited with differing degrees of benzene control
depending on its assigned efficiency. No credit was given for recovery of
7-35
-------
other aromatics for ethylbenzene because it was difficult to assign a
realistic value to these products. In addition, recovery was assumed to
occur for 8,000 hrs/yr.
Fuel recovery credits were estimated by assigning a heating value to
each component of the vent gas streams and then assuming that the combined
heating value would have the same value ($2.00/MMBtu) as the natural gas it
would displace. Total annualized costs for each regulatory option were
obtained by annualizing capital costs using a capital recovery factor of
16.25 percent and adding operating and maintenance costs less recovery
credits from fuel or material, if applicable.
3Q
7.2.4 PI ant-By-Plant Costs03
7.2.4.1 Introduction
Capital and annualized costs for individual pieces of control equipment
were based on the requirements of a model uncontrolled facility. Since
each existing facility practices some degree of emissions control, only the
incremental costs needed to achieve a given benzene reduction will be
considered.
An assessment was made of the equipment that would be needed to achieve
a given level of emissions reduction at an uncontrolled facility and this
was then used as the basis for deciding what equipment would be needed at
existing facilities to achieve the same level of control. Table 7-11
presents the list of equipment that is needed on an uncontrolled plant to
achieve the three levels of control specified. Table 7-14 presents the
results of the comparison between the existing facilities and the model
plant lists and indicates the equipment included in the pi ant-by-plant cost
estimates.
7-36
-------
TABLE 7-13. ANNUALIZED COST PARAMETERS'
1. Operating factors
2. Operating labor
3. Maintenance
4. Miscellaneous materials
5. Utilities:
Electric power
Natural gas
Steam
Refrigeration
6. Operating materials
Polyethylbenzenes for scrubber
7. Liquid waste disposal
8. Capital charge rate
9. Recovery credits
Fuel value VOC and benzene
using boiler option
Recovered benzene from
scrubber and condensers
(returned to benzene drying col
8,760 hours per year for utility re-
quirements assuming 100 percent
capacity utilization credits
8,000 hours per year for recovery
Negligible for flares, condensers,
scrubbers, and boilers
5% of total installed capital cost
4% of total installed capital cost
$8.33/GJ,($0.03/kWh)
$0.071/IT ($2.00/1000 scf)
$0.055/Kg ($2.50/1000 Ibs)
$1.97/GJ ($2.08/MMBtu)
available at zero cost
Negligible quantity generated
16.28% of total installed cost
$1.894/GJ ($2.00/MMBtu)
$0.30/Kg ($0.137/1b)
umn)
Values 1,5, and 8 are used consistently throughout this chapter.
7-37
-------
TABLE 7-14. PLANT-BY-PLANT LIST OF EQUIPMENT NEEDED TO ACHIEVE
GIVEN REGULATORY OPTIONS
Plant & Location
Arco, TX
Arco, TX
Arco, TX
Aaarican Hoechst
New
New
New
Aaerican Hoechst
Old
Old
Old
Cos-Mr, LA
Cos-oar, LA
Cos-Mr, LA
Gulf Chenicals,
Gulf Cheaicals,
Gulf Chenicals,
Monsanto
Monsanto
Monsanto
AMOCO, TX
AMOCO, TX
Aaoco, TX
Dow, TX
Hew
Hew
New
Dow. TX
Old
Old
Old
Old
Sun Oil, TX
Sun Oil, TX
Sun Oil , TX
Sun Oil , Tx
El Paso, TX
El Paso, TX
El Paso, TX
Oxirane
Oxirane
Oxiran«
Arco, PA
Arco, PA
Arco, PA
U.S. Steel
U.S. Steel
U.S. Steel
Oow Midland
(Styrene Only)
Regulatory Partial Complete Incremental 3"
Option Header Header Header Flare
85 „
94 Flare X X
99 Boiler X
85
94 Flare
99 Boiler X
85 „ v
94 Flare X X
99 Boiler X
85
94 Flare X X
99 Boiler X
TX 85
TX 94 Flare X X
TX 99 Boiler X
85
94 Flare X X
99 Boiler X
85
94 Flare X X
99 Boiler X
85
94 Flare
99 Boiler X
85 „
94 Flare X X
99 Boiler X
Compressor X
85
94 Flare X X
99 Boiler X
Compressor X
85
94 Flare
99 Boiler X
85
94 Flare
99 Boiler X
85 „
94 Flare X X
99 Boiler X
85
94 Flare
99 Boiler
85 „
94 Flare X X
99 Boiler X
Control Equipment
In" Partial Comolete Oxyqen
Flare Burner Burner Compressor Monitor Condenser Scrubber
X XX
X X X X
V y
A A
'X
x
V X
A A
X XX
X X X X
A A '»
XV
A
X XX
X X X X
Y X
A A
X X X
X X X X
A A A |,"
V Y
A A
X
x x
v X X
A "
x x
V X X
A i
v X
A A
v X
X A
X XX
X X X X
A A «
Y X
A A
X
x x
y V X
A A A
v y •
X A
x
x
v y
X A
X
x x
v y
A "
x x
Y XX
A " n
v X
X A
x x
XXX
A A A
x x
7-38
-------
7.2.4.2 PI ant-by-PI ant Capital Costs
The pi ant-by-plant capital costs for each regulatory option shown in
Table 7-15 were obtained by using the equipment list shown in Table 7-14 in
conjunction with the plant size and the individual equipment capital costs
presented in Figure 7-3.
The capital costs for each component then were summed and an allowance
was added to the installed capital costs to account for indirect costs,
such as engineering fees, construction overhead, and contingencies.38'39
The magnitude of these indirect costs can vary widely and is dependent upon
the size, type, and complexity of jobs under consideration. In addition,
the contingencies are related directly to the firmness of the cost estimate.
Therefore, the absolute magnitude of the indirect costs used by different
estimates can vary widely.
7-2.4.3 PI ant-by-PI ant Annualized Costs and Savings
The pi ant-by-plant annualized costs for each regulatory option were
obtained by summing the individual operation and maintenance costs asso-
ciated with each piece of equipment shown in Table 7-14 and then adding
the annualized capital cost for the control equipment. Any recovery credits
that could be assigned for either recovery of heating value or recovery of
raw material then were estimated and subtracted from the annualized costs
to yield net annualized costs or savings. In several instances, the recovery
credits far exceeded the annualized cost of the control device under con-
sideration. However, it must be noted that-the amount of savings that
could be realized by a given regulatory option will depend upon the actual
amount of recoverable material in the controlled stream. For the purpose
of this analysis, it was assumed that benzene was the only material of
7-39
-------
TABLE 7-15. PLANT-BY-PLANT COSTS
Plant & Location
Arco, Port Arthur, TX
American Hoechst
Cos-Mar, LA
Gulf Chemicals, TX
Monsanto
Amoco
Dow, TX
Sun Oil, TX
El Paso, TX
Oxirane
Arco, PA
U.S. Steel
Dow, Midland
Option %
85
94 Flare
99 Boiler
85
94 Flare
99 Boiler
85
94 Flare
99 Boiler
85
94 Flare
99 Boiler
85
94 Flare
99 Boiler
85
94 Flare
99 Boiler
85
94 Flare
99 Boiler
85
94 Flare
99 Boiler
85
94 Flare
99 Boiler
85
94 Flare
99 Boiler
85
94 Flare
99 Boiler
85
94 Flare
99 Boiler
85
94 Flare
99 Boiler
Capital Cost
(in $1,000)
121
290
235
164
316
421
268
530
448
130
258
215
66
325
424
72
213
200
142
387
555
66
173
130
33
33
130
243
243
366
74
180
130
No additional
61
173
142
Annual i zed Cost
(Savings)
(40)
6.5
(14)
(46)
(45)
(72)
(167.5)
(96)
(147)
(75)
(40)
(64.5)
(32)
39
(76.5)
(28)
13
7
(78)
(28)
(75)
13.5
45
23
8.5
8
26
(172)
(172)
(119)
9.5
38
27
equipment required.
(1)
30
25
7-40
-------
value in the vent stream and no recovery credit was given for the other
aromatics contained in the vent gases for the model plant. The recovery
credits are based upon an average vent stream analysis and recovery occur-
ring 8,000 hrs/yr. The composition of these vent streams will vary from
plant to plant and, therefore, the absolute magnitude of the recovery
credits also will vary.
Table 7-16 indicates that all three options for controlling the con-
tinuous vent streams can result in net revenues for some of the EB/S facil-
ities. The magnitude of the savings is very dependent upon the amount of
benzene recovered and the value of the recovered benzene as a feed to
process, as well as the heating value of the other organics in the vent
stream.
TABLE 7-16. ESTIMATED INDUSTRY-WIDE COST-EFFECTIVENESS FOR
EACH REGULATORY OPTION
Option
85 Percent
94 Percent
99 Percent
Annuali zed Cost
(in $1,000)
(Savings)
$(608.0)
$(201.5)
$(460.0)
Benzene
Removed (Mg)
1,225
1,588
1,810
Cost-Effectiveness
Dollars per Mg Benzene
(Savings)
$(496)
$(127)
$(254)
This analysis assumed that the benzene which is recovered in the
condenser and scrubber system can be fed back into the process via the
benzene drying column and that the price of the benzene recovered is the
same as its fair market value.
The analysis also assumed that the VOC and benzene that is recovered
and used as a fuel in a boiler can be assigned a value comparable with that
of natural gas ($2.00/MMBtu).
7-41
-------
7.2.5 Cost Comparison and Cost-Effectiveness of the Regulatory Options
7.2.5.1 Costs to Achieve 85 Percent Emissions Reduction
Table 7-16 indicates that, in general, the 85 percent regulatory
option requires the smallest capital outlay of any of the three main regula-
tory options considered. Only one of the 13 facilities under consideration
will not have to install any additional equipment to achieve the 85 percent
emissions reduction. Capital costs of this option range from zero to
$268,000. Total annualized costs of this option range from $13,500 to a
savings of $172,000. The range is due to differences in the treatment of
the various vent streams in each of the facilities. Some facilities already
are controlling those streams which are close to saturation with benzene
and which are cost-effective to treat.
7.2.5.2 Costs to Achieve 94 Percent Emissions Reduction
In most cases, the combined scrubber/condenser/flare system requires
the largest capital outlay of any of the three options under consideration.
This is due to the large amount of piping and equipment required in addi-
tion to the flare system (pumps, PEB chillers, scrubbers, condensers,
compressors, etc.). One of the 13 facilities can achieve a 94 percent
emissions reduction without any capital expenditure. Capital costs for
this option range from zero to $530,000. Total annualized costs of this
option range from $45,000 to a savings of $172,000. It should be noted
that the plant requiring the largest capital expenditure (Cos-Mar) does not
have the highest annualized costs; in fact, Cos-Mar exhibits an annual
savings of $96,000 under this option. The large range in annualized costs
is due to the assumptions regarding the existing degree of control and the
amounts of benzene that can be recovered for use as a raw material. Some
7-42
-------
plants are practicing a high degree of product recovery from the vent
streams which are comparatively easy to control. A large capital expen-
diture is still needed to achieve 94 percent benzene emissions reduction,
but very few recovery credits are realized.
7.2.5.3 Costs to Achieve 99 Percent Emissions Reduction
Capital costs for this option range from zero to $555,000. Only one
plant has existing control equipment in place which is capable of achieving
this level of emissions reduction. Annualized costs of this option range
from $26,000 to a savings of $150,000. This range is attributable to the
amount of fuel recovery credits assigned to each facility.
7.2.5.4 Cost-Effectiveness of Regulatory Options
The cost-effectiveness of a given regulatory option is defined as the
quotient of the total annualized cost of the option and the amount of
benzene recovered by the option. This ratio usually is expressed in units
of dollars per megagram of benzene recovered. Table 7-16 presents the
total annualized cost (or savings), the amount of benzene recovered, and
the cost-effectiveness of each regulatory option considered, on a national
basis. Because all of the existing facilities practice some degree of
control, the amount of benzene recovered by each of the emission options
cannot be derived simply from the uncontrolled emissions levels shown in
Table 7-12 and the percentages emission reduction specified in the option.
Instead, the quantity of benzene recovered is obtained from the incremental
level of control between existing levels of control at each facility and
the level specified by each of the proposed options.
In addition, the*variability in cost-effectiveness between plants for
each regulatory option makes it difficult to determine which option is the
7-43
-------
most cost-effective in aggregate. For this reason, the cost-effectiveness
for each option is presented on a national basis.
7.2.6 Costs for Continuous .Monitoring
The costs for continuously monitoring benzene emissions are the same
for each EB/S plant. The capital costs are based on two gas chromatographs
(GC) to measure benzene emissions from a superheater and a boiler. The
installed capital cost for the two gas chromatographs is estimated to be
$100,000. The total annualized cost of the system is estimated to be
$53,000. This includes annualized capital at 16.28 percent, maintenance
material, 2.5 percent of installed capital, taxes, insurance, and miscel-
laneous at four percent, and operating labor at $30,000 per annum. The
industry-wide annualized cost of monitoring would be $689,000 for the 13
plants for which costs have been derived.
7-44
-------
7.3 OTHER COST CONSIDERATIONS
7.3.1 Introduction
This section presents an estimate of the costs that typical EB/S •
facilities have to assume in order to comply with other environmental
regulations. Currently, in addition to air pollution regulations under
consideration, EB/S facilities are faced with compliance with water
pollution control regulations and Occupational Safety and Health Admin-
istration (OSHA) regulations. In addition, in the near future EB/S
facilities will have to comply with regulations concerning control of
emissions from storage and handling and fugitive sources.
7.3.2 Water Pollution Control
EB/S facilities are required by the Federal Water Pollution Control
Act and Amendments PL 92-500 to apply Best Practicable Control Technology
Currently Available (BPCTCA) as of July 1, 1977. In addition, EB/S
facilities are required to achieve Best Available Technology Economically
Achievable (BATEA) by 1983.
Pursuant to the Act, EPA prepared a document entitled "Development
of Effluent Limitations Guidelines and New Source Performance Standards
for the Major Organic Products Segment of the Organic Chemicals Manufac-
turing Point Source Category." The costs presented in this section were
taken from this document. The base document presented a range of
expected annualized costs for each industrial subcategory.40
Due to the wide variations in flow and raw waste loading attributable
to each industry in each subcategory, no specific costs for compliance
with PL 92-500 were identified for the EB/S industry.
7-45
-------
The compliance costs have been determined with respect to the following
three levels of technology:
• Best Practicable Control Technology Currently Available (BPCTCA) (1977
standard).
• Best Available Technology Economically Achievable (BATEA) (1983
standard).
• Best Available Demonstrated Control Technology (BADCT) (New Source
Performance Standard).
The costs associated with these technologies have been estimated based
on model systems which are considered capable of attaining the reduction
factors associated with each technology. The particular systems chosen for
use in the cost models are not the only systems which are capable of attaining
the specified pollutant reductions.
Activated sludge was used as the BPCTCA model treatment system.
Dual-media filtration and activated carbon adsorption were used as BATEA.
New source end-of-process treatment (BADCT) involves the addition of dual-media
filtration to the biological waste treatment model processes.
The general effect of the control techniques is to reduce both the
pollutant raw waste load (RWL) and the volume of contact process water
discharged for end-of-pipe treatment. It is not possible to recommend a
list of process modifications or control measures and associated costs
which clearly are applicable to the EB/S industry since the data in Reference
33 are for a group of processes into which the EB/S process falls.
The estimated annualized cost for compliance with water pollution
control requirements for the 300,000 Mg/yr (330,000 tons/yr) model plant
has been estimated to be:
• For BPCTCA, $130,000 to $810,000.
7-46
-------
• For BADCT, $165,000 to $1,100,000.
• For BATEA, $140,000 to $870,000.
These costs are in 1978 dollars based on capital recovery at eight percent,
operations computed at two percent of capital, and maintenance computed at
four percent of capital.
This cost information is presented in ranges to reflect the variations
in flow and RWL given for the process types described in Reference 40.
7.3.3 Storage and Handling Control
EB/S facilities may be required by standards currently under develop-
ment to control benzene emissions resulting from benzene storage and
handling. For the sake of this analysis, it is assumed that the model
plant (300,000 Mg/yr) will have two benzene storage tanks of the following
sizes:
Tank Large Small
12.2
Height (m)
Diameter (m)
12.2
31.0 12.2
Capacity (nO 10,000.0 15,000.0
Also, each plant is assumed to have a benzene/toluene tank with the same
characteristics as the small benzene tank. These tank sizes are based on
data provided in "Emission Control Options for the Synthetic Organic .
Chemicals Manufacturing Industry, Storage and Handling Report," by Hydro-
41
science, Inc. The number of tanks at each plant was based on trip reports
4?
and State EIQ's. Also, it is assumed that for the uncontrolled model
plant, the large tank is fitted with an external floating roof with a
single seal and the small tank has a fixed roof.
7-47
-------
In order to comply with the regulations, the small tank will have an
43 44
internal floater and the large tank will have a second seal. ' Table
7-17 shows the costs associated with these compliance activities.
7.3.4 Fugitive Emission Control
EB/S facilities may be required by standards currently under develop-
ment to control fugitive benzene emissions. The draft standards currently
require a leak detection and repair program for pipeline, valves, and pro-
cess drains in benzene service and would require certain equipment, such as
double mechanical seals, for pumps, compressors, sampling connections, and
open-ended valves in benzene service. The estimated capital and annualized
costs for meeting these standards could range from $57,000 to $253,000 and
45
from $14,000 to $50,000, respectively, for an existing facility.
7.3.5 Compliance With OSHA Regulations
OSHA is charged with the responsibility for developing, promulgating,
and enforcing regulations to protect workers against the hazards of toxic
materials found in the work place. To this end, the proposed regulation
specifies permissible exposure of one ppm in any eight-hour time period,
with a ceiling of five ppm averaged over 15 minutes, and additional require-
ments which are specified in the proposed regulation.
Probable compliance activities have been identified in References 46
and 47 for the EB/S industry and the costs of these activities have been
calculated. The computation for costs of compliance under the proposed
regulation recognizes only the incremental costs of moving from compliance
with the existing benzene regulation (29 CFR 1910.1000) to compliance with
the new regulation.
7-48
-------
TJ
a>
N
03 O
3 O
C O
c
«=C r-
•tfl-
"o3
4J
O
co
C_5
-O
C O CU
C U S-
SC CD O
O
O
O
to
O
O
03 O)
3 C
C T~
C +3
•r—
4-
O
S-
-P
CU
S-
^
c
03
•p
<4_
o
o
i-
"O
ai
X
•f—
u_
O
O
£«
•a
r~
cu
cu
5-
u
c
HH
O3 _Q
,
^E
CU
f"
1
1
v^
C
03
t^—
0
o
O5
C
•i— .
j ^
<0
0
4—
a.
o
-P
c
cu
a.
o
c
o
•a
cu
r-—
ta
•P
to
c
•r—
I/I
r—
03
CU
to
^•j
£_
03
C
O
u
cu
to
+j
ijl
0
s_
-p
cu
S-
1
1
to
VX
C
03
•P
tf_
O
o
s-
D3
C
•i—
-P
03
O
(,_
a.
0
i
£=
CU
o.
o
c
o
•a
cu
r—
03
V>
C
•r—
in
03
cu
to
>^
£_
03
C
o
u
cu
to
o -a
7-49
-------
There are alternative methods available to employers for complying
with the proposed OSHA regulation. However, under the scope of this analysis,
only the least cost method of compliance is attributed to the proposed
regulation.
Operating costs for compliance during the first year are those associated
with exposure measurements, respirators, personal controls, recordkeeping,
medical surveillance, employee training and information, and signs and
labels.46 For the model plant, this costs $18,200. At this level, the
total EB/S industry costs will be $308,000. Recurring costs are similar in
nature but reflect reduced usage of respirators and a less vigorous monitoring
program. These costs have been estimated to be $10,200/yr for the model
plant and $173,000/yr for the total EB/S industry. These costs are based
on the assumption that at the model EB/S plant there are 32 persons exposed
to benzene and that industry-wide there are 540 persons exposed to benzene.
47
Additional basis for the costs of compliance are shown in Table 7-18.
The capital costs for compliance with the OSHA regulations are those
to rebuild both pumps and compressors, install rupture discs or similar
devices in vent systems, improve maintenance of pipeline valves, improve
procedures for transferring and storing benzene, install automatic fill
monitoring devices, and initiate procedures for closed system sampling.
These capital costs are included as part of the costs of compliance with
the fugitive and storage and handling regulations and are not included as
costs of compliance with OSHA regulations.
7-50
-------
TABLE 7-18. COST BASIS FOR CALCULATIONS OF OPERATING COSTS
Exposure Measurements
Measurement Equipment
Pump, Sipin, Model
SP2, 20-100 cc/min
Charger
Calibration Kit
Measurements
Charcoal tube
Analysis
Notification
Medical Surveillance
Initial Exam
Incremental Exam
Respirators
Half-facepiece chemical cartridge
respirator without cartridge
(MSA Cat. No. 460968) Comfo Respirator
GMA cartridge for Comfo
Respirator (MSA Cat. No. 459315)
Full facepiece chemical
cartridge respirator without
cartridge
Personal Protection Equipment
Gloves, Apron, Faceshield, Goggles
Training Costs
Preparation
Training
Signs & Labels
Labels
Signs
Labor
Clerical
Employer
Employee
Costs
$328/facility
$ 20/facility
$ 98/facility
$.71/charcoal tube
$ 35/charcoal tube
$ 2/measurement
$ 68/exam
$ 48/exam
$8.85/respi rator
$2.04/cartridge
$63.00/respirator
$13.16/employee
$110/facility
$ 14/employee
$14.75/100 labels
$22.50/signs
$ 8/hr
$20/hr
$10/hr
7-51
-------
7.4 ECONOMIC IMPACT ANALYSIS
Four control strategies exist for reducing benzene emissions in the
EB/S industry. Three options, the 85 percent scrubber/condenser system, the
94 percent flare, and the 99 percent boiler regulatory option, represent
control levels which can be achieved through existing technologies. The
potential impacts of these control strategies will be analyzed in Sec-
tions 7.4.1 through 7.4.3. The analysis, which is based on plant-
by-plant costs presented in Section 7.2, highlights differential
impacts across producing firms. It is recognized that differentials in
production economies exist for these firms. Differences in fuel, feedstock,
and transportation costs per unit of output can be sizable. However, the
focus in this report is on differential impacts of controls, since these.
impacts may either mitigate or exacerbate production cost differentials.
The fourth regulatory option, 100 percent control, cannot be achieved
through any existing technology. Such control would require the closure of
all styrene plants. The potential impact of this option is analyzed in
Section 7.4.4.
7.4.1 Control Costs and Feasibility
7.4.1.1 Costs of Control by Regulatory Option
Tables 7-19, 7-20, and 7-21 present capital and total annualized costs
of control by firm for the 85, the 94, and the 99 percent options, respec-
tively. Cost estimates vary considerably across firms. Capital costs per
firm range from lows of zero (under all options) to highs of $268,000,
$560,000, and $697,000 under the 85, 94, and 99 percent controls, respec-
tively. While Dow Chemical experiences the largest capital costs, its
capital costs per Mg of capacity are moderate. Comparable estimates for Sun
7-52
-------
TABLE 7-19. COST SUMMARY: 85 PERCENT REGULATORY OPTION
Producing
Company
American Hoechst Corp.
Amoco
Atlantic Richfield Co.
(2 locations)
Cos-Mar, Inc.
Dow Chemical, USA
(2 locations)
El Paso Products Co.
Gulf Oil Corporation
Monsanto Company
Oxirane Corporation
Sun Oil Company
United States Steel
Total Industry
Capital
Cost
($1,000)
164
72
195
268
203
33
130
66
243
66
0
1,440
Capital Cost
per Mg
Capacity9
($/Mg)
0.36
0.19
0.98
0.45
0.24
0.29
0.48
0.10
0.54
1.83
0
0.35
Total
Annual i zed
Cost
($1,000)
-46.0
-28.0
-30.5
-167.5
-79.0
8.5
-75.0
-32.0
-172.0
13.5
0
-608.0
Total Annual i zed
Cost per Mg
Capacity9
($/Mg)
-0.10
-0.07
-0.15
-0.28
-0.09
0.07
-0.28
-0.05
-0.38
0.38
0
-0.15
Styrene capacity given in Table 7-1
7-53
-------
TABLE 7-20. COST SUMMARY: 94 PERCENT REGULATORY OPTION
Producing
Company
American Hoechst Corp.
Amoco
Atlantic Richfield Co.
(2 locations)
Cos-Mar, Inc.
Dow Chemical , USA
(2 locations)
El Paso Products Co.
Gulf Oil Corporation
Monsanto Company
Oxirane Corporation
Sun Oil Company
United States Steel
Total Industry
^_-_=r- - — . -— - •
aStyrene capacity given
Capital
Cost
($1,000)
316
213
470
530
560
33
258
325
243
173
0
3,121
=======
in Table 7-1.
Capital Cost
per Mg
Capacity9
($/Mg)
0.70
0.56
2.35
0.90
0.65
0.29
0.95
0.48
0.54
4.81
0
0.76
_...'-
Total
Annuali zed
Cost
($1,000)
-45.0
13.0
44.5
-96.0
2.0
8.0
-40.0
39.0
-172.0
45.0
0
-201.5
Total Annuali zed
Cost per Mg
Capacity3
($/Mg)
-0.10
. 0.03
0.22
-0.16
0.002
0.07
-0.15
0.06
-0.38
1.25
0
-0.05
• • ' "... • ~
7-54
-------
TABLE 7-21. COST SUMMARY: 99 PERCENT REGULATORY OPTION
Capital Cost Total
Total Annualized
Producing
Company
American Hoechst Corp.
Amoco
Atlantic Richfield Co.
(2 locations)
Cos-Mar, Inc.
Dow Chemical, USA
(2 locations)
El Paso Products Co.
Gulf Oil Corporation
Monsanto Company
Oxirane Corporation
Sun Oil Company
United States Steel
Total Industry
Capital
Cost
($1,000)
421
200
365
448
697
130
215
424
366
130
0
3,396
per Mg
Capacity9
($/Mg)
0.94
0.53
1.82
0.76
0.81
1.13
0.79
0.62
0.81
3.61
0
0.83
Annual i zed
Cost
($1,000)
-72.0
7.0
13.0
-147.0
-50.0
26.0
-64.5
-76.5
-119.0
23.0
0
-460.0
Cost per Mg
Capacity9
($/Mg)
-0.16
0.02
0.06
-0.25
-0.06
0.23
-0.24
-0.11
-0.26
0.64
0
-o.n
Table 7-1
7-55
-------
011 Company are somewhat larger, approximately twice that of the next
highest company estimate. Capital cost per Mg is also fairly high for
Atlantic Richfield. Both capital and capital cost per Mg are lowest (at
zero) for United States Steel.
Industry-wide total annualized costs are actually net credits of
$608,000, $202,000, and $460,000 for the 85, 94, and 99 percent options,
respectively. However, under all options some companies experience net
costs, not net credits. Two firms face positive annualized costs under the
85 percent regulatory option. Six firms experience net costs under the 94
percent control as do four companies under the 99 percent option. The
maximum value of such costs is $45,000 for Sun Oil Company under the 94
percent regulatory option.
7.4.1.2 Feasibility of Financing
There are several reasons for believing that EB/S producers should have
little difficulty meeting the capital requirements of proposed controls.
Note that representative company capital expenditures shown in Table 7-22
are very much larger than the capital costs of control shown in Table 7-19
through Table 7-21. Control expenditures are well within the scope of
normal spending.
Moreover, financial information presented earlier (Table 7-9) indicates
that parent companies should have little difficulty with either debt or
equity financing. Debt-equity ratios are sufficiently low that borrowing
should present no problems. Rates of return on equity are sufficiently high
that equity financing should not be difficult either.
Whether the affected companies will choose to finance controls and
continue operating is a more difficult question. The choice to finance will
depend critically upon the age of the plant in question, the impact of
7-56
-------
TABLE 7-22. REPRESENTATIVE CAPITAL EXPENDITURES
28
Company
American Petrofina
(joint owner, Cos-Mar)
Amoco
Atlantic Richfield
Dow Chemical , USA
El Paso Products
Year
1977
1977
1977
1977
1975
Expenditure
(in $1,000)
60,189
269,839
278,000
1,163,016
445,081
Comment
Capital expenditures and
property acquisitions; all
lines of business
Capital expenditures in
Chemical Operations
Capital expenditures in
Chemical Operations
Capital expenditures; all
lines of business
New property, plant and
Gulf Oil
Monsanto
United States Steel
1977
1977
1977
equipment; all lines of
business
174,000 Capital expenditures in
Chemical Operations
607,100 Capital expenditures; all
lines of business
864,700 Capital expenditures; all
lines of business
control costs on plant and company profitability, and the other investment
options available to the company. As all EB/S producers are diversified
companies, investment alternatives may be very important. While it is not
poss.ible to make any definitive statements with regard to investment alter-
natives, it is possible to assess the impacts of controls on company profit-
ability under some assumptions with regard to plant age. These impacts will
be addressed in the analysis which follows.
7-57
-------
7.4.2 Economic Impact Methodology
7.4.2.1 Pricing Scenarios
The impacts of proposed emission control regulations on individual
firms and on the economy as a whole will depend upon the degree to which
cost increases are accompanied by price increases. Consequently, impacts
are analyzed under two alternative firm pricing assumptions. These assump-
tions are full cost absorption and full cost pricing.
Under full cost absorption, an affected firm bears the entire incre-
mental cost of emission controls. Output price does not increase and control
costs cannot be passed back onto resource suppliers. The result is a lower
rate of return on investment. This scenario generally is associated with
strong market competition. Excess capacity in the industry, the existence
of close product substitutes, or the presence of strong international compe-
tition can result in full cost absorption. This scenario generates maximum
impacts for individual companies and minimum (zero) inflationary impacts.
Under full cost pricing, the affected firm sets a new price so as to
maintain a target rate of return on investment. Again, the firm is assumed
to be unable to pass costs back to resource suppliers. However, the firm
can pass costs forward. Higher prices will be accompanied by a smaller
quantity sold and consequently lower industry-wide output and employment.
Full cost pricing generally is associated with a constant cost industry
and/or minimal competition from other products or other producing nations.
Full cost pricing generates maximal inflationary impacts and minimal
individual firm impacts.
Full cost absorption and full cost pricing are intended to represent
extreme situations which may occur; either is actually possible in the EB/S
7-58
-------
industry. The existence of excess capacity would mitigate against cost pass
through (full cost pricing) currently. But to the degree that higher operating
rates are achieved over the next few years, pass through will be more feasible.
Some styrene substitutes exist, but they are not perfect substitutes.
Styrene can be imported, but transportation costs and higher foreign feed-
stock and energy costs exist. The result is some room for price increases.
However, it is clear that increases will be limited by the threat of product
substitution and/or importation.
Thus, depending upon capacity utilization and the availability of
substitutes, actual impacts may range anywhere from full cost absorption to
full cost pricing. The price, output, and employment impacts generated
under full cost pricing and the profitability impacts generated under full
cost absorption are maximum likely changes. All maximum impacts do not
occur simultaneously. Any price increase means a smaller impact on profit-
ability, just as any cost absorption means a smaller impact on price, output,
and employment.
7.4.2.2 Economic Conditions
As noted previously, an industry's ability to pass forward costs depends
upon capacity utilization. Excess capacity, which inhibits cost pass through,
currently exists in the EB/S industry. Consequently, all impacts are assessed
under two scenarios: 76 percent capacity utilization (the current level)
and 100 percent utilization. Price changes calculated with the 100 percent
utilization assumption have a higher probability of occurrence than do those
calculated with the 76 percent utilization assumption. Rate of return
impacts (with price unchanged) are most likely with the 76 percent utiliza-
tion scenario.
7-59
-------
Impacts also are estimated with two different target rates of return:
six percent (as a lower bound) and 15 percent (as a realistic rate). As
impacts are larger in magnitude under the 15 percent target, only these
results are presented.
All impacts are based on the same assumptions with regard to product
price ($463 per Mg), tax rate (46 percent), and equipment life (ten years).
Note that an equipment life of ten years corresponds to the capital recovery
factor used to calculate annualized costs in Section 7.2. If equipment life
is actually greater than ten years, impacts will be overestimated. Since
equipment life is not likely to be less than ten years, impact estimates can
be regarded as worst case calculations.
7.4.2.3 Estimation Procedure: Simple Rate of Return Impacts Under Full
Cost Absorption
Under full cost absorption it is assumed that firms do not pass cost
changes into price changes. Consequently, there are no effects on output or
employment. Instead, cost changes result in changes.in the rate of return
on investment. Assuming simple rate of return analysis, the impact on the
firms' rate of return on investment is given by the following equation:
A _ r - AK + (1-t) ATOC
K + AK
(2)
where
Ar = change in rate of return on investment
ATOC = total annual operating costs of .the control equipment
AK = total costs (acquisition and installation) of the control
equipment
K = precontrol level of capital investment
r = target rate of return
t = tax rate
7-60
-------
Note that total annual operating costs (ATOC) include an allowance for
the depreciation of control equipment in addition to regular operating ex-
penses. Annual depreciation is estimated to be ten percent of total acquisi-
tion and installation costs based on an assumption of straight-line depreci-
ation over a ten-year life. Note also that this formulation can be used to
estimate impacts either on the rate of return to an individual plant or on
the rate of return to firm equity, with the value of K determined accordingly.
This analysis will focus on simple rates of return to equity. An
equity base was chosen for two reasons. Rates of return to specific existing
plants require more detailed information on specific plant values than is
immediately available. Approximations of these values could have been made,
but they would have been very rough. In addition, it is not clear that
rates of return to specific plants are preferable on theoretical grounds
when applied to an integrated operation. As noted previously, many affected
firms produce their own feedstocks and/or styrene derivatives. Theoretically,
base rates of return and impacts should be calculated on the entire extended
operation, not just on the individual styrene plant. However, this requires
a great deal of specific information which is not available. Rates of
return on equity are taken as a proxy for rates of return to extended (ver-
tically integrated) operations. Given the lack of data and the nature of
the estimate (as a proxy), it did not seem valid to apply a more sophisti-
cated methodology than a simple rate of return analysis.
7-4.2.4 Estimation Procedure: Price Impacts Under Full Cost Pricing
Under full cost pricing, the firm is assumed to react to cost changes
by adjusting output price in order to maintain some specific target rate of
7-61
-------
return on investment. The required price change (AP) is given by the following
equation:
AP = ATOC + r • AK/(1-t) (3)
where AP denotes the required product price change, Q denotes the firm's
production capacity, and C refers to capacity utilization. (All other
variables are defined in the preceding section.)
Capacity utilization is assumed to be unchanged by the imposition of
controls. But a price increase results in a smaller quantity demanded from
the industry. Obviously, not all companies will be able to maintain capacity
utilization. The highest cost company may experience the entire output
impact or various companies may share the impact. A sharing of the impact
cannot be analyzed without a much more complex model.
7.4.2.5 Other Economic Impacts
The price and rate of return impacts estimated in accordance with the
preceding methodology are used to make quantitative and qualitative assess-
ments of additional economic impacts. Maximal price changes together with
elasticity estimates (Section 7.1) can be used to estimate maximal output
and hence employment impacts. Rate of return impacts potentially affect new
investment in the industry. These impacts also are considered.
7.4.2.6 Data
Total costs of control equipment (AK) and total annual operating costs
(ATOC) are derived from information in Section 7.2. For rate of return
impacts, equity values (K) are obtained from Securities and Exchange Commis-
sion 10K Forms., Production capacities (Q) are taken from the 1978 Directory
of Chemical Producers (SRI International). Capacity utilization is set at
either 76 percent (the current estimate) or 100 percent (for full pass
through).
7-62
-------
7.4.3 Economic Impacts
7.4.3.1 Rate of Return Impacts
Table 7-23 presents estimates of impacts on rates of return on equity
under full cost absorption for the 85, 94, and the 99 percent regulatory
options, respectively. Note that while all estimated impacts are small,
they may be either negative (decreasing the rate of return) or positive
(increasing the rate of return). Positive impacts can occur because controls
sometimes result in operating cost savings (negative ATOC). If operating
cost savings are sufficiently large relative to capital expenditures, the
net impact on the rate of return will be positive. Note that although the
procedure for calculating rate of return impacts combines operating and
capital costs in a somewhat different manner, the same principle which
results in negative total annualized costs results in positive rate of
return impacts.
All rate of return impacts are small. Under the 85 percent regulatory
option, the maximum negative impact is -0.000011 (-0.01 percent of the 15
percent target) for El Paso Products Company, while the average of firm
impacts is 0.000013 (0.01 percent of the target). The maximum positive
impact is 0.000111 (0.07 percent of the target) for American Petrofina.
Under the 94 percent regulatory option, the maximum negative impact is
-0.000020 (0.01 percent of target) for Monsanto. The average of firm
impacts is -0.000007 (0.005 percent of target). Only Borg-Warner experiences
a positive impact on rate of return (0.000001). Under the 99 percent
regulatory option, the maximum negative impact is -0.000038 (0.03 percent
of target) for El Paso Products while the maximum positive impact is 0.000063
for American Petrofina. The average of firm impacts is -0.000005 (0.005
percent of the target).
7-63
-------
In terms of both averages and maximum impacts it is clear that the
94 percent regulatory option produces the largest impacts on rates of
return. While these impacts average less than 0.01 percent of the target
rate of return, it is clear that impacts differ across firms. Borg-Warner
shows small positive impacts under all regulatory options. Gulf Oil
experiences relatively small impacts under all regulatory options. El Paso
Products sustains relatively large impacts under all regulatory options.
However, these impacts are large only in a relative sense and represent a
maximum of 0.04 percent of the target rate of return. Substantial impacts
on the industry are not anticipated.
7.4..S.2 Price Impacts
Tables 7-24 through 7-26 present price changes required to maintain.a
15 percent target rate of return under three regulatory options and two
capacity utilization scenarios. As with rate of return impacts, price
changes required by individual firms may be either positive or negative.
The direction of the change will depend upon the interplay of capital costs
and operating cost savings.
While all indicated price changes are small in magnitude and percentage
terms, those required under 76 percent utilization (the current figure) are
the largest. These required price changes are estimated specifically to
obtain an indication of maximum impacts. However, one simply cannot take
the maximum of these required price changes (for any control option) and
consider that change to be a reasonable expected price increase. Such a
figure would overestimate the degree to which costs can be passed forward
in the presence of excess capacity.
7-64
-------
•Vs.
Q£ Q£ *****
O O &S
n^ r** \^
"o
-P
C
0
0
Q£
^^ CD
CD Q£
cn o
re ex. a: <->
to o o 53
O r— <] ^
- 0
o- cn <
UJ
o
"Z.
ce.
3 -v,
h- as a: /•-.
£ 'gg^S
•<1
]Jn
C.3 r-"
o
UJ i-
t— * i ^
Q£ O
O O£.
O
• &? as
co in o
CM co
i
• **
1 1 1
i
O
O
O
O
, O
0
d
CO
o
o
o
o
o
1
o
O :
I—~
0
0
o
'~"N
re
i
in
re o
C 0
Petrofi
owner,
C -P
re c
"re
u
•r—
E
CU
J=
0
3:
o
a
CO
o
,
0
' '
00
co
o
o
CD
0
1
o
d
1
o
o
o
o
1
o
d
i
o
o
o
o
1
.
0
o
roducts
a.
o
in
re
a.
P— .
UJ
o
o
f
CD
CM
O
O
0
O
o
o
o
d
. 000000
1
o
o
d
. 000004
c
o
-p
re
o
e-
o
o
t_
CD
3
C3
O
0
d
1
o
CD
O
O
o
CD
1
O
d
, 000020
i
o
o
d
•=1-
0
0
o
o
o
%
re
QL
o
CJ
o
•p
c
re
in
c
o
r_
o
d
00
o
o
o
o
o
,
1 —
o
d
000014
1
o
o
o
1
. 000005
i
re
a.
o
o
0
c
oo
o
o
d
o
o
o
o
o
d
o
o
CD
o
o
d
o
o
d
cu
cu
cu
-p
re
-P
"O
cu
•p
c
re
re
~°
P_
re
u
re
c
4-
_£;
U
J=
•*
S-
o
cu
• f— '
re
a.
u
, ^
{=
cu
re •
Q.-O
cu
in 3
O in
•p re
O •!-
c.
"O t-
OJ Z3
-P -P
re cu
*r*"
j_ji ji
in o
cu
cu
CO -P
s- re
re s~
>>-P
•p cu
•i— O)
3 i-
cr re
CU -P
c: +J
0 C
cu
c u
S- i.
3 CU
-P 0-
a>
s- in
4-
CD ^C
in
cu .
-p cu
re r—
J- JD
re
O •!-
re
in >
•p re
u
re cu
a. s_
E cu
re
7-65
-------
TABLE 7-24. REQUIRED STYRENE PRICE CHANGES: 85 PERCENT CONTROL*
76% Capacity
Utilization
Producing Company
American Hoechst Corp.
Amoco
Atlantic Richfield Co.
(2 locations)
Cos-Mar, Inc.
Dow Chemical, USA
El Paso Products Co.
Gulf Oil Corporation
Monsanto Company
Oxirane Corporation
Sun Oil Company
United States Steel
AP
($/Mg)
-0.05
-0.05
0.02
-0.27
-0.07
0.16
-0.25
-0.04
-0.38
0.91
0.00
AP/Pb
(%)
-0.01
-0.01
0.00
-0.06
-0.02
0.03
-0.05
-0.01
-0.08
0.20
0.00
100% Capacity
Utilization
AP
($/Mg)
-0.04
-0.04
0.02
-0.20
-0.05
0.12
-0.19
-0.03
-0.29
0.70
0.00
AP/Pb
(%)
-0.01
-0.01
0.00
-0.04
-0.01
0.03
-0.04
-0.01
0.06
0.15
0.00
aPrice changes required to maintain a 15 percent rate of return are presented.
bPercentage price change calculations assume a base price of $463 per Mg.
7-66
-------
TABLE 7-25. REQUIRED SYTRENE PRICE CHANGES: 94 PERCENT CONTROL6
76% Capacity
Utilization
Producing Company
American Hoeschst Corp.
Amoco
Atlantic Richfield Co.
(2 locations)
Cos-Mar, Inc.
Dow Chemical , USA
(2 locations)
El Paso Products Co.
Gulf Oil Corporation
Monsanto Company
Oxirane Corporation
Sun Oil Company
United States Steel
AP
($/Mg)
0.03
0.17
0.83
0.01
0.15
0.16
0.02
0.19
-0.38
2.82
0.00
AP/Pb
(%)
0.01
0.04
0.18
0.00
0.03
0.03
0.00
0.04
-0.08
0.61
0.00
100% Capacity
Utilization
AP
($/Mg)
0.02
0.13
0.63
0.01
0.12
0.12
0.02
0.14
-0.29
2.14
0.00
AP/Pb
(%)
0.00
0.03
0.14
0.00
0.03
0.03
0.00
0.03
-0. 06
0.46
0.00
an . ,. . . .
presented.
Percentage price change calculations assume a base price of $463 per Mg.
7-67
-------
TABLE 7-26. REQUIRED STYRENE PRICE CHANGES: 99 PERCENT CONTROL0
76% Capacity
Utilization
Producing Company
American Hoeschst Corp.
Amoco
Atlantic Richfield Co.
(2 locations)
Cos-Mar, Inc.
Dow Chemical, USA
(2 locations)
El Paso Products Co.
Gulf Oil Corporation
Monsanto Company
Oxirane Corporation
Sun Oil Company
United States Steel
AP
($/Mg)
0.00
0.15
0.51
-0.15
0.11
0.56
-0.13
0.00
-0.16
1.67
0.00
AP/Pb
(%)
-0.00
0.03
o.n
-0.03
0.02
0.12
-0.03
0.00
-0.03
0.36
0.00
100% Capacity
Utilization
AP
($/Mg)
0.00
0.11
0.38
-0.12
0.08
0.42
-0.10
0.00
, -0.12
1.27
0.00
AP/Pb
(%)
0.00
0.02
0.08
-0.03
0.02
0.09
-0.02
0.00
-0.03
0.27
0.00
Price changes required to maintain a 15 percent rate of return are
presented.
""Percentage price change calculations assume a base price of $463 per Mg.
7-68
-------
Consider the following example, which assumes implementation of the 85
percent regulatory option. The largest required price change under this
option is $0.91/Mg for Sun Oil Company. Since no other company requires-
this large a price increase, the operative question becomes whether Sun Oil
could implement this increase and still sell its product. At 76 percent
utilization, current domestic styrene capacity of 4,092,000 Mg/year corre-
sponds to production of 3,110,000 Mg/year. This production level could be
met without any output from Sun Oil's 36,000 Mg/year plant. If the other
producers chose not to match Sun Oil's new price, styrene would not sell at
that higher price. The next largest required price increase is $0.16/Mg
for El Paso Products. But El Paso's capacity is only 115,000 Mg/year.
Both Sun Oil and El Paso could be undercut by other firms at this price
without creating capacity problems-.
In the absence of collusion among producers, maximum required price
increases under 76 percent utilization could not hold. As noted previously,
excess capacity generally inhibits full cost pricing. Consequently, all
further discussion of price increases will be based on the assumption of
100 percent utilization.
Required price changes vary with the regulatory option selected. The
averages of required price changes are $0.0/Mg, $0.28/Mg, and $0.17/Mg for
the 85, 94, and 99 percent regulatory options, respectively. Note that
required increases are greatest under the 94 percent regulatory option.
The maximum required price increase under that option is $2.14/Mg (0.46 percent
of product price). The smallest required increases are under the 85 percent
scrubber/condenser system.
7-69
-------
Required price changes vary across companies also. Assuming TOO percent
utilization, price increases required under the 85 percent regulatory option
range from -$0.29 (Oxirane) to $0.70 (Sun Oil). Under the 94 percent option
the range is from -$0.29 (Oxirane) to $2.14 (Sun Oil) and under the 99
percent control it is from -$0.12 (Oxirane, Cos-Mar) to $1.27 (Sun Oil).
Only Oxirane faces negative required price changes under all potential
regulatory options. Three companies (Atlantic Richfield, El Paso Products,
and Sun Oil) experience positive required price increases under all
regulatory options. The maximum increase required under any option for any
company is $2.14 or 0.46 percent of product price.
7.4.3.3 Output and Employment Impacts
Assuming 100 percent capacity utilization and a 15 percent target rate
of return, the 85 percent regulatory option generates a maximum required
increase of $0.70/Mg (0.15 percent of product price). Assuming a long-run
price elasticity of 0.49 (see Section 7.1), this maximum price increase
corresponds to a maximum potential demand decrease of 0.07 percent, which
corresponds to 2.3 thousand Mg at the current production level and
2.9 thousand Mg at the projected 1983 output level. Based on employment
data from Section 7.1, this corresponds to a maximum of one worker dis-
placed.
Under the same assumptions, the 94 percent regulatory option generates
a maximum price increase of $2.14/Mg (0.46 percent of price), which corres-
ponds to an output decrease of 7.0 thousand Mg at the current production
level and 9.0 thousand Mg at the projected 1983 level. A maximum of fewer
than three workers should be displaced. Under the 99 percent regulatory
option, the maximum price increase required is $1.27/Mg (0.27 percent of
7-70
-------
price), which corresponds to an output decrease of 4.1 thousand Mg at the
current production level and 5.3 thousand Mg at the projected 1983 level.
Fewer than two workers should be displaced.
Note that the employment impacts cited above are maximum values corres-
ponding to maximum price increases. Actual increases may not be as large.
Also, these employment impacts do not include jobs created by the regulatory
options. Operation and maintenance of control systems would require more
labor time than would be displaced by the output decreases estimated. Net
employment impacts under each of the options considered probably would be
positive but small in any case.
7.4.3.4 Investment Impacts
It is unlikely that any of the regulatory options considered in this
report would have a substantial impact on investment in styrene production.
A typical new styrene plant is estimated to cost upwards of $60 million.48
The capital costs of control for any single existing plant are less than
one percent of this base plant cost and control costs in a new plant would
probably be lower, as retrofits are generally more expensive. Moreover,
these capital costs are offset partly by operating cost savings. No major
investment impacts are anticipated.
7.4.3.5 Interindustry Impacts
Interindustry impacts would be negligible under full cost absorption.
Under full cost pricing, some demand decrease could be expected. This
decrease would correspond to a decrease in demand for styrene derivatives.
Some impact thus would be felt by the producers who further fabricate
styrene. However, since styrene cost represents only a minor portion of the
market price of the final goods produced, such impacts would be quite small.
7-71
-------
Similarly, there probably will be some substitution of other products for
styrerte-produced goods. But substitution would be small and would be
spread over a number of basic substitute chemicals. Consequently, little
impact is expected on overall plastics or resin material prices.
7.4.3.6 Impact Summary
Impacts on rates of return on equity under full cost absorption are
small. The largest adverse impact is -0.000038 or 0.03 percent of the
target rate of return of 15 percent. Impacts are generally largest under
the 94 percent regulatory option and smallest under the 85 percent option.
Price increases required to maintain a target rate of return (full
cost pricing) are also largest under the 94 percent regulatory option. If
100 percent utilization is assumed, the average required price increase is
$0.28/Mg. The maximum required price increase for any individual firm is
$2.14/Mg (0.46 percent of product price). The output decrease which is
estimated to correspond with this maximum price increase is approximately
0.23 percent. This would represent 7.0 thousand Mg at the current production
level and 9.0 thousand Mg at the projected 1983 production level. Employment
cutbacks associated with decreased output probably would be outweighed by
the labor requirements associated with controls. Substantial impacts on
new investment are not anticipated. Also, interindustry impacts should be
smal1.
7.4.4 100 Percent Control (Zero Emissions): The Closure Option
A zero emission limitation is the only emission standard which would
provide absolute safety with regard to benzene emissions. However, no
currently available control technology will guarantee zero emissions. Zero
emissions would require a ban on domestic production of styrene. The full
impact of such a ban would depend upon whether domestic production of
styrene derivatives could continue.
7-72
-------
If styrene monomer could be imported in sufficient quantities, styrene
derivative production could continue in the United States. Moreover, the
production of final goods containing styrene could remain unchanged,under
such import conditions. While increased material transportation costs would
increase the prices of styrene products, rendering them less competitive,
they could continue to be marketed. Unfortunately, the availability of
large quantities of styrene monomer for import is highly questionable. The
United States traditionally has been a net exporter of styrene, with facil-
ities representing approximately 40 percent of total world capacity. It is
unlikely that sufficient imports would be available in the near future. In
the long term, foreign producers might choose to further process the monomer
themselves, making styrene derivatives rather than styrene available for
world consumption. Accordingly, the potential impact of closing styrene
derivative plants as well as styrene monomer facilities must be assessed.
In the following analysis we consider the direct impacts of closing
styrene monomer and styrene derivative facilities. The indirect effects of
such closures on material suppliers and intermediate consumers are addressed
also. In addition, the existence and availability of substitutes for
styrene in final goods production are considered.
7.4.4.1 Direct Impacts
Styrene is produced by 11 companies at 12 locations in the United
States. In 1978, domestic styrene production was valued at $1,135 million.
Direct employment in the industry is estimated to be 650 to 950 workers.
Production facilities are concentrated geographically in Louisiana and
Texas. In these areas there would be a multiplier effect associated with
decreased output, increased unemployment, and lower overall income if
styrene production were discontinued.
7-73
-------
There is considerable investment embodied in existing styrene
facilities. As the potential loss for the owners of these facilities is
sizable, an estimate of the value of existing investments in styrene pro-
duction must be made. Several general methods of estimation were employed
to arrive at approximations of styrene asset values.
The first method of estimation utilizes published information for
SIC 2865, Cyclic Crudes and Intermediates. (In 1978, the value of styrene
production accounted for 24 percent of the value of shipments for SIC 2865.)
This method of estimation relies upon the assumption that the value of
assets per dollar of output is uniform across SIC 2865. Under this method-
ology, investments in styrene are estimated to be worth approximately
$925 million.
Of course, the value of assets per dollar of output is not necessarily
uniform across the overall industry. A second estimate was constructed
utilizing limited information specific to styrene which is available. This
methodology employs engineering estimates of costs for typical new plants in
197048 which are updated to 1978 through a price index for producer capital
goods. Historical information from the Chemical Economics Handbook is used
no
to estimate the ages of existing plants. Under this methodology, invest-
ments in styrene are estimated to be between $400 million and $500 million.
Note that the SIC methodology produces an estimate which is twice as
large as the more specific estimate. Since assumptions and approximations
are involved in each estimation, it is not possible to say which is closer
to the truth. However, since the SIC methodology is the best available for
some subsequent estimations, it should be noted that this methodology may
result in high-side approximations.
7-74
-------
Styrene plastics materials (the intermediate product derivatives) are
produced by more than 40 companies throughout the United States. In 1978,
production of styrene plastics materials was valued at $1,812 million.
Under a procedure similar to the SIC method described above, employment for
this industry is estimated to be over 11,000 total workers, including more
than 7,000 production workers.
Investments in the production of styrene plastics materials also are
approximated through the SIC methodology. Asset values are estimated to
total $1,236 million. Styrene monomer and styrene plastics materials
together account for $2.16 billion worth of assets under this methodology.
As noted, this may be a high side estimate, but the true value is certainly
likely to be more than $1 billion.
The explicit impact on the producing companies of banning styrene
manufacture is difficult to assess without detailed profit data. Production
values and overall sales data can be used to produce a proxy for the impor-
tance of styrene to overall profitability, but such approximations can be
misleading for several reasons. First, not all sales are equally profit-
able. Percentage contribution to revenues can differ considerably from
percentage contribution to profits. Moreover, as shown in Table 7-8, many
styrene producers market styrene derivatives and produce styrene feedstocks.
If derivative production in the United States is discontinued, the impact on
individual styrene producers may be larger than the impact captured by
styrene production values.
As an example, styrene sales account for only two percent of Monsanto's
total sales. Yet 60 percent of styrene produced is captively consumed,
which means that the value of styrene production is approximately five percent
7-75
-------
of sales. However, Monsanto reports that if derivative production is con-
49
sidered, 25 percent of total sales is the appropriate figure. Even for a
large, highly differentiated, company, a loss of 25 percent of sales revenues
is difficult to absorb.
The question of direct closure could depend critically on the survival
of domestic derivative production. If. only styrene monomer production were
to cease, direct closures most likely would be limited to the two operating
companies, Cos-Mar and Oxirane. These companies produce nothing but
styrene. However, the parent companies of Cos-Mar (American Petrofina and
Borg-Warner) and Oxirane (Atlantic Richfield and Hal con International) are
large, multi-product producers. Their failure is unlikely.
However, the degree of vertical integration in the chemical industry
indicates that substantial reduction in the production of styrene deriva-
tives would increase greatly the impact on styrene-producing companies.
This impact might be mitigated by the fact that a number of the producers of
styrene derivatives also produce potential substitutes for those deri-
vatives. However, loss of revenues due to styrene closures would put them
in a disadvantageous position vis-a-vis other producers of substitutes in
expanding the production of substitutes. Consequently, these companies
actually could experience an additional adverse impact in the form of
decreased market share in substitute production. While closure is not
predicted for any of these companies, it cannot be ruled out. The shutdown
of styrene derivative facilities would have substantial effects on several
firms.
There are a number of firms which produce styrene derivatives but not
styrene monomer. If styrene monomer is not forthcoming from foreign sources,
7-76
-------
derivative plants would close. While some derivative producers are large,
diversified companies, some are vertically integrated firms which take the
styrene derivative forward toward the final good market. Such firms face
potential impacts which are larger than sales percentages would indicate.
7.4.4.2 Indirect Impacts
The indirect impacts of a ban on domestic styrene production would be
felt primarily by feedstock suppliers and fabricators who purchase styrene
derivatives for the production of final goods. In particular, three firms
who produce ethylbenzene from mixed xylenes for merchant sale only would be
affected. Total production for these firms is greater than the demand for
EB for solvent purposes, which constitutes the only use of EB for other than
styrene production. It is possible that one or more of these plants would
not continue to operate. However, since each plant is owned by a petroleum
refinery and is based on mixed xylenes for which there are alternative
markets, firm closure is not anticipated.
Raw materials for EB/S production are primarily benzene and ethylene.
EB production consumes approximately 50 percent of the benzene produced in
the United States. Demand for benzene is currently quite high. Export
potential is considerable at the moment. Should foreign producers wish to
expand styrene capacity in order to supply the U.S. market, they would need
additional feedstocks for some time. But note that should benzene be shipped
elsewhere for styrene production and styrene or a derivative be shipped
back, costs and prices could increase considerably. In some uses, styrene
would no longer be able to compete with substitutes. Some impact on benzene
producers would be unavoidable.
7-77
-------
Some effect may be felt in the ethylene market also. In 1978, ethylene
consumption in EB/S production was less than ten percent of domestic ethylene
capacity. However, unlike the benzene market, the ethylene market currently
suffers from a slight lack of demand. This slack makes the industry more
vulnerable to any further decrease in demand. Ethylene is produced by
approximately 25 companies of varying sizes and degrees of diversification.
Producers are geographically concentrated in Texas and Louisiana as would be
any diverse impacts from the industry.
A large number of firms purchase styrene derivatives for the production
of final goods. Such firms possess varying abilities to switch production
materials. In all cases there are retooling and retraining costs associated
with switching from one resin to another. In some cases those costs would
be prohibitive. For firms which specialize in making one specific product
for a regional market, the result would be closure if they could not import
enough of their basic inputs. Such inputs would be difficult to secure in
the near future.
7.4.4.3 Substitutes for Styrene
Styrene derivatives are used in the production of a variety of final
goods (see Table 7-27). Should domestic styrene production be banned and
the derivative chemicals be unavailable through import, fabricators would
need to find substitutes for the styrene chemicals. Generally, a chemical
is chosen for use for a variety of mechanical, optical, electrical, chemical,
and other properties. Where clarity is of predominant importance, acrylates
and polycarbonates can be used in place of styrene. If impact resistance is
required, polyolefins, nylons, or acetals may be substituted. There are
other materials which are compatible with a variety of fillers and pigments.
7-78
-------
TABLE 7-27. POSSIBLE STYRENE SUBSTITUTES17'50'51'52'53
1974
tyrene End Use 10
ackaqinq
pnstructi on-Related Markets
Construction
Pl'pe
Industrial Rubber Products
Lighting Fixtures, Signs
Paint
Corrosion-Resistant Goods
hectrical Appliance, TV, Communication.
and Office Machine Markets
Appliance and TV
Electrical and Other
pusehold Goods
Housewares, Furnishings
Carpeting and Flooring
Furniture
Synthetic Marble
ranspbrtation-Related Markets
Tires '.
Auto, Truck, Bus Parts
Auto Putty
Miscellaneous Transportation
ecreati on-Related Markets
Toys, Sporting Goods
Marine
Recreational Vehicles
isposable Serviceware
iscellaneous
1) Industrial Products
Paper Coatings, Additives
Ion Exchange Resins .
Medical, Dental Lab Equipment
Impact Modifiers
2) Consumer Products
Novelties
Writing Utensils
Personal Care Items
Luggage and Cases
Floor Polishes'
3) Other
Exports
Total
Styrene Consumption
3 Mg (% of Total)
499 (22)
360 (16)
119
115
30
-
23
10
12
282 (12)
154
128
265 (12)
129
63
54
15
219 (10)
143
64
8
4
136 (8)
147
24
15
104 (5)
246 (11)
108
58
22
23
97
43
24 •
23
5
41
110 (5)
2,272 (100)
Possible Substitute
Polyethylene, Polypropylene, Polyvinyl .
Chloride, Paper
Polyvinyl Chloride, Polyurethane, Mineral
Polyvinyl Chloride, Chlorinated Polyvinyl
Natural Rubber, Polybutadiene,
Ethylene Propylene Rubber
Polycarbonate
Polyvinyl Alcohol, Polyvinyl
Chloride, Acrylics
Polypropylene, Polyvinyl Chloride
Nylon, Acetal , Phenols, Polyvinyl
Chloride, Aluminum, Polycarbonates
Phenol ics, Urea- formaldehyde
Polyvinyl Alcohol, Polyvinyl Chloride,
Polyvinyl Acetate
Wood, Polyvinyl Chloride, Cellulosics
Unknown
Natural Rubber, Polybutadiene
Nylon, Acetal, Metals
Epoxy
Unknown
Wool
Chloride
Polyethylene, Polypropylene, Polyvinyl Chloride
Nylon, Acetal, Phenol ics
Paper Plates, Polypropylene
'"
Polyvinyl Alcohol, Polyvinyl Chloride
•
Zeolites, Activated Carbon, Membrane Materials
Unknown
Unknown
Unknown
Unknown
Polypropylene
Unknown
Acrylics
Unknown
Unknown
7-79
-------
However, where a combination of properties is required, potential sub-
stitutes, if they exist at all, can be expected to result in a final product
which is only similar to the original one. The degree of similarity will
vary from case to case. In addition, many final goods will be more
expensive when produced with substitute chemicals.
One potential problem area is the production of synthetic rubber.
Styrene-butadiene rubber (SBR) has become almost indispensible in the tire
rubber market. While both natural rubber and a variety of synthetic rubber
materials exist, many rubber products such as tires are made from a com-
bination of materials. Such products made without SBR might be in some
sense inferior. In addition, there is the question of the availability of
substitutes for SBR. More natural rubber cannot be had immediately. Pro-
duction of the other synthetics can be increased only with time.
The question of availability applies to all potential substitutes.
Polypropylene is expected to remain fairly plentiful through 1983 and
beyond.* However, both low and high density polyethylene are expected to be
in tight supply through 1983. Demand for polyvinyl chloride is increasing
and a general shortage currently is predicted for 1983. There is extreme
tightness in the phenol market currently, but that tightness is partially
due to plant problems. Markets for polyvinyl alcohol, acetals, poly-
butadiene, and various cellulose resins are characterized by small numbers
of producers with fairly small capacities.
It is likely that substitute chemicals for a variety of uses would not
be available in sufficient quantities to fill the void left by styrene for
several years and in some cases longer. The magnitude of capital require-
ments and the length of construction lead times will limit the availability
*Not including any demand increase as a result of a styrene production ban.
7-80
-------
of domestically produced styrene substitutes and imports of styrene and its
substitutes.
It also must be noted that increased production of substitute chemicals
can have environmental and health impacts similar to those of styrene pro-
duction. In particular, the various processes which result in polyvinyl
chloride can produce vinyl chloride emission. Vinyl chloride is also a
carcinogen.
7.4.4.4 Summary
Complete prohibition of benzene emissions in styrene production would
require closure of styrene production because no existing technology will
achieve a zero emission limitation. Banning production of styrene would
have a substantial negative impact on the producing companies. The magni-
tude of that impact would depend upon whether the production of styrene
derivatives also must be discontinued. If derivative chemicals are no
longer available, small specialized fabricators who further process such
chemicals may face closure. Whatever happens to derivative production,
feedstock producers also will experience negative impacts. Overall, con-
siderable loss of capital investment could be expected.
Employment impacts would depend upon the future of derivatives produc-
tion. Styrene monomer production directly involves fewer than 1,000 employees.
Derivative chemical production may involve more than 10,000 employees.
Regardless of the future of derivatives production, output, employment, and
income, impacts will be concentrated geographically to some degree in Texas
and Louisiana. A multiplier impact can be anticipated in some areas.
It is unlikely that foreign styrene production would be sufficient to
maintain current levels of domestic consumption of final goods based on
7-S1
-------
styrene. Goods which continue to be produced with styrene would be more
expensive for a variety of reasons (e.g., transportation costs).
There are substitutes for styrene in some but not all applications.
Some products may cease to exist under substitution. Where substitutes
exist, they do not necessarily have all of the desirable properties of the
styrene chemical replaced. The consumer goods produced through substitutes
may be in some sense inferior, as well as more expensive. For a variety of
substitutes, availability would present a potential problem in the short
term. Limited capacities for substitute chemicals likely would result in
price increases under increased substitute demand. This would render con-
sumer goods even more expensive.
7-82
-------
7.5 POTENTIAL SOCIOECONOMIC AND INFLATIONARY IMPACTS
The socioeconomic impacts of the 85 percent, the 94 percent, and the
99 percent regulatory options are not expected to be large. These impacts
can be summarized as follows:
(i) Annualized Costs. Total annualized costs in the fifth year
following promulgation are estimated to be -$0.608 million,
-$0.202 million, and -$0.460 million for the 85, 94, and 99
percent options, respectively. Note that these estimates are
based on existing plants. Any new construction planned or
started between now and promulgation would face control
requirements also. However, total annualized costs including any
such new plants would not surpass the critical level of
$100 million.
(ii) Price Impacts (Inflation). The maximum estimated price increase
required to maintain a 15 percent rate of return which could be
passed on is $2.14/Mg, or 0.46 percent of product price.
Potential price increases are, therefore, well below the
five percent requirement for regulatory analysis.
(iii) Employment Impacts. Impacts on employment are expected to be
negligible. Labor time required to operate and maintain controls
and monitors may be greater than labor displaced by potential
demand decreases (due to price increases).
O'v) Demand for Critical Materials. The only critical materials
likely to be affected by the proposed controls for benzene
emissions in the EB/S industry are plastics. If demand for
styrene (a plastic material) decreases, the demand for other
7-83
-------
plastic materials will increase. Such impacts are expected to be
small. Moreover, they involve the substitution of one plastic
material for another. This does not constitute necessarily an
increase in the demand for plastics overall.
The potential socioeconomic impacts of the 100 percent regulatory
option (closure) are much larger. Impacts would be concentrated geographi-
cally. Income and employment effects in Texas and Louisiana could be sub-
stantial. The degree of impact would depend upon whether or not plants
producing styrene derivatives could continue to operate by importing styrene.
If they could not, employment and income impacts would be substantial.
Some businesses could be expected to close under these circumstances.
Shortages of good substitutes for styrene chemicals also could be expected.
7-84
-------
7.6 REFERENCES FOR CHAPTER 7
1.
2.
3.
4.
5.
6.
7.
8.
SRI International. 1978 Directory of Chemical Producers — United
States of America. Menlo Park, 1978. pp. 595, 909.
Paul, S. K. and S. L. Soder. Benzene. In: Chemical Economics
Handbook. Menlo Park, SRI International, May 1977. p. 618.5022Y.
Paul, S. K. and S. L. Soder. Ethylbenzene — Salient Statistics.
In: Chemical Economics Handbook; Menlo Park, SRI International, May
1977. p. 645.3000B.
Sleeth, C. V. Styrene Monomer. In: Chemical Engineering Progress.
November 1977. p. 33.
Soder, S. L. Styrene. In: Chemical Economics Handbook. Menlo
Park, SRI International, January 1977. p. 694.3053B.
Blackford, J. L. Propylene Oxide. In: Chemical Economics Handbook.
Menlo Park, Stanford Research Institute, November 1976. p. 690.8022B.
Blackford, J. L. Propylene Oxide. In: Chemical Economics Handbook.
Menlo Park, Stanford Research Institute. November 1976. p. 690.8022B,F.
Telecon. Moss, J., Research Triangle Institute with Raffray, B.,
American Hoechst; Padilla, E. and R. Arnold, Monsanto; and Mr.
Jackson, Gulf Oil. May 11, 1979.
9. The Chemical Marketing Reporter. 215. March 19, 1979.
10. United States International Trade Commission. Synthetic Organic
Chemicals — United States Production and Sales, 1976. Washington, DC,
U.S. Government Printing Office, 1976.
11. The Chemical Marketing Reporter. April 9, 1979. p. 13.
12. Key Chemicals: Styrene. Chemical and Engineering News. July 24,
1978. p. 11.
13. Soder, S. L. Styrene. In: Chemical Economics Handbook. Menlo
.Park, SRI International, January 1977. p. 694.3053F.
14. Stanford Research Institute. Chemical Economics Handbook. Menlo Park
1976. pp. 694.3052D-G.
15. United States International Trade Commission. Synthetic Organic
Chemicals — United States Production and Sales. Washington, D.C.,
U.S. Government Printing Office. 1960-1977. (Table and page numbers
vary due to reclassifications. )
7-85
-------
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
Soder, S. L. Styrene. In: Chemical Economics Handbook. Menlo
Park, SRI International, 1977. p. 694.3053G.
The Chemical Marketing Reporter. Various issues.
Soder, S. L. Styrene. In: Chemical Economics Handbook. Menlo
Park, SRI International, January 1977, p. 694.3053H.
The Chemical Marketing Reporter. 1979. Various issues.
Letter from Cantrill, J. E., American Hoechst, to Giberson, L., RTI.
April 18, 1979.
Soder, S. L. Styrene. In: Chemical Economics Handbook. Menlo
Park, SRI International, January 1977. pp. 694.3053I-J.
Wright, A. Styrene Prices Should Reflect New Plant Costs.
Chemical Age, October 24, 1975. p. 9.
The Chemical Marketing Reporter. March 26, 1979. p. 11.
Soder, S. L. Styrene. In: Chemical Economics Handbook. Menlo
Park, SRI International, January 1977. pp. 694.3053G-J.
Stanford Research Institute. Chemical Economics Handbook. Menlo Park,
1976. pp. 694.3052D-I.
Sleeth, C. V. Styrene Monomer. Chemical Engineering Progress.
November 1977. p. 31.
Telecon. Coleman, R., EEA, with Giberson, L. , RTI. January 19,
1979.
Securities and Exchange Commission. 10K Forms. Bethesda, MD: Dis-
closure, Inc. 1977.
Soder, S. L. Styrene. In: Chemical Economics Handbook.
Park, SRI International, January 1977. pp; 694.3052E-H.
Menlo
Soder, S. L. Styrene. Chemical Economics Handbook. Menlo Park, SRI
International, January 1977. pp. 694.3053K-M.
Greek, B. F. Plastics Monomers Ride Demand Boomlet. Chemical and
Engineering News, July 24, 1978. p. 10.
Soder, S. L. Styrene. In: Chemical Economics Handbook.
Park, SRI International, January 1977. p. 694.3051D.
Menlo
Sleeth, C. V. Styrene Monomer. Chemical Engineering Progress.
November 1977. pp. 32-33.
7-86
-------
34. NPRA Mulls Aromatics Outlook Through 1980's. The Chemical Marketinq
Reporter, 215(16): 3,34. April 1979.
35. Hobbs, F. D., and J. A. Key. Emissions Control Options for the Syn-
thetic Organic Chemicals Manufacturing Industry. Ethylbenzene and
Styrene Product Report. Prepared by Hydroscience for EPA. Contract
No. 68-02-2577. April 1978
36. Letter from Gurney, A. S., Croll-Reynolds Company, Inc., to Miles A
EEA. January 1979.
37. Letter from Boettner, J.
EEA. February 8, 1979.
44.
45.
L., Corning Process Systems, to Miles, A.,
38. Guthrie, K. M. Process Plant Estimating Evaluation and Control.
Solano. Craftsman Book Company of America. 1974.
39. PEDCo Environmental, Inc. Cost Analysis for Standards Support Document
Prepared for EPA. November 1978,
40. U.S. Environmental Protection Agency. Development for Effluent Limi-
tations Guidelines and New Source Performance Standards for the Major
Organic Products Segment of the Organic Chemical Manufacturing Point
Source Category. Publication No. EPA-440/l~74-009a. April 1974.
41. Hydroscience. Report, Storage and Handling in Synthetic Organic
Chemicals Manufacturing Industry.
42. Responses to Emissions Inventory Questionnaire to Texas Air Control
Board.
43. U.S. Environmental Protection Agency. Control of Volatile Organic
Emissions from Petroleum Liquid Storage in External Floating Roof
Tanks. OAQPS Guideline Series. Publication No. EPA-450/2-78-047
December 1978.
U.S. Environmental Protection Agency. Control of Volatile Organic
Emissions from Storage of Petroleum Liquids in Fixed-Roof Tanks.
OAQPS Guideline Series. Publication No. EPA-450/2-77-036.
U.S. Environmental Protection Agency. Benzene Emissions from Benzene
Storage Tanks — Background Information for Proposed Standards
Preliminary Draft. March 1980.
46. Arthur D. Little, Inc. Technology Assessment and Economic Impact
Study for an OSHA Regulation for Benzene. Vol. I. Prepared for U S
Department of Labor. May 1977.
47. Arthur D. Little, Inc. Technology Assessment and Economic Impact
Study for an OSHA Regulation for Benzene. Vol. II. Prepared for U S
Department of Labor. May 1977.
7-87
-------
48. Guthrie, K. M. Process Plant Estimating, Evaluation and Control.
Solano, Craftsman Book Company of America. 1974.
49. Letter from Pierle, M. A., Monsanto, to Walsh, R. T., EPA.
May 29, 1979. Response to Section 114 letter.
50. Cunningham, E. R. Guide to Plastic Pipes. Plant Engineering. 1976.
51. Frados, J. (ed.) Society of Plastics Engineering Handbook. Van Nostrand-
Reinhold. 1976.
52. Marsland, D. North Carolina State University.
53. Soder, S. L. Styrene. In: Chemical Economics Handbook. Menlo
Park, SRI International. January 1977.
7-88
-------
APPENDICES
/V,
-------
-------
APPENDIX A
EVOLUTION OF THE PROPOSED STANDARD
-------
-------
E
o
•r—
•p
re
o
o
— i
0
ry
>
•P -P
£ C
re re
Q- CL.
i- S-
re ro
i i
IO CO
o o
CJ CJ
p^ p~
kO 00
r— CM
' ^V
to r-~.
X
i—
-P
i.
0
Q.
CO
CO
5-
LU
•P
to
>
^
C
re
p—
CL.
Dow Chemical
r--
CM
\
•P O CO
OSE E EEErore
CL rere rerererecco r™j
a) . -C .c: ^ jz.c£:oco
COS-S- S- J-S-J_4JCO.
5-33 3333re"O+J
U_QQ Q QOC3Q3OCO
£ £ £ £ £
O O O O O
-p ^ 4J +J ^J
re re re re re
E E EEES-S-S-
COS- S- S-S-S-COCOCO
COO OOOO-P-P-P
C0<(- CO
ioo- o- cro-0-iocoio
COCO CO COCOCO£££
^30. « as oc as o o o
re co <^* ^- ^- «^- ^j- to >-p CO-P core
i — U£CO-r-lO4JO
S- EQ.U^U3*E
r— r— 3 CO COECOUCO-OCO
rere-Po ^roo'r-oo-E
_u^ure£ ooxroiics-o
EE •!- OCO£U£ i —
co co o u>j4->co re-r- re o-r-
^Z^Z COCO£ £-P CJ-P U lOQ
oore rerere-r-£.r-re
a_q-a.io.r- i. ro s-a.4-
S35 " — ECTJCOi — CO i —
O Or— 3O OCO E-P Ei— 3
Q Q LU C30 S S -
-------
CO
o
CO
CO
g
LJJ
Q_
•P
CO
U
U
OJ
ro
CO
in a
C 0)
O O5
O
fO
O)
Q-
C_3
(0
X
O)
O)
rd
O
O
in z:
O)
o
ro
ro
ro
ro
(0
O
01
ro
CO
in •>
S- O)
JC T-
o >
in 0)
ro
CJ
(0
ro
(0
(0
rO
ro
rO
0)
QJ
O)
O)
O)
Ol
O)
O)
ro
r— 1— Q
<— i— a
•i- 0)
in in
in c
3 O
u a.
in in
•i- a>
a>
a.
a>
ro
•r- O
u
(U
•r- -C O
OS O O
U
a>
u
O)
ro
u
in
u
a)
-G
o
o in -r-
ro
O)
c -P
ro ro
in •
U
r— c T3
-P o
CM
LT>
rO
O
00
ID
CM
a>
s: r—
o T-
(J
00
00
00
oo
ro
o
a>
^=
o
C3
00
oo
(0
U
O)
5
00
O)
01
oo
u
at
ro
oo
CJ
a)
(0
i— D.
(0 S-
U O
•r- O
(0
U
cu
in
•P
u
a>
CD
o
r- 00
00 00
oo «=r
i— CM
cu
CO
00
CM
O i—
o
O)
to
00
CM
00
o
(U
CO
00
en
O i—
A-2
-------
>*""^
"O
o>
3
c
-p
c
o
o
x»^
o
>
=3 (J
in c
C CU
O O5
o > o
c
re
o.
E
0
O
CU
-p
re
Q
X
h-
f*
0
-p
in
3
0
s-
01
-p
•p
CU
1 —
1—
1—
o
.p
CU
in
c
o
O-
in
CU
a±
,
CL
i_
O
o
r—
CU
CU
.4_2
to
.
I/O
«
^
CD
r^*
^^
LO
CM
i—
^
ft
S
re
S-
0
c
o
w
1/5
3
U
in
•r—
•o
£.
CU
3:
o
+J
CU
*/
0
re
in
S-
o
in
in
re
O
oo
c
•r—
C
S-
o
0
CD
r^s.
\
o
ro
t—
S-
cu
c
J_
in
re
M-
0
t/5
~v^
CO
LU
r—
O
c
o
-p
re
s.
o
<4-
c
HH
a.
s_
o
C_3
C
o
•r~
4->
in
3
-Q
0
c_>
UJ
in
CU
U
o
s-
a_
CD
r-~
\
CM
C
0
•r~~
4J
O-
Q
-Q
re
CU
c
CU
cu
«i
Q
in
o in
•r- O
in u
^ 5_
O CU
in !3
*r~ O
Q -P
l \
d
flj
gg
p|t
*r—
Z5
cr
o
S-
1
E
JZ
0
CD
f-.
CM
CM
re
O)
(jf
S-
4^
C
CU
<£
LJJ
i. . ^
O (O
"O
. i
C S-
0) 0
£ CD
a> u)
t- S—
O f^
"O £
c o
LU U
O)
•r-
"O
t&
—p.
T3
Vf—
, — T
U.
C"
CO
j-J
CO
"O
r"
3
CTi
•s^
,__
CM
CM
>>s: c_> c_> o • o o o
O"O •. . «. »> n rx ^
3
Q
3
Q
3
Q
3
Q
3
Q
C_>
3
o
3
Q
CU
C
o
«^»
re
e
c
o
re
E
C
0
•r—
re
E
c:
o
re
E
C
o
•1—
re
c
o
•r-
re
s-
c.
0
•r—
re
E
+j
••— in
incu
T-.3
> o-
in
cu
3
o-
in
cu
3
o-
in
cu
3
a-
tn
aj
3
o-
in
cu
3
o-
in
cu
S-
o
CU
a;
in
cu
•P
c
o
°
0
O
S
cu
-£=
o
^
S
0
CD
s-
-P
C
cu
o i—
4->
c
CUi —
£•
0
s-
re
cu
o
.p
C
re
ifl
C
-o
in
•—
re
o
o
W
re
CD
CD
CD
CD
o
o o
O in
-P CU
C +J
re re
c -o
CD
o
-------
O>
in
o
in
ftt
0)
cu
to
.c
o
ctt
to
X
0)
fO
en
to
O)
(O
re
O)
10 o eu
O) •!- in
3 -P O
0)
CO r— +J
r— 0)
i— ce
U
S-
O -P
I—I 3
in
cu in 01 in
3 4^
in
Ol
in
in
u
in
CU O CU O Ol •!-
r>^ (j Qi O Q^ O
cu
cu
-a
cu
o -p cu
M- in in
c x: c
•r- O 0)
CD T3
s- o c:
o ac o
4- u
c
•P fO CU
in o o
cu
cu
cu
cu
cu
cu
(0
3 S- t-
cr a) s-
ai E 3
oi
to
X
u i—
U E
•r- CU
to
cu
to
o
a>
(O
cu
to
C r—
to u_
r— C
r— fO
to
to
(J
r- CU
to -c
u o
*l~
E i—
cu
en
cu
-C
C_5
to
•I- J-
tO
C3
en
cu
cu
en
c.
LLJ
ca
cu
•p
to
00
CM
00
00
00 00
00
CM
CO r—
-X, \
r— CM
00
LO
CM
00
co
00
CO
CD
en
I— I— LT>
CM
CM
en
CM
A-4
-------
/•— s
T3
O)
• 3
C
•P
C
o
0
Q
ce.
o
ro
O
X
^
-. >J
O
in
to
X
0)
in
c
ro
r—
£-
cu
c
i — i
Monsanto Chemica
mediates Co.
en
r--.
3:
LO
X
1—
o >>
•1 —
o
E
ro in
.C ro
i- X
3 CO
0 (—
C;
i >
10
S- CD
O -P
H- -P
C 0)
(-H 1
O i—
•P 0
to -p
CO
3 0)
o" in
O) C
Q; o
Q.
"i" in
r— O>
r— Di
D-
o
r 1
Cosden Oil
Amoco Chemicals (
en en
CM CO
CM CM
LO if)
X
1—
>;
o
in
ro
X
1—
Response to 114 letter
<_
CO
-p
I-H
Monsanto Chemica"
mediates Co.
O*V
s^,^^
en
CM
un
<_>
^
Durham
Emission Test Request
j_
01
•p
i— i
Monsanto Chemical
mediates Co.
r--
CD
CM
to
X
. 1 —
ro
in
in
0>
TJ
- O
S-
O)
•p
O)
1 —
o
-p
0)
in
c
o
Q.
in
Ol
d
o
in
•P
U
3
TJ
0
Q_
0
in
ra
Q.
LU
en
00
CM
to
^
o
in
rO
X
o>
Pretest survey
i
cu
-p
c.
I— 1
Monsanto Chemical
mediates Co.
en
LO
CM
vv.
c? o
z, z
rO fO
i- S-
3 3
Q Q
C
O
•r—
•P
ro
•— -P
3 in
D) O)
01 3
o; cr
0)
Qi
0
1 Tj 1^
co in
O)
to c
S- 0
Q T-
l/)
to in
OQ 'i
LU LU
Q
co
in
-P
O
3
-a
o
o_
»— i
1— 0
Qi in
to
Q.
LU r—
LU LU
en
r^ en
en -~v,
r— CO
en ^>
0 X
Z 1—
E" to"
ro in
.c in
S- 0)
3 -a
Q 0
Emission Test Request
Emission Test
O
o
in
-P
u
•o
r- g
o) a.
0)
•P 0
to in
ro
a.
• « i —
rj LU
en
en r*^
CO \
e£ °
x
z
•P ra
in JT
3 S-
O 3
x: a
c
o
+J
ro
i.
O
M- C
C 0
>— 1 •!-
Response to Request for
114 Request for Informat
Q. Q.
O O
O O
O) Ol
C C
rO ro
O O
en en
CM 10
0 0
A-5
-------
cu
Q
UJ
V)
g
§
UJ
UJ
O
LU
Q.
re
u
u
c
C 0)
O D)
o
re
Q
o
o
re
o
re
re
re
re
re
re
x
I— O
D.
CT
•r- -P
CQ' tO
CD
C
0)
CD
.a
CT
a) to
aj-^s
s: ca
UJ
§•*-
CT S-
c re
s- re
o -P
CD
CD
cu
CU
o;
re
to
03
V) X
CU 1—
o o
•P re
in X
0) C
a; -i-
re
•a
re
CD
N
CD
CT
re
re
re
re
a>
(0
0)
CO
o
(0
Ol
O)
re
^
o
re
C_J
o
O)
O)
u
re
o
d)
Ol
CU
CU
o
to
CD
eu
O)
i— i—
re - -r-
cu T- x
CQ Q O
C_5 •!-
CD
CD
jC
O
r- re
re
-------
/•"N
•o
>
3 U
(A C
C 0)
o en
0 > o
c
re
a.
0
0
+}
n3
Q
O
O
c
o
•p
CO
c:
f~^
in
ro
IS
Qt
C
0)
N
C
J3
C
O
en
f*
•1^
-p
0)
0)
s:
cu
a» to
P CD
•i- LU
E
E S-
o o
C_S 4-
O)T3
C S-
•r- rd
J- 73
O) C
fl) O3
•P +•}
to 10
O £
.p
n3 <*^
t— 3 tO
3 O >>.
en s- QQ
a> cnuu
_%/» ^
>> to o
u to
-------
-------
APPENDIX B
ENVIRONMENTAL IMPACT CONSIDERATIONS
-------
-------
S-INDEXED REFERENCE SYSTEM TO HIGHLIGHT
RONMENTAL IMPACT PORTIONS OF THE DOCUMENT
oo i— <
o >
O^ ^"""^
0 UJ
p_
CQ
Location Within the Background Information
Document (BID)
COr— ^-— *
E fO CT>
'si t= 31
to CO 1^
Q. E CO
CO E
s- o o;
Q- S- U_
t- *> cr>
0 C CO
C|- LU
CO E to
CO 0 4->
C ••- E
•r- 4-9 CO
i— O E
CO «^ CD
"O "^
3 ^ I ^
CD O 00
o £ tj
£= 3 to
CO CO Q.
CO CO £
•> t~\ If*
The regulatory alternatives from which standards v
chosen for proposal are summarized in Chapter 5, |
and in the Addendum to the BID on Excess Emission*
3-7 through 3-9.
1
_
CO
CO
q-
o
j^
to •
E CO
to -f^
-J \
•o to
E C
10 co
"O 4->
E r—
O
i- >>
CO i?
is o
O 4-^
to to
CQ i—
i —
^^
^Z
S
**-)
CO •*
-£= co r~»
to x_
Other EPA regulations and, to the extent possible,
regulations that apply to the affected industries
included in the study of economic impact in Chapte
pages 7-45 through 7-51 .
i- (/)
CO
^^ a^
•°5
•a c
•^
CO i—
It ffj
H— fo
*|
*T~*
^/
£?-S
4-» tO
to i —
•— * —^
E C?
i— i S-
P.M .
1 • CM
3 ,— CM
CO CO I
CO E CO tO CO
S- O O .1-
•i— S- ^;
CO 4-> 3 tO CO
-C 0 0 CO 3
4-> CO CO > O
The specific processes and facilities affected by
latory alternatives are summarized in Chapter 1, S
page 1-1. A detailed technical discussion of the
and processes affected by the regulatory alternati
presented in Chapter 3, Section 3.3, oaaes 3-6 thr
*
CO
CO
CO 4->
tj E
CO i-
M- CO
M— 4->
"io
to
X ?•>
CO 1-
tfj O
CO 4^
O to
O r—
*r 3
^^ 01'
O S-
4— co
CO
f~V ^fc.
*-J- ^>
00 J3
CO O i—
-C S- |
E
•a o co
CO E o CO
0 10 CO
3 CO to
•p co to a.
CO 4-> tt)
The control technologies which can be applied to r
benzene emissions in the ethyl benzene/styrene plan
emission reductions which have been achieved by thi
technologies are described in detail in Chapter 4,
through 4-10.
•
CO
CO
•r—
CO
O
o
E
O
CO
40
C3
i_
E
O
O
CO
.a
o
"o.
=§•
. ^
E
CO •<
cD* >~-(
O CQ CO
i. S_
O CO to
C— I % r—
The applicable enforcement approaches include 1) ei
procedures which are discussed in the Addendum to 1
and 2) applicable formats for the test method whid
described in Appendix D, pages D-3 through D-4.
^
CO
CO
o
to
o
a.
Q.
tO
[ i
E
CO
o
i-
o
E
CO
CO
JO
to
o
"cL
Q.
B-l
-------
ii _ *»
GHLIGHT
DOCUMENT
H LU
o= 3:
S
EFERENCE SYSTE
1PACT PORTIONS
[Continued)
I»*rt T^— -— -
ce »-<
o-J
LU
cx: z:
0 LU
r-;
m
r~
d Informat-
£Z
3
O
CO
JK:
u
re
CO
CU
•p
c
•I— /— N
j=. a
•P M
•r- CQ
rS*-'
C P
o c
•i- CU
P E
re 3
u o
O 0
—i a
e ro cr>
.r- 4J r—
S- C «tf-
re cu r-
a. E co
cu c.
s- o az
O- i- U_
t- '> OT
o c co
q- LU v-.
in c in
cu o -P
C -i— C
•I- p <
i— O 1
cu > re
3 J- -P
O O CO
+J
>> re P
U r— U
C 3 ro
CU O) 1
CO CU E
«C D£ I-H
o ~ .C in
co cn c
c . tn cu 3 o i
O LO -P XI C> "r~ C
•r- CU U CU 5- P "i—
+JC"O re x: co c x: o.
ro a> e Q. "P c re p o cu
3NIO E T- O U
,— c T- 4- TJ •— >>M- C
tO ID CM S-O3"lS-S-''~
>.a . r- cu i— >>r- o a. •
cum repcos- P TJ •=*•
CUOtn cre-r-'i-inCUi — 331
.,- o E O Q. ~"O re OVr- 4-
tocp ocs-o-i- &- ro a> co
S- O U S-T-U-I- - E-Q3
Jrt.r-CU T- in P CU CXI CU 0
Q.+3 CO > "O CU Q--C • X: "O C 1-
o o" " §3 reup
0 LO 0 -O >> 0 T3 4-
5^ CU4- CUJ-PCOinOCM
tns_j_ x: r— o re cumLn
-OCU 4JCUT-PW inS-r— 1
urea. M- 4Jr~~w'm"'«in
S "3 x: re T3 O) o O-i— m cu
0) 0)0 CO CUCEOPCO
^- c r-^ ^ cxr— cu 're co o o- a. •
i oci — orecuc
c i. jc -oj-\c re CU-I-EO
OCUCUO) OO5in J-O-P3'r-
'w^'Qo cureoocpocuco
3m«^ -3^ S.* re-S-SSi^
03UP jg »+» Erjg ^ O.e ^
.—.oinrf- cut."^ P jzCT'r— r~^
TS .r- c i -a i o •- c H- c
£_ O LO .C tO •r— O-'r^ ••!" in 5-
re P re
0 -I- C C
o cu o
cu -r- E m
> «t- c: ro
•i— CU O CU
P C i- S-
re cu T-
£- O > r~
re co
a. cu cu re
EX: cu •
0 -P
<4- i- o T-
m o cu +^
> in re
>> C T3 P C
o o re o s-
C •!- CU CU
cu p -a co
•r- 3
P 0 •
re j-
C -C
S- P
cu
•P i —
i — i
re a
cu in
x: cu
P CO
ro
4- a.
o
•*
c a
3 evaluatio
i Appendix
> •!-
T> cu
re s-
i- re
re
r"t *n
; Agency's comj
lent approache;
O) t=
x: cu
s-
o •
4— 00
. C 1
o cu o
cu
e~
.p
it
O
4J
u
ro
f
.p^
,~~
re
p
c
cu
c
0
*l—
^
f—
LLJ
that Air impact modeling shows the primary impacts for each regula-
y to tory option in terms of model plants and for individual emis-
d sion sources in Chapter 6, Section 6.1.1, pages 6-6 through
ants 6-8. Effects of each option on nationwide emissions are
andard discussed in Section 6.1.2, pages 6-8 through 6-10.
,
in
, c
o
•r"
D.
O
O
P
re
3
CO
cu
4->
o
re
f
^_
re
•r~
s-
o_
cu
in
o
P
O)
s-
re
in
P
u
re
Q.
^>^
re
E
•t—
5-
O.
P^
P
O
CU
t-
• ^
T5
•o
CU
P
3
XI
si
P
P
ro
^
X!
P
3
O
XI
re
P
P
ro
P
in
CU
CO
c
ro
<**
(J
r— •
ro
o
•r—
in
Q.
cu
x: .c
rap
3
o
s-
X2
T3
C
ro
ro
•r—
•a
cu
E
in
3
O
•r—
s-
ro
^
cu
x:
P
c
•r—
i-
3
O
O
o
.
c
o
•r—
P
O
3
•o
CU
i-
in
•r*~
x:
p
x:
4->
• f—
S
re
CO
B-2
-------
o ^>
I— I C_>
—I O
3: o
C3
I — I UJ
31 31
UJ CO
I— -z.
in o
>- t— I ^"N
co t— -a
C£ CO
UJ O 3
O Q. C
Z -r-
UJ I— -P
ce: o c
UJ
C£ Z
C_J UJ
£Z
O
•p
re
s-
o
c:
3
o
S-
O)
u
re
CO
cu
•p i—i
•r- OQ
C -P
O C
•i- CO
-P E
re 3
o u
o o
—1 Q
CU ID
S-
re -o
c
c re
£«
H-> •
, a. ID
o
in
>» c
t? o
O -i-
•p -p
re o
r— CU
3 CO
O5
CU •>'
S- ID
-c s-
cj cu
re -P
cu o.
re.
o cj
'in c
-P T-
u
E CO i—
•r- > I
•r- ID
i— -P
re re -c
-P -P O5
C T- 3
cu r- o
E re s-
C 3 JZ
o cr-p
•r- -o o
> co i—
C -r- 1
CO 4- ID
S- 4J m
cu c co
-C CO O)
•p TJ re
O -I- CL
ai
re
•p
Q- •
O C=>
>>v
5? ID
O
-P CO
re D5
r— re
3 a.
as
CO ^
S- CO
CU ID
-P C
O
O -P
u
in cu
•p co
u
re «
•r- S.
CO
1- -p
cu o.
-p re
re j=
So
^•^
re "~
c cu
O -i-
O 4-
CU •!-
in -P
c
co re
in
c
o
'•p -3-
O-r—
O I
2?»
O CO
-P O)
re re
•— Q.
Oi »
CU LD
to'
CO
JC C
-P O
4- +J
O U
cu
•P CO
u
re **
a. ID
•r- S_
cu
c -P
o a.
••- re
•P JZ
Q.O
3 C
in «P-
c
o -a .
u ai UD
>^4- V
O)-r- ID
s- -P
cu c .e
c ro D)
cu 3 3
o- o
cu s-
f— 0- 1 -- r- -P
c ro en
•r- .p r—
re cu .
a. E co
cu c
S- O Q£
Q_ S- U_
S- > CD
O C CO
4- UJ '
in
cu
c in
O -P
C T- C
•r- +J
i— U
CU
cu , re
3 S- 4->
o o co
>» ro -P
o r- o
c 3 re
co en Q.
O3 CO E
<: a; i—i
u
re
a.
s-
re
o
o
CO
CO
-p
(J
CO
S- I
•i- X £=
-o co o
•r- S- !p
o u
CO U. 3
s- -a
re • cu
in s-
•P U >,
u re s-
re D. o
g-e-£
E T- re
•r- ~O
-a c
>» co re
s- o E
re 3
"C3 ' '^J *i
C C CO
O •!- I—
U Q.
CU S- E
O •—
•P 3
C J= O
O OJ O
3
O T3
S- 5-
J3 re
•a
in c
•p re
c -p
ro in
•p
3 3:
i— CU
i— C.
o
a. ro
u >>
r- 4->
(J 3
CU O
CL-a
c:
o i
u s-
co
q- u
o re
x
{= CU
o
•1- -p
-p re
0.4=
o -P
•a
mg5
CO O
•p 'o
c
C JZ
•r- U
CO
•P 4->
E -P
CU O
i— re
-o a.
2£
Q.
C^
o re
•r- -a
•p c
3 O
i— O
r— CO
o in
Q.
re
cu cu
J" -p^^
•p
O T3
C r—
re 3
o
3 r- CO
in o -P -a
co s- re c:
•
in
c
o
•r—
•P
to
s-
0)
•o
•1—
in
c.
o
u
S-
co
-p
o
in w
c: -P
»
•P 4- 0) S- C
O O S- -P CU rO
to cu re >
Q. co -P c -a j=
E U D-'i- re -P
••- c re s -r-
ro -C -i- cu 3:
•— u o r- in
ro «r- o> o en
•P q- T3 -C C
C. T- C 0 1— 0
co c re -P i —
E o> re
C v- CM 4-> •
O in . in "O •>
S- i — T- CO "O
•r- CO X -i- CU
> .c c cu q- -i-
c: +J o T- 4-
CO -f- >, +J -r-
4- -P T3 C -P
cu o u re cu c:
in cu cu T3 cu .
s- c co s- •.- -a E
CO O i — T- CU
> -r- - re cu .C
"O I/) i — S- Q) -P
rO in J= re S-
3 s- o re co
i — o co •!— in N
re in -P _c .p o •!—
•r- -i- a. 3s u in E
•p -a re re r— T-
c .C m a. re c:
cu re o s- E ••-
-P O •!- T3 E
. o -a c -P cu
Q. C -i- O CU T3 0
ro re in -i- 4J
cu -a u. s- o
-c tn c co > c
-p c 3 > re co
_o o • ~a -^
0 -P ~ J3 -P
Q. CO ID i —
c o .a re -P co
o .c: M- o .a
•r- >> C O5 -P C
in s- re 3 c c: c
in o o o a> re to
3 4J S~ -P U U
u ro -P j= o
to ^"** u »^^ Q. -C" _c
•i- 3 ro u u
"O O> O. CM CO -r- «r-
CU E • •-£= -C J=
-C C
re in o
a. •!-
E T3 -P
•i- cu a.
•o o
CO •!-
in o ^^
5™ ^* S—
cu re o
> -P
•a cu re
i-
co o
•1- -p
s- re
-p r—
CU 3
S- D5
S- CO
•r- S-
T3 4=
C U
re re
cu
CO
r— C VD
J3 -r- .
in T3
S- CU C
CU > 0
co "o -P
s- > u
S- C CO
1— 1 •!- CO
1
CU -C 3
r- O O5
.a T- co
ro -c S-
> 3: co
co j= .a
•i— in u
s- cu re -p
-P U CU T-
CO S-
s- 3 c -a
t. 0 -r- r—
•r- in 3
CO T3 0
•a J- co -e
c > in
re 4- i —
o o c
cu > o .
i — I/) C T- T3
J2 +J -r— 4J CU
•i- C Q.-P
in CO CO O C
S- E -Q CO
CO -P >, E
> •!- "O J- CO
CU E i— 0 r—
S- E 3 +J 0.
S- O O ra E
HH u 3: r— -r-
to
B-3
-------
B.2 GENERAL MODELING METHODOLOGY
The ISC model consists of two separate models: a short-term model
(ISCST), which normally is used for averaging periods of from one to 24
hours and a long-term model (ISCLT), which is used to predict seasonal or
annual average concentrations. In this study, the short-term model was
used to calculate one-hour, eight-hour, and 24-hour average concentrations;
the long-term program was used to predict annual averages.
Both the short-term and the long-term models use Gaussian plume dis-
persion equations and Pasquill-Gifford dispersion coefficients. For the
model features used in this study, the ISCST program corresponds to the
Single Source (CRSTER) Model, modified to include the effects of the separa-
tion of individual sources and the effects of area source emissions. The
ISCLT program, which is a sector-averaging model similar to the Air Quality
Display Model (AQDM) or the Climatological Dispersion Model (COM), makes
the same basic model assumptions as the ISCST program.2'3 However, the
ISCLT program uses Statistical Array (STAR) summaries (statistical tabula-
tion of the joint frequency of occurrence of wind-speed and wind-direction
categories, classified according to the Pasquill stability categories) to
calculate annual average concentrations. For the same source data and meteoro-
logical data base, ISCLT calculates annual average concentrations that are
equivalent to those obtained from ISCST.using a year of sequential hourly
meteorological data.
The source characteristics required by the ISC model are stack height
and diameter, stack gas temperature, and exit velocity. These parameters
are given for all sources of the model plant in Tables B-l and B-2,
B-4
-------
co
0
1— 1
co
t— 1
:E CHARACTER
o
N"*
CO
1
CQ
UJ
CQ
cu
E -P
to co
•r- S
O *— <•
CO U
> CO
10
•r- E
X >— '
UJ
Q. .
E D5
CU CU
H- Q
/— N
•P in
OJ CU
•r- +j
CU CU
ro
cu
-p ^
ro u
on cu
to
c
0 S-
•r- CU
m Q.
tn
E^-'
' UJ
CO
u
3
O
co
o
•p
to c
r— 0
3 -r-
O5-P
CU Q.
Qi 0
in E
co co uj to co r— r— t— i£> to vo o oooo in cto
co co o o CM i — i — i — o o o co oor^-m E £ ro i£>
oooooddddddco c\] CM co d s- r-in. x
r— OJ O -P O CU
U C -P
C CU TS
•r- EEC CU
3 .p rO .C
CO 3 t/3 -p
OOOOOOOOOOOO OOOO -P U3O
i— I— Cfta>r— r— r— i— r--OOOr— CM CM r— CO rO rO'r-j. C
GJ ^O ^i" *^r t-^3 ^^j3 OOOUJ "Q^5- "in CO^ U
i — i — i — i — i — i — i— i — , — i — CM COCOCD « CO-PO 3 ECU
"O -C 4- T-J LT) in
C H-> S- "00^
o os- to ro
0 -O "4- CU O E
CU C -P S. (T>
^_Sl/,-^ to to in cu cu || co
<^"'*'1 — ' — CMooooi'^-or-- CDI — ij:>>^- s- >>3to M- cu co-
un in i— i— CMr— i— r— i— 001 — oooo tO'r-"* *• ti,"
^— ^ ^"^ Nfc*^ *^^ NH^ >«.«/ SwX Sh«< UJ ^^ TJ yj ^^
— •. £ C O^ C CU 3s -P
CO CO ^D tO i — C\I CM CM CTJ ^" LO i — i— «^- ^3' i — S_ (>J «r- +— .
00)C CU COCO >, UCU+J tnl-0
•^ *= *- >- s-ccc r— ins-tn incucu
-P -i— -i— +J
COO CU CU CO COCOCOt/5 E E-r-U COT30)
S-tOCinooO 33 CJ5 (0 OJ.C E
Q- co cue o ooccc E i— c £ - ro
CC 3E S- S- J-OOO -i- OinCU T-ET-
0 i— 3 Q. Q. Q.-I- -r- -r- X • O CO J= r- 1- O T3
•>— C O i — 4J^4J fOin 31 — CUOCM
S- 0 1—0 CO CU CUfOrOfO ECU CUr— -a «£. «
co-r--\o > > . > s- s- s. m s-ro one
-C -P 0) o o orotoro tncu 3>. lino
O- tO C E JQ *~* f^ pt CL O »r- f~ (/) | ^ jj j \ ^rt
inr— CU3 to to tocucucu -p mx cu-i-x:
0 >,N3 COCOCO CUC CU-CU ^=+JO5||
E Jt CO i — i — r— -PCU S-T3-P -Pio-i—
^ ^ CUro r-r- ^-CMCMCM rot. O.S- -r-S-P
Q. U 3S -C S- CU O1
C -r- C -P fO -P .r-
2> o e s- 3s . o .. cu
"5 "^ "O in^^ZTS'r-r— CUS-S-
o) -^ 5i - ^j wo Q. cu>rore
rr ST. co f— EO oracu-r- cnoCcu
p cos- ocuuss-in-P s_
s- 22 "a. ^ °> -promu mi— ^i ro
4-> c JD cu s-u -P +Jm ro.e3cu curour—
C- COS- rOCU COCUC in OUC3. 33C1*-
O ~S° b i— • s- o S- s- cummin r--pT-
u cm
-------
gram to calculate hourly concentrations for a year of sequential hourly
meteorological data, the procedure generally followed with the Single
Source (CRSTER) Model. However, from a knowledge of the critical meteoro-
logical conditions affecting emissions from EB/S plants, it was possible to
select 12 24-hour periods representative of worst-case dispersion con-
ditions. ISCST then was executed only for those days to obtain maximum
one-hour, eight-hour, and 24-hour average benzene concentrations.
The worst-case meteorological conditions were selected as follows.
First, the ratio of the vector mean wind speed to the scalar mean wind
speed was computed for each day of 1964 from the Hobby Airport hourly
surface observations. This ratio has a value near unity for days of
persistent wind directions. The days with ratio values near unity then
were examined to select days with stable conditions and light-to-moderate
persistent winds. Based on analysis of the source characteristics and test
runs using the PTMAX dispersion model,7 these were considered to be worst-
case meteorological conditions for the small process sources (i.e., uncon-
trolled vents and condenser/absorption tower) and the fugitive and storage
sources. However, some days also were selected with unstable conditions
and light-to-moderate winds since these were shown on test runs to provide
worst-case conditions for the flares and boiler.
The objective of the dispersion modeling is to estimate maximum benzene
concentrations at any distance and at 0.3, 0.5, 0.7, 1.0, 1.5, 2.0, 10.0,
and 20.0 km from the plant. Therefore, receptors first were placed at these
distances along 30 radial axes originating at a reference point in the model
plant complex (shown in Figure 6-1).
Maximum concentrations produced by fugitive and storage tank emissions
can be expected to occur in the immediate vicinity of the plant production
B-9
-------
area. However, the dispersion coefficients used by the ISC Model (the Pasquill-
Gifford curves) do not apply at a downwind distance less than 100 m from
the source. Therefore, a ring of receptors was placed at 0.16 km which is
the smallest practical placement that would maintain a distance of at least
100 m between receptors and sources and still capture the maximum impacts
of fugitive and storage emissions (Se.e Table B-3).
Test runs were made using ISCST and PTMAX to predict the distance to
maximum concentration for each of the stack sources. These showed that the
ring distances mentioned above were sufficient to determine maximum short-term
concentrations and that an additional receptor ring at 5 km was necessary
for the long-term model runs.
B-10
-------
TABLE B-3. AMBIENT BENZENE IMPACTS FROM INDIVIDUAL AND GROUP SOURCES
(concentrations 1n uq/m )
[Source or Source Group
Process Emissions —
1 Uncontrolled
process Emissions —
1 Condenser/Absorption
1 Tower
1
Process Emissions —
1 Condenser/Absorption
1 Tower/Flare
1
Process Emissions —
Boiler
Fugitive Sources
storage Tanks
1
Hydrogen Separation
Section — TO Boiler
Hydrogen Separation
Section — To Flare
Annual Emission Rate
Hydrogen Separation
Section -- To Flare
Maximum Emission Rate
Hydrogen Separation
Section — Uncontrolled
Annual Emission Rate
Hydrogen Separation
Section — Uncontrolled
Maximum Emission Rate
Uncontrolled Model Plant
Plant with Condenser/
Absorption Tower
Plant with Condenser/
Absorption Tower/Flare
Plant with Boiler
1
8
24
Annual
Average
1
8
24
Annual
Average
1
8
24
Annual
Average
1
8
24
Annual
Average
1
8
24
Annual
Average
1
8
24
Annitdl
Average
1
8
24
Annual
Average
1
8
24
Average
1
8
24
1
8
24
Average
1
8
24
1
8
24
Average
1
8
24
Annual
Average
1
8
24
• Annual
Average
1
8
24
Average
0.16 km
7472
2579
999
338
1470
511
190
65
40
8
3
0.2
0.2
N
M
N
1016
263
121
i
27
1291
326
178
38
1
0.1
0.3
N
N
N
N
N
0.1
N
N
25
14
7
1
20373
10949
5693
7590
3060
1247
400
2410
992
423
127
1846
452
262
63
1846
481
262
63
0.3 km
11317
2777
1350
296
1781
438
218
46
45
12
4
1
0.5
0.1
N
N
501
130
48
10
1241
393
143
24
2
0.3
0.1
N
N
N
N
N
20
3
1
20
10
6
1
16032
7974
5077'
12305
3127
1541
330
3103
788
409
79
1423
491
191
35
1423
486
191
34
0.5 km
6783
1532
593
ISO
1059
236
90
23
29
17
6
2
1
0.1
0.1
0.03
218
52
17
4
763
223
74
13
3
1
0.2
0.2
0.2
N
H
N
124
16
5
16
7
3
1
12526
5234
2541
7350
1754
681
167
1795
458
177
39
912
285
95
19
906
268
89
17
0.7 km
4402
912
333
88
675
139
53
13
32
22
7
2
1
0.2
0.1
0.04
130
28
10'
2
449
128
43
8
2
1
0.4
0.3
0.2
N
N
N
137
32
6
14
4
2
1
10909
3079
1609
4845
1052
378
98
1164
279
95
23
583
177
59
12
562
155
52
10
1.0 km
2560
502
205
48
389
77
31
7
33
21
7
2
1
0.2
0.1
0.04
72
15
6
1
256
72
24
4
2
1
1
0.3
0.1
N
N
N
99
22
7
11
3
1
0.3
9035
2378
1043
2856
581
226
54
685
161
54
13
352
109
36
7
322
87
29
6
1.5 km
1394
310
113
26
211
47
17
4
29
16
5
1
0.4
0.2
0.1
0.03
39
9
3
1
142
43
14
2
2
1
1
0.2
0.1
N
N
N
81
19
6
8
2
1
0.2
6002
1716
600
1573
362
126
29
389
99
33
7
206
68
22.8
4
178
52
17
3
2.0 km 5.0 km
899
235
78
16 4
135
36
12
2 1
24
13
4
1 0.2
0.3
0.2
0.1
0.03 N
25
7
2
0.4 0.1
94
32
10
2 0.4
1
1
0.4
0.2 0.1
0.1
N
N
N . N
69
14
7
5
2
1
0.1 N
4240
1349
450
1018
274
91
18 4
255
74
25
5 1
140
51
16.9
3 1
119
38
13
2 1
10.0 km
108
54
18
1
16
8
3
0.2
5
3
1
0.1
0.1
N
N
N
3
1
0.5
N
12
6
2
0.1
]
0.1
0.1
N
N
N
N
N
28
7
4
1
0.4
O.I
N
651
321
107
123
62
21
1.4
31
16
5
0.4
v 20
10
3
0.2
15
8
1
Q.2
20.0 km
59
28
9
0.4
9
4
1
0.1
3
1
0.1
N
0.1
N
N
N
2
1
0.3
N
6
3
1
N
0.3
0.1
H
N
N
N
N
N
20
5
2
0.4
0.2
0.1
N
352
169
56
67
32
11
1
17
8
3
0.1
11
5
2
0.1
8
4
1
0.1
IN = negligible, i.e., <0.05^«g/in
B-ll
-------
B.3 REFERENCES FOR APPENDIX B
1. EPA, Office of Air Quality Planning and Standards. User's Manual for
Single-Source (CRSTER) Model. EPA-450/2-77-013. Research Triangle Park,
North Carolina. July 1977.
2. TRW Systems Group. Air Quality Display Model. Prepared for HEW, Public
Health Service, Washington, D.C. January 1970.
3. Burse, A. and J. Zimmerman. National Oceanic and Atmospheric
Administration. User's Guide for the Climatological Display Model.
Research Triangle Park, North Carolina. December 1973.
4. Airways Surface Observations (TD-1440) for Station 12918, Houston/
Hobby, Texas, 1964. National Climatic Center. Asheville, North
Carolina.
5. Mixing Heights for Station 03937, Lake Charles, Louisiana, 1964.
National Climatic Center. Asheville, North Carolina.
6. STAR Normalized Data (TD-9773) — Annual for Station 12918, Houston/
Hobby, Texas, 1964. National Climatic Center. Asheville, North
Carolina.
7. Turner, B. and A. Burse. EPA, National Environmental Research
Center. User's Guides to PTMAX, PTDIS, and PTMTP. Research Triangle
Park, North Carolina. June 1973.
B-12
-------
APPENDIX C
EMISSION SOURCE TEST DATA
-------
-------
APPENDIX C
EMISSION SOURCE TEST DATA
C.I INTRODUCTION
The purpose of this appendix is to present and summarize data gathered
during the development of a standard for benzene emissions from the production
of ethylbenzene/styrene. The facilities tested are described and the source
testing methods are identified. Any reference to commercial products and
processes by name does not constitute endorsement by the U.S. Environmental
Protection Agency.
C.2 SUMMARY
Three facilities producing ethylbenzene/styrene were tested to determine
percent reduction and exit concentrations of benzene. Samples were analyzed
for:
• C02, 02, N2, H2, CH4> by GC/TC.
• Benzene, toluene, xylene, ethyl benzene, styrene by GC/FID.
• Low molecular weight hydrocarbons (C-j-Cg) by GC/FID.
• Total hydrocarbons as benzene by FID.
• Moisture content by EPA standard method.
C.3 DESCRIPTION OF FACILITIES AND TEST PROCEDURES
Ethylbenzene for styrene production is made by (1) benzene alkylation
with ethylene and (2) extraction from mixed xylene streams. Styrene is
produced from ethyl benzene predominantly by ethylbenzene dehydrogenation.
The three EB/S facilities tested use benzene alkylation to produce ethyl-
benzene, followed by ethylbenzene dehydrogenation to produce styrene. The
basic process description for this manufacturing scheme is presented in
Chapter 3. ,
C-l
-------
The following sections briefly describe some process details for each
plant and summarize the test procedure followed.
C.3.1 Plant A1
This plant consists of two separate trains of equipment to produce
o
styrene. The total published capacity for the plant is 590 x 10 Mg/yr
styrene from 689 x 103 Mg/yr ethyl benzene. The emission test was conducted
on one train with a styrene capacity of 318 x 10 Mg/yr. At all times
during the test, the unit was running between 90 and 100 percent of capacity.
Testing was conducted at this plant during June 19. to 30 and July 10 to 14, 1978.
The purpose of testing was to obtain data, before and after control
devices for total organics and benzene, from the following systems:
t Benzene drying column vent
t Alky!ate degasser vent
t Catalyst mix tank
t Benzene/toluene column vent.
The benzene drying column vent system consists of a condenser used to
remove organics from the vent streams from the column reflux condenser and
reflux decanter. The exit stream from this final condenser is routed to
the plant flare system. Sampling was conducted befpre and after this
condenser. These gas samples were collected simultaneously into flexible
TedlarR bags at each location. Analyses were conducted on the bag samples
to determine C,-CC nonaromatics and benzene, toluene, xylene, ethyl benzene,
1 b
and styrene. The results of these analyses are shown in Table C-l.
The alky!ate degasser vent equipment consists of a polyethylbenzene
scrubber and a caustic scrubber. The vent stream from the degasser is
first scrubbed with a polyethylbenzene solution and then routed to a
caustic scrubber. After caustic scrubbing, the vent stream is routed to
C-2
-------
TABLE C-l. PLANT A: BENZENE DRYING COLUMN VENT SYSTEM
Species
Before Condenser
Average
Ppmv as Benzene
Average
Ppmv as Specie;
C-l
C-2
C-3
C-4
C-4
C-5
C-6
Benzene
504
17,264
1,512
209,020
478,627
8,740
45,750
23,299
2,157
3,915
2,363
274,340
628,202
8,877
38,608
23,299
C-l
C-2
C-3
C-4
C-4
C-5
C-6
Benzene
After Condenser
4,068
34,579
1,046
89,544
119,857
18,193
379,920
21,420
17,417
78,343
1,634
117,528
157,313
18,478
320,608
21,420
C-3
-------
the plant flare system. Gas samples were collected from the streams before
and after the caustic scrubber. Integrated samples were collected in
flexible bags for organic compound analysis and inert gas analysis. The
results of these analyses are shown in Table C-2.
The catalyst mix tank vent equipment consists of the catalyst storage
vessel and the mix tank where the catalyst is dissolved in a poly-
ethyl benzene solution.
The vent gases are essentially purge nitrogen streams that are used to
inert the system. The vent stream is scrubbed with fresh polyethylbenzene
solution prior to discharge to the atmosphere. Samples were collected to
determine organic species and hydrogen chloride content of the vent stream.
The results of these analysis are shown in Table C-3.
From the benzene toluene column, the vent stream from the reflux
condenser and accumulator is first passed through a chilled brine condenser
to remove organics prior to the steam ejector. The ejector exit stream is
passed through a vacuum condenser to a hotwell. The hotwel1 vent stream is
then passed through a final brine-chilled condenser prior to exhaust to the
atmosphere. Samples were collected from the non-condensible stream prior
to the ejector for organic species analysis. The results of these analyses
are shown in Table C-4.
In each of the analyses, benzene, higher molecular weight hydrocarbons,
and low molecular weight hydrocarbons (C-i-Cg) were determined in the field
by gas chroma tog rap hy with, flame ionization detection. Total hydrocarbons
were determined by flame ionization detection with no separation column.
Oxygen, nitrogen, carbon dioxide, and carbon monoxide were determined by
gas chromatography with thermal conductivity detection. Moisture analysis
was conducted using EPA Method 4.
C-4
-------
TABLE 02. PLANT A: ALKYLATE DEGASSER VENT EQUIPMENT
Species
C-l
C-2
C-3
C-4
C-4
C-5
C-6
Benzene
Before Caustic Scrubber
Ppmv as Benzene
C-l
C-2
C-3
C-4
C-4
C-5
C-6
Benzene
447.
1,928
—
1,760
19,804
5,348
225,139
36,988
1,914
4,368
2,310
25,993
5,432
189,991
36,988
After Caustic Scrubber
100
863
779
3,103
2,926
64,188
33,763
428
1,955
1,022
4,073
2,972
54,167
33,763
C-5
-------
TABLE C-3. PLANT A: CATALYST MIX TANK EQUIPMENT
Species
Average
Ppmv as Benzene
Average
Ppmv as Species
C-l
C-2
C-2
C-3
C-4
C-5
C-6
Benzene
1.74
1.23
11.07
1.28
0.38
7.44
2.78
25.08
2.0
0.50
32.47
32.47
C-6
-------
TABLE C-4. PLANT A: BENZENE-TOLUENE COLUMN VACUUM EQUIPMENT
Species
C-l
C-2
C-2
C-3
C-4
C-4
C-5
C-6
Benzene
C-l
C-2
C-2
C-3
C-4
C-5
C-6
Benzene
Before Steam Ejector
Average
Ppmv as Benzene
Average
Species
5.032
8,520
1,575
6,746
17,166
3,084
5,308
83,724
38,704
rpiiiv a» op<
21,544
19,304
3,569
10,541
22,530
4,084
5,391
79,653
38,704
After Hotwell
1,722
7,536
6,393
4,978
16,692
8,615
78,633
27,151
7,373
17,073
14,484
7.778
21,909
8.750
66,357
27,151
C-7
-------
C.3.2 Plant B
This plant has a maximum production capacity of 115,000 Mg/yr of
styrene from 125,000 Mg/yr of ethylbenzene. Styrene is produced at this
plant by benzene alkylation to produce ethylbenzene followed by ethylben-
zene dehydrogenation. Although this production method is the basic process
identified in Chapter 3, some process variations do exist. The Alkar
process by UOP is used by Plant B with a boron trifluoride catalyst. This
plant uses a dilute ethylene stream (50 percent) to feed the process, and
3
sends the offgases from the alkylation reaction to its boiler as fuel.
This process does not produce by-products or sludge and the catalyst lasts
4
for several years.
The test at Plant B was conducted during the periods of September 24
through September 28 and October 1 through October 5, 1979. Test samples
were taken of the process offgases, the fuel gases feeding the superheater,
and the exhaust gases of the superheater. Also, samples were taken of the
fuel gas and exhaust gas of the oil heater. The purpose of this test was
to determine the destruction efficiency of benzene in these combustion
devices and to determine the exit benzene concentration.
The superheater at Plant B was burning fuel gas from the fuel mix drum
and the dehydrogenation offgas from the styrene unit. The fuel gas from
the fuel mix drum consists of natural gas and the offgas from the ethylben-
zene unit alkylation reaction. The oil heater was burning the offgas/natural
gas mixture from the fuel mix drum.
p
In the actual sampling, integrated grab samples in Tedlar bags were
taken simultaneously at each of the two fuel inlets and at the flue gas
outlet of the superheater. Integrated grab samples also were taken simul-
taneously at the fuel inlet and flue gas outlet of the oil heater. A
C-8
-------
modified Method 110 was used for collecting the samples. The modification
was the replacement of the vacuum pump with an evacuated can. Gas flow
data from the superheater outlet were obtained using Methods 1 and 2.5
An analysis of the samples was made with gas chromatography, flame
ionization, and thermal conductivity detection. The volatile organic
compounds studied were analyzed with flame ionization detection; the
stationary gases were analyzed with thermal conductivity. Tables C-5 and
C-6 show a summary of the test results.
C.3.3 Plant C
This plant has a production capacity of 477,000 Mg/yr of ethylbenzene
and 380,000 Mg/yr of styrene and produces ethylbenzene and styrene by the
basic process described in Chapter 3.
The test at Plant C was conducted during the period of May 7 through
May 18, 1979. This test addressed two areas: vacuum column emissions and
combustion unit control efficiencies. Sampling and analysis were conducted
on the hotwell vents from the following vacuum columns: ethylbenzene recycle
column, polyethylbenzene column, and styrene purification column. Multiple
samples were taken at each vent, and flow measurements were taken both
before and after sampling. The second part of the test focused on sampling
the superheater fuel gas and flue gas. The heater fuel gas consisted of
natural gas, benzene/toluene hotwell vent offgas, hydrogen separation vent
offgas, and other fuel feeds from the process.
The data reported for Plant C may not be representative of the true
benzene emissions from that source. During the field test, the contractor
identified significant residual contamination in the sampling and analysis
apparatus. Several modifications to the procedures were made during the
C-9
-------
TABLE C-5. PLANT B: SUPERHEATER OUTLET TEST RESULTS
Run No.
3
Stack Benzene Concentration
(ppmv at sample conditions)
Stack Oxygen Concentration
(% v/ at sample conditions)
Moisture Content of Sample
(% v/v)
Corrected Stack Benzene
Concentration
(ppmv at 3% 02, dry)
0.3
3.15
6.28
0.33
0.4
4.61
6.45
0.49
6.6
5.42
7.6
5.60
C-10
-------
TABLE C-6. PLANT B: OIL HEATER OUTLET TEST RESULTS
Stack Benzene Concentration
(ppmv at sample conditions)
Stack Oxygen Concentration
(% v/v at sample conditions)
Moisture Content of Sample
(% v/y)
Corrected Stack Benzene
Concentration
(ppmv at 3% 02, dry)
1.3
4.16
5.53
1.52
3.63
6.45
3.98
6.24
C-ll
-------
test, with the overall result being that the final methods used were
significantly different from Reference Method 110. Using the results of
analysis of blank samples, it is estimated that contamination could have
accounted for at least half, if not all, of the benzene detected in the
stack samples analyzed.
Subsequent to this field test, laboratory evaluations were performed
to determine the precautions necessary to eliminate contamination when
measuring low-level sources. These improvements were incorporated in the
contractor's techniques prior to testing at Plant B and no contamination
problems were observed during that test.
Table C-7 shows the results of the superheater flue gas analyses.
C-12
-------
TABLE C-7. PLANT C: SUMMARY OF TEST RESULTS FROM SUPERHEATER OFFGAS
Stack Benzene Concentration
(ppmv at sample conditions)
Stack Oxygen Concentration
(% v/v at sample conditions)
Moisture Content of Sample
(% v/v)
Corrected Stack Benzene
Concentration
(ppmv at 3% 02, dry)
2.8 34.0 11.9
8.6 8.6
4.1 3.7
5.5 4.4 4.23 4.5
8.6 8.79 7.7 7.9
5.7 4.6 3.85 2.9
53.8 18.5 8.9 7.0 6.1 6.5
C-13
-------
C.4 REFERENCES FOR APPENDIX C
1. Hartman, M.W. Source Test at Cos-Mar's EthylI benzene/Styrene Plant,
Carville, Louisiana. TRW, Environmental Engineering Division.
2. Key, J.A. Trip Report for Cos-Mar Plant, Cosden Oil and Chemical
Company, Carville, Louisiana. Hydroscience. July 28, 1977.
3. Responses to EPA Section 114 Letters. Houdry questionnaires and
state emission inventory files.
4. Ethylbenzene (Alkar)—UOP Process Division. Hydrocarbon Processing.
November 1977. p. 52.
5 Hartman, M.W. Source Test at El Paso Products' Ethylbenzene/Styrene
Plant, Odessa, Texas. TRW, Environmental Engineering Division.
6 Hartman, M.W. Source Test at Amoco Chemical's Ethylbenzene/Styrene
Plant, Texas City, Texas. TRW, Environmental Engineering Division.
C-14
-------
APPENDIX D
EMISSION MEASUREMENT METHODS
-------
-------
APPENDIX D
EMISSION MEASUREMENT METHODS
D.I GENERAL BACKGROUND
Benzene will exist in the presence of other organics, except in the
case of systems handling pure benzene. Accordingly, methods for benzene
analysis consist of first separating the benzene from other organics,
followed by measuring the quantity of benzene with a flame ionization
detector. However, non-uniformity in procedures could exist in the fol-
lowing areas: (1) sample collection, (2) introduction of sample to gas
chromatograph, (3) chromatographic column and associated operating para-
meters, and (4) chromatograph calibration.
Two possible approaches for benzene sample collection are grab samples
and integrated samples. Since emission concentrations may vary considerably
during a relatively short period of time, the integrated sample approach
offers an advantage over the grab sample approach because emission fluctua-
tions due to process variations are averaged automatically. In addition,
the integrated approach minimizes the number of samples that need to be
analyzed. For integrated samples, both tubes containing activated charcoal
and Tedlar bags have been used. However, charcoal sampling tubes were
designed basically for sampling ambient concentration levels of organics.
Since source effluent concentrations are expected to be higher (particularly
since organics other than benzene could be present), there would be uncertainty
involved with predicting sample breakthrough, or when sampling should be
terminated. Bag samples also would offer potentially the best precision,
since no intermediate sample recovery step would be involved.
D-l
-------
Based on the above considerations, EPA considers collecting integrated
samples in Tedlar bags to be the best alternative. This conclusion is
confirmed by an EPA-funded report whose purpose was to propose a general
measurement technique for gaseous organic emissions. Another study of
benzene stability, or deterioration, in Tedlar bags was undertaken to
n
confirm the soundness of this approach. This study showed no significant
deterioration of benzene over a period of four days. Consequently, the
integrated bag technique was determined suitable; however, anyone pre-
ferring to use activated charcoal tubes has this option, provided that
efficiency at least equal to the bag technique can be demonstrated and that
procedures to protect the integrity of the sampling technique are
followed.
A collected gas sample can be introduced to a gas chromatograph
through use of either a gas-tight syringe or an automated sample loop. The
latter approach was selected for the reference method since it has a lower
potential for leakage and provides a more reproducible sample volume.
Several columns are mentioned in the literature which can be suitable
O A
for the separation of benzene from other.gases; ' most notable among them
have been 1,2, 3-tris (2-cyanoethoxy) propane for the separation of aro-
matics from aliphatics and Bentone 34 for separation of aromatics. A
program was undertaken to establish whether various organics that were
known to be associated with benzene, in stack emissions interfered with the
benzene peaks from the two columns. The study revealed that the former
column was suitable for analysis of benzene in gasoline vapors and that the
latter column was suitable for analysis of benzene emissions from maleic
anhydride plants.5'6 It should be noted .that selecting these two
D-2
-------
columns for inclusion in Method 110 does not preclude the use of some other
column(s). In fact, the method has a conditional provision for use of
other columns.
Calibration has been accomplished by two techniques, the most common
being the use of cylinder standards. The second technique involves inject-
ing known quantities of 99 mole percent pure benzene into Tedlar bags as
they are being filled with known volumes of nitrogen. The second technique
has been found to produce equally acceptable results; both are included in
Method 110.
D.2 FIELD TESTING EXPERIENCE
During the EPA field testing program, measurement studies were per-
formed at three facilities. A variety of individual process and atmospheric
emission streams were sampled. The measurement programs can be divided
into three major categories: (1) vents to atmosphere from sources such as
hot. wells serving column vacuum producing equipment, (2) inlets to and
outlets from organics recovery equipment, such as scrubbers or condensers,
and (3) fuel to and flue gas from combustion devices utilizing process
offgases as fuel. Each major category required different measurement
approaches and presented differing sampling problems. However, the basic
analytical techniques were the same. A gas chromatograph/flame ionization
detector system equipped with a Poropak Q* column was used to quantify
parafins and olefins in a general boiling point range designated as C1-C6.
Aromatic organics, specifically benzene, toluene, xylenes, ethylbenzene,
and styrene, were quantified using an 0V 101* or SPlZOO/Bentone 34* columns
as described in EPA Method 110 in a GC/FID system. Oxygen, nitrogen,
Trade Name.
D-3
-------
carbon monoxide, carbon dioxide, and, in some cases, methane and hydrogen
were determined by a gas chromatograph with a thermal conductivity detector.
The major measurement differences were in sample collection and pretreatment
of the samples prior to introduction to the analytical systems. Testing
experience with each category are discussed below.
(1) Atmospheric vents from hotwells: This stream generally can be
characterized as a relatively low flow rate source from a small pipe, with
total organics content less than ten percent by volume. The vent size was
usually two to four inches. Samples are easily accessible by placing the
sample probe into the end of the vent pipe. Because the total organics
concentration was relatively low, the gas samples could be analyzed without
any pretreatment. Volumetric flow rates were determined by using an aneometer
to measure flow velocity. The aneometer was installed on the vent pipe
using an adaptor section. Calibrations were performed in the laboratory to
relate flow velocity and pipe size to volumetric flow.
Testing at these sources was relatively straightforward except for
those cases where the hot well vent was at temperatures greater than 1,100°F.
In those cases it was necessary to install a liquid knockout impinger (with
no cooling) prior to the sampling bag to remove condensed water droplets.
Analysis of collected condensate indicated no collection of organics in
this impinger.
(2) Organics recovery equipment: These streams generally can be
described as the vent from a process vessel, such as a reactor, degasser,
or atmospheric or pressure distillation column, where the gas stream is
either all organic or high concentration organics. This type of stream
usually is routed to either a fuel or flare system and samples must be
D-4
-------
collected through some type of drain, or purge, or sample valve existing in
the pipework. Usually, the installation of a special sampling valve would
require a process shutdown.
Gas samples were collected using preconditioning systems in some
cases. At one facility, liquid carryover and condensing water and organics
required that an ice-cooled condenser train be used prior to the gas sample
bag. At other locations, a simple knockout impinger was all that was
required, or no pretreatment at all was necessary. In those cases where
condensers or traps were used, the collected liquids were recovered and
analyzed. In those cases where a cooled condenser train was used, either
as a sample pretreatment or to measure moisture content of the stream, a
two-phase liquid was formed in the condenser during sampling. This was due
to condensed organics in addition to water in the condenser.
Prior to analysis, it was necessary to dilute the samples. Metered
flows of concentrated sample and nitrogen were mixed into a second flexible
bag. Nominal dilution ratios of 10:1 were used. If further dilution was
necessary, the mixture resulting from the 10:1 dilution was again diluted
to obtain a 100:1 final mixture. Dilution ratios were confirmed by either
performing the dilution on a known calibration standard and analy±ing
before and after dilution, or by analyzing a component (CO,,, CH4, etc.) of
the sample and all dilutions.
The major analysis difficulties encountered were: on one test,
calibration standards were not available for the concentration range
encountered and it was necessary to assume linear extrapolation of lower
level calibrations (a dilution apparatus was not available at that test);
and in some cases where C-l through C-6 concentrations were much greater
D-5
-------
than the benzene concentration, there were some peak overlaps that caused
benzene results repeatability to be poor.
Volumetric flow rates were determined by plant process instrumentation
when available. An existing orifice meter was used to determine flow, and
temperature and pressure were obtained from process instrumentation. In
those cases where flow meters did not exist, no direct measurement was
possible since a process shutdown was necessary to install a meter. Also,
in one case it was not possible to use an existing orifice meter because
entrained liquids in the gas stream caused erratic pressure differential
readings.
(3) Combustion Devices: Tests were performed to obtain a complete
fuel analysis and a complete flue gas analysis, including specifically
benzene, at three combustion devices.
Fuel samples were obtained using the standard flexible bag technique,
with the exception that the fuel pressure was used to fill the bags. Due
to the fuel distribution arrangement at one site it was necessary to
collect samples of two different fuels that were fired in the process
heater since no mix drum was used in the system. After collection, the
samples were diluted to a nominal 100:1 ratio and analyzed for organics.
Fixed gases (N2, 0£, CO, C02) were determined from the undiluted sample.
Two major problems were encountered. First, even after a 100:1 dilution,
the methane content was erroneously determined using the GC/FID system.
Apparently the combination of high concentration and fast elution time from
the column causes saturation of the FID. This problem occurred during the
first combustion device test and was not confirmed until after testing was
completed, this difficulty was corrected for subsequent tests by including a
D-6
-------
specific methane analysis by GC/thermal conductivity. Second, it was
observed that where hydrogen makes up a significant portion of the fuel,
analysis for hydrogen and dilutions for organic analysis must be performed
as rapidly as possible. The permeation rate of hydrogen through flexible
bags is high and, after times as short as one hour, essentially all of the
hydrogen present will be lost.
The sampling and analysis of the flue gas stream presented a major
problem during the first combustion device test. After the first test run
at that unit, it became apparent that the Method 110 sampling apparatus had
been contaminated during earlier tests and could not be adequately cleaned.
The sampling system was changed so that essentially grab samples were
collected into glass flasks. However, the subsequent test results included
higher than expected organic results and inconsistencies. During the
analysis program, blank samples were prepared using the sampling techniques
with pure grade nitrogen as the gas source. Significant blank levels were
.observed. Also, it was not possible to sufficiently clean the chromatograph
sample injection system using air or nitrogen. Laboratory evaluations
after the field test identified the following information:
(a) After exposure to high level aromatics, the sampling apparatus
cannot be adequately cleaned by simple purging.
(b) Glassware exposed to aromatic compounds cannot be cleaned using
purging and solvent rinses alone. (This affected both the glass flasks
used for sampling and the glass syringe used for sample injection.)
(c) An unheated sample loop on the gas chromatograph does not purge
to zero quickly.
(d) The syringe used for sample injection was too small to adequately
flush and fill the GC sample loop.
D-7
-------
Because of the contamination problems encountered during the first
site test, changes were made to the preparation, sampling, and analysis
procedures prior to the tests at the next two sites:
(a) New sample bags were obtained and analyzed individually for
contamination prior to the field test.
(b) The sample acquisition tubing arrangement was changed so that no
part was reused from one test to the next.
(c) A heated sample loop and a sample pumping system were added to the
GC system.
During the performance of tests at the next two sites, the ability to
obtain essentially zero system blanks during the field analysis confirmed
that these modifications were adequate to avoid invalid results due to
contamination. These requirements are specifically addressed in Method
110.
The volumetric flow rate of the fuel was determined from plant instru-
mentation. Since in two of the three units tested it was not possible to
measure a flue gas flow rate, a combustion calculation using the fuel and
flue gas analyses was performed. This calculation was used to determine a
dilution factor for combustion so that a mass-basis benzene removal effi-
ciency could be calculated.
D.3 EMISSION MONITORING
No emission monitoring instrumentation, data acquisition, and data
processing systems that are specifically installed to measure benzene
emissions in atmospheric vents at ethylbenzene/styrene (EB/S) facilities have
been identified. However there are commercial systems available that incor-
porate automated gas chromatography with flame ionization detection and are
D-8
-------
equipped with automatic data processing systems. These have been specifi-
cally applied to ambient process area measurements for occupational health
and safety reasons. Various compounds or combinations of compounds have
been measured, including benzene. Since the only major differences between
systems designed for different compounds are the separation column selected
and the operating cycle time, there is no technical reason why currently
available instrumentation cannot be applied to monitoring benzene emissions.
EPA has not developed performance specifications for benzene emissions
monitors. The following criteria are recommended for the purposes of this
application:
(1) The monitoring system should consist of a sample acquisition
system which continually extracts a sample from one or more sources and
these samples are to be analyzed by gas chromatography using flame ioni-
zation detection (essentially an automated version of the reference test
method, EPA Method 110).
(2) If desired, the chromatographic separation step may be deleted
and direct measurement of total organics by flame ionization detection,
non-dispersive infrared spectrometry, or equivalent or alternative pro-
cedures may be used if all organics are assumed to be benzene.
(3) The emission monitoring system may be used to evaluate more than
one emission point but at least one data result for each location should be
available in each one-hour period. Assuming a 15-minute cycle time for
analysis, this would allow the concurrent monitoring of four emission
points. If a faster cycle time or fewer points are monitored, then the
number of results per hour would be greater.
(4) The analyzer system should be zeroed and calibrated daily. A gas
with less than 0.1 ppm benzene content is to be used for zeroing and a five
D-9
-------
ppm benzene mixture is to be used for calibration. The calibration mixture
may be prepared as a cylinder gas or by manual procedures as provided in
EPA Method 110.
The resulting emission data would usually be on a wet basis as present
at the emission point. It is recommended that the facility be allowed to
develop an average moisture content value for the purposes of correction to
a dry benzene concentration basis.
It is recommended that for the purpose of identifying and reporting
excess emissions a three-hour average be used. This is consistent with the
actual sampling time over which a performance test is conducted. The
arithmetic average of the benzene concentration data over three contiguous
one-hour periods represents the three-hour average.
It is estimated that the installed capital cost of an emission monitor-
ing system serving from one to four emission locations would range from
$30,000 to $50,000 depending on the system chosen and the installation
difficulty encountered. The annual operating cost can vary significantly,
depending on the time requirements for calibration and maintenance, and is
estimated to be between $5,000 and $20,000/year. The operating cost does not
represent a total annualized cost.
For those applications where an enclosed combustion device, such as an
incinerator, process heater, or boiler, is used to control the affected
emissions, it is recommended that monitoring of operations be used instead
of direct benzene emission monitoring. Suggested parameters that should be
monitored are the flow of the affected emission streams, the internal
temperature of the combustion device, and the oxygen content of combustion
device flue gas.
D-10
-------
EPA has not developed performance specifications for flow/meters or
temperature measurement devices; therefore, manufacturers' specifications
and installation, calibration, and maintenance directions are the only
available criteria for evaluation of the acceptability of any particular
device. For the purposes of operations monitoring, a flow measurement
device with a stated accuracy of + 5 percent and a temperature measurement
device with an accuracy of + 2 percent are adequate.
EPA has developed performance criteria for oxygen analyzers for those
facilities affected by New Source Performance Standards emission monitoring
requirements. However, these specifications only include stability criteria
and do not require a specified accuracy. Since the oxygen measurement
equipment in this application is an operations monitor and not an emission
monitor, it is recommended that the existing NSPS criteria for oxygen
analyzers not be applied.
The only recommended operating specification is that the analyzer
should be calibrated weekly using either air or a prepared cylinder standard.
The operating standards to be used as references are those existing
during the required emission test. Any time after the emission test
that the average process vent stream flow, combustion device temperature,
or flue gas oxygen content varies by more than a selected amount indicates
an excess emission interval. The time to be used for computing these
averages is three hours.
The estimated installed cost of each process measurement and recording
device combination is from $1,000 to $5,000. Therefore, assuming one
flow, temperature, and oxygen monitor, the total installed cost would be
from $3,000 to $15,000, depending on installation difficulty. The annual
D-ll
-------
operating cost of all process monitors combined is estimated to be from
$500 to $1,500. However, in many EB/S facilities, an oxygen analyzer is
currently installed and temperature measurement devices are installed in
essentially all facilities. In most cases, the only additional operations
monitoring equipment required would be flow devices.
D.4- EMMISSION TEST METHODS
The recommended emission test procedure for determining benzene emis-
sion concentrations is EPA Method 110. An integrated sample from the
source is drawn into a flexible bag over a one-hour sampling period. A gas
chromatograph/flame ionization detector system equipped with a column
selected for separation of benzene from the other organics present is used
for analysis. Unless a different value is known, the moisture content of.
the sample is assumed to be the saturation point at the sample storage
temperature at the time of analysis. The moisture content is used to
correct the measured benzene concentration to a dry basis. Three one-hour
samples are collected and analyzed.
For those applications where enclosed combustion is used as the control
technique, it is necessary to determine the oxygen content of the collected
sample. The recommended analysis procedure is by Orsat apparatus, as
described in EPA Method 3 (40 CFR 60, Appendix A). The oxygen concentration
in the sample is used to convert the measured benzene concentration (dry
basis) to a reference basis of three percent oxygen.
Subpart A of 40 CFR 61 (National Emission Standards for Hazardous Air
Pollutants) requires that facilities subject to provisions of that part
provide sampling ports adequate for the applicable test methods and plat-
forms, access, and utilities necessary to perfonn testing at those ports.
D-12
-------
Assuming that the test location is near the analytical laboratory and
that sample collection and analytical equipment and supplies are on hand,
the cost of field collection, laboratory analysis, and reporting of benzene
emissions from a single source is estimated to be $2,500 to $3,500 for a
compliance test effort. This estimate assumes an overall labor cost of
$25/person-hour and is based on triplicate samples and triplicate analyses
of each sample. This estimate could be reduced by as much as 50 percent
per source if several test programs are conducted concurrently. In
addition, if the facility has established in-house sampling and analyt-
ical capabilities, the cost of a compliance test effort could be lower.
D-13
-------
D.5 REFERENCES FOR APPENDIX D
1. Feairheller, W.R.; A.M. Kemmer; B.J. Warner; and D.Q. Douglas.
Measurement of Gaseous Organic Compound Emissions by Gas Chromatog-
raphy. EPA Contract No. 68-02-1404, Task 33 and 68-02-2818, Work
Assignment 3. January 1973.
2. Knoll, J. E.; W. H. Penny; and M. R. Midgett. The Use of
Tedlar Bags to Contain Gaseous Benzene Samples at Source-Level
Concentrations. Publication No. EPA 6004-78-057. Research Triangle
Park, North Carolina. September 1978.
3. Separation of Hydrocarbons, Bulletins 743A, 740C, and D. Supelco,
Inc. Belleforte, Pennsylvania. 1974.
4. Carle Instruments, Inc., Current Peaks, K) (1). Fullerton, California.
1977.
5. Communication from J. E. Knoll. Chromatographic Columns for
Benzene Analysis. October 18, 1977.
6. Communication from J. E. Knoll. Gas Chromatographic Columns for
Separating Benzene from Other Organics in Cumene and Maleic Anhydride
Process Effluents. November 10, 1977.
7. 40 CFR 60, Appendix B, Performance Specification 3.
D-14
-------
APPENDIX E
METHODOLOGY FOR ESTIMATING MORTALITY AND MAXIMUM RISK
FROM EXPOSURE TO BENZENE EMISSIONS FROM
ETHYLBENZENE/STYRENE PLANTS
-------
-------
APPENDIX E
METHODOLOGY FOR ESTIMATING MORTALITY AND MAXIMUM RISK FROM
EXPOSURE TO BENZENE EMISSIONS FROM ETHYLBENZENE/STYRENE PLANTS
E.I INTRODUCTION
This appendix describes the methodology used in estimating leukemia
mortality attributable to population exposure to benzene emissions from
ethylbenzene/styrene (EB/S) plants. The appendix is presented in three
sections:
• Section E.2 summarizes and references reported health effects from
benzene exposure. The major reported health effect is leukemia.
Mortalities cited include only the estimated leukemia deaths
attributable to exposure to benzene emissions from existing EB/S
plants although other, sometimes fatal effects are known to
result from benzene exposure. ,
t Section E.3 describes the method used to estimate the population
residing within 20 km of existing EB/S plants.
• Section E.4 describes the methodology for estimating benzene emis-
sions from a model plant, calculating expected population expo-
sures, and estimating leukemia deaths attributable to benzene
emissions from 13 existing U.S. plants.
E.2 SUMMARY AND OVERVIEW OF HEALTH EFFECTS
E.2.1 Health Effects Associated with Benzene Exposure
Numerous occupational studies over the past 50 years have documented
severe health effects in humans from prolonged inhalation 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 residing 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
f-1
-------
concern at environmental exposure levels: (1) leukemia (a cancer of the
blood-forming system), (2) cytopenia (decreased levels of one or more of
the formed elements in the circulating blood), and (3) 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 to be established beyond all 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 occur and death may
result from 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. The significance of
these aberrations is not fully understood. However, aberrant cells have
been observed in individuals exposed to benzene who have later developed
leukemia. Furthermore, some types.of chromosomal aberrations may be inherit-
able. In one study (too recent to include in the review ), workers exposed
to 2.1 ppm benzene for four years showed a statistically significant increase
in chromosomal aberrations (as high as tenfold) over those in unexposed
2
controls.
E-2
-------
The review concluded that humans may be the.only species yet observed
to be susceptible to benzene-induced leukemia.1 Evidence for production of
leukemia in animals by benzene injection was considered nonconclusive.
Although other effects, including pancytopenia, have been widely observed,1
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.
E.2.2 Benzene Exposure Limits
The health effects described above are based on benzene exposure at
occupational levels. That is, the benzene exposure levels associated with
the effects have been high (ten ppm up to hundreds of ppm benzene, except
in a few cases of exposures to two to three ppm benzene) or they are un-
known.
Benzene exposure was first associated with health effects in occupa-
tional settings. Therefore, initial attempts to limit benzene exposures
were aimed at occupational exposures. With the greatly expanded use of
benzene after 1920 and recognition of its toxic effects, several occupa-
tional exposure limits were established in the United States.3 These
limits, originally in the range of 75 to 100 ppm, v/ere successively lowered
as more information on benzene toxicity became known. For example, the
American Conference of Governmental Industrial Hygienists (ACGIH) recom-
mended a benzene threshold limit of 100 ppm in 1946, 50 ppm in 1947, 35 ppm
in 1948, 25 ppm in 1949, and ten ppm in 1977.3>4 The National Institute
for Occupational Safety and Health (NIOSH) recommended an exposure limit of
ten ppm in 1974 and lowered it to one ppm in 1976.5 The current Occupational
Safety and Health Administration's (OSHA) permissible exposure limit is
ten ppm (a more stringent limit of one ppm is currently in litigation7).
E-3
-------
Occupational exposure limits were initially established to protect
workers from adverse changes in the blood and blood-forming tissues. The
most recently recommended or pending limits of one ppm and ten ppm are based
C Q
on the conclusion that benzene is leukemogenic in man (NIOSH and OSHA ,
respectively) or a suspected carcinogen (ACGIH ) in humans.
E.2.3 Health Effects at Environmental Exposure Levels
Very little information is available on health effects of nonoccupa-
tional exposures of the general populace to benzene. Virtually all of the
studies cited '2 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 and to the general population
(including infants, the ill, and the elderly) requires extrapolation.
The recent analysis of benzene health effects concluded that the evi-
dence of increased risk of leukemia in humans from 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 leukemia in the
Q
general population from low-level benzene exposure. Data from three
epidemiological studies of leukemia in workers (mostly adult white males)
Q
were used to estimate the risk of developing leukemia. The annual risk
factor derived by CAG for benzene-induced leukemia is 0.34 deaths per year
per 105 ppb-person years. CAG calculated the 95 percent confidence intervals
for this risk factor to be 0.17 to 0.66 if exposure estimates in the three
studies are correct and 0.13 to 0.90 if exposure estimates are within a
factor of two.
E-4
-------
A no-threshold linear model was used to extrapolate this estimated
risk to the low levels (below five to ten ppb) to which some populations
may be exposed. Use of a "linear" model means that the model would predict
the same number of leukemia deaths among three million people exposed to
one ppb benzene as among six million people exposed to 0.5 ppb.
The risk factor was used in estimating the number of leukemia deaths
as attributable to benzene emissions from EB/S plants. Other effects of
benzene exposure (including fatalities 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. :
Four sources of uncertainty exist in applying the risk factor: (1)
the retrospective occupational exposure estimates may be inaccurate (howver,
CAG has estimated the confidence limits for this source of error); (2) 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; (3) the risk factor includes only leukemia
deaths and not other health risks; and (4) the risk factor assumes that the
susceptibility to leukemia associated with a cohort of white male workers
is the same as that associated with the general population, which includes
men, women, children, infants, the aged, non-whites, and the unhealthy. No
quantitative estimate of th error in the risk factor due to the latter
three uncertainties has been attempted.
E.3 POPULATION DENSITY AROUND EB/S PLANTS
The population residing within 20 km of each EB/S plant was determined
from the 1970 Bureau of the Census Master Enumeration District List (Med
List).
E-5
-------
Each plant site was located by latitude and longitude on a grid system
having grids of approximately ten square kilometers (km). The population
was determined from the Med List for each grid block within 20 km of each
plant site.10
Circles of radii of 0.1, 0.3, 0.5, 0.7, 1.0, 1.5, 2.0, 10.0, and
20.0 km were overlaid on each grid and the population within each annular
(ring or doughnut-shaped) area was determined. Where grid blocks over-
lapped two annular areas, the population was assumed to be uniformly dis-
tributed and was assigned proportionately to each area. The population in
each annular area was considered to be the population exposed to the es-
timated benzene concentration at the midpoint of that area, the estimated
total population exposed as a function of distance from the plant site is
reported in Table E-l. (All tables are presented at the end of this appendix.)
The method used contains potential sources of error. First, the
assumption of uniform population distribution within grid blocks and annu-
lar areas may not hold true. For urban areas the assumption is valid, but
it may introduce a degree of error for rural areas (which,is the case with
one plant) 10 to 20 km from the site. Another potential source of error is
the use of 1970 population data. However, these are the latest available
data in the form required for the model.
E.4 POPULATION EXPOSURES, RISKS, AND MORTALITY
" E.4.1 Summary of Methodology
A typical (or "model") EB/S plant was developed based on 8,000 hours
of operation per year for continuous process vents and fugitive emissions
and 8,760 hours per year for storage and excess emissions. Four continuous
process benzene emission rates exist for the model plant and each reflects
the use of specific benzene emission control equipment. These emission
E-6
-------
rates are 69 kg/hr for the model plant with uncontrolled process vents,
9.8 kg/hr for the model plant equipped with a condenser and absorption
tower system, 3.9 kg/hr for the model plant with a combination condenser/
absorption tower/flare, and 0.32 kg/hr for the model plant employing an
existing boiler to combust the process vent waste streams. In all four
cases, fugitive and excess emissions, with emission rates of 0.0009 and
0.84 kg/hr, respectively, were assumed uncontrolled for all four process
emission rates. A benzene storage emission rate of 27.0 kg/hr reflects the
current level of control for this source.
The Industrial Source Complex Dispersion (ISCD) model, urban mode two,
was used to estimate annual average benzene concentrations for the four
continuous process emission rates out to 20 km from the model plant.
Houston meteorological data, which best represent weather conditions for
most EB/S plants, were used in the dispersion model.
The area over which the benzene concentrations are estimated is a
polar coordinate map. The emissions source (plant) is at the center and,
depending on input variables, such as nature of the emission plume, local
meteorological conditions, and terrain, benzene dispersion characteristics
follow predictable patterns. The ISC model predicts approximate benzene
concentrations at any distance from the plant complex up to 20 km. Dis-
tances of 0.16, 0.3, 0.5, 0.7, 1.0, 1.5, 2.0, 10.0, and 20.0 km were chosen
for location of receptor rings. Graphically, they are best described as '
nine imaginary concentric circles surrounding the source along which, at
equidistant intervals, are 30 receptor points at which benzene concentra-
tions are predicted.
Benzene concentrations predicted at each of the 30 receptor points
were arithmetically averaged for each ring. This mean annual average
E-7
-------
benzene concentration for each of the nine ring distances was then plotted
on "log-log" graph paper. A curve fitted to these points shows that benzene
concentrations decrease at a predictable rate with distance from the source.
Population values were given for each of the annular spaces between
the receptor rings and were assumed to be equally distributed throughout
that space. Since benzene concentrations continuously are decreasing over
distance, an average concentration had to be determined for each annular
space. This was done by calculating the area midpoint for each annular
space and interpolating the benzene concentration on the curve correspond-
ing to the calculated midpoint distance. Except for fugitive emissions
which are independent of plant capacity, dispersed benzene concentrations
of that model plant equipped with controls most closely resembling those in
the existing plant were then scaled to reflect actual plant capacity for .
existing EB/S plants.
The population (1970) around each actual plant was correlated with
benzene concentrations of each corresponding plant to yield benzene dose in
ppb-person years. The methods for determining populations around EB/S
plants are described in Section E.3 of this appendix.
From health effects data, the EPA Carcinogen Assessment Group derived
a leukemia risk factor of 0.34 deaths per year per 10 ppb-person years
from exposure to benzene. The methodology for estimating the leukemia risk
factor is described in Section E.2.3 of this appendix. The leukemia deaths
per year attributable to exposure to benzene from EB/S plants were esti-
mated by multiplying "0.34 deaths per year per 106 ppb-person years" times
the exposure in "ppb-person years."
Maximum lifetime (70 years) risk is the probability that an individual
will contract leukemia from lifetime exposure to the highest maximum annual
E-8
-------
average benzene concentration downwind of a,plant. The 70-year lifetime
risk factor consists of multiplying the annual CAG risk factor (0.34 deaths
per year per TO6 ppb-person years) by 70 to yield 2.38 x 10"5. .The product
of this lifetime risk factor, the highest maximum annual average benzene
concentration, and one person exposed to that concentration yields the
maximum lifetime risk of contracting leukemia for that individual.
This appendix provides methodology and examples for risk and death
calculations for existing EB/S plants, assuming current control levels for
continuous process and storage emissions and uncontrolled levels for fugitive
and excess emissions. Leukemia incidence and maximum risk calculations for
each regulatory option can be done similarly according to the methodology
presented by using the appropriate emission concentrations and scaling
factors listed for that option in Tables E-2a, E-2b, and E-3, respectively.
Calculations for this chapter were performed by computer and inter-
mediate steps are accurate to four decimal places. However, the numbers
presented in the tables have been rounded to two decimal places for
convenience. Hand calculations therefore may result in a slight deviation
from tabulated values.
E.4.2 Continuous Emissions
E.4.2.1 Estimates of Leukemia Deaths
The general equation for estimating leukemia deaths attributable to
continuous benzene emissions (including fugitive and storage emissions)
from a particular plant (e.g., plant X) under current controls is:
10-20
> (1)
Dx = (R)
i = 0.1-0.3
in which
Dx = estimated number of leukemia deaths per year due to benzene
emissions from the plant (e.g., plant X) ' Denzene
E-9
-------
R = the CAG risk factor (0.34 deaths per year per 10 ppb-person
years)
P. = population in area (i) around plant X
B = mean annual average benzene concentration (ppb) in area (i) around
1 plant X for current control level (including uncontrolled emissions
from fugitive sources)
i = the particular area in which PI and Q- occur
I = summation of deaths from all areas (i).
Values for P. are found in Table E-l. Tables E-2 and E-3 were used in
calculating values for B which are presented in Tables E-4 to E-7. Table E-8
presents total exposure [ZCPpCB.)] for each source.
Leukemia deaths per year were estimated for each plant, using equation 1.
The total estimated number of leukemia deaths per year attributable to
continuous benzene emissions from all plants was determined by the equation:
'13
(2)
The total number of estimated leukemia deaths per year attributable to
benzene emissions from each EB/S plant at current control levels is presented
in Table E-9.
E.4.2.2 Example of Leukemia Death Calculation
Values in Table E-9 were determined according to the same methodology
as the following example calculations for the Cos-Mar Plant. The following
parameters are assumed:
R =
Dx =
0.34 x 10"5 (CAG annual risk factor)
10-20'km
(0.34 x 10~6)
i = 0.1-0.3 km
P. values for each area (i) around this plant are taken from Table E-l.
Model plant concentrations (B.) are from the plant equipped with condenser/
absorption tower/flare to reflect existing process control levels in the
E-10
-------
Cos-Mar facility (Table E-2). B. values for each area (1) around this
plant are taken from Table E-6.
Total
E.4.2
AREA
(i)
0.1- 0.3
0.3- 0.5
0.5- 0.7
0.7- 1.0
1.0- 1.5
1.5- 2.0
2.0-10.0
10.0-20.0
Deaths D
.3 Estimate
p
(persons)
15
27
41
87
210
310
7,650
60,100
= (.34 x 10"6
B.
(PPD)
15.5
11.6
8.23
5.72
3.36
2.05
0.25
0.07
)(8843) = 3.0 x 10"3*
P v R
Pi X Bi
233
313
337
498
706
636
1 ,913
4,207
8,843
of Leukemia Risk
The methodology used for estimating maximum lifetime risk involves
determining the highest maximum annual average benzene concentration at any
receptor at any ring distance out to 20 km from the model plant. Table E-2a
shows the highest maximum annual average benzene concentration for the
model plant (1) under uncontrolled conditions, (2) with condensers and
absorbers, (3) with a condenser/absorber/flare system, and (4) with boilers.
The highest maximum annual average concentration for the uncontrolled model
plant is 126 ppb. This figure is a composite of uncontrolled process and
fugitive emissions, and existing controls for storage emissions. To obtain
the highest maximum annual average concentration for existing plants the
model plant concentration which best reflects the emission rate and control
equipment used by each existing plant is factored for the styrene capacity of
the existing plant (Table E-3). The factored highest maximum benzene
concentration (Table E-10) is then multiplied by the CAG risk factor of
^Example calculation result differs slightly from that found in Table E-9
due to rounding.
E-ll
-------
2.38 x 10"5 to yield the maximum risk (Table E-ll) of contracting leukemia
assuming that an individual is exposed to that benzene concentration for a
period of 70 years.
Maximum lifetime risk
= (B the highest maximum per-plant annual benzene concentration
inm^b)
x (R, the risk factor (0.34 x id"6)) x (70 years exposure) =
2.38 x 10"5 deaths per lifetime per ppb-person years
= (2.38 x I0~5)(highest maximum B^ in ppb)
E.4.2.4 Example of Leukemia Risk Calculation
For the Cos-Mar plant, a highest maximum annual average benzene concen-
tration of 40.1 ppb resulting from fugitive and storage emissions is predicted
to occur 0.16 km downwind of the plant. Concurrently, a highest maximum
annual average benzene concentration of 9.0 ppb resulting from continuous
process emissions is predicted to occur 0.5 km downwind of the plant.
In order to obtain the combined highest maximum annual average benzene
concentration for the plant, the more significant source contributor to the
combined concentration is determined. (In this case, fugitive and storage
emissions predominate.) The remaining contributing concentration (from
process vents) predicted to occur at that receptor point are added to the
primary emissions.
In this example, the actual process vent contribution to the maximum
combined benzene concentration is less than the 9.0 ppb predicted to occur
at 0.5 km from the plant because, at 0.16 kilometers, the bulk of the
emission plume emanating from the tall flare stack has not traveled a
sufficient horizontal distance to reach maximum ground level benzene
concentration.
E-12
-------
The maximum lifetime risk to an individual residing 0.16 km from the
plant is calculated as follows:
Max lifetime risk = 40.1 ppb
Max B. due to
fugitive and storage
emissions at 0.16 km
0.5 ppb
Max B. due to
emissions from continuous
process vents at 0.16 km
X
2.38 x 10"5 (risk factor) = 9.7 x 10"4
E.4.3 Excess Emissions
E-4-3.1 Estimates of Leukemia Deaths from Excess Emissions
Ambient benzene concentrations resulting from emissions due to malfunc-
tion of control or process equipment, startup, and shutdown were modeled
exclusive of continuous process, fugitive, and storage emissions. These
excess emissions were modeled in a similar fashion as were the continuous
process emissions, i.e., a completely uncontrolled model plant of 300,000
Mg/yr styrene capacity was assumed. The benzene concentrations from the
model plant were factored according to plant capacity to predict benzene
concentrations, at distance, for each existing plant (Table E-7). Unlike
adjustments made for the continuous process emissions, where existing
control equipment in each plant was known, factoring for excess emission
rates for existing plants considered styrene capacity only because current
levels of excess emission control are unknown. Therefore, excess emissions
from existing EB/S plants were assumed uncontrolled for this analysis.
Incidence of leukemia deaths attributable to excess emissions is
derived in the same manner as that for continuous emissions. Average
exposure is calculated for each of the eight annular distances for each
plant to a distance of 20 km (Table E-7). Table E-8 presents total expo-
sure from excess benzene emissions as well as exposure from all sources.
Table E-9 presents estimated incidence of deaths due to benzene exposure
from excess emissions and from all other sources.
E-13
-------
E.4.3.2 Estimates of Leukemia Risk from Excess Emissions
Maximum lifetime risk from exposure to excess emissions is determined
by selection of the maximum annual average benzene concentration occurring
at any one of the 30 model plant receptors on any of the nine rings around
the model plant (Table E-2a). This concentration is factored (Table E-3)
to reflect existing plant capacity and the scaled benzene concentration
(Table E-10) is multiplied by the CAG risk factor to yield the maximum risk
to an individual of contracting leukemia due to lifetime exposure (Table E-ll).
Again, the risk estimated for lifetime exposure to excess emissions assumed
uncontrolled plants. Table E-10 presents maximum risk estimates for an
individual due to a lifetime exposure to excess emissions from one plant.
E.4.4 Validity of Estimates
Incidence and risk estimates presented in Tables E-9 and E-ll are the
actual midpoints of the ranges presented in Tables E-12 and E-13. These
ranges are based on a 95 percent confidence interval that assumes the
estimated concentrations to which the workers are exposed are within a
factor of two.9 The ranges presented here represent the uncertainty of
estimates made concerning the levels to which workers were exposed in the
Infante, Aksoy, and Ott studies that served as the basis for developing the
benzene risk factor.
In addition, other sources of uncertainty exist that are not quanti-
tatively expressed. The number of deaths was calculated based on an extra-
polation of leukemia risk associated with a healthy white male cohort of
workers to the general population, which includes men, women, children,
infants, the aged, non-whites, and the unhealthy. Furthermore, some uncer-
tainty is inherent in the benzene levels predicted by the dispersion modeling
to which people in the vicinity of EB/S plants are exposed. Moreover, only
E-14
-------
one effect of benzene was considered, i.e., leukemia. Benzene has also
been causally implicated in aplastic anemia, cytopenias, and the development
of chromosomal aberrations. Finally, the benefits to the general population
of controlling other hydrocarbon emissions from EB/S manufacture are not
quantified.
E-15
-------
to
•z. c
=D 0
O W
CO
to
00
to
<^
en
^
o
en
CD
0
to
•
0
«3-
CO
) •
- o
•
>
-P
I/)
o
CD
O
00
ci
o
^
to
to
to
to
^
0
en
r-^
en
00
CD
0
to
o
CO
CD
o
o
o
-------
i—
a.
LU
Q
s-
0£.
0
1 1
a.
o.
— *
2:
o
i — i
— i
— i
i — t
ca
§
oo
1—
C£
£
i — i
tr>
z:
o
i — i
fe
LU
C_3
2^
CD
C-'
LU
LU
LU
CO
LU
CD
•=c
LU
_,
^
z:
5
*^.
LU
S
S
h-
t-H
e
(O
CM
LU
LU
_l
CD
O
CM
0
CD*
r—
0
CD
CD
CM
CD
CM
LO
-
LO
*
0
'E' •-*
\s
* — *
LU
O O
to i^
1— 1 »
Q 0
r--.
•
CD
1
LO
O
LO
CD
OO
CD*
OO
CD
CO
CD
CONTROL
co
CD
OO
oo
CM
CO
.
OO
to
CM
^
CM
OO
•*
CM
CM
^
CM
°2
o->
Un control led
Process Emissions
CM
0
,0
•sS-
CD
0
.
O
»
oo
CD
^
oo
•
CM
CM
**
CM
CO
^
Process Emissions
Control led Wi th
Condenser/Absorptiol
Tower
LO
OO
CD
0
CD
OO
r—
O
0
LO
O
<~^
CD
CM
CD
o
CO
OO
0
CD
^
CD
O
^j.
OO
0
CD
LO
O
CD
'o
0
*
CD
Process Emissions
Controlled With
Boiler
o
CD
CO
OO
O
CD
^.
oo
CD
O
(O
CD
CM
0
0
CM
CO
r--.
oo
LO
CD
Is*.
1
to
CO
c
0
CO
"O CO
re'E
LU
cu
>• CO
•tf £
cno
O
0
CD*
O
CD
CM
CD*
CM
O
CO
00
CD
O
CO
CD
LO
0
^_
CD
CO
,_•
Excess Emissions --
Uncontrolled
j_
«<£
O CO
•U *CU
O CU
a t—
(tj O
•r- t-
i. E
OS O
0_ 0
o b
"O CO
CO TD
""~
CO 4->
•r- E
CU
O) S-
•U S-
2 0
C CO
O -l->
•r- 0
co cu
CO i—
cu -a
a> c:
**0 n*5
s-
o «
CO O
cu ^~
1— en
•r-
* re
CU c"i
r~ r—
"o
£» tO
C -C
O 4-^
0 -i-
13
-a "S
CD 4->
ic
CO
03 CO
CU £= .
S- "3 CO
Fugitive emissions i
e.g., some storage
benzene storage tan
E-17
-------
CO
UJ
8
UJ I
UJ ca
C5J O
(O O
i— CQ
0)
0) U_
O) OJ
"c o
LO
CM
in
•
CO
O)
to
O)
o
to
QJ
O
cr>
r-~
10
CM
LO
•
oo
O)
X O
CO CQ
UJ
3 UJ
H-4 O.
UJ
(U S-
10 OJ
C S
OJ O
•Oh-
O E
O O
td
i—
o.
(O
•
oo
O
O)
4->
M- (O
0)
CD
(O
•P c
"O •!— "O -r-
O)
CO
10
to
to
o;
UJ
l-O
O
(O
O O
o c
o o
o o
10
(d 3
o c
o o
o o
o
E-18
-------
CO
Q£
O
£-
•,- O
-P -P
•r- CO
en
LL.
/•—s
in
in c
0) O
(J • !—
O -P
s- a.
Q- 0
CO
in
in c
cu o
0 -P
S- Q.
0. O
^
<3C
in c
0) o
O "1—
O -P
S- Q.
0_ O
i—
O
•s-
•P
C
in o
in o
CO
U -P
o c
i- CU
o. s-
s_
3
CO
»~s
S-
LU >, >>
Z2H +^ ^-S
LlJ *i— &)
Q£ O 5T*
h- CEO
CO 10 O
CO r—
—•
*^~
«=t
i
Q-
in
•—
CD
r—
CD
O
in
r—
en
o
CO
o
o
in
o
cu
o
n:
*
s
o
00
CO
0
in
o
r-
o
CO
0
o
.a
CO
0
o
-Q
**•
0
in
o
in
a.
LU
en
o
CD
0
<*
0
CD
0
-Q
CM
i —
_a
in
CM
CM
q-
"3
CM
CM
O
CM
O
CM
O
0
CM
O
CD
CM
O
o
CM
:-
O
00
co
o
c
IO
in
c
o
m
, —
in
•-
CD
0
in
,—
in
^_
^
0
CD
m
CO
c
to
o
CM
O
CO
o
in
0
o
u
in
0
o
CO
o
co
0
o
co
CO
r_
•r—
O
C
CO
00
p
,—
o
i—
o
0
u
r—
o
0
u
0
0
u
o
o
m
cu
CO
-P
co
CO
ZD
O)
ro
i-
o
co
.
cu
o in
i. r-
-P CU
c >
o cu
U r—
3 i—
O
•a s-
£ S=.
3 O
in u
in
to -P
c.
CU CO
S- i-
(0 i-
in o
U +J
s- o
3 CU
O i—
tn <*-
CO
c
o ~a
•r- C.
in to
in
e at
CU i—
in o
in s-
CU -P
u c
X 0
cu o
in >s
fO i —
i— IO
CO -P
S s-
to
in a.
to
CO
m s-
a> to
u
s- in
3 CO
o o
in S_
3
c: o
o in
in c
in o
•I™ •!•"•
e m
co in
cu 'E
> cu
• \ \y
•r- C
en to
3 -P
i ;
fO
cu
s-
(0
5-
CO
3:
O
•p
c
o
-p
Q.
S-
O
in
-Q
to
"^s^
1-
cu
c
d)
T3
C
O
U
JC
+J
*5
4J
r**
«
"a.
r—
CO
•a
0
e
0
i-
r—
O
J=
-P
'5
•p
c
IB
"a.
CO
TJ
o
E
O
S-
4-
c
o
•p
fO
S-
+J
£Z
CO
u
c:
0
0
CO
£=
-------
CO *—'
z
o co
M C£
z: o
LU —I
O I—I
o
o;
LU
LU OL. >—t —I
CO. HjUJ
LU O CO LU
t3 o£ i—i I
eCLL.Ce:
LU LU Z O
> o o a:
_J 1— (— O
-i
o
i
CD
O
cn
in
co
r—
CD
O
O
CM
o
CD
00
CO
O
I
CO
CM
CD
CO
u
}
7
3
?
J-
D
CM
O
CD
0
0
ISs^
ID
O
"=rt~
0
CD
CD
°
o
d
CO
in
o
CD
CO
0
CD
CM
CD
in
cn
CD
oo
0
o
CD
cn
o
o
CD
cn
o
^^)
CM
CD
f""i
f*"t
f"*^
c^
tn
0
o
o
CD
CD
en
CM
o
o
o
CD
3
3
T
si
0
r>
o
*•.
n
x>
30
CD
0
CM
CD
CO
o
in
CD
CD
CM
,_!
CO
o
in
CD
•
0
o
CM
•—
CM
Cs3
CO
o
<*
•>r
m
CM
o
co
-^
m
CO
in
in
r>>
CD
i —
CM
CO
•^h
CO
CM
CM
*
CD
^
uu
CD
O
in
•
CD
r-
CM
CO
*
CD
CO
cn
^•j
CD
O
CD
CD
O
r~
CD
cn
•~
CM
CO
S-*^
CD
CD
CD
^
CD
CO
CD
CO
CD
CD
O
cn
CD
£
r—
co
1 —
CD
1 —
CD
CO
o
CD
CM
CO
CD
CD
00
CD
0
CD
O
CO
O
CD
CD
cn
r—
*
CM
CO
en
csi
•=*•
CO
CM
°
in
CD
00
CM
CD
cn
CD
co
cn
CD
CD
CO
CO
O
o
o
CD
0
O
0
CD
CO
O
O
O
CD
m
o
o
o
CD
CD
CM
O
U
O
in
co
-------
LU
o
I
co
(—1
Q
o
o
CM
J
o
o
1
o
c
o
c\
o
c\
in
*~~
LO
r—
1
O
O
r—
1
O
iv_
0
1
in
o
o
i
co
t-*\
CO
0
CD
'
O
CM
O
O
CO
O
0
LO
LO
O
ID
cn
o
en
<—
O
CM
CO
CM
CO
ID
CO
•—
'
I/)
U
cu
o
zn
E
CM
O
O
LO
co
CO
5-
0
CM
o
co
i—
0
CM
£
«3-
CO
CO
O
O
o
<
.
o
o
o
o
CO
o
o
CO
r—
o
CO
CO
o
0
CO
,_
CM
CM
•1—
CO
c
CO
o
0
o
o
o
0
co
o
o
CD
o
o
;z
o
o
CM
o
00
CO
o
1—
f-.
0
cu
cu
CO
co'
1
o
u
"ird
'•p
(0
Q.
cu
re
CO
cu
.u
£•»
o
U)
o
•r™
ft\
y^
£
cu
CO
re
S-
o
.J
•T**
CO •
,1^-*
o
T3 -p
cu c
<— o
i— O
o
£— CU
•P C
c. cu
O N
u c
c: cu
3 JO
•a if-
O) O
e
3 V)
LO CU
re >
cu
CU r—
fi-
re -p
c.
in cu
cu s-
u s-
i. 3
3 0
0
in -P
o
O i —
•r- 4-
(/) CU
in s-
cu c.
re
cu
•r- CU
•p 1 —
o)1©
=; s-
E-21
-------
to ^—'
o co
Hoi
)— UJ
•eth-
OS UJ
(— S CO
Soz
UJ —IO
O H-l »-*
CO
CO
o :
UJ:
UJ
UJ
; CD
UJ Q. OS
UJ O CO
CD Oi
• CJ
•
«C CO M
rp i-« I—
Z Q KH
UJ J3 CO
5E O.V5
Q.UJ
Q V--CJ
:§
CM
00
CO
in
in
in
en
CM
CD
CO
ro
u
in
en
co
in
oo
CD
u
o
oo
co
co
00
o
o
CM
CO
00
CM
CO
LT>
cn
en
CO
CO
CO
co
cn
in
CM
CM
o
CD
CO
CM
CM
CO
00
00
CM
CD
CM
CD
O
CD
CM
00
CO
CO
CD
oo
CD
cn
CM
CO
CD
CM
CO
in
CO
cn
m
CM
co
CM
cn
co
CM
cn
cn
ro
CO
o
o
co
co
cn
CD
o
co
cn
co
o
cn
in
in
co
co
CD
O)
LU
xs.
I o
co
o
CM
CM
CD
co
CM
CO
CO
co
0)
CD
co
E-22
-------
*£
•^
CO **-s
O CO'
>— i o;
1— LU
< l—
a; LU
1— S
z o
1 1 f 1
O 1— I
0
1—1
LU
Z 1— Z
LU Z OfO
M «=c HH a
Z _l h- LU
LU a_ za s:
ca ca =>
S 1-1 CO
LU o a: to
C3 Qi H™ ^C
0 >
cn
o
o
0
o
0
o
CM
CD
O
CO
O
o
LO
o
o
cn
o
o
o
r~
-P
to
r"*
O
CO
o
o
f^^
00
o
:-:
CD
CD
to
0
i/J
to
ex.
LU
0
CD
CD
i —
O
O
r—
O
cn
o
CO
CO
o
LO
!O
0
CD
cn
0
to
r-^
|:
cn
0
CD
o
CM
o
CD
CM
o
CD
CM
00
CD
to
CO
r— •
CO
CO
CM'
CO
to
co'
Monsantc
tO
o
0
CD
CM
CD
CD
00
o
CM
CO
o
LO
CD
cn
0
cn
LO
pJ
CM
CM
CO
to
X
o
LO
o
0
o
0
0
o
o
o
CM
o
CD
CD
O
to
CD
0
O
o'
to
0*
• f—
o
c.
3
CO
o
o
o
o
CM
0
CD
CD
CM
O
O
O
o
to
CD
O
.-
0
cn
CD
cn
CM
CD
r— •
CO
CO
-P
to
to
>,
CO
CO 0
'o.-p
E c
O CO
U E
0)
"-P
C 10
o in
C i.
J^ CO
c >
3 O
CO C C
•i- O O
O •!- -I-
•r- +-> W
^ O. tn
tn CO
O to us
is!
CO i.
> . 0
•i- ~a 4-
O E W
co 3 c:
o. in o
in in -r-
co to -P
i- (0
CO i-
c to c:
to co
in u
-P c c
coo
CO •<- o
E -P
Q.T- CO
•r- "O C
3 C CO
cr o N
CO U C
CO
i— "O -Q
to co
3 •— D)
•P r- C
U O -i-
(0 i- r—
•P •!-
co c re
U 0 >
C 0 CO
•r- C S-
CO 3 Q.
re
E-23
-------
a:
eC
LU
>-
z
o
to
C£
LU
a.
ca
a,
Qu
u.
o
LU
to
o
LU
LU
CO
CO
I
LU
ca
u
X
LU
a>
D)
re
s-
o
TJ
c
to
O)
'-p
*5)
3
UL.
in
in
o
s
en
o
CO
cn
0
ID
ID
cn
CO
IT)
o
u
0
cn
CM
CM
cn
ID
CM
CO
CO
CO
cn
(D
1 —
D_
o
o
•=c
CM
cn
CO
ID
1 —
0
cn
CM
r—
ID
^"
X
H-
o
u
i-
^
o
CO
0
1—
LO
CO
LO
LO
s~
s:
in
o
o
r-.
LO
co
r—
CO
o
ID
r~.
0
CO
r^
X
t—
o
a
ID
i —
o
cn
r—
cn
CM
•~
«.
CM
r— '
t— I
O
a
CM
CM
CM
ID
CO
O
LO
1
CM
0
o
in
(0
a.
LU
0
cn
•-
0
cn
0
LO
ID
0
ID
O
r—
f-
3
CJ3
ID
LO
CM
r— •
*J" '
r^
CM
CO
«=*•
r—
Monsanto
cn
CO
CO
CO
ID
'
cn
CO
* '
Oxi rane
ID
(D
-------
LU CO
>- -J
UJ
az>
UJ LU
Q I
CO _l
DC o
I— ee
+j
O5
-3
LJ_
U)
U)
ID
U
0
S_
CM
1
CD
X
10
>=*•
•=*•
•CD
X
CD
•~
CM
0
X
CM
•— '
CM
CD
X
"*"
CO
in
.c
U
ID
O
31
£
CM
|
0
X
•~
CO
1
o
X
•—
CM
CO
1
CD
X
l**t
(JD
CO
O
X
00
o
U
o
CD
X
•**•
->
ro
to
c:
o
CM
1
CD
X
•—
i —
ro
CD
X
i—
,—
CO
CD
X
CD
CO
•o
CD
X
•*
to
a>
c.
ro
*X
CD
CO
0
X
ro
CM
•=i-
'(-,
X
CM
r—
CO
1
CD
X
CM
r—
1
CD
X
<3-
CD
,—
O
c.
CO
no
i
O
X
1 —
'
•3-
X
<3-
LO
no
i
CD
X
CD
, —
1
CD
X
CM
1
a>
O)
CO
•r->
X
CM
r-t
CM
1
X
(-)
r-t
CM
I
X
CM
,-j.
'',_,
X
•=*•
to
— 1
o
C£
|
CO
ZD
CD
1 — 1
E-25
-------
o.
CO
•a
a>
o
CM
CM
cr>
Csl
10
tO
00
LO
00
00
LO
LO
Cvl
00
UJ
to
t/>
10
£
UJ
o
o
o
UJ
UJ
o
az
UJ
0)
O
en
CXI
CO
ro
CM
tD
Csj
CO
co
CO
CM
CO
CO
UJ
m
UJ
UJ
to
01
(0
to
(0
0)
CM
CO
Csl
CM
en
CO
CM
^*^»
3
to
0)
to
•
co
CM
00
CO
co
m
en
(2J
LO
00
CO
CM
u
00 O
CM
ID
o
U
ID
co <2>
CM
CM
10
en o
CO
to
rH
UJ
u
a)
u
o
a.
LU
E-26
-------
o
o
UJ
co
LU
o
o*
O
CO
•z.
o
I—I
CO
co
j—i
S
to
c.
o
•f—
10
to
LU
W
3
O
C
O
o
t/)
C
O
LU
o
X
a>
O5
(0
o
-P
co
fO
o>
3
la-
Process
•a
a>
c:
JQ LO
E 00
O
CO
<*•
CD
OJ
IT)
esj
CM
O
O>
-a Q) i—
c c as
D5 C >,
s= a) ra
'£ M °
o o
e
o
c
" C
O O)
• -Q
E O
« 3 £2
if) E ro
•r- O O) J= -P
r— E -P
3 C TJ C
o o c in a)
r- .r- (TJ ••- O
oxoioretDi— o
, re ^= o a) o
ro4-+J
occ
CO
co
O
CO
o>'i-.pj--pc
•r- £ C (O O)
re re
E O)
s-
0) u»
EtOO)
030C
•!- O
C O) JD C
CT) (SF
o
O LT>
0 r—
• o
o o
0
0
o
rots 4->-a
O)O)a)co)
.c 4J s_ re .p
fOO
io
OJ-P
OS-O)-PO
OOO-P
-r-C'r-C:
«,£ ,2
'r-S-OO)
5 , S W 0 0 JC
S-P-P 4-J +i
tooreroj=ooE
.
a>
a>
CO
CO
re E E O) -P O)
re s- s- o> c s-
• c x o) jz o) re
>> o o) -P ~a -P s-
i— •!— o) c re 10
0 0) 0 0
oos-ow
10
S. -r- S- CO
o-a.p.,-
ro
re
i—
a.
q-
o
c
o
T3
CO
10
O
O
a>
10
o
o
g:
O)
>
O E
O
3
OCD
£=
O)
o
o
10
-^ ro ^: re
CO i— i—
^
UJ
is
g co
O O-P O
1= -r-(3 ."
0 +J +»
o
3
OCD
g
N o to o to
a) o E o 'E
-a o a> o o»
O) 0) "O O) "O
c c
c c
re re
! "
X
ro
I 3
• ja
o
re
x
ro
a».
-a
E-27
-------
t/)
OS
UJ
H*4
Sn
u.
HH
s
g
| |
tfC
i^TS
•
p~"
1
p 1
CD
^
•
UJ
o
o
to
o
I— H
to
to
J— 1
LU
1
Q-
(A
c
0
•r—
in
(A
•r-
LU
(A
3
O
C
'•p
o
o
1
c
•o
cu
c
•r-
e
o
o
in
c
o
in
(A
LU
IA
0)
u
X
UJ
cu
O)
re
j-
o
•p
co
-o
c
re
cu
•r-
5>
3
U_
w
in
cu
o
o
J-
0.
1
O •=
i
O
r~
X
^
CM
*
X
•=1-
r—
^.
0
r—
X
ID
t^
co
i
o
X
LO
r—
•P
in
JZ
u
cu
o
n:
-------
C_3
O
CO
co
CO
Im
in
C
0
in
LLJ
in
O
3
C
-M
O
<_>
-O
O)
C
•r—
.a
o
o
in
c:
o
in
in
S
LLJ
in
in
cu
u
x
LU
CU
O5
re
s_
0
4->
co
-a
re
O>
-i->
O)
u_
in
in
CU
U
0
a_
0
X
CM
C\f
0
X
1^.
r—
=3-
0
X
o
LO
i
f>
o
X
CM
i^
CM
i
CD
X
CO
ro
CD
X
•3-
•=»-
CM
CD
X
0
CT)
CM
CD
X
co
r—
4-»
in
_c
O
cu
o
3Z
CM
CD
X
CO
CM
no
i
CD
X
CD
•=t-
no
i
CD
X
LO
LO
O
U
O
i
CD
X
CM
CM
0
X
•*
no
i
0
X
0
CM
•>t"
1
CD
X
CO
CM
LO
0
X
o
<5f
no
CD
X
CO
<£>
1
•*
CD
X
CO
en
CO
CD
X
CD
r--
i
•o
o
X
0
X
*t
1
CM
I
CD
X
«3-
r—
no
O
X
CD
CM
no
CD
X
r-.
i
<*•
i
CD
X
•3-
CM
no
i
CD
X
CM
**•
1
CD
X
|-~-
•-
CM
CD
X
I
"O^
CD
X
VJD
r—
l—t
E:
*>
-s
o
a
no
i
CD
X
CD
CM
<>r
i
o
X
CO
CM
•*
CD
X
i — •
co
i
f>
i
CD
X
LO
-it-
no
CD
X
CO
r— •
1
<5h
CD
X
cri
—
«=t-
CD
X
f~~
CO
•>
CD
X
•5t-
LO
O
in
re
a_
_LJ
no
CD
X
r~.
i
•=1-
i
CD
X
LO
CM
•3-
CD
X
r»«
i
r>
CD
X
*3-
CM
=f
1
CD
X
cr>
LO
un
CD
X
**•
CO
'i-
CD
X
LO
CT>
I
•=*•
CD
X
•=f
<<-
"3
C3
CM
CD
X
CD
^>
no
O
X
r~
LO
CM
i
CD
X
CD
no
i
CD
X
<*
I—
CM
CD
X
cr>
CM
i
no
i
O
X
CM
•=*•
CO
CD
X
LO
1
=*•
CD
X
CM
Monsanto 2.
CM
,CD
X
CD
CO
1
no
i
CD
X
co
«3-
no
i
CD
X
CD
CO
•si-
CD
X
co
«3-
CM
CD
X
CD
1
no
I
CD
X
LO
•-
CM
CD
X
r»-
i
-o
CD
X
«*•
CM
O)
C
re
X
no
i
CD
X
0
1
•3-
CD
X
«3
CO
*3-
CD
X
co
i
LO
1
CD
X
•3-
«3-
no
CD
X
CM
CO
1
•si-
CD
X
IT)
^
CO
CD
X
LO
CM
«3-
CD
X
l£>
CO
r~
CD
C
3
CO
no
CD
X
CM
«3-
•=1-
O
X
no
i
o
X
«3-
1—
•5J-
1
CD
X
CM
no
i
CD
X
r^
CM
i
•*
CD
X
cn
co
>^-
o
X
CM
co
LO
CD
X
LO
«*
"cu
cu
4->
CO
co
~>
n—
1
CD
X
r— .
CO
CM
1
CD
X
CD
-3-
CM
CD
X
V£>
CM
CO
CD
X
co
co
CD
r—
X
CM
1
CD
X
r~
—
i
CD
X
t^-
1
CM
1
CD
X
<3-
CM
*Z
r> —i
=D
-------
CO
o
CO
oo
CO
o
co
o
•a
cu
co
CM
X
CO
x
in
X
m
in
to
to
co
co
CM
CM
to
in
LU
co
oo
«O
X
IO
CO
en
in
m
cn
co
CM
co
oo
CM
LU
o
in
o
o
LU
o
LU
in
a>
UJ
to
to
UJ
o±
a.
x
LU
to
h-«
02
LU
0)
u
X
LU
LU
to
to
to
in
(A
LU
CO
in
co
•
m
CO
a>
O)
(0
CM
0)
x
Cn
in
CM
cn
00
X
in
in
in
in
CM
CM
in
o
X
CM
in
CM
to
CO
CM
co
CO
co
CM
in
in
CM
co
tn
CM
CO
co
in
cn
Cn
CM
co
CM
00
CM
•
00
CM
in
o
CM
oo
CM
oo
un
co
od
in
in
CM
co
co
CM
*
co
in
m
CM
o
10
co
x
(O
CM
to
CM
x
00
CO
X
co
CO
in
co
co
m
CO
00
CO
LU
LU
en
to
to
LU
a>
0)
a>
to
fd
E-30
-------
E.5 REFERENCES FOR APPENDIX E
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Assessment of Health Effects of Benzene Germane to Low Level Exposure
U.S. Environmental Protection Agency. EPA-600/1-78-061. September
1978.
Picciano, 0. Cytogenetic Study of Workers Exposed to Benzene.
Environmental Research (in press). 1979.
In:
Criteria for a Recommended Standard — Occupational Exposure to Benzene.
National Institute for Occupational Safety and Health. HEW Publication
Number (NIOSH)74-137. 1974.
Threshold Limit Values for Chemical Substances and Physical Agents in
the Workroom Environment with Intended Changes for 1977. American
Conference of Governmental Industrial Hygienists. 1977.
Revised Recommendation for an Occupational Exposure Standard for
Benzene. National Institute for Occupational Safety and Health
August 1976.
Occupational Safety and Health Standards, 29 CFR 1910.1000, Table 7-2.
Occupational Safety and Health Administration Publication OSHA 2206
1976.
A benzene standard with a limit of one 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 1979 term. A decision is unlikely before early 1980.
42 FR 27452. May 27, 1977.
Carcinogen Assessment Group (R. Albert, Chairman).
Population Risk to Ambient Benzene Exposure. U.S.
Protection Agency. September 12, 1978.
Final Report on
Environmental
Energy and Environmental Analysis, Inc. Estimation of the Population
Exposed to Airborne Pollutants: Methodology and Case Study. November
1978.
E-31
-------
-------
AA. ADDENDUM
BACKGROUND INFORMATION FOR PROPOSED STANDARDS
FOR BENZENE EMISSIONS FROM THE ETHYLBENZENE/STYRENE INDUSTRY
-------
-------
TABLE OF CONTENTS
Page
1. ENFORCEMENT ASPECTS -,_-,
1.1 Introduction 1_1
1.2 EB/S Processes and Vent Streams. ............ 1-1
1.3 Routing of the Vent Streams -!_3
1.4 Excess Emissions -|_3
1.5 Monitoring . 1-5
2. REPORTS IMPACT ANALYSIS 2-1
2.1 Legal Authority 2-i
2.2 Background '. 2 ,
2.2.1 Purpose 2_-,
2.2.2 Narrative Description 2-2
2.3 Alternatives 2_3
2.4 Impact Analysis. . 2-4
2.4.1 Respondent Labor Requirements 2-4
2.4.2 Agency Labor Requirements 2_4
3. EXCESS EMISSIONS 3-]
3.1 Introduction 3 ,
3.2 Processes and Their Emissions. 3.]
3.2.1 Startup o-i
3.2.2 Shutdown 3-2
3.2.3 Equipment Malfunction 3_3
3.3 Controls ~.
3.3.1 Flaring 3-4
3.3.2 Boilers 3_5
3.3.3 Recycling of Hydrogen Separation Stream .3-5
3.3.4 Backup Hydrogen Separation Compressor 3-5
3.4 Regulatory Options 3_7
3.4.1 Baseline Case 3_7
3.4.2 Option 1 3_8
AA-i-
-------
TABLE OF CONTENTS (CONT'D)
Page
3.4.3 Option 2 3"8
3.4.4 Option 3 3"8
3.4.5 -Option 4 3"9
3.5 Environmental and Energy Impacts . .3-9
3.5.1 Air Impacts 3"9
3.5.2 Water Quality Impact and Consumption 3-10
3.5.3 Solid Waste Disposal Impact 3-10
3,5.4 Energy Impact 3~10
3.5.4.1 Regulatory Option 1 3"14
3.5.4.2 Regulatory Option 2 3-14
3.5.4.3 Regulatory Option 3 -3-14
3.5.5 Other Environmental Concerns 3-15
3.5.5.1 Flare Noise and Thermal Radiation 3-15
3.6 Cost Impact 3"16
3.7 References 3"21
4. ETHYLBENZENE AIR OXIDATION ANALYSIS 4-1
4.1 Introduction 4-1
4.2 Emissions 4-1
4.3 Control Techniques 4~2
4.4 Regulatory Options • 4~4
4.5 Environmental and Energy Impacts 4"4
4.6 Secondary Environmental Impacts .4-5
4.7 Cost Analysis 4"5
4.8 References • 4"8
AA-ii-
-------
LIST OF TABLES
Page
2-1 Labor Requirements to Fulfill the Recordkeeping, Reporting,
and Monitoring Requirements of the Proposed Standard 2-5
2-2 Labor Requirements to Fulfill the Reporting Requirements
of the Proposed Standard 2-6
3-1 Stack Source Characteristics for Excess Emissions 3-11
3-2 Maximum Annual Average Benzene Concentrations for the
Four Excess Emissions Regulatory Options. . „ 3-12
3-3 Total Energy Requirements of Regulatory Options for
Excess Emissions .3-13
3-4 Flare Noise Levels Relative to OSHA Standards and EPA
Guidelines. 3_17
3-5 Control Equipment Costs .- .3-18
3-6 Capital and Annualized Costs to EB/S Industry per Regulatory
Option. 3-19
4-1 Typical Output of Thermal Oxidizer 4-6
4-2 Alternative Heat Recovery Scenarios and Costs 4-7
AA-iii-
-------
-------
1. ENFORCEMENT ASPECTS
1.1 INTRODUCTION
The proposed standard for the ethylbenzene/styrene (EB/S) industry
would limit benzene emissions from process vents at existing and new EB/S
plants to five parts per million by volume (ppmv). Emissions in excess of
this numerical emissions limit would be allowed only during startup, shutdown,
and malfunctions. During these times, however, these emissions would have
to be flared. This addendum outlines sections of the standard of potential
concern to enforcement personnel. Areas examined include: the applicability
of the proposed standard to EB/S processes and vent streams, the routing of
vent streams, excess emissions, and monitoring methods.
1.2 EB/S PROCESSES AND VENT STREAMS
The two processes used domestically to produce ethyl benzene are benzene
alkylation with ethylene and mixed xylene extraction. The two processes
used to produce styrene are ethyl benzene dehydrogenation and ethyl benzene
hydroperoxidation. The proposed standard would apply to each plant which
produces ethyl benzene from benzene alkylation or which produces styrene
from ethylbenzene dehydrogenation or ethylbenzene hydroperoxidation.
Plants producing ethylbenzene from mixed xylenes are not regulated under
this standard.
The following benzene-containing streams from the three covered pro-
cesses would be regulated under the proposed standard:
• Alkylation reactor section vents
• Atmospheric or pressure column vents
t Vacuum-producing device vents
• Hydrogen separation system vents.
AA-1-1
-------
The alkylation section in the typical EB/S plant will have two to four
vents. In general, one of these vents will be on the alkylation reactor
itself and the others on catalyst neutralization separation vessels imme-
diately downstream of the reactor. The distillation column area typically
will have six to ten vents. In general, there will be one vent on each
vacuum atmosphere and pressure column. The vent streams in the alkylation
reactor section and column area have low flows, generally less than 100
standard cubic feet per minute (scfm). The vents may be piped together and re-
leased at a common point.
The hydrogen separation system typically will have only one vent
stream. Since this stream has a high flow and Btu value, most plants use
this stream as fuel.
The definition of process vent in the proposed standard is writte'n so
that certain vents in an EB/S unit are not covered. These vents are not
covered since they either have negligible benzene emissions or are to be
covered under other standards. The vents not covered may include, at a
typical EB/S plant, catalyst preparation area vents, intermediate product
storage vents, and compressor lube oil vents, among others.
The typical EB/S plant produces ethyl benzene by benzene alkylation and
styrene by ethyl benzene dehydrogenation. However, depending upon process
variations and plant capacity, the production train and the number of vents
at each plant may vary. For example, one plant produces styrene by ethyl-
benzene hydroperoxidation and some plants produce only ethyl benzene or only
styrene. Plants which produce only ethyl benzene will not have a hydrogen
separation section vent and plants which produce only styrene will not have
alkylation reactor section vents. Consequently, it is important for enforce-
AA1-2
-------
ment personnel to be cognizant of these variations in reviewing monitoring
records and during plant inspections.
1.3 ROUTING OF THE VENT STREAMS
In controlling continuous process benzene vent streams, plant operators
will route these streams to a control device. Generally, a boiler* will be
used to meet the numerical emissions limit. However, EB/S plant operators
are not restricted to the use of boilers and can use any control device
which meets the requirements of the regulation.
Because of the number of vent streams to be controlled and the complexity
of EB/S plants, the routing of the process vent streams to the control
device may be equally complex. In addition, in some cases, smaller process
vent streams covered by the standard can be vented to the atmosphere without
a noticeable change in the flow entering a control device or compressor.
Therefore, special attention should be given to determine if the proper
streams are being routed to each control device. During plant inspections,
enforcement personnel should request a process flow sheet detailing all the
process vent streams to be controlled, all relevant control valves, and the
control devices. A process description and a flow diagram of the specific
process, along with the generalized descriptions and diagrams in the Back-
ground Information Document, should help verify proper routing.
1.4 EXCESS EMISSIONS
Emissions in excess of the numerical emissions limit would be allowed
only during startup, shutdown, and malfunctions. During these times,
however, these emissions would have to be flared. Particular attention,
*The term "boiler" includes process heaters and superheaters.
• AA-l-3
-------
therefore, should be given to the definition of these periods of permissible
excess emissions.
Startup involves bringing into full operation all equipment in the
entire plant, the dehydrogenation section, or the alkylation reactor section,
which is originally at ambient temperature and empty of process liquids.
Conversely, a shutdown involves terminating operation of either all equip-
ment in the entire plant, the dehydrogenation section, or the alkylation
reactor section, and allowing the equipment to cool to ambient temperature.
Complete plant startup and shutdown each generally require 12 to 24 hours
and occur on an average of once every two years. The emissions from startup
and shutdown occur because the volume of the hydrogen offgas stream is
below the minimum needed for the hydrogen separation compressor(s) to
operate and therefore cannot be sent to the boiler. The stream therefore
would have to be vented to the flare.
The only other time during which the five ppmv emissions limit can be
exceeded is during periods of malfunction. During malfunctions, the benzene-
containing vent streams would have to be combusted in a flare. Process and
control device failures occur periodically; compressor outages are the most
common process failures. Process or control equipment failures that are
sudden and unavoidable are by definition malfunctions. Conversely, process
or control equipment failures caused entirely or in part as a result of
design deficiencies, poor maintenance, careless operation, or other prevent-
able equipment breakdown would not be considered malfunctions. Although
specific criteria on which this decision is based are not definitive, cer-
tain indicators, such as frequency of breakdown, age of equipment, and the
sequence of events leading to the equipment failure, should provide supportive
. AA1-4
-------
information for enforcement personnel to determine whether or not a prevent-
able equipment breakdown occurred.
Enforcement personnel should pay particular attention to failures of
equipment which is not central to the operation of the process. For
example, the hydrogen separation compressor must be kept in good operating
condition for the process to operate economically. The other vent compressors,
however, are not central to the process operation. Malfunctions of these
compressors may go unrepaired for long periods of time. Excessive delays
in repairing these compressors would constitute a violation of the standard.
Enforcement personnel therefore should obtain a complete description of the
equipment causing the malfunction and the reason for the malfunction. This
information, along with a knowledge of the relative importance of the
equipment to the process, should allow enforcement personnel to determine
the validity of a claim of malfunction.
1.5 MONITORING
The final potential area of concern to enforcement personnel is emission
monitoring and testing. Because the standard is based on known benzene
destruction of 99 percent in a boiler, a quantitative benzene monitor which
would measure the benzene emissions exiting a control device, such as a gas
chromatograph monitor, is unnecessary. An equally effective approach
directly applicable to boilers involves installation of a firebox tempera-
ture monitor and flue gas oxygen monitors, each equipped with a strip chart
recorder, which measures and records critical operating parameters of the
boiler. If these monitors indicate that the control device is operating
within the range specified in the proposed rule, benzene emissions from the
boiler stack should be in compliance with the standard.
AAl-5
-------
An additional continuous monitoring device, flow meters coupled to a
strip chart recorder, would determine if all benzene-containing streams are
ducted from the compressors to the control device. These would be placed
at strategic points on the piping and would indicate if any benzene-containing
streams are diverted from the boiler.
To further check that benzene-containing streams are routed properly
to the boiler, the proposed standard requires weekly visual checks of vent
stream piping, valves, compressors, and other components. These checks
include components not on ground levels and the operator, therefore, may
need to climb to inspect some components. Operators are to keep logs of
their inspections and are to report improper routings if they occur more
than once every 90 days.
For plants which use control devices other than boilers, monitoring
would be done by gas chromatography with a flame ionization detector.
Monitoring devices, in addition to periodic visual checks on valve
positions, will provide sufficient information for determining compliance
with the standard. All monitoring equipment should be in operation before
the emission tests. Enforcement personnel should visually check during the
test to determine if all vents covered by the standard are being routed to
the control device.
AAl-6
-------
2. REPORTS IMPACT ANALYSIS
2.1 LEGAL AUTHORITY
Under the authority of §112(b)(l) of the Clean Air Act, as amended,
the Environmental Protection Agency (£PA) proposes a national emission
standard for benzene emissions from ethylbenzene/styrene (EB/S) plants.
Under §114(a) of the Act, the Administrator may require any owner or opera-
tor of an EB/S plant, to "(A) establish and maintain such record, (B) make
such reports, (C) install, use, and maintain such monitoring equipment or
methods, (D) sample such emissions (in accordance with such methods, at
such locations, at such intervals, and in such manner as the Administrator
shall prescribe), and (E) provide such other information, as he may reason-
ably require."
2.2 BACKGROUND
2.2.1 Purpose
Sections 61.102(a) and (b) of the proposed standard require that own-
ers or operators of EB/S plants limit to five parts per million the amount
of benzene being discharged to the atmosphere from a process vent stream(s)
and combust unavoidable excess emissions due to startup, shutdown, and
malfunction by one or more smokeless flares at all times. Section 61.102(c)
requires that each owner or operator maintain each source, including associated
air pollution equipment, in a manner consistent with good air pollution
control practices for minimizing benzene emissions. To determine whether
an owner or operator is in compliance with the standard, and to determine
whether acceptable operating and maintenance procedures are being used, the
AA2-1
-------
Administrator needs specific information concerning plant operations,
maintenance, and monitoring procedures.
2.2.2 Narrative Description
The proposed standard requires that specific information concerning
emissions,in excess of the standard be reported to the Administrator within
ten days of each incidence of the excess. This includes the identity of
the process vent stream where the excess occurs, continuous monitoring and
operational data which indicate the excess emissions, the cause and descrip-
tion of the excess emissions, the duration of the excess emissions, and
whether the owner or operator believes that the excess emissions were due
to startup, shutdown, or malfunction. If the owner or operator states that
he or she believes that the excess emissions were during startup, shutdown,
or malfunction, he or she must include the steps taken to remedy and prevent
a recurrence of the malfunction and documentation that good engineering
*
practices were followed. Owners or operators of sources which use boilers
as the air pollution control device also must keep daily records of the
oxygen level as determined by the flue gas oxygen monitor, the firebox
temperature as determined by the firebox monitor, the flow level as measured
by the compressor flow meter, visual checks to determine if the process-
vent streams were going to the control device, and emission test results.
Owners or operators of sources which use air pollution control devices
other than boilers must keep daily records of the concentration of benzene
measured in the exhaust gas by the monitor specified in §61.105, the flow
level as measured by the compressor flow meter, visual checks to determine
if the process vent streams are going to the control device, and other
Boilers include process heaters, superheaters, and other similar units.
A A-2
-------
data needed to determine emissions as specified in §61.104. The Adminis-
trator has not requested a specific format for the information required to
be submitted or kept in daily records.
These requirements apply to the owners and operators of 13 plants
which produce ethyl benzene, styrene, or both. Of these 13 plants, 12
produce styrene, 11 produce ethyl benzene, and nine are integrated plants
producing ethylbenzene/styrene. A recent analysis of the EB/S industry
revealed that no new plants are expected to be constructed within the next
five years.
2.3 ALTERNATIVES
To ensure that the requirements of this standard minimize the public
reporting and recordkeeping burden while maximizing EPA program effective-
ness, the Administrator considered alternatives to the recordkeeping,
reporting, and continuous monitoring requirements discussed above.
One alternative that he considered, for sources which use air pollution
Control devices other than boilers, was not to require that records be kept
of the benzene concentration and amount of excess oxygen in the exhaust
gas, or that the records be kept, but not on a daily basis. In control
devices such as boilers (BAT), there is potential for air dilution. These
devices use excess air to aid combustion and the quantity of excess air can
vary. This information therefore is needed on a daily basis to ensure that
each EB/S plant is in compliance with the standard and that the quantity of
benzene emitted is the same regardless of how much excess air is used. For
sources which use boilers, the Administrator considered not requiring all
the monitoring data be recorded and kept. However, many EB/S plants cur-
rently keep records of this information and it therefore would not impact
the EB/S industry's recordkeeping practices.
AA2-3
-------
The Administrator also considered further limiting the amount of
information reported during incidences of excess emissions. The standard
requires that good air pollution control be practiced for limiting benzene
emissions. To ensure that plants are practicing good air pollution control,
the Administrator determined that these data were necessary.
Intermittent monitoring was considered as an alternative to continuous
monitoring. Both types of monitoring require the same equipment, maintenance,
and operating procedures. Intermittent monitoring would not require as
extensive manpower requirements, but would not provide as effective means
of ensuring compliance as continuous monitoring. Given this, the Adminis-
trator determined that continuous monitoring was the best means of maxi-
mizing EPA program effectiveness while minimizing monitoring burdens.
2.4 IMPACT ANALYSIS
2.4.1 Respondent Labor Requirements
Table 2-1 represents the labor requirements for the EB/S industry to
comply with the recordkeeping, reporting, and monitoring requirements of
the proposed standard. Because no new EB/S plants are expected to be
constructed within the next five years, these requirements are only for
existing sources. The labor requirements are estimates through the fifth
year of applicability of the standard.
2.4.2 Agency Labor Requirements
Table 2-2 details the labor requirements for EPA to fulfill the reporting
requirements of the standard. The labor requirements are estimates through
the fifth year of applicability of the standard.
M2-4
-------
n 2~]' LABOR REQUIREMENTS TO FULFILL THE RECORDKEEPING
REPORTING, AND MONITORING REQUIREMENTS OF THE PROPOSED STANDARD
Requirements
NESHAP Compliance Status
Information Report
Waiver Application
Employee Hours
Per Report
2
2
Number Of
Sources
13
0
Total
Employee Hrs
26
0
Notification of Physical
or Operational Changes
Notification of Demonstration
of the Continuous Monitoring
System
Continuous Monitoring
Demonstration Report
Excess Emissions Reports
Recordkeeping Requirements
Notification of the
Administrator before an
Emissions Test
Emission Test Reports
Notification of the
Administrator before
Startup
8
20
40
8
13
13
13
13
13
13
13
91
1,300
2,600
2,600
162.5
1,300
32.5
TOTAL
8,112
AA2-5
-------
TABLE 2-2. LABOR REQUIREMENTS TO FULFILL THE REPORTING
REQUIREMENTS OF THE PROPOSED STANDARD
Requi rements
Review NESHAP Compliance
Status Information Report
Review Waiver Application
Employee Hours
Per Report
2
2
Number Of
Reports
13
0
Total
Employee Hrs
26
0
Review Notification of
Physical or Operational
Changes
Review Notification of
Demonstration of the Con-
tinuous Monitoring System
Review Continuous Monitoring
Demonstration Report
Review Excess Emissions Report
Review Notification of the
Administrator before Emission
Test
Review Emission Test Reports
Review of Notification of
Startup
8
4
8
1
4
13
162.5
130
6.5
162.5
162.5
32.5
13
1,300
572
162.5
650
32.5
TOTAL
2,756
AA2-6
-------
3. EXCESS EMISSIONS
3.1 INTRODUCTION
Existing data indicate that emissions in excess of the numerical
emissions limit are unavoidable during startup, shutdown, and some equip-
ment failures at EB/S plants. Some emissions are due to sudden and
unavoidable equipment failures and are by definition malfunctions. Others,
however, are a result of deficiencies in design, poor maintenance, or
careless operation and are by definition avoidable. Because unavoidable
emissions occurring during startup, shutdown, and malfunction could
exceed the numerical emissions limit for EB/S plants, plant owners or
operators would not be in compliance with the standard which requires that
good air pollution control practice be followed and would be subject to
enforcement procedures. The Administrator therefore has considered alter-
native regulatory options for regulating unavoidable excess emissions.
3.2 PROCESSES AND THEIR EMISSIONS
Unavoidable excess emissions can occur during (1) startup, (2) shutdown,
and (3) as a result of either process or air pollution control equipment
mal function.
3.2.1 Startup
Startup involves bringing to full operation equipment in either the entire
plant, the alkylation reaction section, or the dehydrogenation section,
which is originally at ambient temperature and empty of process liquids. Plant
startup occurs on the average of once every two years and involves gradually
bringing all empty and non-operating equipment, initially at ambient tempera-
ture, to full operating temperatures. The procedure requires an average
AA3-1
-------
of approximately 12 to 24 hours to fully establish reactions and ensure
that the quality of the product meets specifications. In this interim
period, however, benzene emissions in excess of the standard can be released
from certain points in the production train.
The benzene emission potential during one startup period is approxi-
mately 4,200 kg/yr for ten hours per year from the entire 300,000 Mg/yr
uncontrolled model plant, most of which is from the hydrogen separation
vent. Due to a lack of sufficient flow during startup, this stream is not
compressed and sent to the aromatic recovery section as is the case during
normal production. Rather, typical practice involves either venting the
stream directly to the atmosphere, recycling the stream, or flaring it
until the dehydrogenation reaction is established fully. Only after the
reaction is established and the volume of the stream is sufficient can the.gas
be compressed, ducted to the aromatic recovery section where the benzene
and other condensables are recovered, and the remainder of the stream,
largely hydrogen, be sent to the steam superheater for use as supplemental
fuel.
3.2.2 Shutdown
Shutdown involves terminating operation of either all equipment in the
entire plant, the alkylation reaction section, or the dehydrogenation
section and allowing the equipment to cool to ambient temperature. Plant
shutdown is necessary an average of once every two years so that equipment
can be inspected, cleaned, or replaced. The shutdown procedure for a plant
comparable to the 300,000 Mg/yr model plant normally requires about 12 to
24 hours. Actual venting of emissions requires an average of ten hours.
The procedure involves cutting off feedstock to the reactors while concur-
rently decreasing the introduction of steam and catalyst, so that the
AA3-2
-------
temperature of all equipment reaches ambient levels. During this procedure,
emissions in excess of the standard can occur.
Residual benzene in equipment must be purged before maintenance can
begin. During plant shutdown, as in plant startup, the bulk of benzene
emissions occur at the hydrogen separation vent. Based on the 300,000
Mg/yr uncontrolled model plant, benzene emissions due to plant shutdown are
approximately 4,200 kg/yr from the entire plant.
3.2.3 Equipment Malfunction
Unavoidable excess emissions also occur due to either process or
control equipment malfunction. Though benzene emissions due to malfunc-
tions can arise anywhere on the production train, certain pieces of equipment
have been cited as likely sources of emissions. Experience has shown that
the hydrogen separation compressor is inoperable an average of ten hours
per year. The hydrogen separation compressor is necessary for the recovery
of aromatics and for producing the necessary pressure to combust the remaining
hydrogen-rich stream in the steam superheater. During periods of outage,
the benzene-laden stream is either vented directly to the atmosphere or
flared. Some plants employ backup compressors which nearly eliminate
benzene emissions from this source. In the absence of a backup compressor,
the benzene emission potential from this source is approximately 4,000 kg/yr
(based on the 300,000 Mg/yr model plant).
Experience has shown a variable potential for benzene emissions in the
event of a vent gas compressor outage. Compared to the hydrogen separation
compressor, the vent gas compressor services relatively small streams. Its
function is to direct gas streams under low or negative pressure to the
main header. Since this equipment is not central to styrene production and
because the vent streams are small compared to process vent streams, little
AA3-3
-------
incentive exists for immediate repair of this equipment. In addition to
actual malfunction of this equipment, it is, in many cases, more convenient
to turn off the vent gas compressor to avoid more serious or costly opera-
tional upsets in other areas of the production train. Records at one plant
show that a vent gas compressor had been shut off for months, thereby
avoiding more costly repairs. Since the streams directed by this device
contain little recoverable product, they usually are vented directly to the
atmosphere. Intentional emissions such as these would be considered avoidable
and a violation of the standard requiring good air pollution control practices.
Assuming plant operators allow the vent gas compressor to remain inoperable
for one month, benzene emissions of up to 16,000 kg/yr (for the 300,000
Mg/yr model plant) can occur from the venting of the stream at this source.
3.3 CONTROLS
This section describes applicable control strategies for excess benzene
emissions in the EB/S industry and analyzes their efficiencies, advantages,
and limitations. The controls discussed are flaring, incineration in an
existing boiler, recycling of the vent stream, and backup control devices.
3.3.1 Flari ng
Flaring is an important control strategy for limiting excess benzene
emissions from the EB/S industry. Various vent streams can be manifolded
to a properly-sized flare for thermal destruction. The main advantage of
the flare is its ability to control high volume, short duration releases as
normally are encountered during equipment malfunction. For example, flares
can control emissions due to a hydrogen separation compressor outage, vent
gas compressor outage, and streams of highly variable Btu value such as
those encountered during startup and shutdown. Chapter 4 of the BID describes
AA3-4
-------
flare system equipment and design features. A ten-inch smokeless flare
system* is assumed to be used for control of benzene emissions due to
malfunction and is essentially identical to the flare system used to control
process emissions except for stack diameter and height. Smokeless flares
can reduce visible emissions to no more than five minutes or less within a
two-hour period. Combustion efficiency for the ten-inch smokeless flare is
assumed to be 60 percent.
3.3.2 Boilers
Vent streams can be ducted to the burner of a boiler for thermal
destruction of benzene. This control system would use existing EB/S boilers
appropriately retrofitted to accept all vent streams. The boiler control
system can be retrofitted in two ways. In the first, lower pressure vents
first are compressed and then piped along with higher pressure vents into
the boiler fuel header and burned with the natural gas fuel. In the second,
one or several burners in the boiler are replaced by forced draft burners
capable of accepting low pressure streams. The lower pressure vents then
are piped to the replacement burners for destruction and higher pressure
vents to the fuel header. Forced draft burners are available as standard
items for pressures as low as one inch of water gauge.
The chief advantage of this system is its high control efficiency.
This is due to the burner's ability to provide excellent fuel/air mixing,
relatively long residence time in the flame, and high temperatures. A
benzene destruction efficiency of 99 percent is assumed for this analysis.
Sized for the 300,000 Mg/yr model plant.
AA3-5
-------
This assumption is based on both engineering analyses and field tests on
existing EB/S boilers.
A disadvantage of this system involves the complexity of controls,
safety equipment, and manifold design. Due to the variable flow and com-
position of streams during startup, shutdown, and malfunction, installation
of all necessary equipment may be prohibitively expensive.
3.3.3 Recycling of Hydrogen Separation Stream
This method of emission control involves redirecting the hydrogen
separation compressor discharge back to its inlet header. This procedure
limits benzene emissions by retaining and compressing the stream until
conditions are suitable for introduction to a control device. A primary
hydrogen separation compressor can be retrofitted to provide recycle capa-
bility at minimal cost. Approximately 75 percent of the EB/S industry
currently employs this method of control. Its estimated benzene emission
reduction potential ranges from 80 to 100 percent.
3.3.4 Backup Hydrogen Separation Compressor
A backup hydrogen separation compressor serves not only to control
excess emissions but also allows nearly continuous aromatic recovery.
Additionally, it provides nearly continuous utilization of the hydrogen-
rich stream after recovery for use as a supplemental fuel. In the event of
a primary hydrogen separation compressor outage, the backup compressor is
able to come on line within an hour, thereby curtailing potential benzene
emissions from the separation vent by 90 to 100 percent.
Though a backup compressor is a fairly costly piece of equipment, it
requires no additional energy since only one compressor will be in oper-
ation at any one time. Approximately 50 percent of EB/S plants currently
AA3-6
-------
employ a backup hydrogen separation compressor. However, for these plants,
the second compressor was installed for conditions specific to those plants
and these conditions may not exist at other plants.
3.4 REGULATORY OPTIONS
This section discusses the regulatory options applicable to the con-
trol of excess benzene emissions resulting from startup and shutdown opera-
tions and malfunction. These options are based, in part, on the control
devices discussed in Section 3.3. Under each option, avoidable excess
emissions due to fai.lures of process or air pollution control equipment
caused entirely or in part by deficiencies in design, poor maintenance,
careless operation, or other preventable equipment breakdown would not be
permitted. Failure of process or control equipment, regardless of the
cause, would require prompt repair; failure to do so would constitute
avoidable excess emissions.
3.4.1 Baseline Case
A 300,000 Mg/yr, completely uncontrolled model plant is used as a
baseline of emissions in assessing the impact of the regulatory options.
The extent of emissions from the model plant overstates those currently
emitted from existing plants. The sources and assumed baseline emissions
of the uncontrolled model plant are as follows:
a The hydrogen separation stream vents to the atmosphere for an average
of ten hours per year during startup and ten hours per year during
shutdown operations (startup/shutdown operations are assumed to occur
once every two years). Total benzene emissions amount to 8 000 ka
every two years or 4,000 kg/yr.
e The hydrogen separation stream vents to the atmosphere for an average
of ten hours per year during a hydrogen separation compressor outage.
Total benzene emissions amount to approximately 4,000 kg/yr.
• The vent gas compressor goes unrepaired or is inoperable for one
month, during which time the vent immediately upstream of this device
emits 16,000 kg/yr of benzene to the atmosphere.
M3-7
-------
3.4.2 Regulatory Option 1
Regulatory Option 1 would require that:
• Emissions from startup and shutdown would be flared until the stream
can be ducted to the boiler. Assuming a 60 percent benzene destruction
efficiency and an average of two hours emission duration, total benzene
emissions from this source would be approximately 320 kg/yr.
• All emissions from the hydrogen separation vent due to a hydrogen
separation compressor outage would be flared during the entire outage
period. Assuming an average ten-hour period, total benzene emissions
from this source would be approximately 1,600 kg/yr.
Though not specified as a requirement, Regulatory Option 1 assumes that the
following practice is observed:
• The hydrogen separation stream is recycled prior to flaring during
startup and shutdown operations, thereby retaining the stream until it
can be burned by a smokeless flare.
3.4.3 Regulatory Option 2
Regulatory Option 2 would require that:
• Emissions from startup and shutdown would be flared until the stream
can be ducted to the boiler. Assuming a 60 percent benzene destruc-
tion efficiency and an average duration of two hours, total benzene
emissions from this source would be approximately 320 kg/yr.
t All EB/S plants have a backup hydrogen separation compressor (dual
compressor). Additionally, all emissions from the hydrogen separation
vent would be flared until the backup compressor is able to come on
line. Assuming a 60 percent benzene destruction efficiency, total
benzene emissions from this source would be approximately 160 kg/yr.
Though not specified as a requirement, Regulatory Option 2 also assumes
that the following practice is observed:
• The hydrogen separation stream is recycled prior to flaring during
startup and shutdown operations, thereby retaining the stream until it
can be burned by smokeless flares.
3.4.4 Regulatory Option 3
Regulatory Option 3 would require that:
AA3-8
-------
• An existing EB/S boiler would be retrofitted to accept all streams at all
times. Assuming a 99 percent benzene destruction efficiency total
emissions from this source would be approximately 80 kg/yr.
• All excess emissions would be flared for short periods required to
adjust boiler operation for acceptance of the waste stream at the
beginning of an outage. Assuming a 60 percent benzene destruction
80 k/erCy' t0tal benzene em1ss1ons from this source are approximately
3.4.5 Regulatory Option 4
Regulatory Option 4 (100 percent benzene reduction) requires that:
* f11 !fB/S pl?nts achieve 100 percent benzene reduction. Since this
'* "* technica11* ^"^le, the option
3.5 ENVIRONMENTAL AND ENERGY IMPACTS
This section discusses the environmental and energy impacts of the
five regulatory options presented in Section 3.4. These options involve
limiting benzene emissions from the following sources:
* Th! ^rogen separation vent during startup and shutdown operations
and hydrogen separation compressor outages. *
» Emissions from adjusting the boiler or superheater at the beginning of
OU UclycS •
• Emissions due to vent gas compressor outages.
3.5.1 Air Impacts
Air quality impacts were analyzed through the use of the Industrial
Source Complex (ISC) dispersion model.1 The methodology used to model
excess emissions was the same as that described in Appendix B of the BID,
except that only annual average impacts were modeled for excess emissions.
Aside from source data, all other impacts and assumptions for the excess
emissions modeling are the same as for Appendix B. The estimated total
excess benzene emissions for the EB/S industry are 133 Mg/yr under current
controls, 21 Mg/yr under Regulatory Option 1, 10 Mg/yr under Regulatory
Option 2, and 1 Mg/yr under Regulatory Option 3.
AA3-9
-------
Table 3-1 shows the source impacts used in modeling excess emissions
impacts. The 8,000 kg of benzene emissions from process vents under the
baseline scenario were divided evenly among the eight process vents for the
uncontrolled option.
Table 3-2 presents the results of the excess emissions modeling. For
each option, the highest maximum annual average concentration from any
plant at any distance is given. The table shows that impacts resulting
from Regulatory Options 1, 2, and 3 emissions scenarios are all negligible.
Under the baseline scenario, however, excess emissions impacts of up to
7.9 ppb are predicted very near one EB/S plant (0.16 km).
3.5.2 Water Quality Impact and Consumption
None of the regulatory options considered results in any significant
increase in wastewater discharge by EB/S facilities. No water effluents
are discharged from either flares or boilers. The net water consumption
for the boiler is zero. The flares use water in the form of steam which is
injected into the combustion zone of the stack. Water consumption for the
300,000 Mg/yr model plant employing a ten-inch flare for excess emissions
is negligible.
3.5.3 Solid Waste Disposal Impact
None of the regulatory options being considered generates solid waste.
Consequently, discussion of solid waste disposal impacts is not applicable
to this analysis.
3.5.4 Energy Impact
Table 3-3 presents the total energy requirements for each regulatory
option as applied to the 300,000 Mg/yr uncontrolled model plant. Energy
requirements for Regulatory Options 1, 2, and 3 amount to less than
AA3-10
-------
TABLE 3-1. STACK SOURCE CHARACTERISTICS FOR EXCESS EMISSIONS
Scenario
Uncontrolled.
Model Plant0
Regulatory
Option 1 —
Recycle0
Annual Average
Emission Rate
Source (G per sec)
Process Vents:3
Atmospheric Pressure Columns
Alkylation Reaction Vents
Benzene/Toluene Column
Vacuum Columns
•
Hydrogen separation vent
compressor failure-
uncontrolled
H2 Section- Flare
-startup
H, Section-Flare
-outage
Process Vents-Flare
Hy Section- Flare
-startup
.0.063
0.063
0.063
0.063
0.063
0.063
0.063
0.063
0.272
0.010
0.051
0.006
0.010
Height
(meters)
12
12
12
12
12
12
12
12
6
90
Temp.
(Deg.K)
318
318
318
318
318
318
318
318
318
823
Exit Vel.
(m/sec)
0.1
0.1
4.9
4.9
0.1
0.1
0.1
0.1
19.8
0.1
Diameter
(meters)
0.33
0.33
0.06
0.06
0.23
0.11
0.11
o.n
0.50
13.7
Regulatory
Option 2 -- .
Dual Compressor
Regulatory
Option 3 —
No Excess.
Emissions
H2 Section-Flare
-outage
Process Vents-Flare
All vents to flare
0.005
0.006
0.003
aThe atmospheric pressure columns discharge horizontally and the benzene/toluene and vacuum columns dis-
charge downward. Values given are the modeling inputs for these sources, after adjustments, as discussed
in the text of the main BID. Th" =-+••'•' -*--i. i-.-'-u^. —i _,.-—^.... ^_ _ ^,' ° ..». " av.uo«u
respectively.
— •*" ••••3 ...f*««w i w. un^^c jvsui *-ca, QIUCI aujua Uildtlo, ab UlSCUSSeQ
The actual stack height and diameter for these sources are 12.0 and 0.06 m,
Values given are the modeling for the flares, after adjustments, as discussed in the main text
characteristics for the flares are:
3-inch flare: height = 20m, diameter = O.lm, flow rate = 0,05m /sec.
H2 flare: height = 60m, diameter = 0.3m, flow rate = 3.89m /sec.
Actual
AA3-11
-------
(0
Q-
a.
to
ro
C3
LU
OH
to
to
to
to
to
LU
LJU
LU
IO
LU
o
LU
LU
CQ
DC
LU
ra
0) S-
o •»->
c c
CO 0)
-P 0
(A C
•r- O
o o
O -Q
•r- O.
+> d»
<= o
cu e
o m
o -r-
It
CM
CO
nf
c
CD
to
to
o
CM
CVJ
CSI
ro
oo
o
csl
0)
-P- c=
C (Q
O i—
O Q-
cu
esi'
O) O
4-> C r— O
ro o U
"3 43 O ro
O) O. HI 3
D. O
(U O CQ
oi
0)
o
-------
to
0
to
to
1— 1
LU
to
to
LU
O
X
LU
O
LU
to
^^
0
1—4
Q_
O
.
eg
o
5
C3
LU
OS
LU
O
to
"Z.
LlJ
LU
1— \
cr
LU
02
^*"
CD
O£
LU
LU
1
^C
0
1—
co
i
CO
LU
— 1
CO
1 —
,-"•>
•p
c
re
r—
Q_
>
cvs- >,<;
O r— re ^^, r— ^
o
•-V J_
,
0 0 S- >,^x
•f- >» re ^v^ r—
•P O T— r-3 ^J
O. CU U_ S ^Q
O OS v-/
O «r- CU S- S-
O) S; TJ r—
re cu ^2
•P >>-!- *— '
0 O5-P
f— S-
c: o
LU M-
CO
o
X. O
0 00
C^ "Sfc^1
^
LO
oo
0
/""°*s
X 0
CD 00
f"^ ^*^
LO
CO
0
o
X «3-
00
O '*—s
oo
CM
»»
LO
l^^s
tD
0 0
r— i—
X X
CD tr-
tD r-^
.
1 —
re
en
s"*>*
-p
QQ
co
o
••~
X
o
^L
0>
c
'i
re
t ^
c
cu
"re
_-*
^
cr
cu
•r-
o
4-
o
U5
s-
re
CO
re
-------
0.1 percent of total plant energy requirements. All energy impacts are
essentially linear with respect to plant capacity.
3.5.4.1 Regulatory Option 1
The smokeless flare requires energy in the form of steam for smokeless
operation and natural gas both for pilot operation and purge requirements.
c
Total energy for these requirements is approximately 5.23 x 10 MJ/yr for
the model plant. In terms of nationwide impacts, Regulatory Option 1 would
require 71 x 106 MJ/yr.
3.5.4.2 Regulatory Option 2
Energy requirements for the smokeless flare under this option are
nearly the same as those for Regulatory Option 1 except that slightly less
steam is required for the projected three hours of flare operation. Energy
needs for steam and natural gas are approximately 5.14 x 10 MJ/yr. The
backup compressor has no incremental energy impact since only one compres-
sor will be in use at a given time. In terms of nationwide impacts, Regulatory
Option 2 would require approximately 70 x 10 MJ/yr.
3.5.4.3 Regulatory Option 3
Regulatory Option 3 requires that all safety valves used in relieving
surges in process streams be manifolded to a ten-inch smokeless flare.
Energy requirements for the flare under this option are nearly the same as
those for Regulatory Option 2 except that even less steam is required for
the projected 15 minutes of flare operation. Energy requirements for steam
and natural gas are 5.1 x 106 MJ/yr. The nationwide energy requirement for
this option would be 69 x 106 MJ/yr of energy or approximately 11,200
bbl/yr (fuel oil equivalent) for all existing EB/S plants.
AA3-14
-------
Regulatory Option 3 also requires that an existing EB/S boiler be
retrofitted to accept vent streams generated as a result of plant startup/
shutdown and malfunction. Energy requirements for the boiler are negligible
as are credits in terms of steam generated as a result of combusting the
waste stream.
3.5.5 Other Environmental Concerns
3.5.5.1 Flare Noise and Thermal Radiation
The flaring process and the noise it generates may have an impact on
the surrounding environment. Estimated sound pressure levels (decibels (dB))
of the ten-inch flare are approximately 84 dB at the base of the flare
stack. This estimate assumes no steam injection. While the introduction
of steam into the combustion zone of the stack increases its destruction
efficiency, it may increase overall noise levels of the flare.2 However,
it is believed that flares can be designed properly to minimize noise.
Table 3-4 compares flare noise levels with OSHA standards and EPA recom-
mendations for noise levels requisite to protect public health and welfare
in the industrial environment.3'4
Sound pressure levels are reduced as a function of distance.5 Sound
pressure levels of the ten-inch flare require a distance of 340 meters to
fall to 70 dB.
Flares also emit thermal radiation.6 Stack height is determined so
that personnel working in'the vicinity of the flare are exposed to only
harmless levels of thermal radiation. Though the thermal radiation of the
EPA recommendations for maximum dB level to protect against hearing loss
(24-hour exposure). a
AA3-15
-------
ten-inch flare is greater than that of the process flare, stack height of
the larger-diameter flare is correspondingly greater.
3.6 COST IMPACT
Table 3-5 presents a summary of installed capital costs for specific
equipment, used alone or in combination, for control of benzene emissions
due to EB/S plant startup, shutdown, and equipment malfunction. Actual
costs will vary somewhat depending on plant capacity and existing controls;
those presented are based on the 300,000 Mg/yr uncontrolled model plant.
Though neither recovery nor energy credits is possible with the smoke-
less flare, both the boiler and the backup compressor recover some of the
benzene or heat value of the waste streams. The backup hydrogen separation
compressor can be on line within one hour of a primary compressor malfunc-
tion. Assuming an outage duration of ten hours per year, the backup com-
pressor can continue to control the hydrogen separation stream, thereby
providing benzene recovery and utilization of the remaining hydrogen-rich
stream as supplemental fuel in the steam superheater.
Table 3-6 presents the industry-wide economic impact and annualized
costs per regulatory option. Regulatory Option 1 requires the least capital
outlay of approximately $524,000 for the purchase and.installation of
flares at four of the 13 plants. Regulatory Option 2 would require a
capital outlay of approximately $5.5 million for the purchase and instal-
lation of flares or backup compressors at six plants. Regulatory Option 3
would require the greatest capital outlay of approximately $20.0 million
for flares or a boiler retrofit and controls at 13 plants. Industry-wide
annualized costs for Regulatory Options 1, 2, and 3 would be $171,000, $1.6
million, and $5.2 million, respectively.
AA3-16
-------
LU
1— 1
— 1
1 1 1
a
t^J
CD
a.
LU
Q
1— I
•^
—J
LU
D£.
LU
LU
—1
LU
V)
t— 1
O
Z
LU
Z S-
O) l/>
•— o
x) a.
•r- X
(/> LU
10
^^
ca
13
to
a>
o>
0)
10
0
•z.
O)
•£
to
s_
Q.
E
O
0
•i— S-
E 3
S- 0
oi nr
Q- I
00
T-
to
o
0
CD
10
fO
0)
S-
«C C
o
r— »r-
(0 -P
•i— fO /•>
S- S- Ul
•P O) r-
w -a o>
3 ••- >
•O in ai
£= c: _ J
1— 1 O
0 S-
1- 3
o 10 o
t- w in
O 1
C CM
•i- O)
i— C S-
<0 v- 0
-o s- i-
•r- (O
3 0>
CJ3 31
**~s
a.
LU
O
ca
-a
*— ^
ai
u>
o
01
to
r~"
LL.
T3
CD
-P
tO
S
•r—
•P
tn
uu
o>
to
'~"
LL.
o
I— 1
00
O)
to
cr
Ol
-o
fO
c '
to
x:
'5
OI
fO
t)_
'oi
•o
to
x:
-p
to
0)
f—
**
f
r-_
^
~y
a.
flj
x:
•p
+j
u
OI
4->
o
o.
o
4J
O)
•p
to
3
CT
OI
S-
IO
to
•a
o>
f-
O) O
—
•a T-
c: 03
3 S-
O tO
co e
to
oi
U)
0
a.
IO
OI
S-
o
(/)
«ar
CO
T3
r>*.
O)
>
T3
C
3
O
to
M3-17
-------
TABLE 3-5. CONTROL EQUIPMENT COSTS'
(in .thousands of dollars)
Control
Equipment
Ten-inch flare, header,
and all necessary control
equipment
Backup H2 separation
compressor and
controls
Total Installed
Capital Costs
$131.0
828.0
Existing boiler
retrofit and
controls
1,518.0
*Based on the 300,000 Mg/yr model plant.
AA3-18
-------
TABLE 3-6. CAPITAL AND ANNUALIZED COSTS TO EB/S
PER REGULATORY OPTION3
(thousands of dollars)
Option
Regulatory
Option 1
(Flare)
Regulatory
Option 2
(Dual Compressor/
flare)
Capital
Costs
$524
$5.5
Annual ized
Costs
$ 171
$1,600
No. of
Plants
4
6
Regulatory
Option 3
(Boiler/flare)
$20.0 x 103
$5,200
13
Represents costs to indicated number of plants lacking equipment
required by regulatory option.
AA3-19
-------
Regulatory Option 4 would have the most severe economic impact on the
EB/S industry. Because 100 percent control of excess emissions is techni-
cally infeasible, the industry would either be in noncompliance or be
forced to close. This impact is discussed in detail in Chapter 7 of the
BID.
-AA 3-20
-------
3.7 REFERENCES
1.
2.
3.
4.
5.
J. F. Bowers, et al. H. E. Cramer, Inc. Draft Industrial Source
Complex (ISC) Dispersion Model User's Guide — Volume I and II. Pre-
pared for EPA, Source Receptor Analysis Branch. Research Triangle
Park, North Carolina. January 1979.
Straitz, J. F. III. Solving Flare-Noise Problems. National Air Oil
Burner Co., Inc. Philadelphia, PA. p. 2.
Bureau of National Affairs. Noise Regulation Reporter. OSHA Stan"
dardized Procedure for Noise Measurements, Chapter VI. p. 41-3242
October 10, 1977.
U.S. EPA, Office of Noise Abatement and Control. Information on
Levels of Environmental Noise Requisite to Protect Public Health and
Welfare With an Adequate Margin of Safety. March, 1974. p. 29.
Swithenbank, J. Ecological Aspects of Combustion Devices With Reference
to Hydrocarbon Flaring. Department of Chemical Engineering and Fuel
Technology, University of Sheffield, Sheffield, England, p. 553.
May 1972.
6.
Flaregas, Inc.
Issue 2.
Ground Pollution From Elevated Flare Effluents.
AA3-21
-------
-------
4. ETHYLBENZENE AIR OXIDATION ANALYSIS
4.1 INTRODUCTION
In the existing ethylbenzene/styrene (EB/S) facilities being analyzed
for regulation, most benzene-containing process vent streams are similar in
composition and therefore can be controlled by the basic control techniques
described in Chapter 4. However, in the production of styrene from ethyl-
benzene hydroperoxide, the oxidation of ethyl benzene with air to produce
ethyl benzene hydroperoxide results in a benzene-containing vent stream with
a markedly different composition. Because of this significant difference
in composition, the vent is analyzed separately.
4.2 EMISSIONS
The ethyl benzene oxidation reactor vent releases nitrogen and excess
oxygen brought into the process with the reaction air; carbon monoxide,
carbon dioxide, and light organics produced in the reaction; and organics
entrained in the vent gas. The vent stream has a reported flow of approxi-
o
mately 16.4 cubic meters per second (m /sec) (580 cubic feet per second
3
(ft /sec)) and contains approximately 4.4 percent by weight aromatics, or
3,400 kilograms per hour (kg/hr) (7,480 pounds per hour (lbs/hr)).
At the Oxirane facility, the only U.S. plant using ethylbenzene hydro-
peroxidation, the vent gas is scrubbed with oil and water for 99 percent
removal of organics as required by process economics. The resulting stream
is vented to the air and contains 35 parts per million (ppm) benzene or
7.2 kg/hr.
1
AA4-1
-------
Since a high flow stream dilute in organics cannot be effectively
controlled by boilers, condensers, or scrubbers, an alternate control
technique must be examined.
4.3 CONTROL TECHNIQUES
The prospects for new plants using the ethylbenzene hydroperoxidation
process depend almost exclusively on market dynajnics. The supply of pro-
pylene oxide is governed by only five producers nationwide. Given the con-
centrated nature of the existing market for propylene oxide and the predic-
tion of excess capacity in both styrene and propylene oxide in the near
future, it is unlikely that a new producer would enter either market.
Oxirane already had been marketing propylene oxide from a different pro-
duction facility when it constructed its coproduct plant. In this way,
production efficiency for propylene oxide was enhanced without creating a
surplus in supply. Coproducing styrene served to ease a market supply
deficit in styrene at competitive prices. Therefore, the control technique
analysis for the ethylbenzene air oxidation vent need focus only on the
existing facility.
Thermal oxidizers are the most widely employed method to control
volatile organic compounds (VOC) emissions in dilute streams. Thermal
oxidizers destroy VOC through oxidation to carbon dioxide and water.
Temperatures of 1,300 to 1,600°F are sufficient to nearly complete conversion
of most substances in 0.1 to 0.75 seconds residence time. Destruction of
most hydrocarbons occurs rapidly at 1,100 to 1,200"F. but oxidation of carbon
monoxide to carbon dioxide requires the higher temperatures and residence
2
time.
M4-2
-------
A thermal oxidizer system consists of a combustion chamber, a heat
recovery unit, and an exhaust system consisting of blower, fan, ductwork, and a
stack. The combustion chamber can be of any shape or cross-section, although
a cylindrical chamber generally is preferred. It can be either horizontally
or vertically disposed, having a burner located so that it is able to heat
\
the incoming offgas stream to the desired temperature before it exits the
oxidizer chamber.
Burner type and arrangement affect combustion rates and residence
time. The more thorough the contact of flame with the waste organics, the
shorter the time required for complete combustion. Burner placement depends
not only on the burner type, but also on the design requirement for inti-
mate contact of the combustible gases with the burner flame. Maximum
efficiency occurs when all of the combustible matter passes through the
burner. Multi-jet and mixing-plate burners provide the most effective
o
flame contact. The oxidizer may be refractory-lined or it may be con-
structed of heat resistant metal, depending upon the destruction tempera-
ture selected for the organic waste gas.
Thermal oxidizers discharge flue gases at a temperature of 1,300 to
1,500°F. This generally would represent inefficient use of energy if not
recovered. Recovery methods include heat exchange between hot flue gases
and incoming cool offgas stream, recycling a fraction of the hot flue gases
to the process, and the use of the heat in other processing or heating
loads, such as in generating steam for plant or process heating, or power
generation.
Combustion temperature determines whether recuperative heating or
waste heat boilers apply in recovering a portion of the waste heat. Combus-
M4-3
-------
tion temperatures exceeding 1,600°F rule out the use of recuperative heat
exchangers because of construction materials problems as well as secondary
factors such as pre-combustipn occurring in the exchangers. However, waste
heat boilers are alternatives in this range. Temperatures of less than 1,500°F
are compatible with standard recuperative heater designs and waste heat
boilers may be considered throughout this range.
VOC destruction efficiency will increase with increasing concentration
of VOC in the offgas stream. This is partly due to the first order depen-
dence of oxidation rates on VOC concentration and partly due to the signif-
icant local heat of reaction which is released during oxidation. It is
estimated that at incinerator temperatures of 1,600°F and residence times of
0.75 seconds, a minimum of 70 percent destruction efficiency for a 35 ppm
benzene stream can be achieved.
4.4 REGULATORY OPTIONS
Based on the use of a thermal incinerator as the control technique,
two regulatory options are being considered for benzene emissions from the
ethylbenzene air oxidation vent:
• Require that the facilities producing styrene from ethyl benzene hydro-
peroxide reduce benzene emissions to a maximum of five ppm. This
alternative is based on the use of an incinerator operating at a
temperature of 1,600°F and a 0.75 second residence time.
• Do not regulate facilities producing styrene from ethylbenzene
hydroperoxide.
4.5 ENVIRONMENTAL AND ENERGY IMPACTS
Depending on the type of incinerator chosen, energy requirements are
1.3 x 109 MJ/yr for the unit with no heat recovery, 3.2 x 108 MJ/yr for
that with 70 percent heat recovery, and 5.1 x 10 MJ/yr net energy require-
ment for a waste heat boiler. Expressed as a unit of energy per megagram of
AA4-4
-------
benzene destroyed, the energy requirement for the incinerator with no heat
recovery is approximately 4.3 x 107 megajoules per megagram (MJ/Mg); for
the 70 percent heat recovery scenario, the energy requirement is 1.0 x 107
MJ/Mg. The waste heat boiler would require approximately 1.6 x 107 MJ/Mg.
Comparing this energy requirement to that of the other process vent streams,
Regulatory Option C would provide an energy credit of 53 x 106 MJ/yr.
4.6 SECONDARY ENVIRONMENTAL IMPACTS
Table 4-1 illustrates the typical stack constituents of a thermal
oxidizer unit. Relative stack gas makeup should remain constant whether or
not a heat exchanger is used.
4.7 COST ANALYSIS4
Calculations of incinerator costs for the ethylbenzene air oxidation
vent are based on the following production levels and stream characteristics:
(1) 635,000 Mg/yr total production, 454,000 Mg/yr of styrene monomer,
181,000 Mg/yr of propylene oxide; (2) a stream flow rate of 35,000 standard
cubic feet per minute (scfm); (3) a stream thermal content of 1 Btu/scf; and
(4) a stream composition of 95.6 percent inerts.
As presented in Table 4-2, capital costs for the three control configu-
rations are $1,354,000, $2,554,000, and $2,240,000 for the no-heat-recovery
incinerator, the 70 percent heat recovery incinerator, and the waste heat
boiler, respectively. Annualized costs for the same configuration are
$3,461,000, $1,448,000, and $1,756,000, respectively. Cost-effectiveness
for each scenario is $86,000/Mg for no heat recovery, $36,000/Mg for 70
percent heat recovery, and $44,000/Mg for a waste heat boiler.
M4-5
-------
UJ
UJ
o
UJ
03
O (8 in
i- in to
3 CO
o
CE: j= o
•p -P
u>
o t- to
r-
-------
OO
oo
o
oo
o
LU
O.
oo oo
> o
o -a
LU L|_
0:0
I— 00
«s: -o
LU E
00
LU 3
> O
CM
4
CO
oo
CO
c:
1 CU
4-^ ^
OO >
4->
i — 00
•t- O
4->0
"O CO
CU O
N O
i— i—
03 03
3 4-»
c'o.
^
^O
CO
r«
LO
^^*
A
CO
CO
o
0
co
CO
CO
LO
CO
«t
I —
03 CU
CU >
:c o
o
O CU
^ CU
CD
s:
*JD
co
CO
^}"
^3-
A
•—
CM
CO
LO
I-.
LO
1
^"
f-«
LO
LO
ff%
CM
4-» >»
03 S-
CU CU
=n g
^9 O
O CU
I C£
CD
s
^"
^^*
10
CO
p*^
A
1—
LO
CO
1 — 1
0 0
LO 00
O CT>
co r-?
1 — 1
o
LO
to
^J-
CM
CM
as
CU
4-> ^
CO *i~
03 O
,3 CQ
CU
4->
OO
S-
cu
o
o
CU
i.
CU
3
03
AA4-7
-------
4.8 REFERENCES
1. Letter from Fretwell, S., Oxirane to Farmer, J., EPA. November 19,
1979. Concerning ethylbenzene air oxidation process.
2. Rolke, R.W. Afterburner System Study. U.S. Department of Commerce.
NTIS PB-212 560. August 1972.
3. Control Techniques for Volatile Emissions from Stationary Sources.
EPA-450/2-78-022. May 1978.
4. Letter from Towel!, T.W., Halcon to Peterson, C,, Hydroscience.
March 22, 1979.
5. Blackburn, J. W. Emission Control Options for the Synthetic Organic
Chemical Manufacturing Industry. Control Device Evaluation: Thermal
Oxidation. EPA Contract No. 68-02-2577. Hydroscience, Inc.
December 1979.
M4-8
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-450/3-79-035a
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Benzene Emissions from the Ethylbenzene/Styrene
Industry-Background Information for Proposed Standards
5. REPORT DATE
August 1980
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
DAA for Air Quality Planning and Standards
Office of Air, Noise, and Radiation
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
13. TYPE OF REPORT AND PERIOD COVERED
Interim Final
14. SPONSORING AGENCY CODE
EPA/200/04
15. SUPPLEMENTARY NOTES
16. ABSTRACT : ~~~ — •
A National Emission Standard for the control of benzene emissions from
ethyl benzene'/styrene plants is being proposed under the authority of Section .112
of the Clean Air Act. The proposed standard would apply to both new and existing
sources. This document contains background information and environmental and
economic assessments of the regulatory alternatives considered in developing the
proposed standard.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Air Pollution
Pollution 'Control
National Emission Standards for Hazardous
Air Pollutants
Ethylbenzene/Styrene Plants
Benzene
Hazardous Pollutants
Air Pollution Control
13b
8. DISTRIBUTION STATEMENT
unlimited
19. SECURITY CLASS (ThisReport)
unclassified
21. NO. OF PAGES
N3°2
20. SECURITY CLASS (Thispage)
unclassified
22. PRICE
EPA Form 2220—1 (Rev. 4—77) PREVIOUS EDITION is OBSOLETE
-------
-------
-------
£
------- |